Adobe Photoshop Training with Dan Margulis

Hi all. I'm a magazine photographer using a Nikon D1X and Photoshop6.1. I deliver my images in Adobe RGB, because the prepress/ADrequires it. I've been at this a year now. I work very hard on myimages and have the luxury to "interpret" my images as I see fit. This could include slight off color (slight overwarming) or creatinghigh key images etc. The prepress house seems to do a very good jobof conversion to CMYK, retaining a surprisingly consistant and snappyresult.

After reading Dan's book I tried to use targeted curves to increaselocal contrast, but I would always get a flat area of the image whichI guess is a unavoidable by-product.

But I feel, by selecting areas of common brightness, then usingcurves or levels to increase local contrast, I can *seem* to getaround that flat spot in the image. I know this is sort of a no-no,but I'm not sure why. I feel the fuzzy logic of selections make theflat spots easier to hide.

For example, I make a threshold mast to collect the shadows, then feather and do levels to add lots of contrast to the shadows. They I can do the same thing with highlights and maybe midtones. Or I will reverse the shadow selection and work that, Or if it's a product that's one color, I can do color range to get a selection to work on.

A fair bit of further manipulation like saturation, and working on the selections to fine tune needs to be done, but the results really pop on screen and press. I also commonly shoot several exposures of still subjects on a tripod so I can layer them to use parts of different exposures within the same image.

Have others had trouble making targeted curves work? Is there any legitimacy to my technique? Any comments? … Aaron

________________________________________________________________________

Aaron,

Making the adjustments more efficient is one thing, otherwise I don't know why you'd think a technique that works for you is a "no-no". Have you clicked on the curve to "lock" areas and effecting the desired areas.

Lee

________________________________________________________________________

Thanks Lee, Yes, I've tried adding several points on the curve.

"A no-no" I get a little nervous that I might tweak an image in some way that may give me a big surprise on press. But it's been OK so far. I make sure I have adequate shadow and highlight detail and that cmyk gamut warnings are not going off everywhere. The magazines I work for don't send the proofs back unless it's a cover. Several people beyond the art department look at and approve color. ... Aaron

________________________________________________________________________

That's how the "basis" of curves works. You will gain contrast in one area and lose it in another. You can select areas, as well as to lock in areas(on the curve). But remenber "the book"(chapters 2 & 3) is based on global corrections(faster….and looks natural,than using selections). When you start selecting, it takes longer and might not look as natural. Now of course, after you make your global corrections, you can use selections for color and contrast moves, using selections based on the existing channels(remember every file has 10 channels)is a faster way than using "hard" selections with, let's say, the lasso tool.

Their is an article on this at www.electronic-publishing.com. The name of the article is "The first refuge of the fearful". Author: Dan Margulis. Look at these articles as well from the same author. "The mask behind the disguise", "The channel-choice mutiny", "Lighten,Darken,Luminosity,Color"

I find the above methods faster(and better)than "hard" selections. I use P.S., like my Vett'ster......Drive fast. See... As fast I get into a situation, is as fast, as I want to get out of one.

Being a shooter, myself, I find for my photography,using "false profiles", "channel blending in luminosity mode"(dumping the green channel into the composite channel),overlay blending(in L*A*B,to beef up colors,if need be. This method of blending can be used to downplay colors as well. Just invert the blending)is faster(for me,anyway)than curving,at times. Not to say, no need for selections. Which I hate. Making actions to fit the above situation,helps a lot too.

I know their are people (scenic shooters) that do this as well. But their is a time factor involved there,if you scan negatives or use a digital camera for that matter. Also.. Not good on windy days(outdoor scenics). Unless, you want that kind of effect. Too time consuming, when you needed to get a bunch of images out, a week ago. You can do it that way, though. Another way is to scan a single negative twice. Once for the shadows and once for the highlights. For digital, use your blend-if sliders.

No..... Using the eyedropper, click the area(s),where you want to make your correction, while viewing the range where it falls on the curve(s), target that area on your curves.

What I like about the above methods, is that their quick corrections. Example: The "false profile". I showed my comrads(photographers)this method. And, they could not believe it. The method they were using took longer, than the way I showed them. Some can't grasp it, though. One was a graphic artist,as well, and told me: " I never saw this done before". This also is in reference to swash-buckling lab moves,non-kosher "channel blending" for cmyk output as well. Like dumping the red channel into the cyan. Or the "L" of lab....... It's all in "the book". Take the advice from that guy(Karl Malden) on that retro commerical for American Express. "Don't leave home without it".

John Opitz

________________________________________________________________________

Thanks for all the good advice as well as articles to read. I guess I fall into the group that "can't grasp it" I'm fairly new at serious Photoshop work, but a light goes on once in a while. I'm sort of getting what I want, but it is likely the hard way. ... Aaron

________________________________________________________________________

Aaron writes,

It's generally a no-no because the curve is usually designed to emulate how a human would respond to the scene as opposed to how the camera did, and human observers don't make selections. When our visual system adjusts to give more emphasis to some facet of the image other areas become less well defined. Selecting out those areas so that they won't be damaged tends to look unnatural.

Often a selection can be helpful for the opposite purpose--where you're actually trying to tone down some unimportant area, by desaturating or reducing detail in the background typically.

Avoiding selections that are aimed at preserving detail isn't a hard-and-fast rule--you just have to use soft-edged masks and be very careful not to overdo it. For example, I was just in the Sequoia forests of California on a sunny day after a heavy snowfall. The scene consisted of harshly lit snow and parts of trees, coupled with the relatively heavy shadows cast by the big trees. The human visual system adjusts to these challenging conditions easily and has no trouble perceiving color and detail in the shadows. Any camera exposure is either going to have overly dark shadows or blown-out snow. If I had to correct such an image with curves they'd have to be so radical that it might not be possible to control them. So, I would endorse lightening the image through an inverted luminosity mask to try to preserve the lighter areas.

But, as a general matter, such selections are best avoided.

Dan Margulis

________________________________________________________________________

But photographers do, they have always been creating masterpieces in their darkrooms making selections and local adjustments. That is the way Ansel Adams, for instance, made his fine art printings...

Maybe the fact that one knows what is doing is more important than methods...

--

Claudio Corvino

Florence, Italy

________________________________________________________________________

Have faith. Not only reading helps. But keep working at it, as well. I was one of those photographers who was against altering the image with P.S. While their are still photographers who don't believe in using P.S....Even today. I look at P.S. as a tool. Not a band-aid. Even though, those lighting plug-ins look very tempting. The field has changed.

We're a new generation of photographers and imaging specialists. We must seek the experience and the knowledge of those who had and have the experience of the past as well as the present to help us with incorporating the technology they have learned and experienced into today's technology with us......For those people,... who have the experienced of the tried and true, will be our resource to this frontier.......Let these people be our guiding light,...to this new technology. For they will be the gate-keeper that holds the key for this knowledge.

p.s. I better stop now. Susan is rolling her eyes and thinks I'm pompous.

John Opitz

________________________________________________________________________

John, you must listen to Susan more often. She seems very wise...

;-)

J (AKA - "The GateKeeper")

________________________________________________________________________

Claudio writes,

Ansel Adams used the tools he had. He couldn't apply curves in Photoshop, so he implemented the Zone System of shooting, which is a first cousin. Even today, if Photoshop had no curves command, we'd be forced to do a lot of localized enhancing, just as Adams had to in his time. It is the great misfortune of the industry that Adams didn't live in the age of digital enhancement, because he had a driving curiosity about how best to use available tools and the analytical skills to decide which new techniques might work best.

There's no maybe about it. If one person uses the highest quality professional photographic equipment, the latest version of Photoshop and the fastest available computer, and someone else a disposable camera, Photoshop 3, and a Mac FX with a black and white monitor, if one of these persons is significantly more skillful than the other he will get more professional-looking results. But, this doesn't prove that we should all color-correct on black and white monitors.

Dan Margulis

________________________________________________________________________

When it comes to color correcting, to each end, there are dozens, possibly hundreds of procedures, tools, and methods to get there. No one is right. No one is wrong, although some may take you a longer path to get there. Dan's book is an excellent tool to point you in the right direction, and help avoid some of the "less correct" methods.

I do however feel that emphasis on the method of corrections overshadows the reasons for correction. The basic idea of image capture, or scanning is not to have to correct anything. If your original scans are done right, or digital captures (from a studio environment), are as close as they can be, very little correcting or sometimes no correction would be needed at all. I'm very "old school" in this, in that if an original image is not 95% or better, of the way there to start with, it needs to be done over. The less correction the better, in quality and in time. You may be willing to invest 30 minutes, or an hour, or even more, "playing" with an image that you want to make as perfect as you feel you can get it, but what if you had 10 images to do? Or 100? Or 3000 images to go into a catalog. We regularly do catalogs with hundreds of images, and do a couple per year involving 2000 to 4000 different product shots, of a vast array of subject matter. The client demands, and expects the color of these images to match the products they are selling. How many do we have to color correct? Other than conversion to CMYK, and applying USM, (all done with hot keys), about 50% of them are not touched at all, 40% are tweaked (a minutes time or less per image), and the remaining 10% of the "problem children" may need anywhere from 2 to 5 minutes work on them each, or possibly a reshoot. Imagine what we would have to charge, or attempt to charge the client, to spend 10 minutes per image. They either wouldn't pay it, or we would not have gotten the job to start with.

My point to all this, is that I see the newer generation sliding into a "I can fix or make anything in Photoshop mode." Well... yes... they can (almost). But if they could just learn how to do it right to start with, they could save a lot of their time from fixing images in Photoshop, and use that time for more creative purposes, and for the jobs that really do require extensive (and chargeable) Photoshop time.

There is nothing wrong with taking an image and playing with it to see how good you can make it. If it were your hobby, it might be lots of fun. If you are just learning, you need to take all the time you need, and practice, practice, practice. Its the only way you will learn. But, if you consider yourself, or want to consider yourself a color specialist, or a scanner specialist, or a Mac Photoshop artist, you have to be able to turn out work. Quality work, and a lot of it. That means being able to pull a good percentage of your images into Photoshop that don't need any, or very little color correction at all. Its all starts with the image capture.

My .02 cents... plus maybe a little more.

John Rawlins

________________________________________________________________________

Dear Mr Margulis,

I totally agree with you. The issues I wanted to underline with my previous posting are the following:

1) I appreciate your knowledge and the effort you make to share it. I think that this sharing knowledge is one of the most noble actions a man can do.

2) Making local adjustments to a picture not necessary leads to unnatural results.

3) To me, it seems that going curves implies to decide that, in a picture, the most important subject falls, for instance, between 30 and 50%; If I apply local adjustments I can decide that in a portrait, (black, Italian) hairs are so important than hands...

4) Of course, it depends on my workflow: if this is like John Rawlins' in his posting about this same topic, only curves maybe are better; for fine art prints, the two methods can live together.

--

Claudio Corvino

________________________________________________________________________

Dan writes:

Adams' first control of his image involved the number of tones in his shot. He could control this by exposing for the lowest important value, and developing for the highest important value, using an appropriate film developer, typically water bath, D-23, or HC-110. When the image was properly developed, he could use paper grades (gamma control), paper development, and dodging & burning.

Most digital photographers don't shoot an appropriate number of tones for their target. Most shoot too many tones, then struggle with curve corrections that leave overly flat sections of the image. 16 bits and cooled sensor has been the "answer" but it really isn't. Shooting an approriate density range is the answer.

Richard Chang

________________________________________________________________________

Apples and oranges, is it not?

Ansel Adams was a master creative artist who often spent hours on a single exposure and additional hours in the darkroom to achieve an impressionistic work of art from a single negative. This is obvious from some of his notes and sketches for printing some of his known photographs, and for his known penchant for waiting for just the right lighting or weather conditions. Often, the right conditions never occurred and the "right" negative was never created to record what he "saw" in that photograph...so he tweaked it to show his audience what he wanted them to see.

Professional catalog photography calls for "photo mills" putting out large quantities of "correct" images, all taken under carefully controlled and highly repetitive lighting and exposure conditions. Color matching is often critical, obviously.

It would be financial suicide to treat each catalog photograph as an

artistic individual, just as it would have been for Ansel Adams to treat

each negative as a minimal exercise in B&W tonal correction of a technically

"near perfect" shot.

Mike Davis

________________________________________________________________________

The way I approach photography and retouching is as follows: I have many tools available to me, and I have an image to create--and what is the best route to create it. Sometimes this happens all in camera, but, most of the time, what has to happen is a combination of what happens in the camera, and what can happen in photoshop. It is important to get the best quality capture, but its also important for me not to forget the incredible possibilities after the capture. I guess I like to focus on the fact that I am creating an idealized image--which is photographic in part--but is not limited to the a specific set of tools.

________________________________________________________________________

From: Dan Margulis

Just a nit, but actually, the Zone System is a combination of exposure and chemistry techniques that optimize a film's dynamic range for a given scene contrast.

Most people forget -- or never realize -- that at least half the Zone System involves fine-tuning the development process. The "shooting" part is almost trivial, compared to the photochemistry part.

--

: Jan Steinman -- nature Transography(TM): <http: //www.Bytesmiths.com>

: Bytesmiths -- artists' services: <http: //www.Bytesmiths.com/Services>

: ________________________________________________________________________

John Rawlins writes,

These are all entirely different situations. Some types of images are worth the investment, and others, such as the ones you mention, aren't. Plus, in catalog work the originals are typically shot under controlled lighting with objects that don't move very much if at all, with backgrounds that are of limited importance, and with simple scenes which don't emphasize the way in which the human visual system varies from that of a camera. Ansel Adams had to work with lighting that wasn't under his control, ill-defined objects that were often in motion, subtle backgrounds of great significance, and complicated scenes in which simultaneous contrast plays a great part in viewer perception. Is it any wonder his originals needed more intervention afterwards?

If you're interested in getting the best quality out of an image, some correction is inevitable. No method of capture gets the best highlight and shadow, neutralizes all casts, or applies optimal unsharp masking. The question is how much do you have time for. I've asked that question of several classes and several clients. That is (I asked) suppose that I told you I could get a significantly better result than just going with an automated process, but it would cost one minute per picture. Suppose it's five minutes? Fifteen? An hour? How much time would you be willing to invest if it really made the picture look better?

People's responses were all over the place. 15-30 minutes per image was a common response but quite a few said they could only spare one minute. At the other extreme, one client (who produced calendars) said that, if for the sake of argument that was what it took to ensure quality, they would spend 40 hours on an image.

Almost any image, regardless of how good the acquisition process, will benefit somewhat from 2-5 minutes of adjustments. But, sometimes you don't have that time. Very few images benefit from more than 15-30 minutes of correction. Ansel Adams, he was willing to invest the 15-30 minutes.

Dan Margulis

________________________________________________________________________

Jan,

Perhaps a "nit" of my own, but if the "shooting" you refer to is "pushing the button", then yes...it is trivial. However, scene contrast evaluation and appropriate exposure for the low value luminances (on neg film) is not trivial. Without these steps, all the development and printing techniques in the world won't save the shot.

If more photographers understood these concepts, we wouldn't see images that lack luminous shadow detail.

Darren Bernaerdt