UNIVERSITY OF CINCINNATI
Date:______
I, ______, hereby submit this work as part of the requirements for the degree of: in:
It is entitled:
This work and its defense approved by:
Chair: ______
VIRTUAL HERITAGE
RECONSTRUCTION: THE OLD MAIN CHURCH OF CURITIBA, BRAZIL
A Thesis submitted to the
Division of Research and Advanced Studies of the University of Cincinnati
In partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE
In the School of Architecture and Interior Design of the College of Design, Architecture, Art, and Planning
2004
by
Jose Manoel Kozan
B. in Architecture, Universidade Federal do Paraná, 1984
Committee Chair: John E. Hancock Committee Members: Patrick Snadon Elizabeth H. Riorden VIRTUAL HERITAGE RECONSTRUCTION:
THE OLD MAIN CHURCH OF CURITIBA, BRAZIL
Jose Manoel Kozan University of Cincinnati 2004
ABSTRACT
The “virtual reconstruction of vanished heritage” is now a widespread practice around the world, due to the growing capacities of digital media to replicate and interpret lost or inaccessible cultural sites. Each effort within this growing industry has its unique technical, artistic, and interpretive challenges. This effort describes a computer-model reconstruction of The Old Main Church of Curitiba, a demolished 18th Century structure in a city in Southern Brazil. Because there was only a single reliable image of the building, the central argument of this thesis concerns the documentary and interpretive processes necessitated by this situation, involving historical research and digital reconstruction.
The church demolition began in 1875, and among the remaining historical data the most reliable source of its formal features is one photograph taken in 1870. This key image was submitted to a digital rectification process based on the dimensions of a surviving clock from the church. The results provided information that allowed a virtual reconstruction and 3D visualization of the Old Main Church. Additional interpretive support came from old paintings, drawings, and maps, as well as analogies with similar contemporaneous buildings.
General research on the Internet about existing practices in virtual heritage provided examples of approaches for collecting, interpreting and representing information about completely or partially vanished buildings. The analysis of these diverse cases enabled the development of a specific methodology for data elaboration using image edge detection and architectural elements simulation techniques.
Currently, worldwide discourse about virtual reconstruction has shifted its main focus from technological issues and limitations to questions of data access and origin, and to decision- making within the modeling process. In particular, the production of 3D models of vanished structures introduces the problem of bias, since information must often be added to complete missing data. Usually the origin of these elements is not acknowledged. The present research, however, proposes a visualization method for classifying the data being presented, through a colored layer system that defines the levels from certainty to uncertainty in the generated model.
Key words: Heritage, Virtual Reconstruction, Single Image, Rectification, Photogrammetry, Church, 18th Century, Curitiba, Brazil.
i
ii DEDICATION
To my wife Iara, and to my sons Kenneth and Kevin,
for their support and encouragement.
iii ACKNOWLEDGEMENTS
I am deeply in debt to my advisor Prof. John Hancock for his insightfulness, giving me the opportunity and support to develop the Old Main Church virtual reconstruction. I would also like to express my sincere thanks to the other committee members, Elizabeth H. Riorden and Patrick
Snadon, for their valuable comments and insights. My gratitude extends to the University of
Cincinnati for providing the means to conclude this research.
During the long course of the research I encountered many people that I would like to thank for their help and valuable information. Many thanks to Maria Ester Teixeira Cruz, Igor Chmyz,
Key Imaguire, José La Pastina Filho, Suely Deschermayer and Edilberto Trevisan. They were involved with history, archaeology, architecture and restoration. They took their time to enlighten me about details never published and shared their knowledge on the subject without any other interest but the desire to help.
My research would not be so comprehensive without the many hours spent in archives looking for the pieces of the puzzle. Thank you to the Museu Paranaense, Fundação Cultural de
Curitiba, Círculo de Estudos Bandeirantes, Instituto Histórico e Geográfico do Paraná,
Biblioteca da Câmara Municipal de Curitiba and Arquivo Público do Estado do Paraná for making available your collections.
Particular thanks are due to the Arquidiocese de Curitiba, Exército Brasileiro (5ª RM-5ª DE),
Instituto de Pesquisa e Planejamento Urbano de Curitiba, Departamento de Comunicação
iv Social da Prefeitura Municipal de Curitiba, who provided information from their archives. The
vital task of measuring the remaining clock of the Old Main Church was accomplished with the
help of the Departamento de Arborização da Prefeitura Municipal de Curitiba. Many thanks to
Roberson Mauricio Caldeira Nunes for his patience in looking for old photographs, suggesting
sources and helping out whenever necessary, and to Kubit for providing the software for the
rectification process.
I would like to express my gratitude to Oldemar Blasi, for sharing his knowledge and expertise, in a continuous steady guidance throughout my local research. His valuable insights were funda mental to find new directions that led to important contacts and essential sources.
The research has been discussed every day during the past two years with my wife, Iara
Beduschi Kozan, who, besides sharing all the ups and downs of the long process, dedicated
innumerous hours to the 2D and 3D modeling of the church. Every single line of the façade was
patiently analyzed and the results would not be the same without her meticulous and dedicated
work.
v TABLE OF CONTENTS
ABSTRACT...... I DEDICATION...... III ACKNOWLEDGEMENTS ...... IV TABLE OF CONTENTS ...... 1 CHAPTER I INTRODUCTION AND BACKGROUND ...... 3 INTRODUCTION ...... 3 CASE STUDY ...... 7 Background...... 7 The Town...... 9 The Old Main Church ...... 9 The Towers ...... 11 The Demolition ...... 12 The Legacy...... 14 The Lateral Altars and Retables...... 16 The Site ...... 21 DATA ORIGIN...... 22 Character of Information...... 23 Information Density...... 25 Object Properties...... 26 GLOBAL SOURCES...... 26 Internet Research ...... 27 Websites Classification...... 29 CHAPTER II VIRTUAL HERITAGE...... 31 HERITAGE ...... 31 VIRTUAL RECONSTRUCTION ...... 33 VIRTUAL HERITAGE...... 36 UNCERTAINTY ...... 40 CHAPTER III OLD MAIN CHURCH RECONSTRUCTION ...... 44 DATA RESEARCH ...... 44 Texts...... 45 Images...... 46 Primary Sources...... 49 Secondary Sources...... 56 Reproductions ...... 58 Maps...... 60 MAP REFERENCES...... 61 SINGLE IMAGE ...... 64 Image Selection...... 66 Image Geometry...... 67 DATA ELABORATION ...... 69 References...... 71 Rectification...... 73 Edge Detection...... 77 Simulated Data...... 78 DIGITAL 3D MODEL ...... 82 UNCERTAINTY REPRESENTATION ...... 83 CONCLUSIONS ...... 86 APPENDIX A – CHURCH TIMELINE...... 88 APPENDIX B - LOCAL DATA SOURCES ...... 94 APPENDIX C - INTERNET LINKS ...... 96 REFERENCES...... 99
2 CHAPTER I Introduction and Background “Rather than exclude ourselves from recent advances in communications, multimedia, and other interpretive technologies, cultural resource professionals must strive to use them in new and innovative ways.” Nowakowski, 1998
Introduction
Architectural design and visualization have been in a rapid and continuous change in the past decades due to constant development in computer hardware and software technologies. The evolution of CAD, Computer Aided Design, played a vital role in this transformation, acting upon every stage of the design process.1 From the preliminary conceptual schemes to final presentation images, the production of projects using digital tools evolved to prior unthinkable quality levels and completion timing standards. Advanced computer visualization procedures also enabled the simulation of distinct design alternatives, enriching the decision-making and allowing the exploration of conditions previously detected only after building conclusion.
The association of an architectural project’s needs with digital three-dimensional data manipulation led to the development of electronic architectural models: virtual representations of buildings and details yet to be built, with its existence bound to computer screens or media outputs. The visualization of these models evolved from schematic wireframe shots to realistic looking images, featuring attributes such as surface color, textures and lighting casting actual shadows. The process follows a path from the virtual to the real: the digital model precedes the concrete construction.
1 Heuvel (2000) mentions other CAD applications not related to design, such as manufacturing, facility management and quality control. Even though points that CAD is “in the first place, associated with software packages for storage, manipulation, and increasingly for visualization of 3D data” (p. 1).
3
Another approaches began to use architectural visualization in the opposite direction: real buildings foregoing virtual models, in situations where the digital representation displays the reconstruction of partial or completely vanished buildings.2 Many structures that ceased to exist, now enduring solely through drawings, photographs, films or merely written words have benefited from such procedure. Digital reality technologies created an ‘electronic life’ for such buildings, bringing them back to existence through virtual walkthroughs and representations that provided an opportunity to transcend time and space, allowing the exploration of historical artifacts as never before.
Computer graphics applications provided the most powerful visualization tool for understanding the characteristics of architectural heritage (Quintero, Neuckermans, Van Balen & Jansen, 1999).
By managing sets of three-dimensional data and transforming the complex binary code into intelligible images, innumerous virtual reconstructions appeared in the past years, from Greek temples to Egyptian monuments and ancient Indian settlements.
The process to develop an architectural heritage virtual model depends on the available database providing the necessary dimensions. In some situations a sound set of architectural drawings is accessible, but the usual scenario are incomplete datasets that need to be examined to generate additional building information. Theoretical interpretations based on “documentation, archaeological evidence, or conditions known to have existed at similar sites” in addition to the
2 Kent Larson produced a distinct computer visualization composed of realistic imagery of some Louis Kahn’s projects that were never constructed, called “Unbuilt Ruins” at http://architecture.mit.edu/~kll/
4 present situation of the building, are the basic resources for three-dimensional modeling (Van
Scoy, 2000, p. 1).
The interpretative task in digital reconstruction projects remains a questionable issue. Different points of view, contrasting data sources and even distinct technical skills generated diverse virtual models of the same heritage example in recent years. Although divergent versions for the same heritage site also occurred in hand drawn reconstructions, the fast growing dissemination of images “without the normal supporting documentation found in academic papers” (Ogleby,
1999b, p. 1) is leading to attempts in defining regulating criteria for generation, presentation and distribution of heritage virtual reconstruction media.3 The pioneer reconstruction projects in the
1990’s provided the necessary “technological and theoretical preconditions” to develop scientific heritage representations, and now is the time to address accuracy and authenticity to produce models with “historical credibility” (Frischer, Niccolucci, Ryan & Barceló, 2000, p. 5).
The growing number of claims to data sources validity and accurate model representation demonstrates that “scholars active in the field of cultural heritage” (Frischer et al., 2000, p. 1) have gone beyond the first excitement with new technological innovations. More conscious reconstruction approaches are now concerned with data sources validity and aware to the “fact that the virtual models are not ‘transparent’ in respect to the initial information” (Forte, 2000, p.
2).
3 Ogleby (1999b) shows different Stonehenge interpretations, including hand drawn illustrations and computer generated images.
5 For some buildings, as of the Old Main Church of Curitiba, a single picture comprises the unique reliable dimensional data repository available for a virtual reconstruction. In such situations, the only choice is the use of ‘single image’ techniques to restore the probable geometric properties of the vanished heritage. The process proved to be efficient in several cases, as demonstrated by innumerous successful examples worldwide and the support of a special working group dedicated to single image reconstruction at the International Committee for Architectural
Photogrammetry (CIPA).4
In face of the virtual heritage community’s claim for scientifically based results, the present research stresses the importance of document references to establish a solid base for the church reconstruction. Besides providing a thorough list of unpublished images and artifacts of the Old
Main Church of Curitiba, a method is proposed to visualize the levels of uncertainty detected in the generated imagery of the virtual reconstruction.
The intention to provide a comprehensive analysis of the lost dimensions of the Brazilian church led to the use of data derived from analogy and interpretation of other historical evidence, as clearly stated in the method description. Although a photograph was the distinctive data provider for the project, other unclear and less detailed images, such as maps and paintings, added minor information to the database regarding elements not shown in the main photo. The contribution of other sources was minimal and the term ‘single image virtual reconstruction’ can fairly be employed, as the dimensions derived from the photograph rectification prevailed over other data sources.
4 For an updated list of ‘Single Image’ projects visit the International Committee for Architectural Photogrammetry (CIPA) website at http://cipa.icomos.org
6 Case Study
“A nation must have its own history … should not be ashamed of bearing it … Who disdains it does not deserve taking part in the great common experience of modern civilization." D'Andrade, 1880.
Background
The city of Curitiba was settled in the year of 1654, with the official village establishment occurring in 1693. At the heart of the original settlement there was a primitive chapel. It was replaced by a main church, which was completely demolished in 1880.
Figure 01. Curitiba location in Brazil. Figure 02. Plaza in downtown Curitiba where the (Source: Author) Old Main Church stood. (Source: Author)
The church’s origin goes back to 1714, when the community started raising a new building to
replace the crumbling wood chapel. From the outset of its construction up to the year 1875, the
most significant events of local history occurred at the church. Besides being a place for devotion
and prayer, its interior sheltered the Local Council, where crucial debates were held and
important decisions were frequently made. These institutions laid the groundwork for the growth
and consolidation of Curitiba.
7 However, in spite of its eminent role in local events, the available data sources about the Old
Main Church’s formal features are reduced to some exterior images and few written notes left by
European travelers. Some artifacts from the church such as one of the tower clocks still remain but its architectural plans were lost. Even its exact location in the plaza is uncertain, as no archaeological excavations were conducted at the site. Some sacred statues, religious artifacts and the lateral altars are the remaining pieces of its interior.
The historical information about the Old Main Church is scattered through several places. No compilation of the existing data about the building was ever made. As old documents confirm, during its lifespan the church was plagued by a succession of construction problems, which were not properly solved as they happened. Even though the church was the most visible and praised building in town, the local community continually deferred solutions for its construction troubles. After the complete demolition of the building its furnishings were spread throughout several places, receiving various levels of maintenance.
Two photographs taken in 1870 represent unique trustworthy sources of information about the church, and without a closer observation the pictures seem to be the same, as they were shot from almost the same angle in the same day. Some older paintings and drawings provide the initial configuration of the building and its formal evolution through time. The interior and outside dimensions remain imprecisely acknowledged through texts and two old maps.
8 The Town
The patron saint of Curitiba is Our Lady of the Light (Figure 3).5 According to an old legend, in a village to the east, a statue of a saint was constantly found facing west every morning. At that time, the inhabitants were afraid to venture westward because Kaingang Indians dominated it.
However, after gaining support from the Tingui tribe, villagers started exploring western lands.
Astonishingly Kaingang Indians, a known warlike tribe, welcomed the settlers as they arrived, throwing towards the ground their weapons as a peace gesture. An Indian chief established the location for the settler’s new village, and thus went the legendary foundation of Curitiba.
Figure 3. Primitive image Figure 4. Painting by Euro Brandão (1983) showing of Our Lady of the Light the pillory placement in Curitiba in 1668. (Source: Museu Paranaense) (Source: Museu Paranaense)
The Old Main Church
For many years, the villagers went to mass at a graceless wood and clay chapel, erected at the location determined by the Indian chief. In 1714, the Village Council appointed Lourenço de
Andrade to manage the construction of a new church. The local inhabitants were required to
5 The English translation for the city patron saint’s name is ‘Our Lady of the Light’, i.e. the Virgin Mary, the mother of Jesus. The name currently used is Nossa Senhora da Luz dos Pinhais de Curitiba: ‘Our Lady of the Light of the Pines of Curitiba.’
9 provide wood and clay shingles for the building. On November 16th, 1720, after seven years of construction, the image of Our Lady of the Light (Figure 5), brought from Portugal, entered the
Old Main Church.
Figure 5. Second image Figure 6. Cover of the Old Figure 7. First page of the of Our Lady of the Light. Church’s first register book. Old Church’s register book. (Source: Fundação Cultural (Source: Fundação Cultural (Source: Fundação Cultural de Curitiba) de Curitiba) de Curitiba)
By the beginning of the 18th Century, Curitiba was a small village with modest buildings, mostly
made of clay bricks and façades displaying double windows and a single front entrance door.
However, for the construction of the Old Main Church the builder used masonry (cut stones) and
finished it with clay stucco. In 1723, the church received a bell.
Unfortunately the building proved to be precarious from its beginning. The chosen site was a
low, marshy spot. Soon the walls cracked and the church needed continuous restoration work.
The community became frustrated with the constant problems and neglected it. The main
economical activity for most citizens had been gold mining. Once that ended, the village was
very poor, and this is one cause of neglect of the church. Later the yerba maté6 economic cycle and cattleman activity gave the region a new spirit of economical growth.
6 South American shrub that produces leaves used for preparing tea.
10
Figure 8. Curitiba in 1827. Watercolor by Jean-Baptiste Debret. (Source: Secretaria de Estado da Cultura, 2001)
The Towers
The church’s walls fissures and foundation problems became evident in 1825, when a lightning hit the building during a mass service. The cross was destroyed and a wall shattered. In 1852, the church would suffer another lightning hit the roof. To reinforce the construction local authorities decided to add two side aisles and two towers to the building, under the government of Zacarias de Góes e Vasconcellos (Figure 12).
Figure 9. Curitiba in 1855. Watercolor by John Henry Elliot. (Source: Secretaria de Estado da Cultura, 2001)
11 The Demolition
When Adolpho Lamenha Lins (Figure 13) became the President of the Paraná Province in 1875, the church’s destiny was the main concern among local citizens. More than a century of repair work and a major addition did not improve the structural conditions of the building, which remained treacherous. The interior walls started releasing great pieces of stucco under heavy rains and threatening a probable collapse, posing a danger to the population. The towers addition proved to accelerate the appearance of fresh fissures in the façade.
Figure 10. The Old Main Church Figure 11. Four main fissures Figure 12. Provincial President of Curitiba in 1870. in the church façade. Zacarias de Goes e Vasconcellos. (Source: Fundação Cultural (Source: Composite after (Source: Museu Paranaense) de Curitiba) Fundação Cultural de Curitiba)
In an attempt to solve the problem the local government asked to experts and engineers to examine the church and present a document with an analysis about its structural conditions. That the building should be demolished became a common opinion among the experts but promptly the community disagreed, as a strong emotional feeling linked them with the old building. For generations the place held under its roof innumerous and unforgettable ceremonies, and besides baptisms and marriages, the decaying church also held the remains of family members, laid to
12 rest on its ground and walls. A spiritual communion between the people of Curitiba and the Old
Church could not be easily broken.
Figure 13. Provincial President Figure 14. Catholic priest Figure 15. Rosario Church received Adolpho Lamenha Lins. Agostinho Machado Lima, the Old Main Church’s images in (Source: Fundação Cultural church principal in 1875. December 1875. de Curitiba) (Source: Museu Paranaense) (Source: Author)
Despite the outcry of the population, the Provincial President supported the demolition recommendations. In December 19th, 1875, the religious artifacts were removed in a procession of tears and mourning to the Rosario Church (Figure 15), which received the legacy of the Old
Main Church. The demolition started in December 22nd, 1875 and was finished during 1880.
Figure 16. The Old Main Church towers vanished from the urban perspective. (Source: Fundação Cultural de Curitiba)
13
Figure 17. Tiradentes Plaza in downtown . Figure 18. Old building on the opposite The Old Main Church probably was located side of the Tiradentes Plaza. in front of the cathedral. (Source: Author) (Source: Author)
The Legacy
In the plaza where the Old Church once stood, only a few elements bring memories related to what had once been the most important building in town. Outside the plaza information about the legacy of the demolished building is scattered through innumerous sources and a comprehensive and detailed account is not available. The remaining artifacts of the church represent the concrete and visible memory of the old building. One of the tower clocks provided the necessary dimensions to generate the scaled image of the church, the first step for the virtual reconstruction process.
Figure 19. Detail of the façade of an old building Figure 20. Detail of a ceremonial cross from the located near the Old Main Church site. Old Main Church. (Source: Author) (Source: Author)
14 The presented legacy of the Old Church was classified as: clock, lateral altars and retables, sacred images, other remains and site. Next are displayed images of artifacts encountered in each category.
The Clock
The remaining clock (Figure 21) currently is installed on the façade of the Ordem Church
(Figure 22). This clock is labeled number ‘1’ in Figure 23; the second clock in position ‘2’ was lost. An enlarged image shows that the clocks had different lettering layout and size, and this allows us to identify which is the existing clock.
Figure 21. The remaining clock Figure 22. The clock Figure 23. The clocks original from the Old Main Church. at the Ordem Church. location. (Source: Author) (Source: Author) (Source: Composite after Fundação Cultural de Curitiba)
Suely Deschermayer directed the restoration of the existing clock in 1993 (Figure 24), removing it from the Ordem Church tower and applying new layers of painting to the fading numbers, marks and abraded surface (Figure 27). The original clock mechanism still remains at Ordem
Church (Figure 28), although a new system is responsible for the movement of the hands.
15
Figure 24. The old clock before Figure 25. Clock rear Figure 26. Clock removal restoration. view. of the Ordem Church in 1993. (Source: Suely Deschermayer) (Source: Suely Deschermayer) (Source: Suely Deschermayer)
Figure 27. Old clock during restoration in 1993. Figure 28. The original clock mechanism. (Source: Suely Deschermayer) (Source: Suely Deschermayer)
The Lateral Altars and Retables
The two lateral altars and retables of the Old Main Church were made in Portugal during the 18th
Century and brought to the Old Church in 1759 by Afonso Botelho de Sampaio e Souza. After
the authorization for the church demolition in 1875 they were transferred to the Rosario Church,
where they remained until the early 1930’s. The altar segments were reassembled and restored in
the 1970’s, and currently stand side by side in a permanent exhibit at the Memorial da Cidade in
Curitiba. Each set has more than 115 parts, with an approximate overall dimension of 3.0 meters width by 6.40 meters height.7
7 The Ordem Church altar may contain some parts of the main altar of the Old Main Church. The dimensions of the church altar are: 6.2 meters width by 5.90 meters height.
16
Figure 29. Some altar parts before Figure 30. Schematic diagram Figure 31. Lateral altar after the restoration. of the lateral altar parts. after restoration in the 1970’s. (Source: Governo do Estado do (Source: Governo do Estado (Source: Governo do Estado Paraná) do Paraná) do Paraná)
Figure 32. Restoration works. Figure 33. Lateral altar pieces. Figure 34. Altar assembly. (Source: Governo do Estado do (Source: Governo do Estado (Source: Governo do Estado do Paraná) do Paraná) do Paraná)
Figure 35. Lateral altar detail. Figure 36. Both altars and retables at the (Source: Governo do Estado do Paraná) permanent exhibit in the Memorial da Cidade de Curitiba. (Source: Author)
17 The Sacred Images
The majority of the sacred images of the Old Main Church are exhibited at the Museu de Arte
Sacra de Curitiba. Some exceptions are: Nosso Senhor dos Passos is at the local cathedral and
the primitive image of Our Lady of the Light is at the Museu Paranaense (Figure 40). The second image of Our Lady of the Light of the 18th Century is a polychromatic painted wood sculpture. Figures 37 and 38 show the image before the restoration in the 1970’s, and Figure 39 presents a detail of the renovated statue.8
Figure 37. Second image Figure 38. Image base Figure 39. Our Lady of the Light. after of Our Lady of the Light before the restoration. restoration in the 1970’s. (Source: Governo do (Source: Governo do (Source: Author) Estado do Paraná) Estado do Paraná)
Other still extant icons from the Old Church are: Nossa Senhora das Dores, wood painted from
the 18th Century (Figure 41), Senhor Bom Jesus dos Pinhais, 17th Century (Figure 42) and São
Miguel Arcanjo, wood painted of Portuguese origin of the 18th Century (Figure 43). A Christ
image was under restoration at the Museu de Arte Sacra de Curitiba in 2002.9
8 Prefeitura Municipal de Curitiba (1994) describes the worldwide veneration of Our Lady of the Light and how the devotion migrated to Brazil and later to Curitiba. 9 Prefeitura Municipal de Curitiba (1993) mentions an image of São Benedito from the 18th Century.
18
Figure 40. Primitive Figure 41. Nossa Senhora Figure 42. Senhor Figure 43. São Miguel Our Lady of the Light. das Dores Bom Jesus. Arcanjo. (Source: Governo do (Source: Author) (Source: Fundação ) (Source: Fundação Estado do Paraná) Cultural de Curitiba) Cultural de Curitiba)
Other Remains
The following is a list of known preserved objects of the Old Main Church of Curitiba.
1. Support for the processional banner of Our Lady of the Light (Figure 44).
2. Silver ceremonial cross of the 18th Century (Figure 45).
Figure 44. Processional banner support. Figure 45. Ceremonial cross. (Source: Fundação Cultural de Curitiba) (Source: Fundação Cultural de Curitiba)
3. Processional banner made of satin embroidered with silk thread, of the 19th Century
(Figures 46 and 47).
4. Wood Box for the processional banner.
5. Processional cross of the 17th Century in silver with wood support (Figure 48).
19
Figure 46. Processional Figure 47. Processional banner detail. Figure 48. Processional banner. (Source: Author) cross. (Source: Composite after (Source: Fundação Fundação Cultural de Curitiba) Cultural de Curitiba)
6. The porcelain urn in Figure 49 is referred as the ‘tower vase’ (Prefeitura Municipal
de Curitiba, 1993). Magnifying the church’s photographs did not allow identifying
the vase presence in the frontal or lateral façade.
7. Gold painted wood sacrarium (Figure 50).
8. Pair of polychromatic wood supports, appearing in an old photograph of the Rosário
Church at Destefani (1993) (Figure 51).
9. Two gold painted wood vases (Figure 52).
Figure 49. Vase Figure 50. Sacrarium. Figure 51. Wood Figure 52. Gold vase (Source: Fundação (Source: Author) support (Source: Author) Cultural de Curitiba) (Source: Author)
20 The Site
Possibly, the most relevant and yet neglected legacy of the Old Main Church still lies unexplored. The plaza where the church was located remains an important and busy area of downtown Curitiba, and the church’s foundations still may lie underground, as specific archaeological excavations were never conducted in the site.10 In January 16th, 1999 while digging the north boundary of the plaza in front of the cathedral, workers of Copel, an energy provider company, found human bones in a ditch. A local television station, Rede Paranaense de
Comunicação, captured images of the excavation and presented a report about it. Figures 53 through 56 were captured from the video shown on the local television station. Citations in old documents refer to burial procedures carried inside the church nave.
Figure 53. Construction fence in front Figure 54. Workers in the old church site. of the cathedral. (Source: Rede Paranaense de Comunicação) (Source: Rede Paranaense de Comunicação)
10 I. Chmyz (interview, December 2, 2002) affirmed that all attempts to conduct archaeological excavations on the site failed due to a lack of agreement with local authorities. Even following construction works in the area it was not possible.
21
Figure 55. Ditch opened in the plaza. Figure 56. Human bones found in the ditch. (Source: Rede Paranaense de Comunicação) (Source: Rede Paranaense de Comunicação)
Data Origin
The research collected material from local and global sources. To acquire historical data about the Old Main Church of Curitiba local references provided the content while the Internet supplied theoretical issues, technical information and examples of similar projects.11 The gathered database allowed one to:
o Analyze historical pictures, drawings and documents related to the church.
o Review the available literature focusing on the Old Main Church.
o Know the opinion of local historians about the history of the church and its artifacts.
o Discuss with architectural history scholars the building features.
o Evaluate worldwide projects of heritage virtual reconstruction.
o Explore the most recent articles about computer visualization applied to archaeology.
o Form a global view about the worldwide organizational structure related to virtual
heritage.
11 Lagerqvist (1999, p. 1) uses a distinction between data and information. The first is defined as the “input stimuli entering the cognitive process” while the later accounts for a “portion of input stimuli retained as knowledge.” A third step in the process refers to intelligence, or the “portion of information used by the decision process for decision making or action.” A hierarchy is established among the three concepts, where ‘information is a subset of data” and “intelligence is a subset of information.”
22 Locally the investigation led to rare books, old documents, artifacts and interviews, which provided directions to images, objects and facts related to the church.12 The research process was informed by concepts outlined by Lagerqvist (1999), which distinguishes three characteristics of the acquired information: character, density and object properties.
Character of Information
The character aspect of information “concerns the type of data which are to be acquired” (p. 10), and pointed to possible input origins and modes of data representation to be considered:
a. Archives – unpublished.
b. Literature – published data.
c. Verbal – data recorded through interviews.
d. Graphic – photographs, measured drawing and sketches.
e. Analytical numeric – accurate measurements.
f. Sampling – extract small parts of the object.
Archives constituted the basic iconographic data source, as the published literature about the church concentrates on repetitively reproducing only a few images of the old building.13 Even though the primary photograph for the research was published in several occasions, the image used for the photogrammetric transformation was scanned from a negative in a public collection.
Contact with original documents from the 18th and 19th Century of the Biblioteca da Câmara
12 See Appendix B for a complete list of local data sources for the research. 13 See on Chapter III under the topic ‘Reproductions’ a list of published images of the Old Main Church of Curitiba.
23 Municipal de Curitiba and Catedral Metropolitana de Curitiba confirmed later, more accessible versions of historical occurrences, but delivered no graphic material such as drawings or sketches of the architectural plans of the church.
The published literature provided a comprehensive view of the historical evolution of the Old
Church and the basic information for understanding the facts that led to its demolition. A few spatial references about the building provided content that was not applicable to the recovery of dimensions of the Church. The book Pintores da Paisagem Paranaense (Secretaria de Estado da
Cultura, 2001) presented a complete collection of all available historical paintings portraying the
Old Main Church that were produced before its destruction.
Verbal data gathering represented a vital source for the research, allowing building a network of scholars holding unpublished information about the Old Church. Some initial meetings with restorer Maria Ester Teixeira Cruz and historian Oldemar Blasi furnished fundamental directions and further contacts, as archaeologist Igor Chmyz and historian Edilberto Trevisan. The
Church’s analysis by architectural historians José La Pastina Filho and Key Imaguire Jr. supplied functional and technical aspects of the building. The images and information about the restoration of the remaining clock originated from restorer Suely Deschermayer.14
The remaining three “information character” features defined by Lagerqvist (1999) do not indicate where to look for data, but on how the content is represented through graphic sources.
The innumerous photographs, sketches and “measured drawings with relevant level of accuracy”
14 In 1993 I had the opportunity to examine the church’s clock during the restoration works at Deschermayer’s studio. It was the first direct contact with the clock, which later I measured at the Ordem Church’s tower in November 2002.
24 are examples of “data represented by graphic description” (p. 10), 15 and are highly typified throughout the research, as the several photographs, church lithograph and the 1850 map.
“Analytical numeric” data for the rectification process derived from the 1857 map, the measurements conducted in the remaining clock and architectural projects of similar churches provided by architectural historians. Data originating from sampling procedures was not collected.
Information Density
The next major aspect defined, as “information density,” refers to detail levels of the acquired data and its direct relation to recording methods. The topic influenced the research by stressing the importance of data analysis during the process in order to identify segments of the collected material that required additional investigation, i.e., situations with inadequate information. In some instances the data acquisition procedure demanded revision and substitution by a more effective approach, while other situations required repeating data acquisition after redefining the goals for a specific data source. The process aimed at increasing “information density” and its dynamic flow is schematically defined in Figure 57.
Figure 57. Procedure to increase information density. (Source: Author)
15 Lagerqvist (1999, p. 10) makes a subdivision of “graphic description” in two categories: “analogous” exemplified as photographs and “analytical” illustrated by X-ray, infrared (IR) and ultraviolet images (UV).
25
Object Properties
The current research did not address the topic, which refers to “the qualities of the recorded object” (p. 10) categorized by Lagerqvist (1999) as: “material properties” and “intangible properties.”
Global Sources
In contrast with printed literature that currently provides relatively few examples of virtual reconstructions, the Internet is an effective provider of worldwide recent researches and examples of architectural virtual heritage, supplying substantial theoretical issues that allowed identifying the current agenda, understand past evolution, basic concepts and related vocabulary.
The World Wide Web also granted contact with technological topics and reconstruction projects with diverse methodology, accessing a variety of resources and answering decisive questions about photogrammetry related techniques for data image acquisition.
The terms ‘computer graphics’ (CG) and ‘virtual reality’ (VR) sometimes are used as synonymous, and the knowledge of the specific meaning of each one helps clarify the research scope towards the type of computer-generated reconstructions under investigation. Frischer et al.
(2000) makes a distinction between the two often-misinterpreted terms based on ‘interactivity’, a characteristic of virtual reality environments but not necessarily present in computer graphic presentations: “While all VR could be called CG, not all CG constitute VR” (p. 4). Forte (2000) also divided digital reconstructions in archaeology “two macro-classes of applications” (p. 1):
26 1. Computer graphics – comprised of rendered images and “synthetic movies” with
higher degrees of photorealism.
2. Virtual Reality – applications depicting less detailed model but allowing different
levels of user interaction and immersion.
Aware of the above distinctions, the research aimed at finding non-interactive projects using previously generated still images and animations for product presentation.
Internet Research
The initial Internet search used the terms ‘virtual’ and ‘reconstruction’ in the following engines:
Altavista, Excite, Google and Yahoo. After some trials the obtained results were compared and
Google was selected for the next attempts. During the procedure decisive references appeared on internal links of the visited websites, in a process that gradually moved from the ‘search engine based’ to ‘suggested links based’ information hunt. A virtual reconstruction community exists inside the Web and it references itself, pointing to recognized data sources of organizations, educational institutions and even small projects.
The amount and type of information acquired in each website presented diverse results. Some websites proved to offer fundamental content, others were immediately discarded and a final group formed a database of related material for future reference.
1. Initially 21.5 megabytes of digital data presented the basic references, acting as
pointers for a more precise search. The downloaded files were not arranged into
27 categories, but placed into a folder with the webpage or document title as the basic
reference. Some encountered links at the very beginning later proved to be major
references.
2. The amount of information grew exponentially in a second search. A total of 542
megabytes in files was analyzed to detect possible missed links that deserved to be
checked. As some promising websites provided several links, the next step aimed at
checking these sites.
3. The third phase provided valuable sources, as it detected the main websites dealing
with heritage virtual reconstruction content. It was a comprehensive and time
consuming stage, collecting 736 megabytes of data that later was divided into four
categories:
a. Documents - comprising Acrobat and Word files originated from multiple
websites.
b. Webpages - individual webpages of diverse origin.
c. Multiple - total of 66 folders with several webpages of the same website
d. Websites - 55 sites were scanned for data.
4. The last search improved and refined the coverage of the previous steps, providing
328 megabytes in retrieved files.
28 Websites Classification
The research testifies that heritage virtual reconstructions involve a mass of knowledge from different disciplines. The diverse nature of the encountered webpages shows that the process ought to have a historical and archaeological foundation before proceeding to 3D modeling.
Computer graphics knowledge is only one among several pre-requisites and skills involved in the production of images of vanished buildings.
The analysis of the downloaded material allowed visualizing an organizational structure that encompasses worldwide structures related in a more direct or indirect degree to heritage virtual reconstructions. The sources found in the Internet fit in the following categories:16
1. Historical and Technical Support
1.a. Organizations – enterprises dedicated to disseminate information and provide
support to heritage conservation and virtual reconstruction. Although presenting
distinct individual objectives the group presents the common characteristic of
providing some kind of assistance without directly producing virtual reconstruction
content.
1.b. Publications – magazines, journals and periodicals devoted to disseminate
information on historical, archaeological and computer graphics researches. Some
provide an online database, while others allow Internet access to partial content of the
printed edition.
16 See Appendix C for some links to virtual reconstruction relates websites.
29 1.c. Research Groups – usually originating from within universities, present a broad
range of objectives and obtained results. They contribute to specific segments of the
reconstruction process and generate practical and directly applicable knowledge.
1.d. Events – meetings, congresses and conferences held to discuss virtual
reconstruction topics, providing the latest news on researches under development.
1.e. Other – general public oriented media also proved to supply information through
the promotion of major virtual reconstruction projects, usually advocating in favor of
heritage-conscious initiatives.
2. Content Producers
2.a. Single Projects – concentrate on exploring one heritage site or object.
2.b. Multiple Projects – exhibit two or more virtual reconstructions.
3. Production Tools
3.a. Software – the research focused on applications which used data acquisition from
images, mainly packages able to generate the necessary database for the 3D modeling
phase through digital rectification of photographs.
3.b. Hardware – acknowledged and innovative equipments are employed for
acquiring data for reconstruction projects and cited in several webpages.
30 CHAPTER II Virtual Heritage
Heritage
“Our cultural and natural heritages are both irreplaceable sources of life and inspiration. They are our touchstones, our points of reference, our identity.” UNESCO, 2000
Heritage is considered to encompass more than the material archaeological retrieval of past
evidence. Heritage also includes tradition, artistic expression and “cultural evidences” (Roussou,
2000). UNESCO (United Nations Educational Scientific and Cultural Organization) defines
heritage as “our legacy from the past, what we live with today, and what we pass on to future
generations” (UNESCO, 2000, para. 1). In both definitions, the concept is not restricted to man-
made concrete artifacts, but includes natural landscape sites and abstract cultural manifestations.
The ‘Convention Concerning the Protection of the World Cultural and Natural Heritage’ was
signed in 1972 and revised in 1992 by more than 170 countries, regulating the procedures to
choose and maintain the growing list of more than 600 sites, selected for significant historical,
artistic or scientific value (Boehler, Heinz, Scherer & Siebold, 2000). The selected locations are
spread throughout the world and considered to be best instances of human cultural production
and natural events.17 Sites bearing “outstanding universal value” (UNESCO, 2000, para. 7) may be selected to become part of the World Heritage List and receive support for protection.
UNESCO (2000) also stresses the importance of protecting not only the select group of sites on the World Heritage List, but other important national heritage examples as well.
17 In contrast to building focused reconstructions, Boehler et al. (2000) present methods to develop the visualization of landscape sites and a classification of cultural landscapes. Defend the visualization of the context in which the building is inserted, as “no cultural heritage object can be understood without taking the surrounding landscape into account”.
31
Preservation and conservation of cultural and natural sites is the focus of several major organizations with worldwide influence in heritage issues.18 Although presenting a diverse and sometimes overlapping range of scopes, these enterprises have in common the goal to promote and provide resources for the preservation of sites around the world, also endorsing and bestowing recognition upon distinctive projects. Some initiatives deal with more technical content while others are politically oriented; and even though they differ in scale and objectives the following organizations all share in UNESCO support, and in the task of shaping the content of the world’s overall heritage agenda:19
1. World Heritage Committee (WHC) – responsible for selecting sites to be included on
the World Heritage List.
2. World Heritage Center (WHC) – operates at the national level, by expanding the
number of countries ratifying the convention, and encouraging heritage sites
preservation and new additions to the World Heritage List.
3. World Heritage Information Network (WHIN) – acts as a global information
exchange center between worldwide sites and collaborators.
4. World Cultural Report (WCR) – conducts research and policy development, keeping
track of trends and new procedures exerting impact on cultural issues.
5. Organization of World Heritage Cities (OWHC) – provides information at the
municipal level to help implement the World Heritage Convention.
18 Forte (2000) addresses the problem of the representation of archaeological landscapes classifying the archaeological landscapes according to cognitive parameters. 19 See on Appendix C the links for the cited organizations.
32 6. International Council on Monuments and Sites (ICOMOS) – UNESCO’s main
advisor for heritage sites issues, embracing scientific committees to research and
propose standard preservation and restoration procedures.
7. Centre for the Study of the Preservation and Restoration of Cultural Property
(ICCROM) – intergovernmental organization providing technical support for
restoration and conservation.
Virtual Reconstruction
“In a new age, new techniques. It's a simple matter of good sense.” J.K. Huysmans, 1879 (Cited by Shukla, 1999)
The process of creating images for the visualization of historical buildings is not exclusive to the digital age. Recent computer generated imagery represents a modern version of previous hand drawn reconstructions (Barceló, 2000), and likewise old image production techniques aim at producing visual outputs from the acquired or generated three-dimensional information (Forte,
2000). Heritage virtual models disseminated through the Internet, and in numerous websites provide a vast number of examples with diverse objectives and presentation technologies.
Reconstructions such as the Temple Site at Phimai20, the Northwest Palace of Ashur-nasir-pal II
at Nimrud21 or the visualization of the Jerusalem's Temple Mount22 are examples of projects reflecting a multitude of historical interpretations and a wide range of locations and time periods.
20 Virtual reconstruction of a United Nations Heritage Site in Thailand, at http://www.ucalgary.ca/EV/designresearch/projects/phimai/index.htm 21 The palace is one of the several heritage models developed by The Learning Sites, at http://www.learningsites.com 22 Developed by Urban Simulation Team at UCLA, at http://www.ust.ucla.edu/ustweb/ust.html
33 An initial contact with web-based heritage models shows recognizable presentation levels, but according to Addison (2000), few of the pioneer virtual reconstruction projects fulfilled the anticipated goals. This frustrated the desired expectations of professionals and the general public alike, due to extravagant publicity and issues related to data accuracy and non-realistic imagery.
He continues, by stating that the initial digital initiatives were trapped by these constraints; though this did not inhibit the second phase that is currently underway, that has now managed to achieve users’ expectations, thanks to strong and rapid advances in digital technologies.
Frischer et al. (2000) positions the first reconstruction projects as a necessary step towards the present evolutionary stages, affirming that “early failures were as instructive and useful as the successes” (p. 4).23 Recent technological development grew from previous mistakes, now constituting a “new hope” for virtual heritage according to Addison (2000, p. 1), who groups the emerging technologies into the following domains:
1. 3D documentation – related to methods and processes applied to site investigation
and information acquisition, “everything from site surveys to epigraphy.”
2. 3D representation – involves “historic reconstruction to visualization.”
3. 3D dissemination – responsible for providing access to the generated content, “from
immersive networked worlds to "in situ" augmented reality.”
Although the importance of historic preservationists being updated with current technologies is recognized (Nowakowski, 1998), technological advances are not unquestionably incorporated to
23 “But pioneering efforts that fail can be just as valuable as those that succeed” (Frischer et al., 2000, pp. 5-6).
34 the virtual repertoire. Not only new insights and techniques come into play, but also novel
problems and challenges take part in the process. Currently, the difficulties encountered in the
lack of realism of the initial projects have been replaced by the excessive realism of recent
projects, and the historical credibility of the images remains a constant and vital aspect to be
addressed and scientifically defined.
A common agreement among several analyses of the virtual reconstruction process refers to the
initial project phase, or data acquisition. Barceló (2000) defines four necessary stages to produce
a heritage geometric model: data research, pre-processing, parameter estimation and modeling.
Addison (2000) uses a similar classification with different terms, and adds one more step to the
information flow: presentation and dissemination, addressing the problem of maximizing the
public exposure of the content produced by the three-dimensional visualization.24
Even though the virtual model is “a representation of some (not necessarily all) features of a concrete or abstract entity” (Barceló, 2000, p. 1), many positive results are accounted to virtual reconstructions. Arnold (2000b) claims the advantages of the digital access to historical sites as a possibility to “experimentation without risk to the original” (p. 2). Addison (2000) stresses that the importance of virtual heritage applications lies in helping to protect and understand our heritage. Although other advantages are usually embedded in reconstruction project descriptions, one aspect mentioned by Forte (2000) stands out and deserves special attention among the
24 Addison (2000) groups the process in the following stages: “documentation – getting the data in,” “representation – authoring, modeling, and rendering” and “presentation – dissemination.”
35 formulated objectives for a heritage reconstruction: provide “a real increase of information” and
not merely high impact graphics.25
Virtual Heritage
“Virtual heritage can be an invaluable tool, but if not applied wisely has the potential to do as much harm as good.” Addison, 2000
Roussou (2000) describes virtual heritage as a six step process involving “synthesis, conservation, reproduction, representation, digital reprocessing, and display with the use of advanced imaging technology” (p. 1), while Stone and Ojika (2000) relate it to the interactive possibilities offered by computer image generation, or “computer-based interactive technologies,” that allows us to “record, preserve, or recreate artifacts, sites, and actors of historic, artistic, religious, and cultural significance” (p. 1). Containing either more or fewer defining specifications, other definitions stress the possibilities offered by virtual heritage through the use of recent technological advancements. For example, Bernd (2000) accounts for
“a new field of technology,” using “innovative multimedia technologies for acquisition, documentation, research, preservation, restoration and promotion in archaeology and cultural heritage” (p. 23).
The term is similar to ‘virtual archaeology’, according to Forte (2000) a “digital reconstructive archaeology” which is “applied to the reconstruction of three-dimensional archaeological ecosystems” (p. 1). Ryan (2000) relates both concepts by mentioning that virtual archaeology
25 Masuch et al. (1999) regards virtual reconstruction “as a continuous evolutionary process in which the 3D model experiences constant refinement” (p. 2).
36 expanded from its initial uses as a tool solely related to “excavation recording” and presentation, to a wider set of applications covering “visualization and presentation methods to the
‘reconstruction’ of past environments, including buildings, landscapes and artefacts” (p. 1).26
Independent of the term under analysis (‘virtual heritage’ or ‘virtual archaeology’), the common ground for researches seems to be Ryan’s concern “to approach Virtual Archaeology as a means of producing tools that aid understanding” (p. 2).
Directly linked to technological resources, virtual heritage has benefited from the recent fast growing stream of digital advancements originating from academic, government and industry labs (Addison, 2000). Easily accessible websites provide innumerable virtual reconstruction examples, and even general public media publishes articles on selected major projects.27 Arnold
(2000a) confirms that “virtual heritage in practice is growing” (p. 9), and along with the increasing volume of projects the possible gains are spreading through different user groups and types with a widely varied spectrum of goals and expectations.
Historically virtual reconstruction projects basically targeted three separate groups: the preservationist, who expected to encounter documentation, the historian who sought interpretation, and the general public, which required visual realism (Addison, 2000). As mentioned, each user category holds its set of demands, expecting diverse and specific results that determine the degree of success of a reconstruction project. Equally, virtual heritage
26 Ryan (2000) mentions that Paul Reilly presented the ‘initial uses’ for virtual archaeology in 1990. 27 Some articles in news publications were retrieved from: 1. Computers help archaeologists reconstruct lost splendors, USA Today at http://www.usatoday.com/life/cyber/tech/cta691.htm 2. Students 'rebuild' lost German synagogues on video, CNN Home Page at http://www.cnn.com/TECH/science/9809/24/t_t/synagogue.reborn/ 3. Como era há 2 000 anos. Arqueólogos usam realidade virtual para reconstruir monumentos históricos em ruínas. Veja Online at http://www2.uol.com.br/veja/idade/exclusivo/150502/p_067.html
37 contributes in different ways for each group. For example, Roussou (2000) summarizes some possible contributions of public oriented applications that could not be entirely applied to the other user categories:
o Make accessible non-extant and unreachable sites.
o Visualize various interpretations and theories.
o Maintain interest in heritage examples.
o Become a potential distance-learning tool.
o Enhance informal education.
Ultimately the benefits and contributions of virtual reconstructions have a direct relation with the definition of the audiences, which are divided in two main categories according to Brown et al.
(2000, p. 16): “informed audience” with prior knowledge and “people who need to be informed".
Addressing both audiences’ needs, Boukhari (2000) developed an elaborate list of functions and possible benefits that “digitization and animation techniques” (p. 40) may present when applied to heritage reconstruction.
1. “Recreate the appearance of famous historical figures and sites that have crumbled
away.”
2. “Restoration of fragile or badly damaged artworks.”
3. “Provide useful tools for archaeologists and curators of museums and sites.”
Also mentions the testing of possible alternatives in virtual reconstructions and improvement of available information databases, leading to the fourth and fifth functions:
38 4. “Simulation of archaeological hypotheses.”
5. “Acquisition of new knowledge.”
A sixth function relates to the preservation of heritage threatened by human activities such as
“wars, pollution, urban expansion and theft”:
6. “Back-up digital images of monuments and objects.”
The next topics are associated with the contribution of virtual models to real reconstruction initiatives:28
7. “Guide in the real-life reconstruction.”
8. “Fund raising tool for a real reconstruction project.”
A ninth function refers the use of a computer model for structural analysis of the original real building or object:29
9. “Structural simulation.”
And finally the last usage possibility cites general public applications intended to:
10. “Educate and entertain visitors” in museums.
28 "The virtual reconstitution, which drew on a mass of historical documents and photographs, was used as a guide in the real-life reconstruction of the church, which should be complete in 2003. The computer model was also used in a televised appeal to Germans that raised money for the project" (Boukhari, 2000, p. 41). 29 "The best thing," says Coignard, "was being able to test the equilibrium of the statue. With the colossus, whose fragments were very worn, we managed to come up with a stable structure and a base which fitted perfectly and in theory enabled the statue to resist even earthquakes" (Boukhari, 2000, p. 41).
39 It may sound strange, but the growth potential for novel uses of virtual heritage is not virtual, but
real. The previous topics provide a general view of some of the possible applications, which can
easily be expanded as new virtual models and technologies are developed, or even as on-going
projects revise their results. The key feature to be aware of in this whole process relates to the
potential uses for the technology, which are described by Arnold (2000a) and point to the future
direction of increasingly precise visualization and increasingly thorough understanding of the
past.30
Uncertainty
"I was accused of having distorted ‘the truth’ even though nobody knew what the truth was. Scientists are scared of pictures." Génévriez (Cited by Boukhari, 2000)
The historical reliability of the 3D models produced by the growing number of virtual reconstructions constitutes a major concern expressed by several researchers worldwide. The necessity to recognize whether an image portrays a scientifically based version of a historical building or artifact comprises a fundamental question affecting all virtual heritage projects. The search for realistic images “with little or no concern for the inherent uncertainty of the data sources” (Ryan, 2000, p. 1) and the lack of clear references to the data that has been used
(Ogleby, 1999b) have contributed to diffuse skepticism throughout the field, toward the results presented by virtual reconstructions. As Barceló (2000, p. 1) points out, the “visual model is an
‘interpretation’ of archaeological data, and it is not readily apparent how one gets from the dig to the interpretation.” Brown et al. (2000) testifies to the pressure exerted over the computer-
30 A growing potential for virtual reconstruction is related to “web-based tourism and e-visiting” (Arnold, 2000a, p. 9).
40 generated images by radical approaches supporting conventional art methods as more acceptable
for the task, while digital “graphics realism can be perceived as misleading” (p. 17).31
Authenticity problems in virtual heritage originate in two phases of the reconstruction process: production and visualization. Strothotte, Masuch and Isenberg (1999a) address issues arising from the project production, detecting common circumstances in virtual reconstructions that require the addition of questionable data to the 3D models. First is “uncertainty,” defined as “the absence of information due to some reason,” a usual situation in archaeological research.
Uncertainty is divided in two categories: “imprecision” meaning that “the existence of a certain feature can be safely assumed, but not its dimensions,” and “incompleteness,” referring “to the fact that certain information is unavailable” (p. 3). The second issue mentioned is “design decisions” as analogies and deductions, which act as data generators for the process, necessary procedures that permit modeling hidden or missing elements of the building.
The visualization of virtual heritage models raises another set of questions. Primary among them is the absence of image components permitting the viewer to distinguish “what elements of the model are known with certainty and which are hypothetical” (Frischer et al., 2000, p. 4).32 As
Forte (2000) states, “the models are not ‘transparent’ in respect to the initial information” (p. 2).
The problem is aggravated by the fact that photorealistic images tend to transmit the false
31 A comment by Cooper about the presentation of the Ayutthaia virtual reconstruction directly mentions the credibility issue: “you have shown us some marvellous ways of telling lies” (Ogleby, 1999a, p. 19). 32 Ogleby (1999b) points to the problem of historical virtual reconstructions without clear reference to the data used in the project.
41 impression that “the object depicted actually exists” or that the “information is correct and
contains a high degree of certainty and accuracy” (Strothotte et al., 1999a, p. 1).
Aware of these debatable issues regarding virtual reconstructions, that consider heritage
visualizations to be “fraught with biases” and ultimately imply that “there are no images without
agendas” (Tringham, 2000, p. 8), several researchers present alternatives to enhance the
credibility of virtual heritage imagery. Ryan (2000) addresses the “validation and authentication”
of virtual reconstructions and proposes a “consistent metadata standard that could be applied at
all levels from the outline description of a project down to the individual elements of a 3D
model” (p. 3). Van Scoy (2000) also suggests attaching to the results of a virtual reconstruction a
“standard set of metadata elements” which would account for the explanation of the “sources of
information used in constructing a model” and “modeling decisions” related to “unknown
features of the structure” (p. 4).
Ogleby (1999b) proposes the adoption of a “Verisimilitude Index”, i.e., a way of “authenticating
VR reconstructions of historic architecture and landscapes” (p. 4). Strothotte et al. (1999b)
defends the use of “non-photorealistic rendering” to avoid “unintentional visual fixations” (p. 2)
and has developed a method where photorealistic imagery is replaced by less detailed images
that can be rapidly modified and tested.33 Heuvel (2000) treats the issue from a different point view, integrating CAD (computer-aided design) with photogrammetry and asserting that the combination of both processes can contribute to the “acquisition and storage of non-geometric or
33 Strothotte et al. (1999b) address problems when using "off-the-shelf" software to generated photographic quality images from historical buildings. Call the attention to hypothesis testing in virtual reconstruction of lost buildings, asserting that in a text based approach it is possible to leave no doubt about which presented historical data has been scientifically proved and what is supposition, while photorealistic presentations lead the audience to believe in the data shown by the image, making no clear distinction between facts and hypothesis.
42 semantic information” (p. 9). Finally, Masuch, Freudenberg, Ludowici, Kreiker and Strothotte
(1999) regard virtual reconstruction “as a continuous evolutionary process,” that should not
exclude realistic visualizations for public display, and propose “alternative visualization
methods” (p. 2) as line drawing rendered images for scientific presentations.34
These researchers call for the need to adopt a further formatting standards for the presentation of virtual heritage results, and to address the two interested public segments defined by Brown et al.
(2000): “informed audience - prior expertise” and “people who need to be informed” (p. 16). The fact that each visualization is ultimately “an interpretation of information” (Tringham, 2000, p.
8) is recognized, but this does not invalidate the reconstruction process, as “knowledge about the investigated building is gained” (Masuch et al., 1999, p. 3) and transmitted by the rendered images.35 Even though uncertainty permeates the process there is always the possibility of
developing another model, a new interpretation, for a heritage site. As Shirley (2000) affirms:
“We shouldn't lose access to multiple interpretations -- the results are richer” (p. 6).
34 Masuch et al. (1999) mention an interactive system under development, which is intended “to model the source of data (excavation, deduction, analogy, assumption) and for visualizing geometric models taking certainty into account” (p. 3). 35 “The procedures of reconstruction rely less on coordinate precision but instead involve a variety of data sources including archaeological evidence, history, architectural analogy and the expertise of the individual” (Ogleby, 1999b, p. 1).
43 CHAPTER III Old Main Church Reconstruction
Data Research
“No other data source is such an objective document as a historical photograph.”
Wiedemann, Hemmleb and Alberts, 2000
The first and essential task for any virtual or real reconstruction is “to gather data of existing conditions” (Addison, 2000, p. 1). The amount and type of information determines “the level of detail which might be derived by photogrammetric approaches” (Wiedemann et al., 2000, p. 2), ultimately reflecting on the 3D model features. Virtual reconstruction processes have relied on several data sources to retrieve the basic numerical information that computer-modeling processes requires: from complex processes as shading depth deduction, texture gradient information and stereo image analysis (Barceló, 2000), to conventional procedures as site excavations and textual descriptions.
Based on the origin of the information, Masuch et al. (1999) established the following classification for project data:
1. Excavation – provides real artifacts.
2. Deduction – derived from findings on the site.
3. Analogy – deductions originated “from similar buildings of the same architectural
period” (p. 2).
4. Assumption – based on the fact that some building features should be present, had to
exist under determined situations.
44
Artifacts and building remains from excavations were not available for the Old Main Church of
Curitiba virtual reconstruction. Still extant objects portray individual preservation efforts and are
not resultant of archaeological explorations in the site. 36 Even though an extensive research was conducted, two remaining photographs are currently the soundest data source on the Old Church.
To complete the information gaps during the modeling of the church, data originated from other sources, as analogy and assumption were necessary. This is not an uncommon procedure and often is used in virtual reconstructions. Barceló (2000) points out that additional simulated data can be created “to see what archaeologically cannot be seen”, and the project success will depend on “how to add knowledge in a systematic way”, or how to ultimately elaborate and interpret data for defining “the most probable ‘shape’ of the object to be simulated.”
Texts
Textual information describing the church’s dimensions found in historical documents and old publications is vague and incomplete. Dimensional observations as “very big” (Illustração
Paranaense, 1930) or “thirty steps long” (Saint-Hilaire, 1978, p. 71) provide no support for further accurate reference.37 Written opinions were also divergent as Bigg-Wither’s (1974)38
36 Moreira (2000) details the investigation conducted to find out the sacred statue of Senhor Bom Jesus (Figure 42) and the historical path followed by the image since the day it was removed from the Old Main Church in 1875. 37 Prefeitura Municipal de Curitiba (1993) mentions the original Saint-Hilaire’s book as: Saint-Hilaire, A.F.C.P. (1851). Voyage dans les provinces. Quatrieme partie. Voyage dans les Provinces de Saint Paul et Sainte Catherine. Paris: Bertrand. 38 The original title of Bigg-Wither’s book is: Pioneering in South Brazil: Three years of forest and prairie life in the Province of Paraná.
45 vision of a church with modest architectural features and D’Andrade’s (1880) account for a good and appraised internal architecture.
Among publications after the building demolition, three deserve attention due to the amount of information about the Old Main Church:
1. Leão (1994) presents the main historical facts related to the Church’s history along
with his memories of the day the sacred images were removed from the church,
December 19th, 1875.
2. Destefani (1993) treats the Old Church in a type of addendum to the book’s main
topic: the builder of the cathedral that succeeded the Old Main Church.
3. Wachowicz (1993) devotes about one third of its content to narrate the history of the
church, viewed as one of the shelters for the local patron saint: Our Lady of the Light.
Wachowicz39 presents only textual content on the church with references to historical documents, while Destefani centers on a chronology and reproduces two images of the church.
Images
“Images are among the most usual primary data for archaeological research.” Barceló, 2000
The Old Main Church project began researching images with the objective of finding information that could help on the virtual reconstruction, rejecting pictures failing to provide
39 I called Wachowicz to discuss the Old Main Church Virtual Reconstruction Project, but unfortunately he died before scheduling a meeting.
46 such data. At a certain point it there emerged a comprehensive image list about the church’s
formal features, elements and historical evolution. The collected images were chronologically
ordered and according to Schuhr and Kanngieser (1999) classified as existing from historical
origin.40
The crucial act to acquire information about the formal characteristics of the Old Main Church of
Curitiba was performed at the moment photographer Adolph Volk shot the camera trigger to take
the historical photographs in April, 1870. Registering a local event the photographer
unknowingly created a historical document and dimensional vault, which came to be the only
available source to provide reliable data to recover the probable dimensions of the later
demolished church. Two photographs remained of that apparent sunny afternoon in the 19th
Century, taken at 1:20 pm and 2:30 pm, as shown by the right tower clock.41
Viewing Angles
Diverse viewing angle were pictured in the Old Main Church images. In more or less detail each building façade appears in at least one image. Unfortunately the main photograph used for the virtual reconstruction shows only two sides of the church (South and West), and no other reliable view of the remaining façades were found (North and East). The map on Figure 58 establishes an approximate position of the observer for acquiring or generating the presented images of the Old
Main Church.
40 Schuhr and Kanngieser (1999) classify the images used in conservation according to its source in: “existing (historic) and currently taken and/or even planned images” (p. 1). 41 Original issues of the newspaper Dezenove de Dezembro (August 1868 and July 1876) cite other photographers working in Curitiba during the 19th Century: Marcos e Linhares (Photographia de America) and Francisco Heiler.
47
Figure 58. Approximate viewing points of the church’s images over the 1857 map. (Source: Composite after Fundação Cultural de Curitiba)
1. 1827 – Watercolor by Jean Baptiste Debret 2. 1855 – Watercolor by John Henry Elliot 3. 1855 – Lithograph by John Henry Elliot 4. 1865 – Watercolor by John Henry Elliot 5. 1865 – Watercolor by Joseph Keller 6. 1870 – Photograph by Adolph Volk 7. 1870 – Photograph by Adolph Volk 8. 1872 – Watercolor by William Lloyd 9. 1876 – Photograph by unknown author 10. 1887 – Lithograph based on a photograph Date
Innumerous sources were researched aiming at finding Old Main Church’s images produced before its demolition in the 19th Century: paintings, drawings and photographs. As the original
48 architectural drawings are no longer available, these could provide a unique data repository for the lost church’s dimensions.
Based on the church’s demolition date the images were separated in two groups:
1. Primary sources - created during the church’s existence.
2. Secondary sources - produced after its demolition.
Images dating after 1880, year of the completion of the demolition works, were considered to be later interpretations about the Old Main Church, and probably were based on the primary images, which were produced while the church still existed. Most of the material consists of paintings and drawings, valuable for information on the building historical development but providing scarce useful architectural data for the virtual reconstruction.
Primary Sources
Except for figures 64, 65, 66 and 69, all other primary sources are cropped portions of the original images to enlarge the church’s size and facilitate its visualization.
The first recognized image to represent the Old Main Church (Figure 59) is a watercolor by Jean-
Baptiste Debret dated from 1827, developed during an expedition to Brazil to acquire material for a book project Voyage Pittoresque et Historique au Brésil (Secretaria de Estado da Cultura,
2001). The rear of the Church is portrayed in a schematic view along with north and west
49 façades. The building presents a series of four volumes: the nave, two different and consecutive double roof areas towards the back, and a small single roof addendum in the lateral, near the right side of the front façade. Debret did not represent any window in the church’s nave and near volumes.
Figure 59. Watercolor by Jean-Baptiste Debret (1827). (Source: Secretaria de Estado da Cultura, 2001)
A second image appears only almost thirty years later, in 1855 (Figure 60). The watercolor by the American cartographer John Henry Elliot presents a composition of five views from Curitiba.
The upper scene shows the Old main Church as a central element in an area surrounded by single level buildings, different from the position it occupied in a corner of the plaza (Berberi and Sutil,
1997). Elliot’s knowledge on perspective representation was deficient (Secretaria de Estado da
Cultura, 2001), a fact that can be observed in the plaza proportion and church’s position, which differs from the upper lateral location in the plaza. The south and west façades show the evolution of the lateral addition in Debret’s version to a double stories construction surrounding the building. The third rear volume from the 1827’s watercolour is not presented.
50
Figure 60. Watercolor by John Henry Elliot (1855). (Source: Secretaria de Estado da Cultura, 2001)
The magazine Revista do Paraná published in 1887 a lithograph by John Henry Elliot dated from 1855 (Figure 61), showing a reverse view of the church, its north and east façades.
Comparing it with the previous watercolor of the same year, the construction behind the church’s nave seems smaller, and each roof segment reveals one single pane instead of two subdivisions.
Figure 61. Lithograph by John Henry Elliot (1855). (Source: Secretaria de Estado da Cultura, 2001)
Ten years later, in 1865, another watercolor by John Henry Elliot (Figure 62) shows the church
in its final configuration with the two towers. Viewing the south and west façades the painting
represents the church in a very schematic style, lacking a more precise proportion to its elements.
51
Figure 62. Watercolor by John Henry Elliot (1865). (Source: Secretaria de Estado da Cultura, 2001)
The German engineer Joseph Keller produced a watercolor in the same year as Elliot, 1865
(Figure 63), presenting a proportional perspective of the church’s west façade and displaying a
broad city view.
Figure 63. Watercolor by Joseph Keller (1865). (Source: Secretaria de Estado da Cultura, 2001)
The photograph that Adolph Volk shot on April 28th 1870 (Figure 64) is the first known to
register an external view of Curitiba. It shows many of the elements added during the festivities
to celebrate the return of local soldiers from the Brazil-Paraguay War (Leão, 1994). The image is
part of an old postcard, showing that the photograph was modified to enhance some lines and
52 details of the building. The south façade holds the clock in the right tower displaying the local hour of the picture shot, 1:20 PM, while the left tower clock exhibits the same hour in the first and the second photographs, so it is assumed that it was broken, and that the elapsed time between the two shots was one hour and twenty minutes, as demonstrated by the operating clock.
Figure 64. First photograph by Adolph Figure 65. Cropped version of the Volk (1870). second photograph by Adolph Volk (1870). (Source: Fundação Cultural de Curitiba) (Source: Fundação Cultural de Curitiba)
Adolph Volk took another photograph on April 28th 1870, which shows manual addition of two buildings on the right of the image (Figure 66). A cropped version of this image (Figure 65) is the usual published view of the Old Main Church of Curitiba, and basis for the virtual reconstruction project, as this portion of the image presents no signs of later modifications.42
42 A distinction is made in photogrammetry between interior and exterior orientation. ‘Interior’ refers to the knowledge about the camera parameters, and ‘exterior’ “describes the position and the viewing direction of the camera in a superior object coordinate system” (Wiedemann, 1997, p. 1). During the research none registered information was found regarding interior or exterior orientation of the photographs of the Old Main Church taken by Adolph Volk.
53
Figure 66. Second photograph by Adolph Volk (1870). (Source: Museu Paranaense)
The engineer William Lloyd painted a watercolor in 1872 showing the rear of the church with
the top part of the north and west façades (Figure 67). The image confirms the existence of two
opening in the rear of the church’s two towers.
Figure 67. Watercolor by William Lloyd (1872). (Source: Secretaria de Estado da Cultura, 2001)
The last known photograph of the Old Main Church was taken in 1876, when it was already under demolition (Figure 68).
54
Figure 68. Last known photograph showing the Old Main Church (1876). (Source: Fundação Cultural de Curitiba)
An exception should be made in the classification of a singular image of the Old Main Church of
Curitiba. Although published seven years after the church’s demolition, the image is referred to as a “lithographic copy of a photography of the Old Main Church of Curitiba” (Revista do
Paraná, 1887). Some elements of the building that are hidden in the two known photographs are displayed in the lithograph. The date of the photograph referred by the magazine is unknown, and the lithograph influenced the production of other images of the church (Figure 69).
Figure 69. Lithograph based on unknown photograph. (Source: Revista do Paraná, 1887)
55 Secondary Sources
Later interpretations based on old images of the Old Main Church appeared after the demolition
completion in 1880. Since then, several publications and artists have provided different versions
of the previous views of the vanished old building.
The first published image of the Old Main Church in the 20th century found in the research is a
drawing based on the 1887 lithograph put out by the magazine Illustração Paranaense in 1927
(Figure 70). Later, in 1929, the same magazine published a drawing by Levino Fanzeres, similar
to the 1870 photographs but in a mirrored angle (Figure 71).
Figure 70. Illustração Paranaense (1927). Figure 71. Illustração Paranaense (1929).
In the book Iconografia Paranaense (Carneiro, 1950) there appears a drawing similar to the photographs taken in 1870, establishing a new sun direction (morning period) and ignoring two frontal openings near the towers at ground level and a second entrance door in the lateral façade
(Figure 72).
56
Figure 72. Church’s view from the plaza. Figure 73. Church and surroundings. (Source: Carneiro, 1950) (Source: Mazzarotto, 1956)
A Arquidiocese de Curitiba na sua História (Mazzarotto,1956) sets out another image
presumably based on the Revista do Paraná lithograph. The artist added elements to the image
left portion, creating a finish line for the church construction and adding a building just behind
the church (Figure 73).
The Museu Paranaense collection has a large model of the city of Curitiba urban area, as it existed in 1875-76. All the buildings, including the Old main Church, are represented in a reduced scale (Figure 74), making it the only known three-dimensional model produced of the church.43
Figure 74. Model of downtown Curitiba Figure 75. Oil painting by Guido Viaro. in the 19th Century. (Source: Museu Paranaense) (Source: Museu Paranaense)
43 The model was built around 1939 and based on the Curitiba map of 1857 (Figure 79). A modification that occurred only after 1875 in the block just behind the old church already appears in the model.
57 Guido Viaro probably based the church’s oil painting on the lithograph published by the Revista do Paraná in 1887 (Figure 75).
In 1993, to prepare the local cathedral for the festivities of its 100th year and celebrate Curitiba’s foundation, the building that replaced the Old Main Church underwent outside maintenance works. During the process the exterior scaffolding was decorated with the demolished church’s image, commemorating the previous core of religious activities in the city (Figure 76).
Figure 76. Scaffolding in front Figure 77. Church drawing by of the cathedral in 1993. Euro Brandão. (Source: Prefeitura Municipal (Source: Trevisan, 2000) de Curitiba)
In the book Curitiba na Província (Trevisan, 2000) there was published a schematic drawing of the Old Church by Euro Brandão (Figure 77).
Reproductions
All the other encountered images of the Old Main Church were reproductions of the previous images. The following is chronological list of the found published images:
58 1. Illustração Paranaense (1928) - Adolph Volk’s first photograph.44
2. História do Paraná (Pombo, 1929) - Adolph Volk’s first photograph.
3. Illustração Paranaense (1930) - mirrored version of the previously published Levino
Fanzeres’ drawing.
4. Viagens na Comarca de Curitiba em 1820 (Saint-Hilaire, 1938) - John Elliot’s watercolor
of 1855.
5. Viagens na Comarca de Curitiba em 1820 (Saint-Hilaire, 1938) - John Elliot’s lithograph
of 1855.
6. Iconografia Paranaense (Carneiro, 1950) – Drawing John Elliot’s watercolor of 1855
(J.B. Groff private collection)
7. Iconografia Paranaense (Carneiro, 1950) – John Elliot’s watercolor of 1865.
8. As Artes e o Artesanato no Paraná (Carneiro, 1955) – Guido Viaro’s painting.
9. A Cruz do Alemão (Destefani, 1993) – Adolph Volk’s second photograph and lithograph
published at the Revista do Paraná.
10. Boletim do Instituto Histórico, Geográfico e Etnográfico Paranaense (1993) - Adolph
Volk’s second photograph.
Other encountered images are:
1. Museu Paranaense – postcard with Adolph Volk’s first photograph.
2. Museu Paranaense – two reproductions of watercolours by John Elliot (1855 and 1865).
44 The text in the magazine presents in two sections dating mistakes: first states that the photo was taken in 1871, and secondly, the date is cited as April 27th. Leão (1993) confirms the date as April 28th, 1870.
59 Maps
Imprecise overall dimensions and the location of the Old Church were found in two city maps drawn in 1850 and 1857 (Figure 78 and 79).45 Both show the church location in the northeast corner of the main plaza of Curitiba during the mid 19th Century. The street limits around the central plaza still remain similar to the boundaries defined in earlier centuries and portrayed by the old maps (Berberi & Sutil, 1997). The maps proportions and angles differ from a current and precise map provided by a local planning authority, Instituto de Pesquisa e Planejamento
Urbano de Curitiba (IPPUC). Probably the differences are associated with the imprecise and limited techniques available for their production and not with concrete changes in the city blocks around the plaza. Trevisan (1996) states that the plaza’s proportions, size, shape and directions constitutes a legacy from the 17th Century.
Figure 78. Map of Curitiba in 1850. Figure 79. Map of Curitiba in 1857. (Source: Fundação Cultural de Curitiba) (Source: Fundação Cultural de Curitiba)
45 The date originally written in the map is 1857, but verbal sources affirm that the correct date of the map is 1861. As this information was not confirmed by written sources, the research follows using the date shown on the map.
60 Map References
Archaeological excavations did not determine the precise location of the Old Main Church, and one of the historical questions that remain unanswered is its precise location in the Tiradentes
Plaza.46 The two old maps represent the building position and dimensions, and to determine an approximate value portrayed by the maps the contours of a recent plan was overlaid on the historical maps.
The process found problems in the differences between the shapes and angles of the old maps and the current precise map derived from aerial photogrammetry. An exact match could not be done for both maps, and other overlay possibilities are open to evaluation. The 1850 map was scaled, moved and rotated until the plaza contour reached a position and size similar to the present conditions. The 1857 map was first scaled in accordance with a graphic reference available at its right bottom corner (Figure 82) and then moved and rotated. 47
Figure 80. Current map overlaid 1850 map. Figure 81. Current map overlaid 1857 map. (Source: Composite after Fundação Cultural (Source: Composite after Fundação Cultural de de Curitiba and IPPUC) Curitiba and IPPUC)
46 Even if the foundation of the church are found, according to Wiedemann et al. (2000), “experienced archaeologists,” affirm that it will be “possible to determine the position of the former building from the fundament only with a maximum accuracy of a few decimeters, because the fundament and the building may differ” (p. 3). 47 Trevisan (2000, p. 30) asserts that Pedro Taulois drawn the map in 1857 not as representing the current situation of the city plan, but as a proposal to later carried out modifications.
61 Figure 80 shows in red lines the current shape of the city blocks over the approximate scaled version of the 1850 map. The approximate dimension of the church derived from the map is
24.07 meters width by 29.00 meters length. Converting the values to the old measurement system used in the 19th Century called ‘braces’,48 the dimensions are 10.94 braces width by 13.18 braces length. The church position is shown as blocking a street, which was called Rua Fexada
(Closed Street).
Applying the same process to the 1857 map (Figure 81) the obtained dimension of the church is
22.59 meters width by 37.38 meters length, or 10.27 braces width by 16.99 braces length. The church position appears almost parallel to the present church, the Metropolitan Cathedral, but the fact may be due to the arbitrary rotation applied. A differential factor in the second map is the provided scale reference, which although it does not respond for the applied rotation angles in the overlay, supplies relatively reliable information about the Old Church’s dimensions.
Figure 82. Graphic scale at the bottom of the Figure 83. Aerial view of the plaza showing 1857 map. the Old Main Church possible location. (Source: Fundação Cultural de Curitiba) (Source: Composite after IPPUC)
48 The measure is called braça, in English brace. One brace is equivalent to 2.2 meters.
62 Both overlays are not precise. The process tried to accommodate the more appropriate fit among the lines of the old and recent maps. The many differences in the maps in conjunction with the lack of reference points and historical notes make it difficult to establish which areas should be considered to match the present blocks perimeter. Nevertheless the origin of the 1857 map, drawn by an engineer, turns it into a more reliable historical document.
The width of the present cathedral is described as 22.20 meters (Moreira, 2000), similar to the measure of the Old Main Church in the 1857 map: 22.59 meters. It may be a close coincidence, as no precise written historical record about the Old Church dimension was found and the map- acquired dimensional values depend on interpretative factors. Figure 84 explores the possibility of a precise numeric distance between both churches and the parallelism of the walls. Figure 85 establishes a probable area were the rear wall of the Old Church could be located in accordance with information provided by the overlay on Figure 81. The presented hypothesis should be used with care, as the correct answer remains open to archaeological research.
Figure 84. Possible integer distance between Figure 85. Probable original location of the rear the Old Church and the present cathedral. wall of the Old Main Church. (Source: Author) (Source: Author)
63 Single Image
“The task of reconstructing objects such as buildings from photographs is receiving increased attention in the effort to create models of valuable architectural sites.” Liebowitz, Criminisi and Zisserman, 1999
Among all types of images, photographs represent a reliable procedure for documenting and retrieving building dimensions and formal features. Its use as an architectural data container is not a recent approach. Historically photographic methods hold an established tradition for building documentation, as demonstrated by the Meydenbauer Archives in Germany. Dating from 1885 to 1920, the German collection presents metric images of about 2,000 buildings
(many of which have been destroyed during and after the Second World War) and places in
20,000 negative plates (Wiedemann et al., 2000). It represents the introduction of architectural photogrammetry by the civil engineer Albrecht Meydenbauer (Wiedemann, 1997).49
A special type of photogrammetric applications refers to situations where a single image comprises the unique source of information about a historical building. 50 Bräuer-Burchardt and
Voss (2001) provide a list of several researches proposing mathematical models and algorithms to solve “the problem of metric reconstruction using single images” (p. 1), and in such cases, a- priori knowledge about usual architectural properties such as “linearity, parallelism,
49 See at http://www.geo.tudelft.nl/frs/architec/Meydenbauer/ examples of rectification projects using the Meydenbauer archives. 50 Ogleby (1999a) presents the virtual reconstruction of the ancient Asian city of Ayutthaya, and describes the importance of photogrammetric methods to generate data for the computer models of vanished heritage: “Photogrammetry offers a rapid, accurate method of acquiring three-dimensional information regarding cultural monuments. Combining the measurements obtained from the photogrammetric record and 3D CAD models offers the scope to recreate historic environments” (p. 664).
64 perpendicularity, and symmetry” (p. 11) allows one to complement missing elements as the
camera parameters and reference points in the building. 51
Although alternatives to obtain results from a single image allow proceeding to a virtual
reconstruction, Streilein and Heuvel (1999) alert one to inherent problems in a 3D reconstruction
based on a two-dimensional single image, stating that a “ monocular image alone does not
contain sufficient information to uniquely retrieve 3D information” (p. 2), and point to the most
outstanding problems: “incompleteness of the 3D object model” and “need for additional object
information” (p. 6). They also affirm that alternatives were developed to overcome the single
image limitations using “certain visual cues” (p. 2) as size, shade and distortion, with continued
efforts being directed to researches about 3D reconstructions based on a single image.52
Schuhr and Kanngieser (1999) present a draft of a systematic approach to single image in
conservation, recommending basic parameters to image acquisition, generation and uses. They
assert that the usage of single images in conservation has a wide spectrum of applications, and
makes a distinction between using images for:
1. Documentation purposes.
2. Partial or complete restorations.
51 Among the cited examples, Liebowitz et al. (1999) addresses the problem of data acquisition exclusively from photographs in cases where it is not possible to acquire any building reference measurements. Proposes algorithms for establishing data based on proportions and ratios, not on absolute scalable dimensions. The real dimension is not the goal of the presented process, and affirms that the generated relative dimensions may provide all necessary data for a 3D modeling. 52 According to Streilein and Heuvel (1999) the techniques to recover three-dimensional data from a single image are: shape- from-shading, shape-from-focus, shape-from-texture and shape-from-geometry.
65 They continue by saying that photogrammetry and remote sensing can help protect historical monuments, and classify the images in the following categories according to its source:
1. Existing from historical origin.
2. Currently taken.
3. Planned using conventional and nonconventional sensors.
Image Selection
Even though several images of the Old Main Church were found, only two photographs shot from nearby positions provide reliable data for acquiring the building dimensions. By superimposing the images it is observed that the photographer slightly changed the camera position between the shots. As the first photograph (Figure 64) evidences manual addition of lines and elements to the church, it was discarded for the rectification. Figure 66, a small photograph, showing a wider view and also manually drawn buildings, is also useless for the geometric transformation process due to its reduced original size.
The alteration problems encountered in the first photograph guided to the adoption of a high- resolution version of the second photograph as the basis for the photogrammetric process (Figure
86). The image seems to be the most faithful reproduction from the original glass plate, with several scratches and dark spots. Nevertheless, the first photograph and other images classified as primary sources were useful for visualizing hidden details in the image chosen for rectification.
66
Figure 86. Selected image of the Figure 87. Enlarged view of the clock in the Old Church. selected photograph. (Source: Fundação Cultural de Curitiba) (Source: Fundação Cultural de Curitiba)
Image Geometry
Waldhäusl and Ogleby (1994) established basic parameters to be observed when documenting buildings using non-metric cameras, defining several rules to obtain photographs with adequate dimensional data.53 Among the nine recommendations, four are observed in the selected Old
Main Church photograph, two geometric and two photographic rules:
1. The geometric aspect of the historic photograph shows a “long distance between well
defines points” and some vertical lines extending from the roof to the bottom of the
building. Waldhäusl and Ogleby (1994) consider these “minimum metric
information” and the basics for information recording control.
2. Another suggestion to “take shots from half the object’s height” is almost achieved,
as the church’s photograph was shot from about one third of the church’s height.
53 According to Waldhäusl and Ogleby (1994), the rules called “3x3 Rules for Simple Photogrammetric Documentation of Architecture” are classified in three categories: geometrical, photographical and organizational.
67 3. The hour in which the photograph was shot (2:30 PM) presents a homogeneous
illumination and probably, as with all other camera of that time, it was used with a
tripod.
4. Fortunately, due to technical limitations in 1870, the photographic process used a
large format glass plate as the negative for registering the shot.
Considering that three of the nine rules recommend providing additional information by other media that could not be portrayed in the image, the historical photograph satisfies two thirds of the six possible recommended procedures.
Hemmleb (1999) offers a simplified version of the above rules, presenting three usual setups presented by Meydenbauer photographs, and two desired geometric features observed in single images of the same archive. Analyzing the Church’s photograph according to Hemmleb’s remarks, the following characteristics and advantages are present:
1. The church photograph was not shot from the ground level, but near the desired
leveled camera position.
2. Like the German archive’s photos, the church was photographed from a corner.
3. The Old Church presents rectangular corners.
4. “On the building there exist vertical and horizontal edges” (p. 3).
The selected photograph of the Old Church could present more details and information about the building, as an ideal situation is often desired. But analyzed in accordance with the previous
68 recommendations, it demonstrates the presence of some fundamental recommended conditions for current historical buildings image registration.
Data Elaboration
"The amount of information which is recorded in photographs surpasses each construction plan or drawing." Albertz and Wiedemann, 1995
Different from a three-dimensional model drawn by hand where lines may be freely adapted to match any desired position, a digital model demands precise dimensional values to form the model surfaces. The necessary point coordinates values may be acquired from images by applying a photogrammetric process, which is “an indirect technique to acquire 3D geometric data without touching, but using images of the object” (Wiedemann, 1997, p. 1).54
Photogrammetry is not a recent process, but digital photogrammetry is recent. The main photogrammetric procedures are stereophotogrammetric restitution, bundle adjustment and generation of orthoimages. The amount of data acquired from photographs exceeds information derived from drawings and the task of data acquisition and three-dimensional digital modeling is a “reverse engineering" process (Albertz and Wiedemann, 1995, p. 1).55
Among the photogrammetric alternatives the research generated an orthophoto of the Old
Church, which is “a fully rectified base map with x and y (or east and north) coordinates for each
54 Eiteljorg (1994-1995) describes the process used in Pompeii to record and retrieve object dimensions using photographs. 55 Albertz and Wiedemann (1995) present a diagram with the digital photogrammetric process data flow in a CAD system, from image acquisition to object reconstruction.
69 pixel. This mean that if we measure distances, areas or angles on the image they will be correct”
(Fowler, 1998, p. 1).
Bräuer-Burchardt and Voss (2001) describe pre-processing procedures applied to historical
images as ‘radial lens distortion correction’ and ‘principal point transformation’.56 The operation was skipped in the church reconstruction project as the selected photograph does not show noticeable line distortions, and the applied software does not require determining ‘vanishing points’ or any other specific parameter.
Even though digital rectification relies on mathematical equations and geometric rules, the user interpretation plays a decisive and variable role in the procedure. Viewing the process through a logic perspective where a unique solution is suitable for each case is a wrong assumption, as reality delineates an interpretative operation with variable results.57 Albertz and Wiedemann
(1995) comment that “each photogrammetrically derived plan is the result of an interpretation
process, where the complex image information is reduced to some simple lines" (p. 6). Schuhr
and Kanngieser (1999) recognize that a “real obstacle for a broad application of single images in
Archaeology is the competition between objective photogrammetric mapping and subjective
Archaeologic interpretation” (p. 2). Aware of the problem, the research explains the
56 Fowler (1998) comments about lens distortion: “because of the shape of the lens, there is radial distortion on every photograph. This is simply a function of the length of distance light travels from the lens to the film” (p. 2). 57 Eiteljorg (1995) describes the application of a basic geometric property of the rectified image: The single-photo photogrammetric process permits all points on a plane to be related to the known points on the same plane. That is, with four known points on a plane, the location of any other point on that plane can be calculated. An important assumption is that one is working with a plane, a flat surface. (para. 5)
70 interpretative decisions along the process that generated the probable original dimensions of the
Old Main Church of Curitiba.58
References
When the building under analysis for an architectural image rectification no longer exists,
another source is needed to supply the minimum references to the photogrammetric process.
Hemmleb (1999) describes several possible origins for the required information:
1. Dimension of existing neighbor buildings.
2. Camera calibration data.
3. Information about the viewpoint.
4. Additional rules kept during the photography.
If the building still exists, points directly measured in the façades provide scale reference.
Unfortunately for the Old Main Church of Curitiba reconstruction none of the cited examples
may be applied. The only measurable reference that appears in the 1870 photograph is the clock
in the church’s right tower.59 A direct measurement of the clock was made in November 29th,
2002, using a lift to reach its position in the Ordem Church (Figure 88). The acquired dimensions
were used as reference to generate a rectified image of the clock (Figure 89), as it appeared
58 El-Hakim (2000) comments about diverse applications for computer reconstructions and concludes that no single solution can solve the technical necessities of all situations. "Methods to acquire the 3D data and reconstruct shapes have been progressing rapidly in recent years. However, many problems remain to be solved and no approach is suited for all applications and all types of object and environment" (p. 1). 59 The research did not encounter any type of information about the left tower clock.
71 before the restoration in 1993; providing the necessary true scale data for the digital transformation.
As in the original 1870 photograph the clock frame seems to be partially hidden by the tower wall (Figure 87), the dimensions of the clock internal white contour were defined as the main references for the front façade rectification.
Figure 88. Clock at Figure 89. Rectified clock image. Figure 90. References for the the Ordem Church. (Source: Composite after Suely lateral façade rectification. (Source: Author) Deschermayer) (Source: Composite after Fundação Cultural de Curitiba)
For the lateral façade the scale reference derived from the front façade, as no other source of information could provide dimensions of the church lateral segment shown in the photograph.
Wiedemann et al. (2000) addresses the problem found in situations where “the required data to extract 3D information directly is not available,” and considers the use of “additional techniques” as a source of “indirect 3D information.”60
The building dimensions in the 1857 map refer to the total depth of the church, but the photograph does not show the ending portion of the wall. The solution was to use already
60 Wiedemann et al. (2000) mentions the acquisition of the façade depth values by using the length of projected shadows.
72 acquired measurements from the rectification of the front façade and apply by analogy to similar
elements in the lateral elevation. For the horizontal reference the average width of columns ‘A’
and ‘B’ was applied to column ‘C’ in the lateral (Figure 90). The necessary vertical dimension
for the rectification derived from the clock’s internal white ring projection on the façade corner,
‘D’.
Rectification
The objective of the digital rectification is the production of a “geometrically rectified metric
image, from which the measurements of the façade and architectonic details can be taken in a
desired scale” (Hemmleb, 1999, p. 2). Usually the process encompasses two steps as defined by
Streilein and Heuvel (1999):
1. Creation of a rectified 2D model of the image.
2. Application of a scale factor to resample the image to a desired scale creating an
orthophoto.61
Depending on the available type of information on the object, different rectification processes
may be applied for building reconstruction from single images.62 Hemmleb and Wiedemann
61 For Wiedemann (1997), “orthoimages combine the geometric properties of plans with the contents of images by rectification” (p. 4). 62 Streilein and Heuvel (1999) mention the following “methods for the reconstruction of 2D information from a single image” (p. 3): projective transformation, combination of projective transformations, non-parametric, parametric and differential. Hemmleb and Wiedemann (1997) provide another classification for rectification methods, overlapping some of the previous procedures.
73 (1997) affirm that several factors influence the choice of an appropriate rectification procedure and the first step is to describe the surface to be rectified:
1. Plane surface – projective rectification.
2. Small uneven areas within the surface – polynomial rectification.
3. Parametric volume – unwrapping techniques.
4. Complex surfaces – differential rectification methods.
Based on the above guidelines the most suitable method to rectify the Old Main Church’s photograph is ‘projective transformation’, as the church’s façades represent a planar surface.
Another remarks by Streilein and Heuvel (1999) confirm the approach for a method where: none of the camera parameters need to be known, only the position of four control points in two dimensions of the façade are necessary and image distortion caused by wide-angle lenses is not excessive. Also, a depth variation of around 30 cm in the façade is acceptable for projective transformations.63 Based on these assumptions the software called ‘Photoplan’ was used for the rectification process. Besides providing a user friendly interface the software allowed testing geometric hypothesis in the image and delivered trustable results without requiring the knowledge of previous photogrammetric procedures.64
As the dimensions were already known from the clock measurements, the next step aimed at selecting the most appropriate location for the group of four points. Geometric relations among
63 Jansch, R.D., 1976. Genauigkeitsansprüche bei der photogrammetrischen Bauaufnahme. Arbeitsheft 17, Landeskonservator Rheinland, Reihnland-Verlag Köln, 1976, pp. 103-106, cited by Streilein and Heuvel (1999). 64 Photoplan website may be reached at http://www.kubit.de
74 the church’s elements appeared as the initial reference lines were positioned over the image. For example, depending on the chosen set of points the towers in the final rectified image could have the same height or not. The points’ location turned to be a major variable without any kind of historical information to confirm one option or the other. The solution ought to derive from the process itself
It was decided to abandon the relation between the towers as a reference, although it was the recommended option due to the outstanding distance between the points, and to attempt to use a single tower to accommodate the points. The alternative also provided imprecise and instable results, once graphically valid interpretations for positioning the four reference points led to different values for the towers height.
To avoid choosing a clearly questionable reference based on an uncertain assumption about the relation between the towers or the top friezes height, another position independent of the towers was selected. Two pairs of points separated by a small vertical distance were chosen over the façade’s central frieze (Figure 91). The reduced upright dimension between the points made it difficult to match the vertical lines in the photograph, leading to an adjustment process where small variations in the position of any selected point resulted in exaggerated changes in the vertical lines angles. Figure 92 shows the reference points in the lateral façade.
75
Figure 91. The selected horizontal reference and four Figure 92. The four control control points on the frontal façade. points on the lateral façade. (Source: Composite after Fundação Cultural de Curitiba) (Source: Composite after Fundação Cultural de Curitiba)
Figures 93 through 98 show the rectification results, presenting the original photograph, the rectified façades and the distortion applied to the initial image to obtain the rectified image.
Figure 93. Frontal façade. Figure 94. Rectified frontal Figure 95. Rectified image with (Source: Fundação façade. frontal façade parameters. Cultural de Curitiba) (Source: Author) (Source: Author)
Figure 96. Figure 97. Rectified lateral façade. Figure 98. Rectified image with. Lateral façade. (Source: Author) lateral façade parameters. (Source: Fundação (Source: Author) Cultural de Curitiba)
76 Edge Detection
Heuvel (2000) defines photogrammetric processes that generate a CAD model as ‘CAD-based’, making a distinction between approaches with different automation levels. The research did not address automated methods, choosing to produce a ‘vectorial representation’ (Wiedemann, 1997) of the rectified façades of the Old Main Church through a manual edge recognition procedure in a CAD environment (Figures 99 and 100). The resulting detected edges for each façade are shown in Figures 101 and 102.
Figure 99. Drawing over rectified Figure 100. Drawing over rectified image image of the front façade. of the lateral façade. (Source: Author) (Source: Author)
Figure 101. Detected edges on the Figure 102. Detected edges on the lateral front façade. façade. (Source: Author) (Source: Author)
77 Simulated Data
The rectified images did not provide all the necessary geometrical information to build a complete 2D model of the church, with limitations caused by the fact that several areas of the frontal façade are hidden, parts of the lateral façade are missing, and the opposite and back elevations were not available. It is recognized that “counting on images for modeling is limited because the features that can be extracted are usually fewer then the required level of details”
(El-Hakim, 2000, p. 1).
Barceló (2000) points out some issues related to information incompleteness in historical virtual models, affirming that the “archaeological record is most of the times incomplete,” and even though some reconstructions consciously target an incomplete model, the usual approach tends to complete missing historical data through “induction, deduction and analogy.” He also mentions cases where due to deficient local data the gaps were filled “with information that does not proceed from the data.”
A method towards data completion was applied for the Old Main Church virtual reconstruction, as contemporary architectural references, the existing 1857 map and non-metric views of the church provided grounds for simulating data that allowed the visualization of the church exterior.
Unlike simplified textured 3D models originated from a schematic model using the building rectified images as mapping, the Old Church architectonic elements were individually identified
78 and evaluated to accomplish a detailed database of its geometry.65 Complementary information added to the church façades originated from:
1. Architectural plans and site measurements from the São Benedito and São Francisco
das Chagas churches (Figures 103 through 108).66
2. Rectified images of the lithograph published by the Revista do Paraná in 1887
(Figures 109 and 110).
3. Rectified image of the1857 map (Figure 111).
Figure 103. São Benedito Figure 104. São Benedito Figure 105. Detail of São Benedito Church. Church front elevation. Church. (Source: Author) (Source: Instituto do Patrimônio (Source: Author) Histórico e Artístico Nacional)
Figure 106. São Francisco das Figure 107. São Francisco das Figure 108. Detail of São Chagas Church. Chagas Church front elevation Francisco das Chagas (Source: Author) (Source: Instituto do Patrimônio Church. Histórico e Artístico Nacional) (Source: Author)
65 Streilein and Heuvel (1999), Bräuer-Burchardt and Voss (2001) and Hemmleb (1999) present examples of textured 3D-models. 66 Both churches are located in Paranaguá, city located around 100 kilometers from Curitiba. The churches were chosen to complement the architectural data of the Old Main Church of Curitiba due to similar construction dates and geographical proximity.
79
Figure 109. Rectified front Figure 110. Rectified lateral façade Figure 111. Church location in façade of the lithograph. of the lithograph. the 1857 map. (Source: Author) (Source: Author) (Source: Composite after Fundação Cultural de Curitiba)
Figures 112 and 113 present the church’s rectified images after the addition of the hidden and missing lines in the front and lateral elevations. The final 2D models, and base for the three- dimensional modeling, are shown in Figures 114 and 115. Figures 116 and 117 provide a probable version for the rear elevation and roof plan of the church. The existence of windows in the rear of the church could not be confirmed. The total length of the church relied on the dimensions provided by the 1857 map, as the photographs and lithograph do not show the rear edges of the building. Simulated sections based on the rectified images defined the roof heights.
Figure 112. Simulated data for the front Figure 113. Simulated data for the lateral façade. façade. (Source: Author) (Source: Author)
80
Figure 114. Final front elevation. Figure 115. Final lateral elevation. (Source: Author) (Source: Author)
Figure 116. Rear elevation. Figure 117. Roof plan. (Source: Author) (Source: Author)
Figure 118 presents a comparison among the acquired dimensions of the church from the maps and rectified images. The difference between the front façade width obtained from the rectification and the 1857 map is 60 centimeters, or 2.65% of the building width in the map. The total lateral length was based on the map dimension, as the photograph does not show the rear portion of the church. The Old Church dimensions are close to integer values in braces: 10 by 17, or 22 meters width by 37.40 meters length.
OLD MAIN CHURCH DIMENSION SOURCE METER BRACE WIDTH LENGTH WIDTH LENGTH 1850 MAP 24.07 29.00 10.94 13.18 1857 MAP 22.59 37.38 10.27 16.99 RECTIFICATION 21.99 * 37.38 9.99 * 16.99 Figure 118. Dimension acquired from historical maps and rectification process. * The dimension derived from the 1857 map. (Source: Author)
81 Digital 3D Model
“A virtual world should be, then a model, a set of concepts, laws, tested hypotheses and hypotheses waiting for testing.” Barceló, 2000
The three-dimensional model of the Old Main Church of Curitiba relied on the final 2D model generated after simulating the missing elements in the rectified image of the photograph, allowing the visualization of “the building after a possible reconstruction” (Hemmleb, 1999, p.
5). Different methods and interpretation about the hidden features of the church, may lead to other results for the virtual model. Nevertheless it should be stressed that the presented model was based on a photogrammetric process applied to the most reliable source about the building, and as such constitutes a “virtual representation of the reality and the product of an objective deduction process” (Wiedemann, 1997, p. 4), not a “subjective fantasy of the modeler” (p. 4).
Figure 119. 3D model wireframe render. Figure 120. 3D model shaded render. (Source: Author) (Source: Author)
82
Figure 121. Similar viewing angle of Figure 122. 3D model allows new viewing angles Adolph Volk’s 1870 photograph. of the demolished Old Main Church. (Source: Author) (Source: Author)
Uncertainty Representation
“The more 'photorealistic' a synthetic image appears, the more easily it is interpreted to represent the 'truth'.” Ogleby, 2001
Strothotte et al. (1999a) mention the need of virtual reconstruction projects to “encode systematically more information about the model than just its geometry” (p. 2). The present research addresses this problem by proposing a method for the visualization of certainty levels in the images produced by virtual reconstructions. The system uses the animated GIF format to sequentially view the origin of the information presented by the model. A collection of sequential still images is grouped in a single image file, linking the final model with graphic data about its construction process. The format allows the insertion and visualization of several layers of information in a single image.
The following are the images selected for composing the animated GIF of the Old Main Church, sequentially presented in Figure 123:
1. Rectified image (Figure 94).
83 2. Detected edges over the rectified image (Figure 99).
3. Detected edges drawing (Figure 101).
4. Simulated data over the rectified image (Figure 112).
5. Simulated data drawing (Figure 114).
6. Color-coded certainty levels (Figure 125).
Figure 123. Image layers to be assembled in the animated GIF sequence. The bottom bar provides information about the project and the displayed image. (Source: Author)
The final layer contains information about the inferred data necessary to complete the model, classified according to Masuch et al. (1999) division for data input in a reconstruction project, making it clear the degree of ‘theoretical interpretations’ (Van Scoy, 2000) applied to each region of the model. Figure 124 shows a gradient chart presenting the certainty level codification to be mapped onto the image according to the decision making process involved in each reconstruction. The information is coded in a color scale ranging from red (less certain) to blue
(more certain), associated to a certainty level from 0 to 1. The method was applied to two- dimensional images of the Old Main Church (Figure 125 and 126), but also could be used in a three-dimensional model.
84
Figure 124. Certainty levels codification applied to the Old Main Church 2D model. (Source: Author)
Figure 125. Color coded frontal Figure 126. Color coded lateral façade showing certainty façade showing certainty levels. levels. (Source: Author) (Source: Author)
85 CONCLUSIONS
The goal of this research, to determine the probable dimensions of the Old Main Church, was attained, even though it is hard to determine how precise the final results are in the absence of any substantial measured remnants of original building, and with no camera parameters or archaeological information available throughout the process. The difference that was encountered in the measurements of the front façade, when comparing the rectified photograph and the 1857 map, was minimal, even though the process employed only a small detail in the façade (the clock) to provide the real scale for the rectification. This has demonstrated that it is not necessary to count on archaeological remains or prominent architectural elements such as windows and apertures to supply the scale; it can be done from seemly minor elements. The decision to use in the rectification four control points positioned close to each other, as opposed to the recommended principle of locating them as far apart as possible, proved to be valid, even though it derived from an uncertainty about the geometric relations among other building elements.
The applied methodology obtained satisfactory results for the reconstruction based on a single image, with additional sources of information detailing hidden elements of the building. The reconstruction of vanished heritage often deals with missing data and the methods tested here could be applied to similar cases with very worthwhile results. In such applications, special attention should be devoted to demonstrating the origin of each architectural element depicted by the model, as suggested by the system employed here that represents the levels of uncertainty in the resulting images.
86 Further research could investigate the reasons for the formal changes in the building throughout time, interpreting the cultural and historical layers of evolution; or explore the development of an interior virtual reconstruction of the Old Main Church based on analogies with other contemporaneous Brazilian churches. The dimensions acquired from the edge detection process would allow the production of an architectural plan of the building, which could in turn suggest probably sectional configurations, knowing that the Old Main Church structural system was probably similar to other existing, contemporaneous religious buildings.
Previous attempts to undertake exploratory archaeological excavations on the site, in an effort to locate the foundations of the Old Main Church, were not allowed by the local authorities. This situation may now be reversed once the results obtained by this research are made available to the academic community, because they can add considerable reliability to the initial starting points for such an investigation.
The historical research that is presented here forms a unique iconographic database about the Old
Main Church of Curitiba, providing a comprehensive organization of all the imagery depicting the building exterior throughout a time span of more than 170 years. The content includes pictures of the church’s remaining artifacts and sacred images that are presently scattered through several locations, presented for the first time in a single investigation. Potential exists for further outcomes using these collected resources, such as narrative documentaries or a multimedia disc, Website, or exhibit.
87 APPENDIX A - Church Timeline
The following chronological sequence provides a list of the main historical facts related to the
Old Main Church of Curitiba. The data compilation originated from several publications cited in the references. Even though some date discrepancies were observed among the sources, it was not the research objective to investigate historical documents to elucidate such differences, opting for dates cited by acknowledged authors.
1714 Lourenço de Andrade is chosen to administer the church construction.
1720 November 16th – The second image of Our Lady of the Light is placed in the church.
1721 January 5th – Lourenço de Andrade receives the last payment for the construction of the
Main Church.
1723 The community acquires a bell for the church.
1727 April 23rd – Joseph Palhano, Gaspar Teixeira and Gonçalo Soares are designated to build the church’s atrium.
1735 Auditor Judge Manoel dos Santos Lobatto compiles a document explaining that the church is without the necessary decorum, having just the tiles and no ceiling, presenting leaks, and also in need of a tribune, a niche for the statue of Our Lady. States the need of a ceiling for the main chapel and asks for community help.
1736 The Town Council inquires of treasurer Captain Miguel Rodrigues Dias about the funds application on the Main Church construction.
1737 November 11th – Judge Manoel Lobatto confirms that the construction of the Main
Church is still pending, and asks for the necessary procedures to resume it.
88 1738 June 14th – Council demands all family men to contribute for the construction of the church.
1747 Father Francisco de Meira Calassa describes the building in the Main Church’s chronicles as made of stone and clay and with a cedar wood ceiling for the tribune and main chapel.
Compiles a complete list of objects and ornaments and observes the lack of a proper place for the chorus.
1752 Visiting Father Manoel José Vaz writes about several church problems, mainly in the account of the vicar’s carelessness.
1756 Visiting Father Dr Antônio de Medeiros Pereira determines to tile the chapel due to a great number of holes, and also demands the removal of a wood deposit from the facilities. In his writings states that it is a pity that the building presents a low ceiling, and notes that several structural items are rotten. Poses a thereat to forbid the building use if nothing is done after a period of eighteen months.
1756 March 15th – To avoid the church’s blockage, auditor Jerônimo Ribeiro de Magalhães urges the requested repairs to be done.
1757 May – Reparation works begin, but with otiose results.
1758 May 16th – To avoid the shame of the church blockage, auditor Jerônimo Ribeiro de
Magalhães decides to fine all citizens who did no fulfill the promise to help the church financially. The same fines would apply to wagon owners who do not transport stones.
1758 May 28th – Town Council fixes a fine to builders for discontinued work. A deadline of six months is set for the completion of the building.
89 1761 June 27th – Town Council receives a letter from Sao Paulo bishop, Dom Frei Antonio de
Madre Deus, regarding the construction of the Main Church. The bishop establishes a one-year deadline for the works conclusion.
1761 September 19th – Councilors demand to remove cattle from inside the church or a fine will be applied to owners.
1762 São Paulo bishop prorogues for another year the deadline for the works conclusion in the church.
1762 Visiting Father Faustino do Prado organizes a list of citizens to help the works in the church, or else be obliged to pay a fine.
1773 Father Fermiano Xavier writes that the Main Church still has only external walls and no internal decoration.
1779 The work continues, but for every problem solved, another rises.
1784 January – The Town Council asks for the installment of a stone fountain for an existing headspring in the church plaza.
1799 July 25th – In a town meeting vicar Father José Barbosa de Brito decides that each family would provide a worker for a week period to help in the construction.
1820 Saint-Hilaire visits Curitiba and describes the church: a building without towers, 30 footsteps in internal length, situated in an elevated level and without ceiling or arches.
1825 Vicar Camello states that the church has fissures, caused by a three palms wall enlargement without proper reinforcement of the structure.
1825 July 25th – A lightning hits the Main Church during a mass, shattering the large cross and affecting a wall. The council decides to add the lateral halls and towers.
90 1829 March 20th – The Town Council moves its headquarters to the São Francisco de Paula
Church due to the bad existing conditions in the Main Church.
1833 Visiting Father Padre João Chrisostomo de Oliveira Carvalho Salgado Bueno, Paranaguá
Vicar, advises to change the church’s floorboards and finish the tower to avoid further damage.
1838 May 28th – The council decides that all revenue from the Barreira da Campina factory will be applied in the reconstruction of the Main Church.
1839 After examining the overall conditions of the construction, a committee advises to tear down a portion of the East wall, the Left side of the frontispiece and a portion of the West wall of the church. All the church’s objects and masses celebration are transferred to the Ordem
Church, except masses celebrated in the main chapel of the Old Main Church.
1840 The administrator writes that reconstruction is going on, with part of the roof beams and the portion of the frontispiece already removed. The vicar helped buy stones and lime with personal funds.
1843 The Town Council acknowledges problematic conditions in a section of the building, and no financial means to carry necessary repairs.
1848 Citizen Silva Carrão suggests using the factory revenue to repair the damaged structure.
1852 January 12th –Manoel da Cunha Bittencourt is appointed as new administrator of the
Main Church and the cemetery construction.
1852 Another lightning hits the church and urgent repairs are conducted on the roof. The
Provincial Governor acknowledges the building conditions.
1853 Father Agostinho Machado de Lima is the new vicar for the Main Church.
91 1855 Counselor Zacarias reports the beginning of the repair works. The community funds the main building and the legacy of Dona Maria Clara do Nascimento Guimarães the finishes on the right tower.
1857 With the two towers addition, the Town Council and its archives move back to the church.
1860 Halfway through the construction, the towers fell down and caused several fissures to the frontispiece.
1862 February 1st – The inner chambers of the Main Church house jury meetings.
1865 Father Agostinho writes an important document describing the church conditions to the
Provincial President. States that: the frontispiece has two cracks, the right tower is not internally completed, the left hall has no ceiling, the left tower shows only the external walls without flooring and wall finishes. The document finishes by recommending either retiling or rebuilding the roof due to several leaks.
1865 The community installs a clock at the church.
1867 Province President Polidoro Cezar Burlamarque authorizes the province engineer
Francisco Antônio Monteiro Tourinho to repair the frontispiece and finish the left hall.
1872 April 5th – Engineer Paulo Freitas de Sá reports the church conditions to the Province
President Venâncio José de Oliveira. Observes the excessive pressure over the terrain exerted by the towers and suggests the construction of a new church.
1872 April 6th – Antônio Pereira Rebouças Filho, future engineer of the Paranaguá-Curitiba railroad, makes a report about the Main Church. Suggests some structural repairs and the possibility of a new building construction at the same site, employing materials from the demolition of the Main Church.
92 1872 April 8th – In a report, Goettleb Wieland advises about some repair techniques.
1875 March – Adolfo Lamenha Lins is elected province president.
1875 December – Inclement weather causes to fall plaster pieces from the walls during the
celebration of the novena of Our Lady of the Light. Engineer Tourinho is appointed to verify the
cracks on the walls and frontispiece.
1875 Tourinho declares that the fissures are wider and alerts to the possibility of a building
collapse. President Lamenha Lins nominates a committee to analyze the state of affairs. The
committee members are the engineers Tourinho, Chalréo Jr. and Goettleb Wieland, and the final
advice is for the demolition of the Main Church.
1875 December – Lamenha Lins accepts the advice and authorizes the demolition of the
church. Engineer Charléo Jr. prepares the Rosário Church to receive the Main Church legacy.
1875 December 19th – A religious procession carries the objects to the Rosário Church.
1875 December 22nd – The demolition begins.
1876 February 15th – The fundament stone of the new church is blessed by vicar Father
Agostinho Machado de Lima
1880 Conclusion of the demolition of the Old Main Church of Curitiba.
93 APPENDIX B - Local Data Sources
The following subdivision is based on data types defined by Lagerqvist (1999). Source repetitions occur in the list, due to multiple contributions by some references that contributed with different sorts of data about the Old Main Church of Curitiba.
1. Data available in archives.
Arquivo Público do Estado do Paraná
Biblioteca da Câmara Municipal de Curitiba
Catedral Metropolitana de Curitiba
Exército Brasileiro – Comando Militar do Sul 5ª RM – 5ª DE
Fundação Cultural de Curitiba
Instituto do Patrimônio Histórico e Artístico Nacional
Maria Ester Teixeira Cruz – Restorer – Fundação Cultural de Curitiba
Museu de Arte Sacra de Curitiba
Museu Paranaense
Secretaria da Cultura do Estado do Paraná
Suely Deschermayer – Restorer
2. Data available in literature.
Biblioteca Pública do Estado do Paraná
Cúria Metropolitana de Curitiba
Fundação Cultural de Curitiba
Instituto Histórico e Geográfico do Paraná
Núcleo de Estudos Bandeirantes
94 Oldemar Blasi – Historian
3. Verbal data.
Edilberto Trevisan – Historian
Igor Chmyz – Archaeologist – Universidade Federal do Paraná
José La Pastina Filho – Architect – Instituto do Patrimônio Histórico e Artístico
Nacional
Key Imaguire Jr. – Architect – Universidade Federal do Paraná
Maria Ester Teixeira Cruz – Restorer – Fundação Cultural de Curitiba
Oldemar Blasi – Historian
Suely Deschermayer – Restorer
4. Data represented by graphic description.
Fundação Cultural de Curitiba
Igor Chmyz – Archaeologist – Universidade Federal do Paraná
Museu Paranaense
Secretaria de Comunicação Social – Prefeitura Municipal de Curitiba
Suely Deschermayer – Restorer
5. Data recorded through analytical numeric description of properties.
Fundação Cultural de Curitiba
IPPUC – Instituto de Planejamento e Pesquisa Urbana de Curitiba
José La Pastina Filho – Architect – Instituto do Patrimônio Histórico e Artístico
Nacional
Key Imaguire Jr. – Architect – Universidade Federal do Paraná
Secretaria do Meio Ambiente do Estado do Paraná
95 APPENDIX C - Internet Links
The following are some website links related to heritage virtual reconstruction of historical buildings. The list subdivision is based on the website classification criteria previously explained.
Historical and Technical Support
Association for Computer Aided Design in Architecture (ACADIA) - http://www.acadia.org/ Computer Applications and Quantitative Methods in Archaeology (CAA) http://www.caaconference.org/ International Centre for the Study of the Preservation and Restoration of Cultural Property (ICCROM) - http://www.iccrom.org/ International Committee for Architectural Photogrammetry (CIPA) - http://cipa.icomos.org/ International Council on Monuments and Sites (ICOMOS) - http://www.icomos.org/ Organization of World Heritage Cities (OWHC) - http://www.ovpm.org/main.asp SIGGRAPH - http://www.siggraph.org/ Virtual Heritage Network - http://www.virtualheritage.net/ Virtual World Heritage - http://www.virtualworldheritage.org World Heritage Committee (WHC) http://whc.unesco.org/nwhc/pages/home/pages/homepage.htm World Heritage Center (WHC) - http://whc.unesco.org World Heritage Information Network (WHIN) http://whc.unesco.org/whin/enindex.htm World Culture Report (WCR) http://www.unesco.org/culture/worldreport/index.html
Production Tools
Photoplan - http://www.kubit.de Archline - http://www.cadline.hu/arch-cd Image Modeler - http://www.realviz.com/products Photo 3D - http://www.photo3D.com/eindex.html Photomodeler - http://www.photomodeler.com
96 Single Virtual Reconstruction Projects
Aboa Vetus - http://www.aboavetusarsnova.fi/aboavetus/virtuaalitodellisuus/e_index.html Aloria - http://tempus3d.com Amiens Cathedral Project - http://www.learn.columbia.edu/Amiens.html Ancient Rome - http://www.unicaen.fr/rome/anglais/projet.html Calahorra - http://www.unirioja.es/Proyectos/Calagurris/english.htm Carnuntum - http://www.multimediaplan.at/carnuntum/Englisch/Project-Info/project-info.html Ceren - http://ceren.colorado.edu Colloseum Web - http://www.colosseumweb.org/english/progetto.html Corinth - http://corinth.sas.upenn.edu/corinth.html Didaskalia - http://www.didaskalia.net Dudley Castle - http://www.imint.freeserve.co.uk/heritage.htm ENAME - http://www.ename974.org Fatehpur Sikri - http://rohini.ncst.ernet.in/fatehpur Forum of Trajan - http://www.getty.edu/artsednet/Exhibitions/Trajan/index.html Great Temple at Petra - http://www.lems.brown.edu/~vote Greek Olympics - http://www.phm.gov.au/ancient_greek_olympics Isthmia - http://www.accad.ohio-state.edu/~japley/html/isthmia.html Istanbul - http://www.princeton.edu/~asce/const_95/const.html Jerusalem Archaeological Park - http://www.archpark.org.il/intro.asp Lost Labyrinth - http://www.casa.ucl.ac.uk/digital_egypt/hawara/index.html Louis I. Kahn: Unbuilt Ruins - http://architecture.mit.edu/~kll/www_compton/exhibit.html Notre-Dame Cathedral - http://www.vrndproject.com Olympia Project - http://www.sli.unimelb.edu.au/~cliff/olympia Palace of Santa Rosa Xtampak - http://www.cis.tugraz.at/IAM/iam_heine_maya-srx-text.htm Phimai - http://www.ucalgary.ca/EV/designresearch/projects/phimai/index.htm Pompeii - http://www.warwick.ac.uk/fac/arts/Theatre_S/pompey/Home.htm Pompeii Forum Project - http://jefferson.village.virginia.edu/pompeii Propylaea Project - http://propylaea.org Roman Canterbury - http://www.ncl.ac.uk/~napm1/dublin_core/index.html Rouen Revisited - http://acg.media.mit.edu/people/golan/rouen/index.html
97 Sagalassos - http://www.esat.kuleuven.ac.be/sagalassos/index.html Temple of Mithras - http://museums.ncl.ac.uk/archive/mithras/intro.htm Temple of Zeus - http://www.ime.gr/projects/olympics/classical/zeus/zeus.html
Multiple Virtual Reconstruction Projects
Anazi Architecture - http://sipapu.gsu.edu/index.html CERHAS - http://cerhas.uc.edu Charismatic Project - http://www.charismatic-project.com Cultural VR Lab - http://www.cvrlab.org Dennis Holloway - http://www.taosnet.com/architectVRe Earthworks - http://earthworks.uc.edu Foundation of the Hellenic World - http://www.fhw.gr/fhw/en/home/index.html Learning Sites - http://www.learningsites.com Museum of Reconstructions - http://www.reconstructions.org Oriental Institute - http://www.oi.uchicago.edu/OI/DEPT/COMP/Computer_Lab.html Pixelle - http://perso.wanadoo.fr/pixelle Synagogen - http://www.cad.architektur.tu-darmstadt.de/synagogen/inter/menu.html Taisei Corporation - http://www.taisei.co.jp/cg_e/cg.html Theatron - http://www.theatron.co.uk Theatron Project - http://www.theatron.org Urban Simulation Team - http://www.ust.ucla.edu/ustweb/ust.html
98 REFERENCES
A Cathedral de Curityba. (1930, August). Illustração Paranaense, Anno 4, 8. Curitiba.
Addison, A.C. (2000). Emerging Trends in Virtual Heritage [Electronic version]. IEEE Multimedia, Vol. 7, 2, April-June, pp.22-25. Retrieved August 18, 2002, from http://dlib.computer.org/mu/books/mu2000/pdf/u2022.pdf
Albertz, J., Wiedemann, A. (1995). Acquisition of CAD Data from Existing Buildings by Photogrammetry [Electronic version]. In Pahl, P.J., & Werner, H. (Eds.) Proceedings of the 6th International Conference on Computing in Civil and Building Engineering, pp. 859-866. Berlin: Rotterdam & Brookfield. Retrieved August 15, 2002, from http://www.fpk.tu- berlin.de/~albert/lit/CADtex.pdf
Arnold, D. (2000a). Computer Graphics and Archaeology: Realism and Symbiosis. ACM SIGGRAPH and Eurographics. Campfires: Synergies for the Future. Snowbird, Utah, USA. Retrieved August 19, 2002, from http://www.siggraph.org/~fujii/campfire/archaeology/website.pdf
Arnold, D. (2000b). Computer Graphics and Archaeology: Realism and Symbiosis. Retrieved August 19, 2002, from http://www.charismatic-project.com/Computer.pdf
Barceló, J.A. (2000). Visualizing what might be: An Introduction to Virtual Reality Techniques in Archaeology [CD-ROM]. In Barceló, J. A., Sanders, D.H., & Forte, M. (Eds.), Virtual Reality in Archaeology. Oxford: Archaeopress.
Berberi, E., & Sutil, M.S. (July, 1997). Tiradentes: a praça verde da igreja. Boletim Informativo da Casa Romário Martins, Vol. 24, 120. Curitiba: Fundação Cultural de Curitiba.
Bernd, E. (1999). New Media for Cultural Heritage. INI-GraphicsNet. Retrieved August 18, 2002, from http://www.inigraphics.net/publications/brochures/cult_broch/cultur.pdf
Bigg-Wither, T.P. (1974). Novo caminho no Brasil meridional: a província do Paraná. três anos de vida em suas florestas e campos - 1872/1875. Curitiba: Universidade Federal do Paraná.. (Original work published 1878).
Boehler, W., Heinz, G., Scherer, Y., & Siebold, M. (2000). Topographic Information in Cultural and Natural Heritage Visualization and Animation. ACSR, 21st Asian Conference on Remote Sensing, Taipei, Taiwan. Retrieved August 26, 2002, from http://www.gisdevelopment.net/events/isprs/2001/ts2/isprs2004.shtml
Boukhari, S. (2000). Computers Rebuild the Past [Electronic version]. The UNESCO Courier, pp. 40-42. Retrieved August 14, 2002, from http://www.unesco.org/courier/2000_03/uk/pdf/00_03_40.pdf
Bräuer-Burchardt, C., & Voss, K (2001). Image Rectification for Reconstruction of Destroyed Buildings Using Single Views. VAA, International Symposium on Virtual and Augmented Architecture, Dublin, Ireland. Retrieved August 15, 2002, from http://pandora.inf.uni-jena.de/cbb/Vaa.pdf
Brown, D., Addison, L., Solomon, A., Bernardes, P., Cain, K., Christel. R. (2000). Showing Off: Presenting the Results. ACM SIGGRAPH and Eurographics, Campfires: Synergies for the Future,
99 Snowbird, Utah, USA. Retrieved August 19, 2002, from http://www.charismatic- project.com/Computer.pdf
Carneiro, N. (1950). Iconografia Paranaense (Anterior à fotografia). Curitiba: Impressora Paranaense.
Carneiro, N. (1955). As Artes e o Artesanato no Paraná. Curitiba.
Corytiba em 1855. (1887, October). Revista do Paraná. Curitiba: Typographia Luiz Coelho.
D'Andrade, A.R.L. (1880). Breve História da Igreja da Ordem 3ª de São Francisco das Chagas. Curityba: Typ. do Paranaense.
Destefani, C.D. (1993). A Cruz do Alemão. Curitiba.
Eiteljorg, H. (1994) AutoCAD Single-Photo Photogrammetry at Pompeii. CSA Newsletter, Vol. VII, 3, November. Retrieved August 26, 2002, from http://www.csanet.org/newsletter/nov94/nl119406.html
Eiteljorg, H. (1995). Single-Photo Photogrammetry at Pompeii in 1995. CSA Newsletter, Vol. VIII, 3, November. Retrieved August 25, 2002, from http://csaws.brynmawr.edu/newsletter/nov95/nl119505.html
El-Hakim, S. (2000). A Practical Approach to Creating Precise and Detailed 3d Models from Single and Multiple Views [Electronic version]. IAPRS, Vol. XXXIII, pp 202-209. Amsterdam: Book 5A. Retrieved August 15, 2002, from http://www.vit.iit.nrc.ca/elhakim/el-hakim_290.pdf
Faculdade de Filosofia, Ciências e Letras, Departamento de História. (1968). Arquivo da Câmara Municipal de Curitiba. Arquivo da Sé Metropolitana e Paróquia de N. Sra. Da Luz de Curitiba (1968). Boletim da Universidade Federal do Paraná, 6.
Friend, T. (1999, February 28). Computers help archaeologists reconstruct lost splendors. Retrieved August 11, 2002, from the USA Today Web site http://www.usatoday.com/life/cyber/tech/cta691.htm
Forte, M. (2000). The Invisible Landscape: Behind the GIS and towards the Virtual Archeology. Harvard GIS Colloquium Series, 2000. Retrieved August 17, 2002, from http://icg.harvard.edu/~maps/hgis/maurizioforte.pdf
Fowler, R.A. (1998). Concepts and Applications of Digital Orthophotos. Retrieved August 28, 2002, from http://www.intermaptechnologies.com/PDF_files/bobfowler.pdf
Frischer, B., Niccolucci, F., Ryan, N., & Barceló, J.A. (2000). From CVR to CVRO: the Past Present and Future of Cultural Virtual Reality. CAA2000 Conference, CVRO Workshop 2002: Scientific Credibility and Authentication in Cultural Virtual Reality and Virtual Archaeology, Crete, Grece. Retrieved August 17, 2002, from http://www.cvro.org/frischer.pdf
Guarinello, A. (1929, March). Illustração Paranaense, Anno 3, 3. Curitiba.
Hemmleb, M., & Wiedemann, A. (1997). Digital Rectification and Generation of Orthoimages in Architectural Photogrammetry. CIPA International Symposium 1997, Gotenborg, Sweden. Retrieved August 15, 2002, from http://www.fpk.tu-berlin.de/~albert/lit/CIPA97r.pdf
100
Hemmleb, M. (1999). Digital Rectification of Historical Images. CIPA International Symposium 1999, Olinda, Brazil. Retrieved August 14, 2002, from http://cipa.icomos.org/papers/99c806.pdf
Heuvel, F. A. (2000). Trends In Cad-Based Photogrammetric Measurement [Electronic version]. International Archives of Photogrammetry and Remote Sensing, Vol. 33, Part 5/2, pp. 852-863. Retrieved August 16, 2002, from http://www.geo.tudelft.nl/frs/papers/2000/heuvel_CAD_measurements.pdf
Igreja Matriz de Corytiba (1887, November). Revista do Paraná. Curitiba: Typographia Luiz Coelho.
Illustração Paranaense (1927, December), Anno 1, 2. Curitiba.
Lagerqvist, B. (1999). A System Approach to Conservation and Cultural Resources Management. Photogrammetry as a Base for Designing Documentation Models. CIPA International Symposium 1999, Olinda, Brazil. Retrieved August 14, 2002, from http://cipa.icomos.org/system.pdf
Leão, E.A. (1928, April). Regresso dos Voluntários da Pátria. Illustração Paranaense, Anno 2, 4.
Leão, E.A. (1994). Diccionário Histórico e Geográfico do Paraná. Curitiba: Instituto Histórico, Geográfico e Etnográfico Paranaense. (Original work published 1926)
Liebowitz, D., Criminisi, A., & Zisserman, A. (1999). Creating Architectural Models from Images [Electronic version]. Proceedings Eurographics '99, Vol. 18, 3, pp.39-50. Retrieved August 16, 2002, from http://www.robots.ox.ac.uk/~vgg/publications/papers/liebowitz99.pdf
Masuch, M., Freudenberg, B., Ludowici, B., Kreiker, S., Strothotte, T. (1999). Virtual Reconstruction of Medieval Architecture [Electronic version]. Proceedings of EUROGRAPHICS 1999, Short Papers, pages 87-90. Retrieved August 17, 2002, from http://isgwww.cs.uni- magdeburg.de/~bert/publications/Masuch-1999-VRM.pdf
Mazzarotto, M. J. (1956). A Arquidiocese de Curitiba na sua História. Curitiba.
Moreira, J.E. (2000). As Imagens de Nossa Senhora da Luz e do Bom Jesus dos Pinhais e a Igreja Matriz de Curitiba. Boletim do Instituto Histórico e Geográfico do Paraná, Vol. LI, pp. 305-371.
Madov, N. (2002, May 15). Como era há 2 000 anos. Arqueólogos usam realidade virtual para reconstruir monumentos históricos em ruínas. Veja Online, 1751. Retrieved August 17, 2002, http://www2.uol.com.br/veja/idade/exclusivo/150502/p_067.html
Nowakowski, H.M. (1998). Bringing the Past into the Future. CRM Bulletin, Vol. 21, 5. Retrieved August 18, 2002, from http://crm.cr.nps.gov/archive/21-5/21-5-2.pdf
Ogleby, C.L. (1999a). From Rubble to Virtual Reality: Photogrammetry and the Virtual World of Ancient Ayutthaya, Thailand [Electronic version]. The Photogrammetric Record, Vol. XVI, 94, pp.651- 670. Retrieved August 17, 2002, from http://snap.lut.ac.uk/PhotSoc/abs/phroct99/Ogleby.pdf
Ogleby, C. L. (1999b). How Real Is Your Reality? Verisimilitude Issues and Metadata Standards for the Visualisation of Cultural Heritage. CIPA International Symposium 1999, Olinda, Brazil. Retrieved August 14, 2002, from http://cipa.icomos.org/papers/99c207.pdf
101
Ogleby, C.L. (2001). Virtual World Heritage: More Than Three Dimensional Models. ISPRS 2001, International Workshop - Recreating the Past -Visualization and Animation of Cultural Heritage, Ayutthaya, Thailand. Retrieved August 25, 2002, from http://www.gisdevelopment.net/events/isprs/2001/ps/ps6.shtml
Pereira, L.L. (1993). A Catedral. Boletim do Instituto Histórico, Geográfico e Etnográfico Paranaense, Vol. XLVIII. Curitiba.
Pombo, R (1929). História do Paraná (Resumo Didactico). São Paulo: Companhia Melhoramentos de São Paulo.
Prefeitura Municipal de Curitiba. (1993). Exposição Curitiba: Tempos e Caminhos – catálogo.
Prefeitura Municipal de Curitiba. (1994). O Culto de Nossa Senhora da Luz. Boletim Informativo da Casa Romário Martins. Curitiba: Fundação Cultural de Curitiba.
Quintero, M.S., Neuckermans, H., Van Balen, K., & Jansen, M. (1999). Accuracy in Affordable Technology for three-dimensional documentation and representattion of built heritage. VSMM 99 5th International Conference on Virtual Systems and Multimedia, Dundee, Scotland. Retrieved August 22, 2002, from http://www.asro.kuleuven.ac.be/rlcc/vsmm99-msq/asro.htm
Roussou, M. (2000). Virtual Heritage: From the Research Lab to the Broad Public [Electronic version]. Virtual Archaeology, Proceedings of the VAST2000 Euroconference, Arezzo, Italy, November 2000. Franco Niccolucci, (Ed.). Oxford: Archaeopress, 2002, pp. 93-100. Retrieved August 18, 2002, from http://www.makebelieve.gr/mr/research/papers/VAST/vast_00/VAST_093.pdf
Ryan, N. (2000). Documenting and Validating Virtual Archaeology. CAA2000 Conference, CVRO Workshop 2002: Scientific Credibility and Authentication in Cultural Virtual Reality and Virtual Archaeology, Crete. Retrieved August 17, 2002, from http://www.cs.ukc.ac.uk/people/staff/nsr/cvro/ryan.pdf
Saint-Hilaire, A. (1938). Viagens na Comarca de Curitiba em 1820 (D. Carneiro, Trans.). Curitiba: Edição dos Amigos do Livro e das Belas Artes. (Original work published 1851).
Saint-Hilaire, A. (1978). Viagem a Curitiba e Santa Catarina. São Paulo: Itatiaia. (Original work published 1851).
Schuhr, W., & Kanngieser, E. (1999). Single Images in Conservation. CIPA International Symposium 1999, Olinda, Brazil. Retrieved August 14, 2002, from http://cipa.icomos.org/papers/99c803.pdf
Shukla, U. (1999). Beyond Realism. Computer Graphics World, December 1999. Retrieved August 21, 2002, from http://cgw.pennnet.com/Articles/Article_Display.cfm?Section=Archives&Subsection=Display&A RTICLE_ID=54575
Secretaria de Estado da Cultura (2001). Pintores da Paisagem Paranaense. Curitiba: Solar do Rosário.
102 Shirley, P. (2000). Open Problems for Archaeological Visualization. ACM SIGGRAPH and Eurographics. Campfires: Synergies for the Future. Snowbird, Utah, USA. Retrieved August 19, 2002, from http://www.siggraph.org/~fujii/campfire/archaeology/website.pdf
Stone, R. & Ojika, T. (2000). Virtual Heritage: What Next? [Electronic version]. IEEE Multimedia, Vol. 7, 2, April-June, pp. 73-74. Retrieved August 18, 2002, from http://dlib.computer.org/mu/books/mu2000/pdf/u2073.pdf
Streilein, A., & Heuvel, F.A. (1999). Potential and limitation for the 3D documentation of cultural heritage from a single image. CIPA International Symposium 1999, Olinda, Brazil. Retrieved August 14, 2002, from http://cipa.icomos.org/papers/99c804.pdf
Strothotte, T.,Masuch, M., Isenberg, T. (1999a). Visualizing Knowledge about Virtual Reconstructions of Ancient Architecture [Electronic version]. In Proceedings of Computer Graphics International '99, pp. 36-43, Los Alamitos, USA. The Computer Graphics Society. IEEE Computer Society. Retrieved August 16, 2002, from http://isgwww.cs.uni- magdeburg.de/graphik/pub/files/Strothotte_1999_VKV.pdf
Strothotte, T., Puhle, M., Masuch, M., Freudenberg, B., Kreiker, S., Ludowici, B. (1999b). Visualizing Uncertainty in Virtual Reconstructions [Electronic version]. In Proceedings of Electronic Imaging & the Visual Arts, EVA Europe '99, Berlin, p. 16. Retrieved August 17, 2002, from http://isgwww.cs.uni-magdeburg.de/~bert/publications/Strothotte-1999-VUV.pdf
Students 'rebuild' lost German synagogues on video. (1998, September 24). CNN Home Page. Retrieved November 29, 2001, from http://www.cnn.com/TECH/science/9809/24/t_t/synagogue.reborn/
Trevisan, E. (1996). O Centro Histórico de Curitiba. Curitiba: Gráfica e Editora Vicentina Ltda.
Trevisan, E. (2000). Curitiba na Província. Curitiba: Gráfica e Editora Vicentina Ltda.
Tringham, R. (2000). Virtual Reality, Real and Surreal Places in the Past. ACM SIGGRAPH and Eurographics. Campfires: Synergies for the Future. Snowbird, Utah, USA. Retrieved August 19, 2002, from http://www.siggraph.org/~fujii/campfire/archaeology/website.pdf
UNESCO, World Heritage Committee (2000, May). Heritage. A Gift from the Past to the Future. Retrieved August 26, 2002, from the UNESCO Web site http://whc.unesco.org/nwhc/pages/doc/main.htm
Van Scoy, F. L. (2000). Documentation of Virtualized Architecture: A Proposal. VSMM2000, 6th International Conference on Virtual Systems and Multimedia, Gifu, Japan. Retrieved August 14, 2002, from http://www.vsmm.org/vsmm2000/review/papers/763-682-3.pdf
Wachowicz, R. C. (1993). As Moradas da Senhora da Luz. Curitiba: Gráfica e Editora Vicentina Ltda.
Waldhäusl, P., & Ogleby, C.L. (1994). 3x3 Rules for Simple Photogrammetric Documentation of Architecture. Symposium of the ISPRS Commission V "Close Range Techniques and Machine Vision", Melbourne, Australia. Retrieved September, 23, 2003 from http://www.univie.ac.at\wgv\3x3.htm
Wiedemann, A. (1997). Digital Architectural Photogrammetry for Building Registration. Internationalen Kolloquium über Anwendungen der Informatik und Mathematik in Architektur und Bauwesen,
103 Weimar, Germany. Retrieved August 15, 2002, from http://www.fpk.tu- berlin.de/~albert/lit/IKM97.pdf
Wiedemann, A., Hemmleb, M., & Albertz, J. (2000). Reconstruction of Historical Buildings Based on Images from the Meydenbauer Archives. IAPRS, Vol. XXXIII, Amsterdam 2000, B5/2, pp. 887- 893. Retrieved August 15, 2002, from http://www.fpk.tu-berlin.de/~albert/lit/isprs00.pdf
104