DAN MARGULIS APPLIED COLOR THEORY

PPW: The Luminosity Step

From: Dan Margulis

Date: August 19, 2013 8:51:42 AM EDT

Subject: [colortheory] The Luminosity Step

A couple of months ago a reader who isn’t AFAIK a member of this list sent me an offline message expressing frustration with the problems he was having with the initial curves step of the PPW. I answered, but I also forwarded the message to the list and several of you made useful suggestions on how he could make his life easier.

Now comes a similar message with respect to the luminosity step. Many of us think that that’s a relatively easy step, but I daresay there are others who are in the position of this reader. So before I answer him privately, I wanted to ask whether the list has any advice for him.

Dan Margulis

Begin forwarded message:

From:

Date: August 15, 2013 2:05:07 PM EDT

To: Margulis Dan

Subject: Long time student

Dan,

We have never met, but I have read, and re-read and re-read nearly all of your books and work. I have always had an affinity to getting my images correct and of the highest standard. That being said, I have never really been able to achieve this on a consistent basis. Over the last two years I have been working with your PPW workflow, first through Kelby and now through your book. To me it seems straightforward and a great process. I have read the chapters many times, but when I actually get to the point of applying it to my own images, I run off the rails.

What I would ultimately love is to attend your class, but unfortunately I live in [DELETED], so I have to learn by book and application of trial and error. The issue that I run into is when I start working on my own images, I can’t tell if they are correct. I know the colors are because I am measuring them. Where I get consistently stuck is within the step of luminosity. To me there seems to be so many options, that I get overwhelmed. I have a full composition book of my own notes from your books and training videos. There I understand it, in my own space it’s as if I forget. After that point it seems the entire process runs off the rails for me.

Color correction finally made sense to me with the use of LAB. It is such a more desirable color space to diagnose images. My question to you is do you have any advice or suggestion for me when it comes to the step of luminosity. I hate to bother you with such an inane issue, but I am meticulous and if something isn’t right, I can’t move forward which prevents me from succeeding with this.

I want to let you know that I greatly appreciate your contribution to color correction and I view your opinions as gospel. In my mind, if I can achieve consistent results using the PPW technique, then I have achieved the pinnacle.

Take Care,

___________________________________________________________________________

From: James Gray

Date: August 19, 2013 1:17:01 PM EDT

Subject: Re: [colortheory] The Luminosity Step

It is not clear what specifically is going wrong for Long Time Student.

His metaphor of running off the rails suggests he is getting ugly results

rather than merely failing to give the images the snap he wants them to

have. I will give a few comments that might be helpful or might not

address his specific problems. I should point out that as a beta reader I

really misunderstood a component of the luminosity step until I viewed the

videos provided at the web site. In most respects the luminosity step is

two steps. First is the possibility of channel blending and the second is

curving of the green and red channels to improve contrast and to

preliminarily set the shadow and highlight. Channel blending is difficult

and requires a lot of practice to see what can be accomplished. I would

suggest being very conservative with channel blending. It might be the

trap running him off the rails. As has been said many times, even though

channel blending is done on a luminosity layer, it can mess up bright

colors. If instead of being conservative, one wants to be brave, try

channel blending using a mask. Making a copy of the image and converting

it to LAB mode can give you access to the A and B channels to use as masks.

I am often disappointed in what curving does for my images because most of

them already have a dark enough shadow and light enough highlight yet do

not have enough snap or pop. However, that sounds different from running

off the rails. Maybe this will be more helpful to somebody trying to learn

this step. Initially I found it very difficult to work on the luminosity

layer in normal mode. So one solution is to actually do the work with the

layer in luminosity mode. It will not be as effective as working with the

layer in normal mode, but might prevent bad results. You can actually see

your results if you work on a curves adjustment layer set to luminosity

mode. Maybe doing that for a while will give him an intuition for what to

do when he changes to working in normal mode.

James Gray

___________________________________________________________________________

From: Hector_APR

Date: August 19, 2013 3:44:41 PM EDT

Subject: Re: [colortheory] The Luminosity Step

To put it simply...it just takes ‘Practice’.

Even knowing what steps to take, you still need to practice to perform

those steps comfortably.

It’s like watching a magic trick being performed.

http://www.youtube.com/watch?feature=player_embedded&v=dmhmKCQIBDM

Then watching the magic trick tutorial.

http://www.youtube.com/watch?v=NZqrdww4yls

As you can see, performing the magic trick even if you know the steps

will still take practice.

Once you have practice it, you should be able to perform it the way Dan

Margulis does his magic.

But it does take practice. Practice makes perfect.

Hector Davila

___________________________________________________________________________

From: Campo Weijerman

Date: August 19, 2013 7:26:40 PM EDT

Subject: Re: [colortheory] The Luminosity Step

Long time student confesses “being overwhelmed” by the many options

available in the luminosity step. If that includes not just curving and

channel blending, but also the false profile and overlay techniques. Use a

little of every possible technique, then add some hiraloam sharpening to

the mix and indeed chances are that the resulting image looks worse rather

than better than the original.

James’ suggestion to be conservative about channel blending is excellent

advice. Personally, I found it very enlightening to discover what

channel-by-channel curving on a luminosity layer could do to an image.

Make up your mind about what’s the most important object in the image.

Find its range in the R and G channels. Apply luminosity curves to improve

its contrast. Often times this works even if the original already has full

range. Of course there are images where this doesn’t work. On the other

hand, it is also possible to allocate too much contrast to a subject and

this is another way to ruin an image.

Hope this helps, Campo

Campo Weijerman

___________________________________________________________________________

From: John Bongiovanni

Date: August 19, 2013 7:55:48 PM EDT

Subject: [colortheory] Re: The Luminosity Step

--- Campo Weijerman wrote:

Make up your mind about what’s the most important object in the image.

This is what hit me reading Dan’s recent book and the associated videos. It wasn’t so much the technique, but looking at the image (and at the individual channels) and seeing what needs to be done. This is quite difficult, as it involves both an aesthetic sense and an understanding of how the various channels come together. I’m trying to work on it in my own images, and it’s a challenge.

I would expand on Campo’s statement a bit, and say make up your mind about what details you want in the different parts of the image.

John Bongiovanni

___________________________________________________________________________

From: “twinkle_max”

Date: August 26, 2013 5:31:28 AM EDT

Subject: [colortheory] Re: The Luminosity Step

In this respect, I find very useful Channels Power Tool by Giuliana Abbiati (You should know her...)

http://www.bigano.com/index.php/en/channels-tool/273-channels-tool.html

Massimo Calvani

___________________________________________________________________________

From: Dan Margulis

Date: August 25, 2013 9:32:32 AM EDT

Subject: Re: [colortheory] The Luminosity Step

On Aug 19, 2013, at 8:51 AM, I wrote:

I put together what the list said and set it off to the guy who was asking advice. FWIW, I append my own note to him.

Dan Margulis

***************************

If I understand you correctly the problem is that you are looking for the single “right” answer when no such thing exists.

The luminosity step is similar to converting an image to B/W. There are at least half a dozen valid ways to optimize contrast. If you don’t like the options I’m showing, you could try adding a Black & White adjustment layer, setting it to Luminosity mode, and adjusting settings to make certain colors lighter and others darker. It’s the same general idea, and in some cases it might be better.

Even if you stick to the techniques shown in the book, however, so many options are possible that if you do the work twice, one day apart, it’s almost inconceivable that you would get the same result, Presumably both would be better than doing nothing.

The other issue that plagues some people is expecting a huge improvement when none is available. These luminosity techniques work best in something like the Grand Canyon image where there is distinct, yet subtle, color. In images that are full of bright colors or that are basically gray, you don’t have much to gain and if you try to force something spectacular to happen you may live to regret it.

If you check out my Chapter 2 video where I am putting together one unified version out of eight versions that I did at separate times, you’ll see that I’m not particularly consistent myself. That suggests the value of doing more than one quick version.

___________________________________________________________________________

From: Dan Margulis

Date: August 27, 2013 10:20:00 AM EDT

Subject: Re: [colortheory] The Luminosity Step

Here is the response I received offline from the OP.

Dan Margulis

**********************************

From:

Date: August 26, 2013 12:43:53 AM EDT

To: Dan Margulis

Subject: Re: Long time student

Dan,

I greatly appreciate the response and I fully understand what you mean. Throughout your last book, I have mostly read, re-read, and taken notes without actually working on the images. What I was trying to gain was a full understanding. With that I took copious notes which actually started with the video series that you did for Kelby which I believe pre-dates the panel.

I have a special set of images that I have taken in [a tropical setting] on my parents 50 wedding anniversary that is serving as my pilot group. What I am finding with this step is a constant desire to blend nearly every image. After working on one 4 different times, I completely understand why there is no “one” way to do this. After going back through the material again, you make every effort to make that point. Finally the bulb went off over my head.

You are also correct in people expecting big change at the luminosity step, but I have also relented on that after seeing the transformation after the process is complete. I think that is another moment of epiphany, when you just relax and trust the process and judge when finished.

To me, step 2 has always been the hardest and still trips me up. I think that I am trying to blend too much and curve both red and green on every image. I end up with some major shifts in saturation with some objects like red shirts and blue/green umbrellas. When that happens, I know that I am bending something too much and start again. It’s funny, some of the most advanced moves and concepts in the book I understand without a second thought, but it is this one that I stub my toe on. I know that it will all come down to perseverance and failure. You have to fail a few times before you succeed. I also like using my own images because it really helps me to focus on the process and clues within the image.

Again, thank you so much for the response.

Take Care