Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Photoshop LAB Color  (Read 5933 times)
jdemott
Sr. Member
****
Offline Offline

Posts: 434


« on: November 11, 2005, 12:13:52 PM »
ReplyReply

Dan Margulis has written a new book, Photoshop LAB Color, which has quite a few interesting suggestions on how to use the LAB color space in Photoshop to get better results than can be obtained in traditional RGB or CMYK editing.  Of particular interest to landscape photographers are techniques for enhancing the separation and vividness of colors in landscapes that appear to have washed out or flat colors, improving the color of forested scenes, correcting overall color to remove color casts, and restoring color to highlight areas that are nearly blown out, such as in sunsets.

This is a book that will make you think.  It is not just a collection of techniques or recipes for image modification.   Margulis is interested in how things work and why.  He looks at how we see color, both in nature and on the printed page, and how color is represented and reproduced in the digital world.  He talks about imaginary colors that can be represented in LAB but not reproduced in the real world (e.g., dark, dark, purely saturated yellow) and then shows how imaginary colors can be useful in a real world workflow.

Margulis’s orientation is not as a fine art photographer; he comes from the world of Photoshop professionals who spend their days color-correcting, retouching and enhancing commercial photos for publication in catalogs, magazines, books, newspapers, web sites, etc. For that reason, not everything he recommends will work for landscape photographers or find application in a fine art workflow.  Nonetheless, thinking about what he says and the explanations for why it works is a valuable learning experience.  In that sense, the book is similar to his previous book, Professional Photoshop, which teaches about color correction “by the numbers” with a strong emphasis on the CMYK color space.   Like many other readers, I learned a lot about color correction from the earlier book but found that I ultimately disregarded much of what he said about CMYK because I was working mainly in RGB.  With this book, I expect it will be much the same.  Some parts of it I am finding to be quite useful and I intend to retain them in my repertoire; other parts I expect I will soon forget.

Margulis tends to be a little controversial.  For example, in the past he has made fairly blanket statements that there is little practical benefit to working with 16-bit files and has drawn a fair amount of flack as a result.  No doubt some parts of this latest book will attract some of the same attention—but they will make you think.

A word of warning—this is NOT a book for Photoshop novices.  The first six chapters cover the basic LAB techniques that Margulis advocates. Each of those chapters is in two parts, a straightforward description and then a more advanced explanation.  Even the simpler parts of those chapters require at least an intermediate knowledge of Photoshop (and photography) and the second half of each chapter assumes a greater level of familiarity.  After the first six chapters, things get even more advanced.

The book is generously illustrated with lots of examples, showing before and after images and screenshots showing the curves that are necessary for a particular move.  All the images in the book are also included on the accompanying CD so you can try everything for yourself.
Logged

John DeMott
Paul Williamson
Jr. Member
**
Offline Offline

Posts: 68


WWW
« Reply #1 on: November 11, 2005, 01:53:15 PM »
ReplyReply

Excellent description of the book, thanks. I would add a few comments:

Potential readers might wonder how a whole book could be written on just the LAB color space. Indeed, this concern is brought up by David Biedny in the Forward of the book. The answer seems to be lots of examples with big illustrations, plus a fair amount of repetition. The actual ideas in the book could be stated concisely in just a few pages; the rest of the book is tutorial and exploration of the consequences. Whether that's a problem or not will depend on your reading preferences.

The illustrations carry a lot of the weight. They are allowed to spill out to the edge of the page, so they're bigger than they otherwise could be, which is a nice touch. Many of the differences shown in the illustrations are pretty subtle, and the color printing in the book is good enough to see what he's talking about, but just barely.

The technical explanations for many of Margulis's arguments are basically mathematical, but he shies away from the actual details and doesn't go into the math. In my view that weakens the explanation, and partially obscures the fact that he's mostly talking about Photoshop's implementation and not fundamental differences between color spaces.

The author spends quite a few words on warning the reader about how difficult the material is to understand. Perhaps he's trying to leave the impression that his explanations are better than they really are. If the user comes away confused, he blames the supposed difficulty of the subject; if he comes away enlightened, he credits the author's brilliance. It's a cheap trick and a waste of space, and Margulis does it consistently throughout. Great technical writing makes the difficult seem easy, without calling attention to itself. The writing here doesn't quite reach that level, but it's actually quite good. It would stand on its own without the apology.
Logged
jdemott
Sr. Member
****
Offline Offline

Posts: 434


« Reply #2 on: November 22, 2005, 12:52:13 PM »
ReplyReply

As I mentioned in the original post, this is a book that makes you think.  I've been continuing to work with the ideas and techniques in the book and  I decided to share a couple thoughts I've had in case anyone is checking back in the new posts.  While there are quite number of specific examples and techniques in the book, I come away with at least three "big ideas" that give me added insights in how to look at an image in Photoshop and how to approach post-processing:

1. Color separation.  As photographers, we are familiar with evaluating and correcting the contrast in photos, something many of us first learned in the black and white darkroom.  A low contrast photo is often described as flat and we routinely use a variety of techniques, such as Levels or Curves, to enhance contrast.  Margulis points out that many images suffer from lack of color separation, such as desert scenes where the camera sees washed out colors even though the eye saw vivid hues of red and yellow.  Likewise, scenes with a lot of vegetation appear to the camera as nearly uniform green even though the photographer saw a variety of subtle shades and differences.  The solution for lack of color separation, as with lack of tonal contrast, is to steepen the curve.  Because LAB color has channels of pure color, it is an easy matter to steepen the curve of the color channels and achieve greater color separation, with colors that are not only more saturated but also more vivid because the differences in their hues have been accentuated.

2.  Understanding how to color correct in LAB.  Margulis's background emphasizes doing commercial color correction work and the book devotes quite a bit of space to this topic.  Since I have begun shooting digitally, rather than scanning film, I find that color correction per se is much  less of an issue.  Nonetheless, understanding this topic is key to appreciating some of the power of the LAB colorspace.  LAB has only two color channels, A and B, in contrast to the three color channels found in the familiar RGB colorspace.   Both the A and B channels have their mid-points at zero, which is an absence of color or a neutral color depending on how you look at it.  The A channel tends toward magenta with positive values and toward green with negative values.  The B channel is yellow and blue.  So in a sense, LAB has four primary colors arrayed around its color wheel, with pure neutral in the center.  Any two values specify both a hue and how "saturated" it is (how far it is from the neutral center).

3.  Approach color and contrast separately.  At first, the idea of dealing with color and contrast separately in LAB would seem to be fairly trivial since LAB has two channels that contain only color information and one that has only contrast.  Some of the techniques will be familiar to many users of RGB who routinely use a luminosity blend mode for sharpening and local contrast enhancement.  Margulis takes it quite a bit farther, and this is where some of the more powerful (and more advanced) techniques come into play.  Essentially these techniques take advantage of all the different ways visual information is presented in Photoshop, extracting the most useful parts as needed to construct the color and contrast for the final image.  For example, most photographers are familiar with the idea of using a red filter to enhance the appearance of the sky in a black and white photo.  A similar result can be obtained in Photoshop by taking the red channel from the RGB colorspace and applying it to the L channel in LAB.  More surprisingly, one can sometimes extract very useful contrast information from A and B (pure color) channels using the Overlay blend mode.  The idea is to work with color channels in either LAB and/or RGB, as well as using the L channel in order to build the best contrast channel, ignoring what happens to the color.  Then adjust the color using either RGB or LAB as appropriate without  worrying about contrast.  LAB lets you put it all together.  The amount that can be done merely by substituting and blending channels is quite remarkable.
Logged

John DeMott
dandill
Jr. Member
**
Offline Offline

Posts: 84


WWW
« Reply #3 on: November 22, 2005, 01:39:02 PM »
ReplyReply

There is an active reading group on the book at http://www.dgrin.com/showthread.php?t=18203
 and detailed descriptions, with examples, of each chapter. I have found the level of expertise illustrated by some of the participants quite high, and I have found the examples very helpful.
Logged

Dan Dill
jdemott
Sr. Member
****
Offline Offline

Posts: 434


« Reply #4 on: November 22, 2005, 03:19:41 PM »
ReplyReply

Quote
There is an active reading group on the book at http://www.dgrin.com/showthread.php?t=18203
 and detailed descriptions, with examples, of each chapter. I have found the level of expertise illustrated by some of the participants quite high, and I have found the examples very helpful.
[a href=\"index.php?act=findpost&pid=51938\"][{POST_SNAPBACK}][/a]

Thanks for the link.  I just spent a little while reading the comments on Chapter One and I'm quite impressed by the level of the discussion, not to mention the amount of effort some of the participants have put into writing and illustrating their comments.  I plan to spend some more time browsing that thread.
Logged

John DeMott
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6945


WWW
« Reply #5 on: November 25, 2005, 09:11:40 AM »
ReplyReply

I'm on Chapter 9 of the book, and I basically agree with John DeMott's review. I think the book is valuable. The content is very rich with all kinds of useful insights - not only on LAB mode, but other things as well.

If you want to get much more deeply into colour control and colour management, it is worthwhile to subscribe (free) to Dan's colortheory discussion forum at Sterling Ledet. It is a Yahoo group. The discussion there ranges from the easy to the very deep. All the controversies are well-exposed.

When reading Dan's material, it is important to separate the controversial issues (e.g. his views on bit depth and color working spaces) which occupy a very small fraction of his material, from the solid technical information that we can actually use to improve our printing. The latter is much more abundant  and important than the former.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad