Easy-to-follow video tutorials help you learn software, creative, and business skills.Become a member
Here I'd like to investigate in the image that offers a challenge. in fact, a couple of challenges. This is a kind of image that has really no neutrals in it and really doesn't have a usable highlight in it. In most images of vast majority and we can usually find a highlight, we can find something that's neutral to help us with our color correction. But there are images like this that really don't offer much help. Luckily, we've got a really important tool to help us. First, image evaluation is always first. And this is a slide, by the way, just to let you know, so we've set this on film and a Positive slide back to Frame.
And we've set our frame here and we'll scan this for print, so 300 pixels per inch at a 150 line screen. And then let's go ahead and change the name to Dark Woman. All right, And then we can rename her to maybe Light Woman later on. So we've renamed it Woman, we've put down the logical name, a color space of RGB, and then the final resolution which would be 300 pixels per inch. And what's the key portion of this image? Well, clearly, it's the skin tone. There's no doubt about it, and the hair. We're not quite sure how much detail is in this image, but we're going to find out exactly how much we have to work with in just a moment.
But this image is all about the skin tone. So the one portion of the image that we have to help guide us is the skin tones, and we want to get that right. It would also be nice to improve the overall brightness and contrast of this image, no doubt about it. Well, let's begin as we always do by at least looking for the lightest portion of the image. I'm going to Shift and click there and sure enough, it's in the background here, so that's not going to be super-helpful to us. And then let's find the shadow, Shift +click there in the foreground, and we've got 18, 29, 31. pretty dark, but still the shadow detail, and then watch the Densitometer, let's move this down just a little bit.
What I always like to do is move my cursor over the image and see if I get much change in the Densitometer or Info readings, right, the tonal range readings. And I'm not getting huge variation, but at least it's not just flat line. What that means is there is some detail there that can be brought out. So that's a valuable tool and we can go in the background and look at the same thing. We're not getting as much variation back there as we are here. Okay. So we know we've got some shadow detail to bring out here and as we move towards the head, we see even more detail.
So that's good news. We don't really need this point back here, we can go ahead and leave it at 1 if we want to. We'll come back and chat about that in just a second. But we certainly want to set a couple of skin tone points here and here to make sure that we get the skin tone correctly. Well, let's look at the skin tone even before we've done anything else to this just to see what we've got for starting point. All right, Red is greater than Blue is greater than Green. So that's good news. At least, we're working with humans here, and notice that the Red is right about 50 points higher than the Green, and the Green is about 50 points higher than the Blue, at least right now.
So there's about equal separation there. And here, we've got about 40, 35, or 40 points of separation, and big separation over 50 points here. It's actually the ratios, and particularly in point number 4, it's actually pretty good. The Red to Green to Blue ratio is pretty good. So let's attack the overall brightness and contrast in the tone compression. We can do this with just a master histogram if we want, or we can do it with individual histograms. Either way will work. I'm going to be inclined to go take a look at the individual histograms. Red and Green and Blue.
And what this tells when I look at these histograms is so much information here. See all this Red here in the histogram? Not nearly as much Green and even less Blue. Well, of course, we have all this Red in the histogram here. Why, because that matches up with the skin that we've got here. And just visually, this looks fairly Red to me, although it's hard because it's so dark. I'm going to go ahead and do a correction on the individual histograms. And I'm going to move this and we've got this long flat line area right here. And I know from experience that probably the first half of this is likely to be noise.
So I'm going to move it halfway in here and then we'll look at the values that we're going to get. Shadow, no, we're not going to move that at all because we've already got a spike in the shadow there. Let's move to Green. I'm going to do the same thing about halfway through that data. We're going to assume this is mostly noise. We do have some shadow adjustment here. We've got a pretty sharp change from flat line to data, so we'll move that right up to the data. And Blue, we'll do the same thing as well. The reason why this works here, we don't have to worry too much about the highlights, is that we don't have any critical neutral diffuse highlights to worry about blowing out.
If we just quickly refer back to this here and we look at our value, we are blowing it out to 255. But I'm not going to worry about it because I'm really much more concerned about what's happening here. We can fix something like that in Photoshop, we're just filling that in using the Stamp tool to get rid of that later on. So I'm really focused on the skin tones. All right, So I'm pretty confident that we've set our highlights and shadows where most of the data is in the image. Let's go back and take a look at our values again. Look at our Shadow value first. 23, 15, 27, all right. Well, we don't want to go any darker than that, do we, because we've one at 15 here.
So that's good, that's fine. We'd lightened that up. And let's look at skin tones. 86 to 133 to 202, 87 that's almost 90. 90 to 130, so that's 40 points there. And then we've got 30 to almost 200, so that's about 70 points there. That's more than double. So notice by the correction that we've done so far, this image does have a fair amount of Red in it. When we move those skin values into the midtones or at least get in there, to the midtones, the Red is very, very high, particularly on point number 3 which is on her forehead.
It is not quite as bad on point number 4. So I'm going to click OK. I'm going to be pretty happy with that as far as setting the upper and lower tonal ranges. And then I'm going to go to Curves and let's click on the Master Channel, and normally, I would work on the individual channels and work on the skin tone, but this time, I'm going to make an exception. Why, because this image was so dark, just moving the Reds from like the three quarter-tone into the midtone has already changed the RGB ratios pretty significantly.
It's raised the Reds, and I'm thinking as we move this up, as we lighten the image, it's going to happen again. So to prevent us from having multiple rounds of back and forth, I'm going to go ahead and first do a master channel lightening of this image and then we'll take another look at the skin tone values and then make our skin tone adjustments. And this is a great image to demonstrate the difference between a linear and a logarithmic adjustment of the master curve. Let's start with the Linear and I'll just click here and drag this. And remember, you can do the same thing by just holding down your Command key and just dragging.
But here I'll use the slider because I want to show you the difference in the curve shape. And that clearly lightens our image, no doubt about it. But watch this. When I go to Logarithmic, watch the shadow area. Do you see how the shadows lightens more with the Logarithmic curve? And that's what, that's the difference. Typically, scanner's response to grayscale is faster and better and more response from highlight to quarter to the midtone and it tends to be less from midtone down to shadow. So we make these Logarithmic Curve adjustments and it's a great tool.
We get preferentially more lightning in the shadows and it's perfect for images like this where it's a low key image, where we need more lightening in the shadows and less in the midtone and even less in the highlight. Okay. So good! So let's go ahead and drag this over until we get an overall brightness that kind of suits us and let's revisit our sample points. Notice in sample point number 2, now that we've lightened it, we're up to 44, 31, 55. So we could darken that some more if we wanted to. It's at about a three-quarter tone now which is fine with me. And now let's look at the skin tones and see what we've got. Okay.
Point number 3, 118, that's about 120 and 156, that's about 155. So that's about 25-30 points of movement there, and then from 156 up to 211, that's 45-55, about 60 points. So that's about double and we look here, 117 to 157, so that's going to be 40 points of change, and 157 to 213, it's 50-60 points of change.
So particularly, in point number 4, the Red values are a bit high on the Red values. And that means I'm going to go to my Red channel here and normally, I would just drag this down manually, but because we've already placed a Logarithmic curve using the Master Channel, I'm going to use the Logarithmic curve to lower the Red channel a little bit. And I'm just going to lower it at about 5 or 6 points. They tick out some of the real obvious Red color cast on her skin. So if we just make a real fine-tune adjustment, but when you've got that much Red in the image, it doesn't take an awful lot.
And this is a creative adjustment at this point. When making it by the numbers, there's no real target value necessarily because the Red is greater than Green is greater than Blue and the separation of the Red is greater than separation of the Green and the Blue. So that's all good. But I see that the Red is a little bit higher than I would like it to be. Now even though this woman has definitely got red hair or auburn red hair, so she is likely to have a little bit redder skin, but I'm going to back it off just a little bit so it doesn't appear so red. So what we've done now is we've used both our histograms and then some specific RGB values.
Without these specific values, we'd be just shooting in the dark in terms of how we should be adjusting this image. And sometimes I go back and forth, I'll come back to the Master Channel on this, I might lighten it just a little bit more, and then go back to the Red channel and I might bring the Red channel down just a little bit more because it kind of depends on, I might go back and forth a little bit till I get what I like. I'm constantly referring to my values here. I'm looking to point number 2 here, I'm just checking out the Shadow value once more and I'm going to leave that at a three quarter-tone. There's so much darkness back here.
I don't want to make it any darker. I'm going to leave it a three quarter-tone. And as far as this little white area back here which is 255, I'm not going to worry about that either if my scanning software gives me the ability to edit that. And SilverFast actually does, we're not going to go there. But that can be taken care of in Photoshop. I can just use the Stamp tool and take that right out if I want to, if that's distracting. So there we go! I'm going to click OK. Am I going to sharpen this? Not on your life. One is this image is very soft to begin with and I don't want to lose that soft feel.
This is a low-key soft image. If I'm going to apply any sharpening at all, I'm likely to apply it on a mask. Then I'm going to make in Photoshop the woman. I'm not going to be sharpening the background at all and it's going to be very, very slight. So no sharpening here, so None, 48 Bit -> 20 Bit, named it. Okay, we're going to keep it. Let's see at the same dimension, but why don't we scale this up centered to slide probably to let's take it at 500%. So it's about a 5x7, maybe take it up 550. And here's where you can use scaling percentage to help you determine if you don't want to use output.
So if you know you want to print this at a 5x7, so take it up and take the scaling up until this moves up above the 5x7 that you see here. And then we'll do it at 300 pixels per inch for printing and then we'll hit the Scan button. And there we go! There is our finished image. And just to review, we used our histogram to help us evaluate, we used our Info tool, our Densitometer, and SilverFast speak to see that we have data back here so we knew we could some of it out, and then we used Logarithmic curves to help lighten the image. And then we used an individual Red channel curve to pull the Red down a little bit to just take some of that hot red out of the skin.
We didn't apply any sharpening to keep it nice and soft. Again, I mentioned if this was distracting to you, we can take the Rubber Stamp tool and we can just take that puppy right out. And then we've gotten rid of that distracting white highlight afterwards in Photoshop.
Get unlimited access to all courses for just $25/month.Become a member
119 Video lessons · 50107 Viewers
117 Video lessons · 37453 Viewers
113 Video lessons · 81230 Viewers
65 Video lessons · 10751 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.
Your file was successfully uploaded.