Modern Photoshop Color Workflow - A review of the book
Imagine a classification of retouching work based on the amount of time spent per image. On one side of the scale we see jobs that must be executed in as short time as possible, say five minutes at most. On the other side we have work that in theory can take an unlimited amount of time. In practice, the upper limit will be something like multiple days work for one image.
The "multiple days" jobs are those where budget plays no significant role, where quality of the outcome is the only important factor. The obvious example is a cosmetics ad. The final image will be used thousands of times, on billboards, in magazines, on television, etc. There is no reason to save on retouching costs if that would compromise quality. Every pore must be softened, every eyelash straightened, every unwanted hair removed. Every pixel must be right, no matter the amount of time needed to accomplish that.
The "five minutes at most" jobs may be found where many images must be worked on and not much budget is available. Think of a booking company offering many accommodations via their website, each accompanied by a dozen of photographs. I suspect they don't have the money to spend even an hour on each photo.
Or think of professional photographers, trying to spend as much time as possible shooting and not sitting in front of their computer. If they can improve their images in a few minutes each, fine, but more... no thanks.
Figure 1: Book cover
I believe that the available time defines to some extent how a retoucher approaches an image.
Beauty retouchers usually have a background in arts. Given a photograph and anything to enhance, they pick a suitable brush and start painting over whatever seems wrong. Is the result not accurate enough? Take a smaller brush and keep painting until everything looks right.
Retouchers working on quick jobs don't have time for all this painting. Instead, they search for good alternatives. Adjust the whole image and then find a way to limit the effect to where it is needed. Move to a different color space if it better suits the job. Write an action that does the repetitive steps automatically.
Speed is an important parameter of the 2013 book "Modern Photoshop Color Workflow" (MPCW) by Dan Margulis. As such, it may primarily serve "five minutes at most" retouchers. But there is just as much focus on quality. How the two are related, and what it means for the reader searching for "speedy image enhancement" is the main subject of this review.
Figure 2. Example before and after image. In the book these are called "postcards"
Focus on speed...
Author Dan Margulis has his roots in the prepress business. This is highly evident in his earlier books, where much of the correction work was done in CMYK and where at least a few chapters were devoted to things like dot gain and grey component replacement.
Most of this doesn't apply to the MPCW book. Practically all processing is done in RGB or LAB. Not the print but the digital photograph is the primary focus.
The book presents a workflow for image enhancement, called "Picture Postcard Workflow" (PPW). Margulis states that the whole process can be finished within as little as three minutes on average. He keeps emphasizing that throughout the book, often by clocking the individual steps. "This step, I do this and this, 40 seconds spent, then I make some judgement and decide to skip a step, 10 seconds passed, next I run an action to improve that and that, etc."
But the actual time spent on one run of the workflow is not that critical. The book offers plenty advice for cases when even less time is available, or when (much) more than just a few minutes can be spent.
For images that are important enough for a longer treatment, Margulis recommends an unusual routine: run multiple passes of the workflow, compare the resulting versions and combine the best elements of each. This procedure gives better - and quicker - results than meticulously working on one version's weaknesses. The PPW lends itself well to this approach, yet it's a revolutionary concept to say the least.
A few things to keep in mind:
One may expect sub-optimal results to meet the speed requirement. But Margulis claims the contrary. His workflow gives "better results than traditional methods, in less time", so he states. This position is a central theme of the book, backed up by many examples. The workflow is not just about speed, it's about speed resulting in better quality.
But let me speak for myself. Does a photograph deserve more than just a quick retouch in Lightroom, I open it in Photoshop and run the main PPW steps, always for the better. Do I wish to spend even more time, then I consider local retouching work, but still apply the PPW as well, with more attention to each step, or even redoing parts of the workflow. In each case I get better results than ever before.
The PPW is definitely a good workflow for time-constrained retouching work, but in fact is recommendable for anyone wishing to enhance their photos. Just be aware that not all retouching work is covered by it.
... but it's not a breeze
I already hinted at an important aspect of the book: it is not easy.
The demand for quick results implies that the techniques are not always intuitive. A simple example is the case where one image area, say a bunch of background trees, is considered too dark. The obvious way to handle that is to create a selection for that area and work on that. Not so in this book. Outlining a selection - or painting a suitable mask - is considered too time-consuming.
Instead, other ways are advertized. The desired selection may easily be extracted from one of the color channels. Great, but why not use the channel itself to create a suitable mask? That's even quicker, and it may work equally well. If this causes unwanted side effects in other image areas, invoke blend-if sliders to block out the effect where it's unwanted.
Yes, this may be quick and effective but requires good understanding of channels, layers and blending options. Definitely not for the Photoshop beginner. Of course, all is explained in the book, but learning an advanced technique from a written text and being able to apply that technique fluently are two different things. This is why so many readers - see the reviews on amazon - remark that Margulis' books are worth multiple reads. It's true, and the simple reason is that the presented techniques are not easy to become familiar with.
The full workflow demands a lot from the reader. Step 1 prescribes to correct any color cast "by the numbers". This is the best way in my opinion, but it requires a basic feeling for numbers. As a mathematician, I have that, but I know other people who don't.
To make things worse, Margulis suggests (in a sidenote in his book) to "Work in RGB, think in LAB". That comes down to: assess color by checking the AB values of the LAB color space, but perform the correction by RGB curves. That may be a problematic task for many readers.
Step 2 is the channel blending. I was already familiar with this technique from older books by Margulis, but always saw it as some kind of black magic. The steps to perform are not difficult per se (understanding Apply Image is practically enough) but being able to judge which blend will be effective for a given image, is another story. A good bit of abstract thinking is required: channel blending is anything but intuitive. Not many beauty retouchers will be inclined to apply it, although a proper channel blend may definitely be benificial for skin retouching work.
But arguably the hardest aspect to master is the philosophy of the workflow. The traditional, more intuitive way of working is: correct the weak parts of your image, improve whatever you're not content with. Carefully keep your layers just in case something must be reworked.
The PPW is radically different. Intermediate steps of the workflow can worsen the image, but they prepare for better end results. Also, there is no need to keep all layers. If you think you can do better, or if you are willing to spend more time, do a second run from scratch. Pick the best or blend the results.
The good news is that Margulis knows his stuff to be difficult and splits most chapters (all except the introductory first two) in two parts. The first half to present the concept, the second half for the advanced material. Certainly, the first halves are enough to learn the basics of the workflow, but be assured that they are still not easy at all.
In the end, what's most difficult is not reading and understanding the book, but learning and becoming proficient in the techniques themselves.
Figure 5. The above "postcard" progressing through the channel blend phase
Beta reading and quality
Early versions of the text have been extensively reviewed by a selective group of so-called beta readers. Their feedback has caused many updates, and no doubt this has significantly improved the quality of the book. Some of the comments appear in the final text, functioning as different views on the subject under discussion.
The book's introduction contains an interesting overview of the whole beta reading process - from the selection of the group to the processing of all the comments. If I understand well, a good part of the introduction was rewritten as a result from an argument about the choice of an example image (!). The group discussions that arose from the beta readers' comments, must have been a lively affair.
But why doesn't the author rely on his own writing and retouching skills? (Enough available, one would say.)
I believe that it indicates how much he is striving for quality. He knows the subject matter like no one else, but still organizes a dedicated group around him with the sole purpose to improve his product.
This self-critical attitude applies to his writing as well as to his image retouching. On many occasions in the book, he emphasizes that one should never trust the virtue of an enhanced photograph. Compare the final result with the original: didn't we do a perfect job? Yes, until we see another final result, either by ourselves or by someone else.
This is an important lesson to learn from the book. Margulis, widely recognized as a leading expert on color correction, doesn't hesitate to admit and show his mistakes and warn us for similar faults. To me, this makes his work all the more convincing.
Writing and presentation
Those who own any of Margulis' earlier books will immediately recognize his writing style. Notwithstanding the focus on speed, do not expect many step-by-step tutorials. The author always utilizes a conversational style, even when a technique is laid out, well, step by step. Reading the book feels like someone takes you by the hand, explains what he is doing and why. The text is interspersed with side notes, question/answer sequences ("Why don't you ...") and a good joke here and there. The reader is guaranteed to pick up a lot of knowledge along the way.
The conversational style is not so good when the book is used as a reference source. How again should I use the Apply Image command to perform the channel blend that I have in mind? The pages where that is explained are quickly found, but then it takes some effort to distill the exact steps from the text.
To make things worse, the order of the chapters differs from the order of the workflow steps. That is pretty annoying, even though the book contains lists that map the workflow steps to chapter numbers. (See figure 7.)
Figure 6: Dan Margulis portrayed by some of his students
Figure 7: Overview of PPW steps. Note the mismatch between workflow and chapter order.
Not all of the text is about photography or Photoshop. Margulis, erudite as he is, picks examples from literature, music and visual arts to illustrate his points. For me the most striking case is when a certain action is compared to tempo indications in a Robert Schumann piano sonata. I won't provide the details, but let me state that whenever I apply this action, the comparison comes to my mind. Highly effective.
Printing quality is excellent, as expected from a prepress expert. The book is full of photos: originals and final results juxtaposed, but also many intermediate versions (see figure 5), or multiple alternatives when choices have to be made. Contrary to most other Photoshop books, not many Photoshop screenprints can be found. Some Layers and Channels panels, a bunch of curves, an Info panel here and there and some more obscure dialogs. The author relies on the written explanations instead of illustrating each step by a small screenprint.
Finally, as you may know, Margulis has a strong aversion to histograms: there is none in the book. All right, he has a stubborn side too.
The book "Modern Photoshop Color Workflow" is not a typical Photoshop book. The very short summary is: it is better and it is less accessible than the large majority of Photoshop books on the market.
Read all of it, and you probably learn many useful techniques, no matter how much of the workflow you adopt. Much advanced Photoshop comes along: curving, channels, Apply Image, blend-if, RGB and LAB coding principles, to name just a few. Even the old "equalize" command gets a place. (Tell me, do you know what it does?)
Then, if you want to go for it, start implementing the workflow in your image retouching work, first by using parts of it. Judge the results and ask yourself, if and how does this help to make your images better? Be prepared to reread and re-reread parts of the book, because applying the workflow, even if you use the panel (and you should), will certainly provoke questions.
If you are like me, persistent and ambitious, you will move your retouching skills at least two levels up.
The hard work is the paradox of this book. The workflow is supposed to give quick results. However, it is a long way, months or even years, before you fully master it.
But isn't the way towards the goal at least as fascinating as the goal itself?
Gerald Bakker, 25 May 2015 - rev. 28 May 2015
Thanks to Dan Margulis and Marc Prevot for giving their valuable comments on an earlier version of this review.
Picture Postcard Workflow
Copyright © 2015-2017 Gerald Bakker. All Rights Reserved.
Figure 3. The free PPW panel
Figure 4: Example curve. To be applied to a mask that is an inverted copy of LAB's B channel.