The End Game – the second half

Dave Montizambert's version of events!

Having made it back, safe and sound from the SWPP Coventry Convention to very rainy Vancouver, I sit wide-awake at 4 am (jet-lag is still haunting my sleeping hours) reflecting on that action packed, fun-filled event. During my morning talk at the convention, I tried to pack the essence of a two-day workshop on digital standardization into two hours. As you can imagine some of the info ended up on the “editing room floor” so in this article, after a brief synopsis of my talk, let’s pick up one of those edited “bits” – pre-visualizing printing values from Photoshop densitometer read-outs – and see how you might implement it into your digital repertoire for a healthier, less fearful workflow.

My two-hour talk consisted of a discussion on how one might take the positive aspects of a JPG workflow (speed and standardized processing) and marry it to a RAW-only workflow. Starting with an exposure test done in 1/10th stop increments, I demonstrated how one could find the true sensitivity or ISO of a digital SLR. From these test shots, the best exposure is selected [2]. This is a rather tricky procedure since you are dealing with unprocessed RAW data [1] and so run into that age-old conundrum “which came first, the chicken or the egg”. The crux of the matter is, you cannot decide on the best processing settings until you know which exposure is best, yet you cannot choose the best exposure until you have processed the files. Anyhow after determining a preliminary processing setting we overcame that hurdle and moved on to figuring out the optimum processing settings for your system and for your way of shooting. With JPG you are pretty much stuck to the processing, set at the factory, by a technician, who has never seen how you shoot, hence the blown out wedding dress and blocked-up tux syndrome. After establishing your optimum process settings in Photoshop’s Camera Raw we learned how to create an Action that recorded processing a raw file in Camera Raw then creating a Droplet (an automation feature found under the File/Automate menu in Photoshop) that would allow you to create a drag and drop batch process using the new processing Action. In the end one would simply drag a file, or a folder of files, or several folders of files on to the processing Droplet thumbnail on your desktop. The Droplet fires up Photoshop, if it is not already up and running, then processes each image in just seconds and saves them out as Tiffs, JPGs, PSDs, or whatever, to a “processed only” folder. Now your images are ready to go just like your old JPGs would be. The advantage of this Raw automated workflow is, with only slightly more time spent (that’s computer processing time, not human time) you end up with files that are processed to your way of seeing, and you get to keep the Raw data – a real bonus just in case you need to bail-out an important image with bad exposure/contrast/colour. Raw data, which is high bit data (typically 12 bits per channel or 4096 levels of tone for digital SLR cameras), is like a negative, it contains way more data than the print or in the case of digital, the JPG. A JPG can only exist in low bit (8 bits per channel or 256 levels of tone) therefore Raw provides you with much more editing headroom. As you can see, at my talk, we covered off a lot in two hours, if you missed the talk or want to review, you can order a tutorial CD from me at montizambert@telus.com that covers all this and more, plus it contains files for you to work along in Camera Raw and Photoshop with the tutorial movie. Dave Montizambert's

One area of my talk that I did not get to cover last January, due to time constraints, was a thing I gave the baffling name “Reproduction Value Recognition”. Its really a very simple concept, however I believe you should always give all things photographic impressive confusing names so that you can impress people allowing you to charge more money – I learned that from my lawyer who has a degree in verbal flatulence. When you check-out an image in Photoshop to see how it looks and to perhaps get some idea of how it will look when it prints, we are told not to trust what we see on the screen, but to read the brightness and colour values sampled by your cursor and displayed in the Info Palette. From these numbers, they say, you can predict quite accurately how your image will look when printed. Now how on earth do you get that from a bunch of numbers? Besides, who is the body of people called “they” and how is it that they can see through numbers? Well “they” aren’t talking; in fact they aren’t even listed in the telephone book or even the yellow pages under “They”. After a period of pondering this “seeing through numbers” I recalled a conversation with my local pre-press guru, John Lau. I would often ply John with a few pints of beer, so that I could bombard him with questions on how to colour correct in CMYK. After practicing John’s teachings for a couple of weeks I found that I was not getting any better at it. At the next bout of beer swilling with John, he explained to me that it is very difficult to learn by the numbers when you have no way of seeing the results. For him it was easy, he always got to see the results of his corrections in fairly short order – John tested his corrections by printing the corrected image on a proofing device such as a Fuji One Print or a Color-Match proofing system that simulate press conditions. With this he could begin to correlate what a series of numbers actually looked like. Really no different than how you know what a Zone 6 (Ansel Adam’s Zone system) looks like – you originally identified Zone 6 with your meter then were able to see the result after you printed it – cause and effect. So what we need is a “Read-it and See-it” system so that we don’t have to live in fear of questions like, “Will the highlights burn-out or will the Shadows block-up losing all detail and if so will it look okay printed or what does slight detail look like in a shadow?” Dave Montizambert's

By looking at the numbers it is easy enough to learn if you within the safe zone for pleasing flesh-tone densities, and for shadow and highlight detail, but what do these values actually look like? That is an artistic decision, a visual decision. In my early days of shooting, long before digital, I used to swagger around bragging to fellow photographers that I never needed to shoot Polaroid test prints to get my exposure and contrast range right – I could do it all with meter readings prior to exposure. I was correct, however once I started shooting Polaroids my images became much more powerful because I could better see the meaning of the meter reading values and alter my lighting so I could better manipulate those values for more impact. There is nothing like having a hard cold 2 dimensional preview on paper in your hand to better judge your art rather than an impression floating around your head that you are trying to attach numbers to. And once you have seen a few prints displaying metered values or densitometer values from the Info Palette, you soon develop a feel or a memory for what values will look like printed. For this I have devised a simple test that will help you to pre-visualize what those numbers will look like printed.

To start the test, set up something that is typical of what you shoot. For a wedding or portrait photographer, include a standard flesh-tone (someone’s face), a neutral white such as a white shirt, and a neutral dark tone such as black wool jacket – see my test image, [3]. Also pop in a greyscale like the Macbeth 24 swatch colour chart or the Kodak Separation Guide, if you have it. This test can be done anywhere: outside in available light, in studio with strobe/flash lighting, or in your front room with window light. Warn your model that they will have to hold the pose dead still for 5 minutes. Quickly draw a sketch on paper of what you see in the camera frame, and then using a spot meter (a reflective meter not incident and if you don’t own such a device then rent or borrow one) take readings off the subject’s face, shirt, jacket, in fully lit areas and in shadow. If you have a greyscale in the shot, meter its swatches too. Now label the sketch with the metering values you just took. You could even place sticky notes with the values written on them and with arrows pointing to the metered spot onto the subject. Take an exposure. If you applied sticky notes, remove them and shoot again so that you have an unobstructed image too. Is your model still holding still?

Metering Values

I always use metering values rather than f-stop shutterspeed combinations. F-stops and shutter-speeds are image specific whereas values are not. To calculate a metering value, simply compare the reflective reading with the camera setting. The camera setting always represents middle grey. If my camera was set to f11 (so f11 equals mid-grey) as in the test image of Sylvianne (see image [1]), and if a reflective meter reading off her jacket reads f5.6, then that value is a minus 2 or a Zone 3 – f5.6 reflective reading relative to an f11 aperture opening will render a tone two stops darker than middle grey. I prefer the plus/minus system to the Zone system. In the Plus/Minus system, Zero equals middle grey (Zone 5), Minus 2 equals two stops below middle grey, and Plus 1 equals one stop brighter than middle grey (Zone 6).

Once you have processed the image with your standardized processing – be it a JPG or Raw workflow – open it into Photoshop. Read the same-metered spots of the subject in Photoshop by placing your cursor over the appropriate areas and noting the number values displayed in the Info palette. Now we need to label the digital image with the metered values and with the Info palette readings:

Select the Line tool from Photoshop’s tool bar (or type the letter “U” with the shift key held down as many times as it takes until the Line tool is selected). From the tool options bar at the top of the Photoshop window set arrowhead size to taste. Click and drag arrows that point to the metered areas. Double click on the Foreground colour swatch at the bottom of the Tool bar then select a suitable contrasting colour for your arrow. Make the Path palette visible then click on the “Fill Path With Foreground Colour” thumbnail (white to black gradated circle) at the bottom of the palette. With the Type tool selected, label the “arrowed” spots with the metering values and with the colour values from the Info Palette read-out.

I use this test for pre-visualizing densities only. To develop this ability for colours would be near impossible since the combinations of values in the three channels is staggering. Reading luminance or brightness values only would be a much easier method than reading RGB values, however Camera Raw and Adobe’s new application for image processing and viewing called “Lightroom”, read only in RGB. If you want to read brightness/luminance values only, and will only be doing this in Photoshop, then set your Info palette to Lab or HSB colour mode. Read only the L channel in Lab or only the B channel in HSB. You could use Greyscale too, but keep in mind it reads ink density percentages not brightness or luminance values. For this reason 100% in

Greyscale mode equals pure black and 0% equals pure white. Once your image is properly labeled with arrows and values, have it printed out using whatever media you sell your work on. If you are a wedding photographer you may have to send the file out to the lab, or you could print it out on your inkjet. If your work is usually printed on a printing press you may want to settle for seeing a “soft-proof” print from your inkjet since printing this test on a press would be cost prohibitive. I would suggest printing the image out to a handy size like 5x7 inch. Place this print in a small binder that you can carry around in your camera bag and/or next to your computer display. Next time you are spot metering prior to shooting and are wondering what a minus two value on a tux will actually look like when printed, or next time you read a bride’s white dress in Photoshop and wonder how bright a 238 red / 238 green / 238 blue will look when printed, you can refer to your test image. In fact, over time, collect labeled test images from various shoots that you do and add to your book, after a short time you will find that you will not need the reference book any longer because you will have developed a feel, when metering or when reading the Info palette, for what values will look like in print. You will be able to see through numbers, you will be a crack print value recognition expert. That ought to be worth a free pint or two!

Dave Montizambert's

Image [3]: Sample of test printout image labeled with reflective meter reading values and RGB values. Image [4] & [5]: The digital Cross-processed image of model Chelsea Walker in the black leather jacket may pose some printing problems for the printing press in this publication because of the very high contrast that my Cross-processing curves introduced (digital Cross-processing CD tutorial is available through me at montizambert@telus.com). Problems like: What happens when only one channel burns out, as in this case the red channel on Chelsea’s cheek (compare upper uncorrected image to lower corrected image on Image 5)? The shadow side of her black jacket (Image 4) straddles that fine line between slight detail and no detail – is that going to be okay and what do those values look like in print? With a little Print Value Recognition practice I’m able to make a fairly accurate guess and make corrections if needed. Check out the RGB readouts on Image 4 and Image 5 to see how they reproduce in this magazine. Are the results acceptable? This is somewhat subjective. My lighting mentor, the late Dean Collins, taught me to always go with the money – in other words, all subjective decisions ultimately lie with the cheque book holder. Also keep in mind that presses can be rather inconsistent, the last copies coming off press can be different from the first – consistency is very much in the hands of the pressman.

Dave Montizambert's

On Chelsea’s image, Image 4 & 5, I am not correcting the colour because I purposely shifted it for a Cross-processed look. However I am concerned with brightness values and will make sure that essential highlight detail sits no higher than 245 levels or less in all three channels. I also believe that all shadow detail will be lost below 15 levels. If this image were headed to a good lab for chemical based printing I could probably push that 245-highlight detail figure (which is an average limit for offset press) to 249, same for good inkjet printing. Often the three channels will not read the same because they are not truly neutral, so I average them. On Chelsea’s jacket at the arm pit it reads Red: 4 Green: 20 Blue: 21. This averages out to approximately 15 levels (4+20+21=45÷3=15 levels).

Dave Montizambert's

The SWPP 2008 Convention was an outstanding success,
we have 129 days to get ready for the 2009 convention - which starts on January 14, 2009

Photo Quote: I dream that someday the step between my mind and my finger will no longer be needed. And that simply by blinking my eyes, I shall make pictures. Then, I think, I shall really have become a photographer. - Alfred Eisenstaedt