Easy-to-follow video tutorials help you learn software, creative, and business skills.Become a member
In this exercise I'm going to introduce you to the Calculations command, which allows you to merge two channels together to create a new alpha channel that will serve as the basis for your mask. I'm still working inside Hair in flight.tif, found inside the 26_masking folder. And the first trick is to figure out which two channels you want to merge together. And here is the way I reason it, you start with a base channel and that would be the channel that provides the best overall contrast, and then you look for a support channel that is as different as possible from that base channel.
So, for example, in our case, we already know which channel provides the best contrast. that's the red channel inside of this image. We discovered that in a previous exercise because we have these bright skin tones set against the darkish background and the hair is darker still. So now that we know that it's the base channel, we need to figure out whether the Green or Blue channel is going to be the best support channel. So if I click on Green, I can see that the skin tones darken up slightly and the background gets brighter. And then if I click on the Blue channel, slightly darker flesh tones once again and even a brighter background.
Now, it's a little bit difficult to figure out who is the most different than red because really the differences are very subtle throughout the channels inside of this image. So when in doubt, if you're starting things off with the red channel, the best support channel is probably going to be blue. And the reason is that red and blue come to closest to being color complements. So that's the way we're going to work inside of this image. I'm going to click on RGB in order to make the composite view of the image active, and I'll go to the Image menu and choose the Calculations command. That brings up an old school, remarkably, unfriendly dialog box.
It's been this way for years and years now. And it's talks to you in this, kind of, engineering language where we've got Source 1, and Source 2. Well, those here two channels by the way. And if you want to think of one channel is being stacked on top of the other, then Source 2 is in the background and Source 1 is on top. So just like we're seeing them here and then we're going to take those two channels and blend them together using a Blend mode plus an Opacity value. Now, 99% of the time you're going to leave the Opacity value set to 100% just as we are now.
And you're going to let the Blend mode do the heavy lifting, but first we need to establish the channels. Now, notice that we have three options inside Source 1 and Source 2. First, we're able to specify the image which most of the time is going to be the image you're working on. On rare occasions, you might bring a channel in from a different image and I could do that in this case. I could choose Elaborate Composition, which I still have opened, because it's the exact same size. The two images, if you're going to merge them together here, the two images have to be the exact same pixel dimensions, not one pixel off.
Next, you can blend the contents of different layers if you want to. Now, this is a single layer image, so we just have a Background layer to work from and then you decide the channels. So most of the time you just make sure that the images are set to the image you're working on, that the Layer is either set to Background or if you're working on a larger composition, then you might set it to merged so you can see a merged view of all the layers and then you select the channels you want to use. I typically put the base channel, that is the high contrast channel, at the bottom of the stack. And then I layer the support channel on top of it.
So I'm going to leave the Source 2 Channel is set to Red and then I'm going to change the Source 1 Channel to Blue. And by default, we're merging the two channels together using the Multiply mode. I rarely use that mode for creating a base alpha channel, by the way. I'll usually select one of the others and I'll show you how that works in just a moment, but if you want to get a sense of how the channels look at any moment in time, then go ahead and change the Blend mode to Normal for a moment. And then you assuming that the Opacity value is a 100% you will see the top channel by itself. So the Source 1 channel is now visible by itself because it's opaque and it's on top and that would be the Blue channel.
So this is what the Blue channel looks like. If you want to compare it to the Red channel, then you would take the Opacity value down to 0% so that you're seeing through the top channel down to the bottom channel which in our case is red. So that's how you compare the two. Just FYI. All right, I'm going to reinstate that Opacity value to a 100%. Now, comes to the tough part, choosing the proper Blend mode and, of course, one way is to just sort of cycle through the Blend modes and see what works. I could choose the Screen mode and go, hmm, no that's not going to work at all. And then I could try, for example, the Multiply mode that we just saw a moment ago and say nope, that's really wrong.
That's not the best approach because you're going to be at it for an awfully longtime. What you want to do instead is remember that you have three really great Blend modes to work with where developing a mask are concerned and that's Add, Subtract, and Difference. And so I tend to start with those and see what I come up with. I'm going to start with Difference right out of the gate here so that we can see the differences inside of the image. Because what Different does is it subtracts one image from another and then it finds the absolute value. That is, if a luminance level goes negative it switches back to positive.
You may not find that description very helpful. So here is another way to think of it. You're using one channel to invert the contents of the other and wherever that channel is white, you're totally inverting, wherever it's black, you're not inverting at all. You might also want to go ahead and invert the channel while you're at it because the idea is you want to create as many differences as possible between Source 1 and Source 2. So for example I might turn on the Invert checkbox for blue and see what I come up with and now I'm seeing a lot of differences where the hair is concerned.
And just for larfs, I might try the opposite effect. I might Invert the Red Channel as well and then turn off the inversion of the Blue Channel and see what I come up with. Pretty similar effect, although slightly different. So that's one way to work. That is the Difference mode. Another way, I'm going to go ahead and turn the Invert checkbox off for both of these options, another way is to try Add. Now, the Add mode is identical to Linear Dodge (Add). That's why Add appears in parentheses next to Linear Dodge, with one difference, you can control Add to a larger extent.
So it does the exact same math as Linear Dodge, but it adds some options as well. So, for example, if I just chose Linear Dodge (Add), then I would go ahead now the luminance levels from the two images and I would end up blowing out luminance levels all over the place, leaving black in the background as well. I could turn on the Invert checkbox for both of these channels here. And I could get an opposite effect where the hair is very bright and the background is somewhat dark, but it's not dark enough because as I say, we've blown out all those highlights like crazy.
Well with Add, if I go ahead and choose that mode, notice initially I get the exact same effect, not a single pixel on the screen has changed, but I have two additional options, Offset and Scale, and I'm going to explain exactly how those work in the next exercise.
Get unlimited access to all courses for just $25/month.Become a member
120 Video lessons · 57582 Viewers
119 Video lessons · 67151 Viewers
84 Video lessons · 16563 Viewers
125 Video lessons · 29472 Viewers
Access exercise files from a button right under the course name.
Search within course videos and transcripts, and jump right to the results.
Remove icons showing you already watched videos if you want to start over.
Make the video wide, narrow, full-screen, or pop the player out of the page into its own window.
Click on text in the transcript to jump to that spot in the video. As the video plays, the relevant spot in the transcript will be highlighted.