The Art of Printing Your

The A rt of Printing Your Art

An Introductory Course in Digital Ink-Jet Printing for Photographers Lesson One

Why print your own images?

Adobe Lightroom or Photoshop 1… and you no longer have to deal with Digital photography and the internet have fundamentally changed the working in the dark with smelly, toxic chemicals to achieve your artis- way we display and share our images. We have traded family photo tic vision. albums and shoeboxes of 4X6 prints for Flickr galleries, smart phones and tablets. Finally, your images deserve to be printed and shared, rather than horded on hard-drives. Holding one of your images, expertly printed So why should you print your images instead of just viewing them on on heavy weight fine art paper is a much more pleasant tactile experi- your computer screen? ence compared to passing around your iPhone. The problem, at least as I see it, is that these now common forms of display are small and ephemeral… the impact of viewing an image What do you need? large, filled with all of the fine detail and subtle nuances of tone and colour present in your original image is lost on these devices. If you do not already have one, a photo capable printer is the first obvious necessity. I will not recommend a particular brand or model but Ever since the dawn of photography, the print has been the ultimate Epson, Canon and HP all make excellent photo capable printers in var- expression of the photographic image. To see an Ansel Adams “Moon- ious sizes. rise” on a computer monitor is a distinctly inferior experience when compared to viewing an original 16X20 printed by Adams himself. If you are printing just for yourself, or are on a budget (aren’t we all?) an A4 or US letter size printer is a good starting point. If you want to And here’s something else to consider: printing your images can help sell or display your work publicly, then a printer capable of at least 13” you can become a better photographer. Print your work large and hang X 19” (Super B or A3+) is a better choice. Several models in this range it in an area you pass by every day. Live with it. Look at it every time are available at only a modest premium over a smaller letter size you walk past it. What would make it a stronger image? What ele- printer. ments in the image distract the viewer and therefore weaken the image? Printing your images large reveals all of the flaws that you miss In my opinion though, the sweet spot in printers for serious amateurs when looking at a smaller version. Do this consistently and I guarantee and many pros, is the 17” wide market. For just a few hundred dollars your photographic vision and the images you produce will improve. more than the cost of the smaller 13” printers, you can produce excellent 16X24 prints. Secondly, printing your work, particularly in the current digital era, provides you with unprecedented artistic control. Local adjustments to small areas of an image are incredibly easy with modern tools like

1 Since I’ll be referring to them so often, I’ll use a shorthand for Photoshop and Lightroom from here on: “PS” and “LR”

© 2013 - 2017 Mark English Page | 2 Whichever printer you buy, look for ones with these features: for example sells a paper called “Ultra Premium Photo Paper Luster”. While it doesn’t have the luxurious tactile qualities of a fine art paper, • More than 6 colours in the ink-set. Printers with 6 or fewer it does have an excellent colour gamut and dMax 2. It also matches quite colours are really intended for business graphics, not photo- closely, the colour reproduction characteristics of several high-end fine graphic images. Also, for Fine Art printing, I prefer printers art papers. I use it constantly for work prints and day-to-day printing. with at least 3 distinct shades in the Black to Gray range. Be It’s also very inexpensive as papers go. aware that many printers have provision for two distinct types of pure black ink: one for printing with gloss papers and an- There is a bewildering array of photo quality papers available for ink- other one for printing on matte papers. These printers would jet printers. The first and most obvious difference between them is the therefore advertise four different Black to Gray inks. surface: gloss or matte. Gloss papers, and variations on this like satin or luster can reproduce a wider colour gamut and produce a higher • Individual ink cartridges for each colour: ink is expensive. dMax than matte papers. They are well suited to images with a wide You don’t want to pay to replace all the inks in your printer range saturated tones. Matte papers generally have a narrower gamut when only one colour is exhausted. and a lower dMax. Many fine-art photographers insist on printing only • Printers using larger ink cartridges, although requiring a larger on matte papers, preferring the more tactile aesthetic qualities of these initial outlay, are generally cheaper to operate in the long run. papers. For example, at the time of writing, 14 ml carts for the Canon For me it comes down to the image: some images work well on matte Pixma Pro-10 printer cost $US15, while the 36 ml carts for the papers, others need a wide gamut luster or gloss paper to achieve their Pixma Pro-1 go for $US30. On a per ml basis this represents a maximum artistic impact. Insisting dogmatically on one or the other significant 28% saving. Over a short while, this will add up. merely limits the artistic choices available to you. • Some way to load thicker fine art papers. To start with, you With that in mind, here are a few of my favourite papers as of late 2013: may well stick to the printer manufacturer’s papers. Nothing wrong with that, but at some point, just for sheer aesthetic ap- • Epson Ultra Premium Photo Paper Luster: already mentioned peal, you will want to see your images printed on high quality above and hard to beat for everyday use. fine art paper. • Harman by Hahnemühle Gloss Baryta: Very wide gamut, excellent dMax, and a surface texture very like the premium F- Since we are talking about paper… surface papers we used to use in the chemical darkroom. The Starting out with inexpensive papers, perhaps just those from the man- only drawback is its cost. Significantly, this is the only paper ufacturer of your printer doesn’t mean giving up image quality. Epson, I am aware of, which is available in 17 X 25 inch sheets. This allows you to print full-frame 16 X 24 inch prints which is not

2 “Gamut” refers to the range of colours and tones from light pastels to dark and/or saturated tones that a device such as a printer can reproduce. “dMax” measures the deepest black that can be reproduced.

© 2013 - 2017 Mark English Page | 3 possible on the muchg more common 17 X 22 inch size. • Epson Legacy Baryta: This is a relatively new paper (as of http://harman.hahnemuehle.com/site/en/819/about.html ) late 2016) in Epson’s lineup. It has the highest dMax and widest gamut of any paper I have used. It produces rich • Ilford Prestige Gold Fibre Silk: a current favourite with a sim- blacks, wonderful saturated colours full of subtle gradations. ilar gamut and dMax to the Harman paper above. It has a It is a first-rate fine art paper and priced accordingly. As of very slightly warm base tone that works particularly well late 2016 it has become my go-to choice for the type of images with warm toned black and white images. http://www.il- I create. ford.com/en/products/photo-inkjet/galerie-prestige/gold-fibre-silk/ ) • Moab Entrada: when I decide that an image works best on a matte paper, I reach for this one. A heavy cotton rag paper, it • Epson Exhibition Fiber: A wonderful paper: images printed has the widest gamut of any matte paper I have tried and just with it have an almost three-dimensional quality to them that feels great to hold. is hard to describe and even harder to explain. It doesn’t (http://moabpaper.com/entrada-rag/ ) reach quite as far into the saturated greens as other papers like the Harman Baryta or Epson Ultra-Premium Luster, but These are just my current favourites, I may find others tomorrow, and these differences are more easily seen using software de- you should not feel tied to these. I do suggest though, that you pick one or two and print with them exclusively for a while to learn their signed to test these qualities in a lab, rather in actual printed characteristics well. images.

I’m a big believer in understanding the why of something, and not 4. Choose a paper appropriate to the image and your aesthetic simply the how . Knowing why something is so gives you the tools to goals, and print using the correct settings in your image-editing deal with new situations with a minimum amount of frustration. The program (we will discuss only Lightroom and Photoshop, here.) same is true in any endeavour, not just fine-art printing. So to start things off, we need to spend some time understanding the challenges 5. Lastly, and this is critically important, work with a monitor that in creating a really good print. These challenges include limitations has been properly calibrated and profiled. We will spend a lot of the print process, our camera sensors, and even of our human vis- of time understanding this last point. ual system. Since you are taking this course, I know you want to cre- In this first part of our two-part course, we will cover these elements ate prints that are not just good enough , but excellent prints that you in basic detail: sufficient to ensure that you will be capable of consist- would be proud to hang in your living room or office (perhaps even ently producing a good print. In the second installment, which you in a gallery!) may register for separately later, we will cover each item in much What follows may seem a bit technical side, but we’ll only hit the im- greater detail and look at some advanced ways to maximize the qual- portant points. ity of your prints

The basic steps in creating an excellent print are simple, start with an So in our effort to understand the why , let’s start by looking at some image that is: limitations of the printing process itself.

1. Well composed, properly exposed and sharp where it needs to be. This may seem obvious, but it needs to be said right up front. A poor image expertly printed is still a poor image. 2. Saved in an appropriate file type to ensure that the maximum Pushing the “Print” button has no range of color is captured and that we retain maximum flexibility effect on the quality of your print. for adjustments in post-processing without degrading the image file, and then, It’s everything you do before

3. Properly process the file, including appropriate tonal correc- that makes a difference .

tions, appropriately applied sharpening, and noise reduction if needed.

The most fundamental issue to understand is that no printer, paper and ink combination yet devised can reproduce the range of brightness levels and color found in a typical outdoor scene. Outdoors in natural light, the range of tones confronting you can go from featureless black shadows to blinding bright specular highlights all in the same image.

On the other hand, the brightest highlight you can reproduce in a print is limited to the white base colour of your paper. The darkest shadow is limited to the blackest black of your printer’s ink set. That’s it: ink black and paper white . That’s all you have to reproduce the full range of tones in any image you try to print.

The Real World vs. the Print

Let’s consider the tonal range in a typical sunlit scene. From the brightest sunlit highlights to the darkest tones in the shadows, the dynamic range here can easily exceed 100,000:1 — roughly 20 stops! Now, the example on the right is a little less extreme. Even so, in this image from Cassis in southern France, the sunlit highlights measured The dynamic range of both your camera and the printer is much smaller than the range of brightness levels in a typical outdoor scene about 12 stops above the shadows. This represents a range of 4096:1

The newest and best of the current crop of dSLR cameras may just be of slide film we were working with a useful range of perhaps 4 stops… capable of capturing this range of brightness levels in a single expo- at most . sure … measured on a laboratory bench. Most “prosumer” level cam- Now, for even the whitest paper available, paired with the best ink era sensors will not get close to this in any practical sense. sets available, you can only expect to print a brightness range of about Therefore, the first limiting factor in the process from scene to print is 125 or 150 to 1. This means that the white of the paper will reflect the sensor itself. Most all prosumer grade dSLRs today can capture about 125 to 150 times more of the light falling on it than will the black- about 6 – 8 stops of useful dynamic range. As we said, some may be est black that can be printed with this paper and ink combination. capable of more, but 6 – 8 stops is a good approximation to keep in mind. This might be disappointing, but consider that in the “old days”

© 2013 - 2017 Mark English Page | 6 This limited range is all we have to represent the entire brightness range 3 of your scene. In contrast, a typical sunlit landscape will easily exceed 100,000 or even 1 million to 1 when specular reflections and deep shadows are present. Obviously, something has to give.

Printmaking is fundamentally an attempt to reproduce reality using a medium that is just not up to the task

3 We will call the total range of brightness levels that can be captured or reproduced the “dynamic range” of from now on

A limited dynamic range is not the only problem to deal with; we also have to worry about colour reproduction.

The first thing you must understand is that “colour” is not real; it is a figment – a creation of your eye-brain visual system. It is just the way your eye and brain together interpret different frequencies in a narrow band of electro-magnetic energy, which we call “light”. Moreover, because colour perception is created in the human brain, it is highly subjective.

Our perception of a particular colour is also highly dependent on a number of external factors: An example of colour perception influenced by nearby colours Which square contains the brightest most saturated red circle? • the other colours present around it at the time, and The answer is both of them. Your perception of the colour of each circle is heavily influ- • the colour of the ambient light illuminating the coloured ob- enced by the colour surrounding it. Blue, being the complement of Red, reinforces the saturation of the red circle within it. The orange background on the other hand, being ject we are looking at. close to red on the colour circle, effectively de-saturates your perception of the same colour, making it appear muddy in comparison Complicating this further is the fact that our human visual system is constantly adapting to changing lighting conditions, and in particular The second thing you must understand is that no two devices record to a wide range of colour balance differences in the ambient illumina- or reproduce colour in the same way: no two monitors, no two print- tion. All of this happens without us being aware that these adapta- ers, no two cameras, in fact no two people perceive, capture or repro- tions are going on – the result is that we perceive a reasonably uniform duce color in the same way. In other words, you must give up any palette of colours in the world around us. A white object appears hope of exactly matching the colours you thought you saw in the orig- white to us under warm household tungsten lighting, just as it does inal scene to the colors on your computer screen, or to those you can under the bluish light of open shade. reproduce in a print. The best you can hope for is a pleasing approx- Because our eye-brain visual system is constantly making uncon- imation. scious adjustments to our color perception, we are all in fact really Since colour perception is highly subjective, and since no two devices lousy judges of color. (in fact no two people) perceive colour the same way, what we really

© 2013 - 2017 Mark English Page | 8 need is a way to objectively measure and describe colour, so that we blend in the middle to create white . As you move away from the white can have a uniform objective way to compare the colour reproduction centre the colors change from pastels to the most saturated colours the of one device to another. average person can see, out toward the edges.

Beginning with work as far back as the 1930s methods exist to define color in terms of its spectral frequency distribution and intensity; in other words, in precise measurable scientific terms that do not depend on the subjective interpretation of our human vision. Now, to make great prints, we don’t need to understand anything about spectral frequency distributions. All we need to know is that it is possible using this method to precisely define the way any device reproduces colour: whether the “device” is an average person, or a particular monitor or printer.

When we use this method to map out all the colours that the average person can perceive, we are using it to define all the colors that together make up something called the human “colour space”. When we map it out on paper it produces a graph like the one on the right (it’s known as the CIE–XYZ colour space, if you’re interested) and as

I said, it represents roughly all the colours that an “average” person The CIE-XYZ colour space describes the limits of human colour can perceive. Don’t be put off by the seemingly complex nature of the perception in objective scientific ter ms graph and all the numbers around the edges: they don’t matter to us The important thing to grasp here is that this is just a way to map out here. Just look at the shape of the colored area. This roughly triangu- the limits of all colours the average person can see. Later, we will use lar area is organized with the three additive primary colours: red, this diagram to illustrate how other devices, such as printers and mon- green, and blue at or near the corners of the triangle. These are the itors can only reproduce a portion of this total human color space. By same three primary colours your camera sensor or your monitor blend the way, other colours exist outside of this shape (remember that together to create all the different colors they can reproduce. color” is just the way your eye-brain visual system interprets different In between the corners, where these colours blend, are the so-called frequencies of energy), but since we can’t see them, there is no point subtractive primaries: magenta, cyan and yellow. All the colours talking about them here. 4

4 Bees for example can see up in to the ultra-violet range beyond human perception. This would be represented in the lower left corner of the graph above, outside of the coloured area.

How Big Is Your Crayon Box?

Some people have a hard time understanding the idea of a “Color Perhaps now you can begin to see the problem facing us. Let’s say Space”. The easiest way to visualize a “colour space” is to liken it to your camera records a particular shade of red. This shade of red is a box of crayons. The “colour space” is the box… the crayons inside right out at the limit of the reddest red your camera can record. Un- represent all the colours that can be displayed in that “colour space”. fortunately, this shade of red is not present in your monitor’s crayon box. It has one that is close, but not exactly the same and just different enough that you will easily perceive them as different colours. What are you going to do? You can’t display the particular shade of red, so you really can’t show a viewer who wasn’t there, what it looked like. To add insult to injury, you discover that your printer does not have the particular shade of red in its crayon box either; in fact, the printer’s closest red is quite different from both the monitor and the original scene.

This triangle represents the color space limits of a typical computer monitor. It can’t display many of the saturated greens, reds, and yellows that you can see. "Crayola" is a registered trademark of, and the package designs shown here are copyright Crayola LLC This smaller rounded area represents the colour space limits of a Obviously, the bigger your crayon box, the more crayons you can typical ink-jet paper suitable for printing photos. hold, and the more colours you can represent.

The human colour space is a very big crayon box. No device yet created can reproduce all the colours in this colour space: your monitor’s crayon box is substantially smaller, as is the crayon box of your inkjet printer. On top of this, the crayon box of each device in the colour reproduction chain, aside from being much smaller than your own crayon box, also contain different, overlapping ranges of colour compared to another device’s crayon box. The colours that can be reproduced by your monitor or by your printer are far less than those you can see.

To sum up, the human colour space represents the limits of human Translation dictionary. colour perception, All other devices, including all printers and com- The process of converting the colour language of one device to colour puter monitors, can only reproduce a portion of this colour space. language of another is part of the “Colour Management System” Furthermore, they all reproduce different, overlapping portions of the which is incorporated into the operating system of your computer total human colour space; and therein lies the central problem with (likely Windows or MacOS) and “colour aware” applications like colour management: if we send the appropriate set of instructions to Adobe Photoshop and Lightroom. This happens on-the-fly, seam- your monitor to reproduce a particular shade a color, and then send lessly in the background. Once you have properly set up Photoshop the same instructions to your printer, each will reproduce different and told the colour management system something about how your shades of that colour. It’s as if your monitor is speaking French while “devices” reproduce colour, the whole process basically flies on your printer only understands Italian! “auto-pilot”. Now, if you will allow me to carry the So where to begin? We use our monitors analogy a bit further, if we could figure to evaluate and adjust the color and tone out a way to understand the “French” in- If you are not prepared to calibrate in our images before we print them. structions for a particular colour our and profile your monitor, you may Therefore, we should probably begin by monitor needs, and then translate and as well stop right here understanding how our monitor repro- send those instructions in “Italian” to our duces colour. This brings us to the single printer, we could end up with both de- most important step in consistently cre- vices understanding the other’s colour ating great prints: calibrating and profiling your monitor. language. To do this we would need a “French-to-Italian” colour dictionary of sorts. To create this dictionary, we would need to start by For this course, we will learn to measure the colour response of our understanding how our monitor “speaks colour”, and how our monitor and create the information that PS or LR need. The result of printer “speaks colour”. We would do this by measuring the way our this exercise creates a small file called a “profile” 5 (a monitor profile in monitor and our printer reproduce colour using that objective scien- this case) which tells PS or LR how you particular monitor reproduces tific method we referred to on page 9, from there it would only be a colour. Photoshop or Lightroom uses the monitor profile on the fly to small jump to finding a way to create our French-to-Italian Colour correct the colours on your monitor to be as close as possible to those

5 You will see these referred to as ICC profiles as well, after the name of the organization responsible for colour management standards: the Inter- national Color Consortium

© 2013 - 2017 Mark English Page | 11 in your image files. We will also need a similar file (called an output First, we will be asked to choose some starting “aim points” for mon- profile) for our particular printer-paper combination. PS and LR use itor luminance (brightness), color temperature and something re- the information in the output profile to reproduce the colours in your ferred to as monitor “gamma”6. Gamma and Color Temperature are print as closely as possible to what you see on your monitor easy, since there are accepted industry standards for our situation. Monitor luminance on the other hand takes a bit more thought; we Creating good output profiles is beyond the scope of this course, so will look at that in a minute. In the second step, we simply sit back we will rely on profiles already created for us by the manufacturer of while the software presents a set of coloured patches to the monitor, your paper, which you can download from their web sites. which are measured by the colorimeter. The software compares the Profiling your monitor is so important, that if you are not prepared to measured value of each coloured patch and compares it to the ex- take the time, and invest in the correct tools for this step, you may as pected value. The software uses this information to build a picture of well just stop right here. how your particular monitor reproduces colour and stores it in the Fortunately, it’s neither hard, or terribly costly to perform this critical profile for your monitor. step. All you need is a small hardware device called a “colorimeter”, The first step is sometimes referred to the “calibration” step, since we which is used to measure the colour response of your monitor, and are going to choose settings to which we are either going to manually the associated software to run it. These are sold as a package, and adjust our monitor, or allow the software to adjust the monitor for us. good examples include the X-Rite ColorMunki Display, the X-Rite Follow the instructions that come with your calibration device, and i1Display Pro or the Datacolor Spyder4Pro. Any of these will serve when asked by the software, choose 6500K for color temperature , and you well, and will only set you back between $150 and $250. 2.2 for monitor gamma . (Have a look at page 15 for the setup dialogs and a few more optional settings for a popular X-Rite product.)

We are still in the Calibration step, and now we must choose a brightness level (more correctly referred to as a luminance level) for our monitor. This setting takes a bit more thought since there is no specific standard to offer you; this setting depends heavily on the ambient il- Two versions of the X-Rite colorimeter device used to calibrate and profile you monitor. lumination in your work area. The newer i1 DisplayPro on the left the older version on the right

Using one of these is easy, but you first need to understand that the process happens in two steps:

6 ”gamma” is an adjustment that needs to by applied to signals sent to your monitor in order for the tones it displays to appear normal to human vision… say no more

© 2013 - 2017 Mark English Page | 12 Setting up your work area it’s just a number. It is, however a number that your monitor profiling and choosing an appropriate monitor luminance level package will want you to choose:

Remember our earlier discussion about the subjectivity of human vi- • If your workspace is a typical residential environment, and sion? Lighting levels and the colours in your surroundings heavily most of your work is in the evenings by artificial light, start influence your perception of colour and tone. If you want a reasona- with a luminance of 110 cd/m2. ble shot at fairly evaluating your prints, you need to think about the • If you work in typical bright office-like environment, then a area in which you will be editing and printing your images. value of 140 cd/m2 will be more appropriate If you view a dark grey tone under a low level of illumination, it will These are good starting points that should get you “in-the-ballpark”. appear black or almost black. Viewed under high levels of ambient I suggest you choose one of these (depending on your circumstances) illumination, this same tone may appear to be almost middle grey. for your initial calibration and profiling run. Later, if you wish to The point is, your perception of relative light and dark depends really finesse your monitor luminance level, try the simple procedure greatly on the level of illumination under which you view an image. below. If this is your first time through the calibration step, skip ahead Now, the unfortunate truth is, most monitors are simply far too bright to page 15. “out-of-the-box” to provide an accurate portrayal of the tones in your images. One of the most common complaints I hear from beginning Setting a luminance level that is right for you: printers, is that their images print too dark. The problem is not that the printer is printing too dark, but rather that their monitor is too bright . To make to image appear correct on screen, they adjust the image in Photoshop or Lightroom to make it appear darker… so that it appears normal to the eye . When this darkened image file is printed, it prints darker than what appeared on screen: the view on the monitor was not an To finesse your monitor luminance level in relation to your particular accurate representation of the tones in the image file… the monitor ambient illumination level, download the file called “MCLStep- incorrectly portrayed them as very bright, so they edited them to ap- Wedge.jpg” from the Lesson 1 download area on the PPSOP web-site. pear correct on the monitor, and sent an artificially darkened image to This file (shown above 7) contains a step wedge of tones beginning the printer. with pure black through several steps of dark charcoal grey. The start- Monitor luminance is measured in something called “candelas per ing black step is marked with a grey “0” and the others progress to the square meter”, or cd/m2. We don’t really care what that means: to us right from this point. Use this step wedge to determine if your luminance is set correctly for your work environment. Follow these steps:

7 I have lightened the image quite a bit for visibility on this page. The file you download will appear darker, and it may be difficult to see any separation between the steps.

© 2013 - 2017 Mark English Page | 13 • First, calibrate your monitor with an appropriate starting point be sure to choose to “adjust the luminance manually” in the startup for your luminance level from my suggestions above. options for your software. If you don’t, the software will attempt to • Adjust the ambient light in your work area to your preferred reset your monitor to the original value you chose. If any of this is lighting conditions, making sure that no light falls directly on confusing, have a look at the short video in the Lesson 1 download your screen. Screen hoods are available from several sources, area. and are a good idea if you can’t prevent direct overhead or task lighting from falling on your monitor screen. Neutral 5000K am- Ideally, once you have set your luminance level you should see a dif- bient lighting is best for colour evaluation, but the practical real- ference between the 0 step and each step to the right of it. However, ity is that most of us don’t have a space we can dedicate to the if you can’t see any difference between 0 and the next step or even the sole purpose of print evaluation illuminated with 5000K lighting one after that, don’t despair. Many monitors, particularly consumer (typical residential tungsten lighting is roughly 2800K: much grade LCD or LED monitors will not be able to resolve these subtle warmer than ideal). differences in the very deepest shadow tones. With some monitors, • Load the step wedge file into Photoshop. If the program asks if you may not be able to see a difference between the background of the you want to convert the file's colour space to your working space, monitor and any of the steps in the step-wedge, regardless of the set- click “No” and allow it to open in the Adobe RGB(1998) colour tings you use. If you own a calibration and profiling package like the space. X-Rite i1 Display Pro, which adds a feature allowing you to measure • Hit “F” twice to turn off the Photoshop UI, and make the back- the ambient light in your workspace, you may want to give this a try. ground pure black. If you use it, and find that it is suggesting a monitor luminance much less than 100 cd/m2, you will be better off raising the ambient light in Now, looking at the tones in the step-wedge, you should NOT be able your workspace and working with something closer to 110 cd/m2. Af- to see any difference between the “0” step and the surrounding screen ter profiling, you should still be able to produce great prints. area. If you do, then your luminance is set too high. Manually adjust you monitor’s brightness control until you just lose the difference be- Lastly, if possible, the walls in your work area should be free of strong tween the “0” step and the surrounding area. The goal is to adjust colours since these could influence your perception of colour. Your your monitor brightness until you just lose the difference between the work area will ideally be in a room with relatively neutral walls. If this surrounding screen area and the “0” step… not too dark and not to is not possible, then at a minimum, the area behind your monitor light. Now rerun the profiling routine with your new luminance level. should be as neutral toned as possible. When you rerun the profiling routine with this new luminance level,

Here are the settings dialogs for a popular X-Rite monitor profiling solution: i1DisplayPro with i1Profiler Software

Products from other companies will at ask you choose values for each of these settings, as well.

To fully access all the actions below, choose the “Advanced ” option on the opening Home screen.

The choose Profiling, under Display on the top right

In the next screen, the “Display Settings” dialog (first screen shot on the right)

White Point: Medium White (6500) Gamma: 2.2 is set for you by default Luminance: Choose “Custom” from the dropdown and select 110 for residential settings.

Once these are set, click the “Profile Settings” button at the bottom of the dialog, and adjust the settings as shown on the right (second screen shot).

Next, click the “Patch Set” button at the bottom of the dialog and choose how many coloured test patches i1Profiler will use to profile your monitor. (last screen shot on the right)

Generally, the more patches you allow it use the more accurate will be the resulting profile. The trade-off is time; more patches take more time to read. On my computer, it takes about 10 minutes to complete the profiling process with the largest patch set… an acceptable trade-off in my mind.

Finally, click “Measurement”, at the bottom of the dialog, then “Start Measure- ment” on the next screen, and sit back while i1Profiler measures your monitor’s colour response. When this finishes, follow the instructions to save and set your newly created monitor profile

© 2013 - 2017 Mark English Page | 16 more recent versions of PS and LR, AND you are getting weird results, The i1Profiler software includes some great instructional videos to switch back to ICC profile version 2. help you along with the calibrating and profiling process. Also in the Profile Settings, we chose the “Table based” profile type. Now, a few points on the selections we have made. This is a more accurate profile than the simpler “Matrix” based profile, Generally, if you are using a computer manufactured after say, 2010, but here again certain versions of OS and application programs do not and the latest versions of Windows or MacOS together with a recent like to play in this sandbox: Mac OSX Lion and Final Cut Pro X to- version of PS and LR, the settings I have suggested here will work gether are two examples that come to mind. Once again, if you are well, and will provide the best results, however…. getting weird results such as overall muddy colours, just switch to the Under Profile Settings, we chose ICC Profile Version 4, this is a newer matrix profile type and re-run the profiling process. standard that among other things is designed to produce better colour Having said this, I suspect that virtually all of you will get excellent transformations from one device to another. Not all programs are de- results with the settings suggested above. signed to use this correctly. If you are using software other than the

Creating a great print means taking care of all the details that come depth”, and currently we have two choices for this: “8-bit”, or “16- before we press the print button. Here are a couple of more issues to bit”. We won’t go in to the detail of why 16-bit files can capture more consider before we get to setting up PS and/or LR for printing subtle details and can withstand greater post-processing adjustments here, but if you are interested have a look at the appendix at the end File types for Fine Art Printing of this lesson (which is by no means required reading). Suffice it say, we want to capture and store our images as 16-bit files. Now that we have our monitors properly calibrated and profiled, let’s think about the appropriate file type to use for saving our images. Now let’s have a look at the specific file types you might use to store your images. Most of today’s digital cameras offer a choice of two or more file types for the images they capture. The differences between these file for- JPEG: These files have a file extension of either .jpg or .jpeg. If you mats boil down to two major issues: choose to, you can set your camera to record images using this format. The JPEGs use a lossy compression. It’s quite possible to take a 50Mb • The compression method used to squeeze the image infor- image file and reduce it to about 8 – 9 Mb or less, and still retain rea- mation into a smaller file, and sonably good image quality. However, each time you open and re- • How many computer ‘bits” are used to describe each image save a jpeg image you re-compress the file, and this will eventually lead pixel. to visible degradation of the image. The upper image on page 19 is a Compression methods used in digital image files are either “lossy” or modest crop from a single frame. It shows excellent sharpness and “lossless”. Lossy compression methods throw away part of the infor- smooth tonal gradations; the image would make a great print. On the mation contained in the original image each time it is saved in an ef- other hand, the image below it shows the impact of multiple saves fort to minimize the overall file size. If you go easy with this compres- using high jpeg compression. Degradation is clearly visible… this im- sion, high quality images are still possible. Compress the file heavily, age file would not make a very good print. and you will be trading image quality for more free disk space: not The final problem with jpeg files is that they are inherently “8-bit” and an acceptable tradeoff for a Fine Art printer. because of this have a more limited capacity for enduring large adjust- Using more computer bits to describe colour in your image translates ments later on in programs like Photoshop. into a tonal scale that captures and retains more subtle tonal gradations and allows you to make larger or more extensive adjustments in PS without the image falling apart. A file that uses more “bits” to describe the colour and tone in an image is said to have a higher “bit

© 2013 - 2017 Mark English Page | 18 TIFF : TIFF stands for Tagged Image File Format, and is a popular format for storing high-resolution images. It is a highly flexible format and has provision for a loss-less compression methodology known as “LZW” and can handle 8 or 16-bit images. It is also compatible across Macs and PCs. Few if any newer digital SLR cameras output files as TIFFs, although you will likely come across mention of these in your reading. It is often used as a preferred file format for the delivery of high resolution image files by professional photographers to their commercial clients.

PSD: this is the native format for files created in Adobe Photoshop. It’s essentially a TIFF file with a Photoshop wrapper around it, and has all the same features and capabilities of a TIFF file.

RAW: Although not a technically correct description, think of the raw file format as the unprocessed information your digital sensor cap- Original crop from a RAW file tures. A raw file captures all of the colour and dynamic range that your sensor is capable of capturing. Raw files are inherently “16-bit” and because of this and other reasons can tolerate much larger adjustments in post-production than jpeg files.

The downside to raw images is two-fold. Raw file formats are for the most part, proprietary; each manufacturer creates their own unique brand of raw file. Secondly, raw file data is not in a form that can be interpreted as an image. It must be “developed” to produce a viewa- ble image. Each camera manufacturer supplies software to perform this task, and Adobe Photoshop comes with a plug-in known as “Adobe Camera Raw” or more simply as “ACR” for this purpose. Adobe Lightroom uses the exact same “processing engine” that is used in ACR.

Developing a raw file in ACR or Lightroom allows you to convert or “render” the raw sensor data into image pixels. This rendered data Same crop, but from a JPG file which has been opened and re- saved multiple times using high compression

© 2013 - 2017 Mark English Page | 19 can then be opened in Photoshop for further processing. or used by ers often opt for jpeg, since the files they produce are ready to go with- LR to create a print or exported TIFF or Jpeg. If the image is passed to out any further processing in “post”. And, as a delivery medium , jpeg PS you can save the it as either a PSD file, a TIFF or JPEG: your choice. files are fine - in fact they are often optimal. Once all the processing is In all cases, the original raw file remains unchanged. Think of a raw complete, saving an image as a high quality jpeg produces a smaller file as your digital negative, from which you can make an infinite file more suitable for transferring over the internet. If you go easy number of different “prints”. The advantages of dealing with raw with the compression, choosing a jpeg setting of say, 10 or better, a files far out-weigh the drawbacks. high degree of image quality will be retained. If you still want to capture your images as JPEG files, set your camera to produce the largest The conclusion from this? For maximum creative flexibility and high- JPEG file that it can, and be aware that you really need to exposure est quality printing, set your camera to create raw files and learn to carefully in camera to avoid the need for large corrections later in post. properly develop them in ACR or LR Actually, paying attention to good exposure in camera is good advice JPEG does have a place: high volume shooters and those working un- no matter what file format you choose. der tight deadlines, such spot news photographers and sports shoot-

Recall back to page 6, where talked about the limited dynamic range Think of the individual sensor elements in your camera as little light of your camera sensor. To understand why digital cameras are limited buckets. As light strikes each sensor element, these little buckets in this way and to understand the effect this has on our image making, begin to fill up. Each sensor element tells your camera how “full” it let’s look a little deeper at the problem by way of an analogy. Your is: the brighter the light falling on it, the “fuller” it says it is. Once camera sensor contains millions of discrete light collectors referred to they are full of light, these light buckets overflow… and no greater as sensor elements . Each of these collectors measures the intensity of amount of light will have any further effect… it cannot tell your light falling on it, and together they measure the light intensity at camera that it is any more than just “full”. The sensor element is said millions of discrete points throughout your image. Taken together, to be “saturated”, and this represents a hard ceiling beyond which it these discrete points form an image that your eyes will perceive as we cannot go: no matter how much more light strikes the sensor, it continuous tone. will only ever say that it is just “full”

The Impact of Overexposure Over-exposure is the worst technical mistake you can make with your highlights that give our images delicacy and ‘sparkle’ beyond the digital camera. Overexposing an image overflows the “light buckets” maximum level which can be recorded by the sensor. Over-exposure in your sensor, starting with the brightest highlights. When this pushes all those wonderful highlight details to maximum white, and happens, we ending up with featureless highlights… even a one stop we end up with featureless bald highlights instead. No amount of over exposure in some situations will push many of the subtle work in PS will recover blown highlights.

Let’s look a little closer at the effect even a small amount of overexposure can have Correct Exposure on your images, and therefore on your prints. Let’s use the same image from Cassis as an example of a situation where proper exposure is crucially important: outdoors in bright sunlight. Since this situation encompasses a very wide brightness range, exposure errors can easily push your camera sensor beyond its capabilities to capture the full range of tones present. I’ve highlighted three tones to illustrate the effect of overexposure: a mid tone, a lighter tone and a very bright highlight (one where we still want to retain detail). Below the image, using the “Light Bucket” analogy, I’ve included pictures to represent how a sensor element “seeing” only that part of the image might appear. On the left I’ve shown where these tones would appear on a scale from pure black to pure white in a print.

This represents a properly exposed image. Even the brightest tone is still well within the sensor’s ability to record, and none of our little “light buckets” are full. Further, all tones are in the proper relationship to each other on the tonal scale.

Now let’s see what happens when we over expose by even a modest amount: Slight Over Exposure All of the tones move up the scale, as you would expect.

• The brightest tone is now at the top of the scale and the sensor element seeing it is “full”, any more exposure, and this sensor element will not respond further. • The light tone has also moved up the scale, but the sensor element seeing it is not yet full. There is still some tolerance of additional exposure. • However, the light tone has moved closer to the brightest highlight: we have reduced the apparent difference between these two tones. In photography terms, we have reduced the contrast in our highlights, distorting and compressing the separation between them. It may still be possible to recover from this exposure error in post-processing, - but it is not ideal

Unrecoverable Over Exposure So, let’s overexpose a bit more (perhaps a stop or so in total by now):

• Now there is now virtually no separation between the light tone and the bright highlight. The sensor elements seeing each tone are reporting back to the camera, that they are “full”, and of course the brightest highlight element is now overflowing.

No amount of Photoshop wizardry will recover the lost highlight detail in this over-exposed image. This image will not produce a good print

Underexposure presents a completely different set of problems. Mod- forming signal, particularly in the areas of our image which are dark erate underexposure (say, a stop or so) is relatively easy to correct and to begin with; areas where the image forming light falling on the sen- still produce a very good print. Much more than this however and we sor is quite weak. will start see negative effects in your printed images This represents what an engineer would refer to as a poor ‘signal-to- When we underexpose images, we don’t allow enough image forming noise’ ratio, and when this happens the result is little flecks or blips in light to reach the sensor. We compensate for this by electrically am- our image that appear much like film grain. It will first be apparent plifying the signal produced by the sensor. The catch is that we also in the shadows of your images, but will eventually become visible in amplify any electrical noise present. In the case of significant under- the mid-tones and highlights as you increase the ISO setting. All dig- exposure, this electrical noise may be almost as great as the image ital cameras display this tendency, some more than others, particularly when we turn up the ISO to shoot in dim light. Camera manufacturers have made great strides in reducing noise in high ISO images over the last few years, so you will have to do your own testing to see where the limits are for your own camera.

The solution is to avoid underexposing if possible. Image editing programs have some ability to remove these noise artifacts, and given that the damage from over-exposure is completely irreversible, underexposure is always preferable to over-exposure if you have to err at all.

Technology is improving all the time and each new generation of cameras is better than the last. . Frankly, I am often amazed at the clean detail that can be pulled out of underexposed shadows, even at higher ISO settings, from the newest generation cameras.

The shadows in this interior image taken in a church in the town of Santa Mar- gerita Ligure in Italy are underexposed. The image was shot a high ISO. The underexposure and high ISO combination has introduced noise in the darker areas, which shows up as small random coloured flecks clearly visible in the enlarged section of the image on the right

© 2013 - 2017 Mark English Page | 23 So What Does This Mean For My Prints? exposure. You should not fall in to the trap, as I often see many pho- It is these combined problems of sensor saturation and noisy shadows tographers doing, of constantly relying on this display. Shooting, and that creates a practical limit to the range of tones that your sensor can immediately dropping the camera to stare at the back panel display, record. All we have to work with is what lies within this range. in a never-ending cycle is comical to watch when taken to the extreme. As far as the sensor is concerned, by far the most difficult limit to deal Believe me, you will miss more great images than you will rescue by with is the upper one imposed by sensor element saturation. Once this constantly pouring over the camera histogram after every exposure. happens, you begin to compress the highlights together at the top of However, it is one more tool we have available to us, and it is a good the tonal scale, losing separation in these highlights and eventually training aid as we learn to judge exposure on our own. Learning ending up with featureless bald white patches in your prints. The about histograms will also help us to go a bit deeper with our discus- practical implication of this is that overexposure becomes THE sion of digital camera exposure issues. cardinal sin in digital imaging, and therefore, THE cardinal sin in creating a high quality file from which to print

If you began your photographic career shooting transparency film, you will recognize this problem immediately … overexpose slide film and your highlights end up as clear film base!

From a practical perspective, particularly if you are shooting RAW, your camera’s meter will produce decent exposures in the majority of situations. You should never feel paralyzed or feel a need to over- analyze most exposure decisions. In general, pay attention to the highlights in your exposure decisions and will end up with excellent files almost every time. If you’re unsure, you can always bracket a bit, and then pick the best exposure later.

Lastly, the digital world provides us with a unique tool to “nail” our exposures every time; so long as we understand how to use it. This tool is the Histogram display on the rear LCD of our digital cameras.

Now, before we dive in to the camera histogram, let’s get this one thing out of the way. The histogram is just one more tool to judge

You may already understand that a histogram is a graphic picture of how the tones in an image are distributed from dark to light. Dark tones beginning with pure black are represented on the left side of the chart, progressing through to lighter tones to pure white all the way over on the right. The height of the chart depicts the relative number of pixels in an image that occupy a particular tonal value. This histogram is from the image on the right, and is one

example of a histogram well exposed image. This particular histogram is from Photoshop; your digital SLR will have a similar display. Check your camera manual if you don’t know how to turn on this display.

To see how this helps to achieve a good exposure, let’s look at a few examples from exposure errors.

Overexposure:

Let’s look at an image with a lot of light tones in it. Let’s further assume that some of the lightest tones would be delicate highlights in say, the highlights on a young girl’s hair and the white collar of her Next, let’s over-expose this image a bit, and have a look at the result- dress, as in the studio portrait on the right. Part of the success of this ing histogram … image will depend on retaining separation in the various shadings of white and near white in the collar and her blonde hair.

© 2013 - 2017 Mark English Page | 25 Over exposing this image pushes the highlight values up the tonal scale. Even with a modest amount of overexposure, the subtle white values in the collar are pushed together, losing separation, and leav- ing us with featureless white patches

The Histogram of this image would look like this:

A substantial portion of the pixels are now crowded to the right side of the histogram, with the lightest values being pushed up the right side of the graph. These represent those “overflowed” highlight pixels.

Under Exposure

Here is an example of a histogram from an image that is probably underexposed

I say that it is “probably” underexposed, because images with a preponderance of dark tones will always display a histogram similar to the one on the left, even when properly exposed.

© 2013 - 2017 Mark English Page | 26 This brings up an important point. The image will dictate the shape Expose to the Right of the histogram. While some images may produce an idealized his- Using your camera histogram correctly involves remembering one togram like the one page 25, with relatively few pixels at the black simple rule. “Expose to the Right”: Adjust your exposure until you end, a “hump” in the middle, and relatively few pixels at the highlight have a histogram with the highlight pixels just approaching the right end of the scale, many properly exposed images will not. High-key side of the histogram, but not “climbing the wall”. This will maximize images, which by definition contain a preponderance of lighter tones, the available dynamic range of your sensor, avoid overexposing the will have a histogram shifted to the right. Low-key images, which by all-important highlights, and keep you as far out of the noisy “shad- definition contain a preponderance of darker tones, will have a histo- owlands” as possible. Despite the differences in their histograms, each gram shifted to the left. However, there should never be any signifi- of the images below are correctly exposed. cant pixels “climbing the wall” on the right (highlight) side of the histogram. Specular highlights, such as those created by bright sunlight One final note: Never rely on the image from your camera’s LCD to reflecting of water, glass or chrome for example, however will always judge your exposure. If you practice exposing to the right, the image show up (appropriately) as “blown” highlights in a histogram. You will appear quite light on your camera’s rear LCD. No matter: the should not be concerned with specular highlights in the histogram exposure, regardless of what you see on the LCD will be correct and will be perfect for creating a great print.

Differences in the distribution of tones in different images will produce different looking histograms. Each of the images above is correctly exposed

At this point you may be wondering when I am going to get around technically inferior image file. The expression, common in the com- to talking about actually printing images. puter world: “garbage in, garbage out”, is never more true than for digital printing. I’ve spent a lot of time talking about getting a well exposed image saved in an appropriate file type. My reason for this is simple: you cannot expect to create a great looking print from a poorly exposed

Remember back to page 9, where we introduced the concept of colour Way back in the day (early 90’s) at the dawn of digital imaging, the spaces? Recall that a “colour space” is just a way of describing the whole idea of “colour management” was only in the early stages of limits of all the colours a particular device can record (in the case of a development. The colour reproduction capabilities of most all con- digital sensor), or reproduce (in the case of a monitor or a printer). sumer grade monitors and printers were limited and, then as now, When we load an image into Photoshop or import it into Lightroom, every different monitor and printer-paper-ink combination repro- the program will want to (in fact need to) load the colours from your duced colour differently. Without a way to know just exactly how image file into a specific colour space. Lightroom will make the choice each device reproduced colour, the only way to ensure there was a for you (more later), but Photoshop will ask you to make a choice.

We have a few choices. The profiles we created for our monitor or downloaded for our printer-paper combinations describe the colour spaces of each of those devices. Should we setup PS to use one of these Colour Space of a typical digital camera to edit our images? The problem with each of these is that they are Colour space of several photo papers specific to a particular printer-paper combination, or a particular monitor. If we use a colour space specific to a particular device, we are limiting ourselves to just the colours that device can reproduce. If SRGB colour space we edit an image in the colour space for a specific paper-ink combination, what happens if we decide to print it on some other paper, or on a different printer? What happens if we decide to publish it on the web (which has its’ own colour space) or send it to a commercial printer?

Each “colour space” we have been talking about up to now is specific to a particular device: to a specific printer-paper-ink combination, or to a specific monitor. Because of this, none are really appropriate for editing mages that may be used in several different ways, printed on different papers, published to the web or in printed books. What we need is a sort of universal colour space that does not depend on a particular device. Turns out we have a few choices. reasonable chance of predictable colour, was to work with a restricted colour space that would accommodate most every device then in use.

© 2013 - 2017 Mark English Page | 29 The result of collaboration between Microsoft and HP, the sRGB col- Fortunately, there is another colour space we can choose, one that our space was a good match to most consumer cameras, monitors and closely matches the colour space of our cameras: ProPhotoRGB. Orig- printers in the mid 1990s. It’s still in use today, but any modern dSLR inally developed and later abandoned (somewhat ironically) by Ko- and many photo-capable printers can easily reproduce colours out- dak, this is a very large colour space, and is particularly well suited to side of this colour space. The diagram on page 29 shows the sRGB digital photography with today’s cameras. This space is so large in colour space overlaid on the colour spaces of several modern wide- fact that it includes some colours that we can’t see, but it does cover gamut papers and the approximate colour space of a modern dSLR about 90% of what we can see. The only problem, if you can call it (the largest triangular area). Any of these papers can reproduce col- one, is that to use ProPhotoRGB effectively you must work with 16- ours outside of the sRGB colour space, as can your camera. If you bit images. If you decide to shoot only in RAW format, this won’t be choose to edit your images in the sRGB space, you give up the possi- an issue. bility of reproducing many colours that you camera can record, and To sum up our choice of ProPhotoRGB as our editing colour space of which can be printed with many current printers. choice: As digital imaging evolved and printing capabilities improved Adobe • Modern photo capable inkjet printers (such as the Epson 3880 or defined a new colour space that was a good fit to most printing pro- the Canon Pixma Pro series) can easily reproduce colours outside cesses available at the end of the 90s: Adobe RGB (1998) is a signifi- of sRGB and Adobe RGB(1998). • ProPhotoRGB offers the largest colour space and is a pretty good cantly larger colour space than sRGB, and is a good fit to many mod- match to our digital camera’s capabilities. ern papers. It still represents a much smaller space than would be • Large colour spaces like ProPhotoRGB really demand that we required to describe all of the colours that your dSLR can record, and use 16-bit image files for maximum flexibility in editing. that can be printed with some printer-paper-ink combinations. I ha- • Using a smaller colour space like sRGB results in throwing away ven’t shown AdobeRGB on the diagram on page 29, but suffice it say colours that our camera may capture and that you may be able that it is larger than sRGB and but smaller than your dSLR camera’s to print. In this way, using a smaller colour space represents a colour space. one-way trip. There is no way to recover lost colour information Now, here’s the problem with using a colour space smaller than that that originally lay outside the gamut of the smaller space we of your camera: when we load an image into either the sRGB or Ado- chose to use in PS by converting the image back to ProPhotoRGB. beRGB(1998) colour spaces, we may have to throw away many of the Let’s pull the whole colour management process together with a quick colours your camera captured. Remember that a colour space defines overview of where we are and then move on to setting up PS and LR the limits of the colours that can be defined using it. Throwing away for colour managed printing. perfectly good colour information just doesn’t sound like a good idea to me.

Here’s what we have learned so far: • Each device in the photographic chain, whether your camera, monitor, or printer reproduces colour differently than every other device. • Each device can be said to have its own colour space which defines the limits of all the colours it can reproduce. Monitor Profile.icc PS or LR • ICC Profiles are files that describe the precise limits of a particular Converts colours from camera device’s ability to reproduce colour. • Photoshop uses these ICC profiles to perform the conversions Photoshop from one device’s colour space to another, on the fly. or Lightroom

• We chose ProPhotoRGB as the colour space we will use to edit Editing Space and save our images because it preserves as much of our digital ProPhotoRGB camera’s colour information as possible. • PS and LR reinterpret the colours from your camera raw file, and converts them into whatever colour space we have chosen for the raw conversion; ProPhotoRGB for example. When you set up PS Output Profiles for specific Printer, Paper and Ink work on your images in PS or LR, you are working on them while Example: “Pro38 PLPP.icm” they are in this colour space. • PS and LR use “Output Profiles” to convert the colours from your editing colour space to the colour space of your chosen Printer- Paper combination.

This process is shown diagrammatically on the right.

Setting up Photoshop CC or CS6 for Colour Management

Now that we have all the theory out of the way, let’s set up PS for it’s defined within the AdobeRGB space for example, Photoshop has good colour management. We start by selecting ProPhotoRGB as our the choice of opening it in the Adobe RGB space, or converting it to editing space. The dialogs shown here are from Photoshop CC or CS6: our working space, or ignoring the space all together and just map- earlier versions are similar. ping the images colour numbers to whatever colours they represent in our chosen working space (NOT a good idea, by the way). Gener- Select EEEdit>Edit>dit>CCCColorolor SSSettingsSettings …The Color Settings dialog appears: ally, we will want to preserve the embedded profile and allow PS to If your dialog doesn’t look like open the document in its own this, then click the “ more options ” space. Here again we will ignore button on the right of the dialog. the other working environments Let’s look at each of these settings, (CMYK and Gray) after which we will save these as The next three check boxes tell PS our standard color settings as I to warn us before proceeding if it have done (note the “settings” encounters images that are not in dropdown contains “Paclight Std” the default RGB colour space. I these are my saved standard set- generally leave the first one un- tings. checked, and the last two Under “WWorkingorking SSSpaces Spaces ” we can checked. choose the standard space for dif- Next, we will leave the color ren- ferent working environments, dering engine under Conversion RGB is really the only one we are Options set to Adobe (ACE). Un- concerned with here: choose Pro- der rendering intent , leave it set to PhotoPhotoRGBRGB . The remaining dropdowns deal with other working envi- “Relative Colorimetric ”, “Black Point Compensation ” set to on , “ Use ronments, such as CMYK for offset printing, none of which we are Dither for 8-bit images ” checked . concerned with here. You can choose to save these settings as your default colour settings Under Color Management Policies, we tell PS how to handle files we (Choose “ Save ” from the buttons on the upper right, and give your attempt to open that may not be in our preferred colour space. If we settings a name) try to open an image that is tagged with an embedded profile saying

© 2013 - 2017 Mark English Page | 32 That’s it. Now that your monitor is calibrated, your colour settings properly use output profiles for your printer and various papers. are done and saved, you’re all set to start working in a colour man- We’ll cover that later. aged environment. The only piece missing are the printer settings to

Setting up Adobe Lightroom for Colour Management

Lightroom has become an indispensable tool for many photogra- Photoshop for some reason (complex selections and localized correc- phers, including me. From ingesting large numbers of images and ap- tions, merge to HDR or Panoramas, for example). When we need to plying metadata such as copyright and keyword information, to the go to Photoshop from Lightroom we have to tell Lightroom how our exceptionally flexible non-destructive image editing capabilities of the image should be opened in Photoshop. RAW processing function, to the ability to quickly create slideshows, In Lightroom, go to Edit>Preferences (on a Mac this is located un- and web galleries, I can’t imagine working without it. At $US149, it der Lightroom>PreferLightroom>Preferencesences ) and click the External Editing tab at the has to one of the best deals in digital photography top of the dialog. Set each of the items as you see below under the It also holds exceptional promise for the workflow of all photogra- block titled “ Edit in Photoshop CCCCCC” (or whichever version of PS you phers who produce prints for sale. The major benefits for the fine-art are running. Set each of the drop-downs as you see here. 8 printer are efficiency, ease of use, and repeatability. Ease of use in this case also translates into reduced potential for error compared to the comparable Photoshop workflow.

This will not be a comprehensive course in Lightroom. For that, I refer you to the courses here at PPSOP by Jon Canfield and Rob Sylvan. We will however look at a workflow for fine art printing.

To start with, one of the major advantages of Lightroom is simplified colour management. Lightroom is essentially setup for good colour management right out of the box

Internally, when you open a RAW image from your camera in Light- room, it opens into a colour space known as MelissaRGB, which for our purposes is essentially ProPhotoRGB. So there is really nothing to set up… unless you want to take your image from Lightroom into

8 You will only see the “additional external editor” choices if you have another photo-editing program installed on your computer.

Getting the colours we see on our monitors matched as closely as pos- computer as part of the process of installing the printer software. sible to the same colours on paper is the job of the Output Profile. However, if you are printing to a fine art paper made by someone Output profiles (also referred to as ICC Output Profiles) are created in other than your printer company, you will have to go to the paper a manner similar to profiling your monitor. In this case though, you manufacturer’s web-site to download the profile for this paper on print out a standardized test image, and measure the colours in the your printer. Installing the profile is simple: just follow the instruc- resulting print to determine how close they are to what they should tions that came with the downloaded profile. be. The program running the measuring process then creates a profile Basically, most output profiles are installed in a similar manner: that tells Photoshop how to adjust the colour instructions it sends to the printer. Output profiles are unique to a particular paper, printer For Windows users: right-click on the downloaded ICC profile and and ink combination. You can purchase the equipment to create your select "Install Profile“. After a few seconds, the profile is installed. own output profiles, but it is quite a bit more expensive than monitor For Mac users: Move the downloaded ICC file to your ColourSync profiling equipment, and the results you get may not always be as folder, which will be one of: good as you might like. You can also contact a service that produces • For Mac OS 8-9x - System Folder > Color Sync Folder custom printer profiles. They will send you a test image file with in- • For Mac OS X - HD > Library > ColourSync > Profiles structions on how to load into PS and print it. You mail it back to them, they measure it and produce a custom profile for your particu- Now, let’s load an image to Photoshop and set it up to use the output lar printer, which you install on your computer. You must do this for profiles we have just installed. We may as well use an actual image each paper you want print with… so this option can become expen- you would like to print. For this exercise, pick an image without a lot sive very quickly. Alternatively, you can just visit the web-site of the of strong contrast or heavily saturated colour. Load a sheet of 8½ X manufacturer of your chosen paper and download the free ICC output 11 paper of your choice into your printer. Now, we will talk more profiles available there. about correctly sizing images for printing next week, but for now, in PS go to Image>Image Size and under Document Size first set the Res- Installing and using output profiles olution to 360, and then set the long side dimension of your image (Height or Width , as appropriate) to 9 inches. If you are printing with a paper made by the same company that made In PS CC or CC (2014) set the dropdown at the bottom of the dialog to your printer, then the output profile was probably installed on your “Automatic ”9 if it is not already, and click “ OKOKOK ”

9 If you are using Photoshop CS6 this setting will be “Bicubic Automatic”

Open the Print Dialog. (File>Print). Near the top of the dialog in the block labeled “Printer Setup”, there are two Layout buttons: one for Portrait and the other for Landscape orientation. Choose the one appropriate for your image

From within the Print dialog, choose “ Colour Management ”, you may have to expand this section by clicking the small black triangle just to the left of the section titled: “ Color Management ”.

Under Colour Handling , choose “ Photoshop Manages Color ”.

Under Printer Profile , choose the profile appropriate to your chosen paper (this one is for Epson Premium Luster Photo paper on an Epson 3800).

Choose a Rendering Intent; refer to the installa- tion instructions for the output profile you are using as a starting point. It is however, quite all right to experiment with this setting. There are only two choices among those available to you that are appropriate to fine art printing: Relative Colorimet- ric and Perceptual . Try either or both to see which looks best with your image.

If you settle on Relative Colorimetric , check “Black Point Compensation ”. This setting has no real effect with Perceptual, so on or off is fine.

This second step is different for a Mac or a PC. We’ll cover the PC first…

In the top section labelled “ Printer Setup ”, click “ Print Set- tings ”… this will access the Printer Driver dialog (shown on the right)

In this dialog we want to do one thing if nothing else:

Turn off printer based colour management .

Trust me: Photoshop does a much better job of managing colour in your prints than your printer ever will. Forgetting this one- step messes up more beginning print efforts than any other single cause. To turn off printer based colour management in an Epson 3800 - under MMModeMode , select “ Custom ” and “ No Color adjustment ”.

This one setting is critical. If you fail to do this, you will effectively be applying colour management twice: once by Pho- toshop, and once more by the printer driver.

Under media type choose the type appropriate to the paper you are using. For third-party papers like those from Moab and Hahnemuhle, consult the instructions from the manufacturer. This setting is important, since it sets the maximum ink load de- livered to the paper (gloss papers can generally accept more ink on paper than matte papers before they begin to block up).

Select the appropriate paper size, click “OK” to return the Pho- toshop portion of the Print Dialog

That’s it… load up some paper and click Print… unless you are using a Mac, in which case, read on.

© 2013 - 2017 Mark English Page | 37 The Print Settings dialog looks a little different on a Mac. It’s also a lot easier. If you have chosen “Photoshop Manages Color” in the first step (page 36) it will already be set up for you, but let’s walk through it anyway.

Click the drop-down showing “Layout” and choose “Color Matching”. Confirm that this is set to Color Sync. In fact, this setting will be greyed out, preventing you from changing this to anything else.

In the same drop-down where you chose “Color matching”, now choose “Print Settings”.

Click “Basic” and confirm that the Media type and Ink type are correct for your chosen paper according to the instruc- tion that came with your paper or its output profile

Accept the remaining settings for now, and click “Save”

Setting up Lightroom for Colour Managed Printing in Windows

We have to set up the same items in Lightroom that we did in Pho- describe the PC version first. toshop so that it can use your output profiles. The difference in LR is Select an image to print and head over to the Print Module by select- that we have the ability to save a group of settings for a particular ing it in the upper right corner of the LR interface. Access your paper as a preset, and then re-call them later. This not only saves time, printer’s setup dialog by clicking on the PPPagePageage----SSSSetupetup button in the bot- but also minimizes that chances of missing a particular setting or mak- tom left of the program and then clicking on the properties button for ing a wrong choice. your printer. This is the same familiar print dialog we saw in Pho- I will assume that if you are planning to use LR as your home base for toshop. Adjust the dialog for your chosen printer, paper type, and printing, you are already familiar with the interface. There are differ- size, just as described for PS. ences between the Mac OS and PC interfaces, and here again, I will

© 2013 - 2017 Mark English Page | 40 All that’s left is to tell LR to use the correct profile for your chosen paper. Close the printer setup dialog if you haven’t already by clicking OK. Look at the left side of the LR interface, and let’s run down the choices you have:

For normal single-image-per-page printing, you can ignore the Layout Style and Image Settings Tabs and just accept the defaults.

Under Layout , adjust he margins to your taste. Adjust the Cell Size so that the image size on the page is to your liking.

Move down to the Print Job Tab.

Leave Draft Mode Printing unchecked

For now, check Print Resolution and enter 360 in the field to the left.

Set Print Sharpening to Standard

Set Media Type according the paper you are using. Any coated paper (those described as glossy, luster, satin or similar terms) should be set to Glossy, uncoated matte papers should be set to Matte

Under ColColoror management : this where we will select the appropriate output profile for the paper we will use for this print. Selecting from this dialog is a bit different than in PS. Click the field to the right of “Profile ” to open a list of profiles available for use with LR.. You will likely not see the profile you are after the first time through. At the bottom of the fly-out that appears when you click the PPProfileProfile field, you will see an option for “ Other ”. Click this. The “ Choose Profiles ” dialog appears. (Next page) Scroll Through this until you find the profile you are after, and select it by clicking the box on the left.. Click OK, and you’re back to the Lightroom Color Management section.

© 2013 - 2017 Mark English Page | 41 Fig 1: The Choose Profiles dialog in Lightroom. All profiles installed on your computer that are suitable for printing are listed here. Click the check box next the ones you wish to use for your print jobs

The Mac interface for LR printer set up is a bit different from the PC. In the lower left corner of the Print module, you will see two buttons. Start by clicking Page SSetupetup to open the " Page Setup " dialog. Select your printer, paper size and orientation. Leave scaling at 100% and click "OK"

Leave " Print Settings " for now.

Move to the right side of the LR Print Module. Set your margins and image size in the Layout tab as described in the lesson, and move to the bottom tab labeled, " Print Job ".

Set print resolution and sharpening as described in the lesson. OS X has the option to use 16-bit printing, which you may select, or not: your choice. Theoretically, 16-bit output should produce smoother gradients, so you may as well check it and move on.

Now, select the appropriate output profile for your paper, and select a rendering intent.

Lastly, we have to go to the printer dialog and turn off printer-managed colour, so let’s move to the next page.

In the page setup dialog, choose paper size, orientation

This opens the “ Print ” dialog

Click the dropdown currently displaying " Layout ", and choose " Print Settings ”. Under “ Color Settings ”, choose “ Off (No Color Adjust- ment ”. You may also check “ 161616-16 ---bitbitbitbit ” if you chose this in the LR Print Job panel.

Click the dropdown currently displaying " Print Settings ", and choose "Colour Matching ". Note that all options under this dropdown are greyed out, and no editing is possible. This confirms that we have turned off printer managed colour. Select “Color Matching” here, to see …

This is one of the great features of OS X: by selecting an output profile on the right side of the Print Module in LR, you automatically turn off printer managed colour. If you wish, click the dropdown that is now displaying “ Print Settings ”, and select “Color Matching ”. You can confirm that “Color sync” is selected and other choices are greyed out, preventing any further adjustment

That’s it; you’re done.

© 2013 - 2017 Mark English Page | 44 allowing you to save these settings. Give your new LR print settings Now here’s the best thing about print setup in LR. a descriptive name and click “create ”. On the left side of the screen, locate the small “ +++” sign next to the Sec- Now, every time you want to print an image of this size on Epson tion named “ Template Browser ”, and click it. A new dialog opens up Ultra-premium Lustre paper, you simply select it from the User Tem- plates section of the Template Browser tab on the left of the LR interface. All settings, including those required in the Printer Driver Dia- log are recalled and set for you.

Wrapping Up Week 1

That’s all for this week. To recap, we are now working with a properly calibrated and profiled monitor. Photoshop and Lightroom are both set up for good colour management, and setup to play well together when the need arises to move an image from one to the other. We know how to find output profiles for the papers we want to print on, and how to set up PS and LR to use them correctly.

Next week we will look at approaches to correctly sizing your images for output, for sharpening them for different image types and output media. We will also look at soft-proofing; a useful technique to find that last bit of colour and dynamic range in your prints.

Computers only think in terms of 1’s and 0’s (ones and zeros). 00000000 00000001 Imagine you are given the task of devising a way to represent images 00000010 digitally. Since you are devising this method for use with computers, 00000011 you can only use 1’s or 0’s to represent all the tones in an image, since 00000100 00000101 this all that computers understand. 00000110

In computer terms a single digit, a “1” or a “0” is called a bit ( bi nary And so on, all the way to 11111111. digit). Programmers group these together into groups of eight bits to create a “byte”. Bytes are like letters of the alphabet in computer Turns out there are exactly 256 possible combinations of 1 or zero speak. Bytes can be further grouped into eight byte “words”. when grouped in clusters of eight. So, we could for example use the code with a value of 00000000 to represent pure black. Code value Now that we know that, let’s think about our digital code that we have 11111111 could be used to represent pure white. All of the intermedi- to develop. Let’s just think about black and white images for the mo- ate values would then represent all of the shades of grey in between. ment. We recall that B&W images printed in books are made up of thousands of dots, and when they are assembled together our eye and So far so good, but we have somewhat boxed ourselves in to a corner. brain sees all these dots as a cohesive image with all tones from black Because we are limited to representing the tonal values using combi- to white. nations of one and zero only , when we use groups of eight we can only have these 256 levels to describe our image. We can’t have say, a value So from this we know that that if we break an image into tiny little of 43.6 or 134.9. So: will this code with 256 possible values be enough pieces, too small to be seen on their own we could measure the bright- to describe the dots in our B&W picture? Will there enough different ness or tone of each of these image pieces, assign it a value in our code shades of grey in the resulting picture so that there will be the appear- to represent this brightness (that is, its shade of grey), and then tell the ance of smooth tonal gradations? Or, will it appear more like a “paint computer to assemble all these pieces into the dots of our image. by numbers” painting? Ok… what if we used just a single bit to create our code? Then we Fortunately, it turns out that if you break down all of the shades of would have two choices: 0 or 1, “black” or “white”. Not great, but it’s grey possible in a photographic print between the blackest black and a start. What if we used 8 bits (a single “byte”)? How many possible codes would we have to represent all the tones in an image? whitest white that can be reproduced on the paper into evenly spaced The first few codes in our scheme might look like this:

© 2013 - 2017 Mark English Page | 46 divisions, by the time you have divided the possible tones into some- we could refer to a tone with a value of 178, rather than its binary thing between 250 to 350 evenly spaced steps, most people will no equivalent 10110010: much easier to relate to and to remember. longer be able to discern the difference in tone between adjacent steps. They will instead see the whole as a smooth gradient from black to Expressing colour data white. As you probably already know, your colour television creates full col- Depending on the lighting conditions, many people will see fewer our images by blending red, green and blue dots together to create the than this, and under typical home living room light, most people can illusion of all other colours. A few moments thought will suggest that only discern roughly 150 steps. So it seems that our code is at least a we could extend our new “Digital Code for B&W Images” to provide fair match to human vision under the most common print viewing us with a way to encode colour. conditions. If we were to put little red, green and blue sunglasses on different sen- We can now safely go ahead and represent each dot in our image us- sors, and arrange them in clusters, each composed of a red sensitive, ing one 8-bit code value. If we measure the brightness of each dot in green sensitive, and a blue sensitive sensor, this would provide us a our image, and assign it the appropriate value in our code, we have way to measure the relative amount of red, green and blue at every found a way to encode a B&W photographic image. point in our image.

But remembering and talking about all of those eight digit combina- Now, instead of a single 8-bit code to represent each dot in our picture, tions will get tiresome quickly. So why don’t we just assign a simple we will need three 8-bit codes, one for the red component, the green number to each 8-bit code? Like this: component, and the blue component. Since we have 256 possible 0 - 00000000 tones from darkest to lightest that we can code for, in our colour image 1 - 00000001 2 - 00000010 code we would have 256 X 256 X 256 = 16,777,216 possible colour hues 3 - 00000011 that we could describe. 4 - 00000100 - Every colour possible is therefore represented internally in groups of - three numbers, for example: 250,200,0. By convention the first num- - 178 - 10110010 ber represents the intensity of the red component, the second refers to - - green, and the third refers to the blue component. Digital images de- - scribed this way, using 8 bits to describe the intensity of each of the 255 - 11111111 three primary colours, red, green and blue, are referred to as “8-bit images”. Perhaps somewhat confusingly you will also see them re- Now, in this way black becomes value “0” in our code, while pure ferred to as “24-bit images”, referring to the total number of bits used white is “255” and when we want to refer to a certain light grey tone, (8 bits X 3 primary colours = 24 bits total)

© 2013 - 2017 Mark English Page | 47 Is Eight Enough? We can’t for instance, say a tone falls on value 235.7, or value 46.3. The We have seen that using 8 bits of tone must be described as 235 or computer data to describe all 236, or 46 or 47. the tones in a black & white im- When we start to adjust the tones age, or of a single primary col- in our images using programs like our, provides us with a way to Photoshop, we move them up or describe 256 possible tones. We down the tonal scale by calculat- also know that 256 discrete ing new values for them (or rather tones from black to white (in a Photoshop calculates new values photographic print) are just based on our inputs). These new about at the limit of what hu- values can only fall on whole man vision can perceive. numbers from 0 to 256 in an 8-bit But are eight bits really enough? world. Multiple large adjust- Research in human visual per- ments to the tones in our images ception suggests we can discern cause Photoshop to introduce more than 256 levels at some mathematical rounding errors level (perhaps just subcon- that will eventually leading to sciously), particularly under Image #1 Image #2 “clumping” of the tonal values. good lighting conditions. In a It’s easier to show you the results typical home, were illumination of this, than it is to describe it… so levels are lower than say a well lit art gallery, you will lose the ability read on. to distinguish a good chunk of the available tones, mostly in the Image #1 on this page shows subtle gradations in the blue tones from darker shadows of a print. Outside under bright sun, you will also dark to light. Image #2 shows what happens if we apply several large lose the ability to distinguish tones, but these will be on the highlight adjustments to the tones in an 8-bit image. The subtle tonal gradations end of the tonal scale. So there may be some advantage to having in the wall and in the umbrella are lost. Instread what is left is just a more than 256 possible tones in a photographic print. few tones with large gaps in between, a condition known as “posteri- But there is another more significant advantage. zation” or “banding”. Recall that in the 8-bit digital world, any given tone can only be de- The problems with Image #2 are a result of errors introduced as vari- scribed with one of 256 possible values from pure black to pure white. ous tones are assigned to one of only 256 possible values.

© 2013 - 2017 Mark English Page | 48 The solution is to find a way to allow more than 256 possible tones in Since we are slicing the available colour pie much finer than before an image. This will provide more possible tones within the tonal scale quantization errors are greatly reduced. of our image, and allow us to finally get those intermediate steps we The moral of this story is of course to create and edit your images us- need to prevent or at least reduce the number of rounding errors re- ing 16 bits per primary (Note: 16 bit images are also referred to as 48 sulting from tonal adjustments. bit images [16 bits X 3 primary colours = 48 bits total]). Rather than 8 bits, let’s use 16 bits (two bytes, instead of one). How 65,536 possibilities per colour consume considerable digital pro- many combinations of 1 and 0 are there when we group them into cessing power in your camera and are difficult to move quickly from clusters of 16? The answer is 2 16 = 65,536 possible combinations, and the camera sensor to your memory card. As a result, most high end 65,536 possible divisions from black to white. This is clearly a much SLR cameras actually use 12–bits (some models can work with 14 bits finer way to slice the available pie. When we extend this concept to per colour channel) to describe each colour primary rather than 16. colour, we find that we now have 65,536 red tones x 65,536 green tones This still leaves us with 212 = 4,096 possible values for each primary x 65,536 blue tones = well, billions and billions of possible hues we can colour, which in practice works very well. describe.