The Next “Photoshop Masking & Compositing” Course Goes Live @

Ask anyone at Adobe what distinguishes Photoshop from every other image-editing program, app, or digital blip on the planet, and they’ll tell you “masking and compositing.” Apparently you agree, because my video course Photoshop Masking & Compositing: Fundamentals is tearing up the planet over at It’s N-to-the-1-to-the-L-D-C, as the dope kids say. As if I’d know.

Naturally, I’m gratified. (Thanks very much, btw!) Plus, it emboldens me to report: Today I and my beloved video publisher release another installment in the series, Photoshop Masking & Compositing: Advanced Blending. The image below might make it look exacting and academic. Which it is. But it’s also expansive and practical. Because it tells the ultimate post-processing story: How to paint without permanence, create without consequences, and, in the end, mask without masking. In short, how to assemble photorealistic artwork through the pure power of artistic thought. It really is that good.

Photoshop Masking & Compositing: Advanced Blending

This is a short course, just 4 hours, and yet it manages to comprise 9 chapters. Here they are:

Chapter 1, The Power of Blending.

Here’s where I introduce you to everything, including how blending works, where you find blend modes, the difference between Opacity and Fill Opacity, the “Fill Opacity Eight,” how to blend with adjustment layers, and the blending shortcuts.

Chapter 1: The Power of Blending

Chapter 2, The Revelation of Blending Math.

Some people like math, lots of people don’t. If you’re a math fan, this chapter explains the formulas behind all the blend modes except one (Soft Light, for which there is no reliable documentation). If you don’t like math, skip this chapter! I promise, you won’t miss a thing.

Chapter 2: The Revelation of Blending Math

Chapter 3, The Normal Modes.

The “normal” modes include Normal, Dissolve, Behind, and Clear. This chapter explains how they work and shows what you can do with them. For example, here’s an effect created with Dissolve.

Chapter 3: The Normal Modes

Chapter 4, The Darken Modes.

Every blend mode fancier knows that the best of them all is Multiply, which uses the active layer to darken the layers below. Watch this chapter and learn why Multiply and the other darken modes are even more powerful and day-to-day useful than you thought.

Chapter 4: The Darken Modes

Chapter 5, The Lighten Modes.

The next best mode is Screen, which uses the active layer to lighten those below. But that’s just the beginning, as this moody businessman (masked with the help of the ultimate lighten mode, Linear Dodge) will attest.

Chapter 5: The Lighten Modes

Chapter 6, The Contrast Modes.

Overlay, Soft Light, Hard Light, and the other contrast modes are great for adding clarity, diffusing focus, and wrapping textures onto surfaces, as in the case below. Even Hard Mix is your friend.

Chapter 6: The Contrast Modes

Chapter 7, The Inversion & Cancelation Modes.

The next group of blend modes includes two inversion modes (Difference and Exclusion) and two cancelation modes (Subtract and Divide). Difference is especially useful, permitting you (among other things) to create text that inverts anything below it, as shown here.

Chapter 7: The Inversion & Cancelation Modes

Chapter 8, The Component (HSL) Modes.

Ancient Assyrian carvings aren’t normally this rock-musical resplendent. Unless you colorize them with a host of festive layers in Photoshop. Which is something you can do using the final group of blend modes, known as the component (HSL) modes.

Chapter 8: The Component (HSL) Modes

Chapter 9, The Luminance Sliders.

It’s all advanced stuff, don’t you worry. But this final group of options actually goes by the name Advanced Blending. The luminance sliders (This Layer and Underlying Layer) let you drop out and force through the visibility of pixels based on their luminance levels. And the Blend If option lets you control visibility based on the composition of a single channel. Which is all I did to “mask” this lightning.

Chapter 9: The Luminance Sliders

In just 4 hours (53 movies), you’ll learn how to blend in Photoshop with absolute confidence. Which translates to more efficient compositions, more effective masks, and better looking artwork. Check it out now and let me know what you think.

Next entry:Turning a Photo into an Ink Drawing or a Pencil Sketch (in Photoshop)

Previous entry:Deke’s Techniques 047: Tracing an Image with Path Outlines


  • Deke, this is great

    Deke, this is great stuff!

    And THANK YOU FOR THE FORMULAS!!! I literally want to kiss you for including them in the course!!! smile)

    I mean, one formula speaks better than paragraphs of babbling and mumbling that one so often comes across in books and video courses that cover blend modes. Blend modes are INHERENTLY mathematical - so they are best explained via (very simple, almost elementary school level!) math. There really is no way around this.

    “A+B-1” (for Linear Burn) is SO MUCH MORE CONCISE AND “GRASPABLE” than “Looks at the color information in each channel and darkens the base color to reflect the blend color by decreasing the brightness” (to quote Adobe Photoshop CS5 help).

    Now, there is a Russian guy by the name Andrei Zhuravlev who teaches Photoshop and related topics at a Moscow photography school. He has a blog where he has a series of posts on blend modes (he also has another series of posts on sharpening - which received very positive feedback from Dan Margulis). The blog is here ( The series of posts on blend modes is here (Режимы%20наложения). You could use Google Translate to translate from Russian into English.

    Anyway, what I’m getting at is this. Instead of (or in addition to) relying on documentation and second-hand information, which often contains errors, he analyzed the blend modes himself: he derived formulas for how they work based on overlaying two grayscale images (with known numerical values) and then analyzing the resulting numerical values. That’s very basic - but somewhat tedious - work.

    He discusses the “Soft Light” mode in this post ( Even without understanding any of the Russian text you can immediately see the formulas.

    In my opinion, he does not always give the best possible version of the formulas (for example, your “A+B-AB” for the “Screen” mode is way better than his “1-(1-A)(1-B)”), but it gives you something to work with.

  • PS. His posts on sharpening

    PS. His posts on sharpening can be found here:Повышение%20резкости

    They include some of Dan Margulis’ feedback.

  • I agree

    Investigating and factoring the formulas to their simplest expression (A+B-AB, for example) has helped me better understand the blend modes.

    Andrei’s Soft Light formula may be correct. It works for 50% gray on the active layer, which is more than I can say for the other ones I found. I believe there’s a minor error in his Overlay formula (C>0,5; I think he means S>0,5), but it looks as though he’s spent a good deal of time testing the luminance values, which gives me hope. Still, if true, Soft Light is easily the most challenging formula to wrap your brain around. For example, 1-A (his 1-C) is a straight invert. But what does it ultimately mean to take the square root of a luminance value? Clearly that brightens it; to a mitigated extent perhaps?

    Of course, the practical value of this kind of inspection is to find practical uses for the modes. For those who don’t dig math—-or find the formulas to be a pointless exercise in Poindextering—-creating efficient, successful, eye-popping compositions is the ultimate purpose of the course.

    Now if only I could find the math behind the Fill Opacity Eight. Unless I’m very much mistaken, I was the first one to disclose the damn things. The upshot begin that I can’t find anything that didn’t originate with me.

    If anyone finds evidence to the contrary, or any additional info, my feelings will not be hurt. My jobs are to 1) learn and 2) teach.

  • 1) I’m sure the Overlay

    1) I’m sure the Overlay thing is a typo, I’ve just commented on it in Andrei’s blog.

    2) I agree that the formulas for Soft Light aren’t intuitive. Perhaps factoring the differently would remedy this to some extent. I have to think about that… But I think his Curves diagrams are very useful - and come in handy here, where the formulas aren’t intuitive.

    First, in his first post ( he has a set of Curves diagrams whose application to an image produces the result equivalent to blending that image with itself using a given blend mode.

    More to the point, he has animated Curves diagrams illustrating each blend mode. These help better understand the formulas - and illustrate the latter in case they aren’t intuitively graspable. What these animates Curves diagrams show is this: they show the result of applying a uniformly gray image to an underlying image, for different intensities of gray (0, 32, 64, 128, etc).

    One can use these to compare the Overlay and the Soft Light modes. One immediately sees that the Overlay mode is equivalent to piecewise liinear curves, whereas the Soft Light mode is equivalent to similar-looking, but parabolic curves. That’s where the formulas come from - they describe that family of parabolas. (For C

    <0.5; for C>

    0.5 the parabolas are reflected around the y=x axis and, naturally, become a family of square roots.)

    From these diagrams one also immediatey sees that the effect of Soft Light is gentler than that of Overlay. Also, a little bit further down he compares the Curves diagrams for these two modes and shows that the difference between them is a little bit more subtle than that: the effects that these two modes produce are pretty similar for the highlights of an underlying image and more distinctly different for that image’s shadows (Soft Light affects the shadows more strongly than it does the highlights, while Overlay affects these to a similar degree).

    Anyway, what I am saying is that the formulas together with the Curves diagrams are more instructive than the formulas alone and that sometimes the latter help understand the former. smile

    3. BTW, you could write to him, maybe he knows about the difference that opacity and fill opacity produce for some of the blend modes… He speaks English pretty well, I guess, since he was able to sustain (and win smile ) a meaningful discussion with Mr. Margulis.

    But I’m surprised that Adobe guys didn’t tell you all about it! I mean, you of all people must have lots of friends and connections at Adobe - wouldn’t they respond to your request for information?..

  • By the way, he also writes

    By the way, he also writes that Soft Light is very similar to gamma correction: i.e. applying a uniform gray image to an underlying image in the Soft Light mode is similar (but not equivalent) to gamma correction:

    - applying a uniformly black image is exactly gamma correction with gamma = 2 (because we get R=S^2),

    - applying a uniformly white image is exactly gamma correction with gamma = 0.5 (because we get R=sqrt(S)),

    - applying intermediate shades of gray produces intermediate results (although these are different from gamma correction, which would produce functions of the family R=S^alpha).

  • Andrei says that the correct

    Andrei says that the correct formula for Soft Light has been sitting here ( for a while.

    By the way, I think he and I have understood how “fill opacity” works for these eight modes! Once he has the time to check my results (and if they check out, of course), I will write to you here.

  • Ok, I think I found a way to

    Ok, I think I found a way to look at Soft Light formulas.

    We have (in Andrei’s notation, where S is source and C is correction), when C < 0.5 :

    R = 2C*S+(1-2C)*S^2.

    We can view this as A BLEND BETWEEN R=S (the source image) AND R=S^2 (the source image multiplied by itself, or, for that matter, gamma-corrected with gamma=2) WITH OPACITY 2C. This is especially apparent if we introduce alpha = 2C (when 0 < C < 0.5, we have 0 < alpha < 1), then our formula becomes

    R = alpha*S+(1-alpha)*S^2.

    So we have a blend between the original image and its gamma-corrected version with gamma=2.

    The same holds for when C > 0.5. In this case, we introduce beta = 2-2C (when 0.5 < C < 1, we have 1 > beta > 0) and get

    R = beta*S+(1-beta)*sqrt(S).

    Again, we have A BLEND BETWEEN R=S AND R=SQRT(S) WITH OPACITY 2*(1-C), which, again, can be viewed as a blend between the original image and its gamma-corrected version, this time with gamma=0.5.

    I think this makes it easier to make sense of the formulas for Soft Light, doesn’t it?

  • The Lighten Modes

    lessons 7 and 8 under The Lighten Modes are quite confused and confusing. Nowhere is there a mention of a lighten mode being used. Do you intend to update - revise these lessons.

    Thanks, Joe (cippy)

  • PSW

    I normally take your classes at PhotoShop World but do not see you listed for the March convention. Will you be there?

    Thanks. You make a difference. A good one.

  • Don’t mean to confuse

    But the effect hinges largely on a Linear Dodge (Add) layer introduced about a minute into the first video that you mention above.

    Does that help?

  • I’ll be whooping it up elsewhere

    Happily, I was invited. And I always have a blast attending. Photoshop World is a great conference packed tight with a whole lot of great people.

    But the next event coincides with my 50th birthday. (I know, who’da thunk that’d ever happen?) And I have a few personal plans that, this one time, take precedent.

  • Good for you!

    As one who has passed that milestone,
    (It didn’t even hurt as it passed!!)

    Enjoy your Birthday!

  • Get right on that, Deke…

    Andrei sez jump…

    Tell Andrei to apply for a gig at Adobe where he can tell them everything he knows.

    I had no idea that math, and apparently Andrei, (not Photoshop) made my photos better.







  • Blown Away Again


    I am again blown away buy your tutorials. I have gotten to chapter #4 of this course and just finished the “Cleaning up scanned line art” & “Comping line art against a photo” movies.

    The results you got from just utter crap is pretty awesome.

    I had the same reaction (being blown away!) from a similar tutorial you did in your Photoshop CS5 One-on-One: Mastery course. In chapter #28 the “Dropping out a white background” movie.

    Deke, you are THE MAN!

  • I’m afraid you are missing

    I’m afraid you are missing the point of the discussion here, friend. wink

    In any case, nobody’s forcing you to read this, so I don’t see what the problem is.

  • Hey there again, Deke! Well,

    Hey there again, Deke!

    Well, I guess you got bored by my comments down below. smile)

    If you are still interested, Andrei and I compiled a few comments on your blending mode table. It’s in a MS Word file with a psd example, so not easily postable here. I would love to email that to you so that you could, hopefully, read it some day. smile I obviously don’t know your email, but my email is “alexander-kayumov at”, so if you drop me a line, I’ll be able to write you back.

    Also we have found a simple formula for how fill opacity works for the “fill opacity eight” modes. That’s in the file, too.

    Merry Christmas and happy New Year to you and your family!

  • math behind blending modes.

    accurate equations behind blending modes can be found. you don’t have to be mathematician
    any one who has sufficiently good knowledge of A LEVELS mathematics can do it. But i don’t
    think knowledge of mathematical farmula behind the blending mode would make someone
    very good in post processing.the colors, hues contrasts in any picture are very complicated
    if you run eye droper on an image even on an area where you feel there is a single color,
    the rgb values will differ so greatly that it would make it very difficult to apply your knowledge

    of math behind blending mode .

Share your feedback, work, homages, questions, wisecracks, advice, critiques, riffs, derision (within reason), frustrations, and love of all things graphical. Log in (or register) to lend your voice.