6-5 Adjust channel curves individually – Beginner’s Tutorial

Until now, we have all adjusted in the RGB integrated channel of the curve. In fact, for images in RGB format, the curve of a certain channel can also be adjusted separately.
The concept of curves in the individual channels is the same as in the integrated channel, and is also divided into highlights, midtones and shadows. In the color principle of course #01, we learned that pure white is composed of RGB at the highest luminous level, and pure black is composed of RGB that does not emit light. After mastering the principles of RGB color composition, it will not be difficult to look at individual channel curves. And if you haven’t fully mastered the color principle, it will be quite difficult.

For a given image, changing the curve of a channel alone will cause color shift. It’s very simple. If you increase the middle adjustment of the red channel curve toward the Y axis, just like the adjustment we just made on the learning curve, it means that the red color is increased in the image, and then the image must be biased toward red. If you add green, it will look greener, and if you add blue, it will look bluer. You should be able to figure out all three of these. What if we reduce the red color? What color will the image gravitate towards? See, why I just reminded you to master the principles of color first, it comes in handy here. Now that you understand what goes into color mode, you can now quickly tell that reducing red will make the image appear cyan, reducing green will make the image appear pink, and reducing blue will make the image appear yellow. This phenomenon is called the waxing and waning of inverted colors (complementary colors).

If you don’t understand what I’m talking about above, or you’re a reader who joined in the middle, it’s best to read Course #01 first, and be sure to master it before continuing. The course arrangement is planned, and it is best to learn it from scratch and master it completely to avoid gaps in knowledge.

OK, let’s continue. Now we will not describe the curve adjustment method in detail, but will learn through experience through an example. We still use the previous picture, as shown below. I took this shot on a cloudy day, so it looks average. Now we’re going to adjust it to a brilliant dusk.

Let’s analyze the idea first. Since it is dusk, the sky should be golden, so the sky part should be golden. Golden is a mixture of red and yellow. Then you should make the sky reddish and yellowish. Reddish color means more red, yellow color means less blue color.
Furthermore, the sky part is close to the highlight area in the curve, so this operation can be summarized as: increase the red highlight and reduce the blue highlight. As shown in the 2 pictures below.

The display of the histogram palette is shown below on the left. It’s easy to see how the red highlights are increasing and the blue highlights are decreasing. The picture on the right is the changed image, which already has a very obvious golden effect.

Then make some adjustments in the RGB comprehensive channel to make the dark and mid-tone parts lower and darker to set off the sky. Brighten the highlighted part of the sky some more. As shown below on the left. This results in very brilliant dusk colors.

This effect is more gorgeous, but it looks too fake. Note that the green mountains in the distance in the picture are all biased toward yellow. What should we do if we want to retain the true color of the green mountains?
Let’s analyze it. The green mountain is in the mid-tone (you can press the mouse on the green mountain when adjusting the curve, and an indicator will appear on the curve). When you adjust the red channel and blue channel just now, the mid-tone part also follows. The highlight endpoint has moved some distance, so we should keep what originally belonged to the middle area in the middle area. Therefore, make the following adjustments to the curves of the red channel and blue channel. Leave the midtones of both curves where they are. As shown in the left and middle pictures below.
The image adjusted in this way will look more realistic. As shown below on the right. But we see that the lake surface has also returned to its original color. This is because the lake surface also belongs to the middle tone. But with such bright clouds in the sky, there should also be reflections on the lake. If we want to retain the color of the green mountains and adjust the color of the lake, we can’t do it now.
It cannot be said that it is completely impossible. You can control the adjustment effects in various places through very fine selections, but it is very troublesome. Don’t worry, after you learn about layer masks in later lessons, solving this problem will be a piece of cake.

Does everyone have a feeling here? That is, it doesn’t matter if the photos are bad, as long as you know how to adjust colors, you can create beautiful scenery. Haha, that’s true. Especially for digital photos, it is very convenient to use Photoshop to make color adjustments. Many of the images in this tutorial were taken with a digital camera. In the following tutorials or on the website, I will also explain some basic photography skills so that everyone can shoot the material themselves. Photoshop can create beautiful scenery, but it cannot replace your mood and memory when facing real beautiful scenery.

I love traveling by bike and most of the scenery is taken along the way. This is why bicycles often appear in the previous pictures. Hey, look carefully, you might see it one day. If I’m not around, it means the car has been stolen. Tell me quickly.

In the previous curve adjustment, if the histogram is set to RGB, you will often see the discontinuity phenomenon caused by the histogram after adjustment. The figure below demonstrates the RGB histogram before and after adjusting the RGB curve. It can be clearly seen that the color level in the adjusted RGB histogram has a “fence shape”. Turn on the statistics and use the mouse to measure the “fence gap” and you will find that the pixel is 0, indicating that there are no pixels on this color level. This is the so-called color gradient fault.

So how does the color gradient fault occur?
0610It’s actually very simple. As you can see from the picture above, the adjustment of the curve expands the originally narrow range of color levels. Then the pixels that were originally in a narrow range must adapt to the distribution in the new range according to the expanded ratio. But this expansion ratio is not an integer. The calculation accuracy of the Photoshop image processing core is ten thousandths, which means that the expansion ratio is a value like 1.3795. Then let’s calculate the expanded values ​​of the three consecutive color levels that were originally at 114, 115, and 116:
114×1.3795＝157.263
115×1.3795＝158.6425
116×1.3795＝160.022
We know that the value of the color scale is an integer, so the calculation result must also be rounded. So the expanded color levels should be 157, 159, 160. It can be seen that 158 ​​color levels have been skipped. Then level 158 becomes the “fence gap” in the adjusted histogram. The originally continuous color scale became discontinuous after expansion, forming a color scale fault.

This is why faults always seem to occur at intervals, because errors occur at intervals. And the image core of Photoshop also actively distributes these errors as much as possible in the full color range to avoid errors concentrated in a certain range. Therefore, the discontinuity in the color scale is caused by the expansion of the color scale range.

In the above example, it can be seen that the color level of the original image does not reach the full range, that is, it does not reach the 0 and 255 endpoints. So if the color level of the image itself has reached the full range, will there be no errors? No, even if the original color scale has filled the full range of 0 to 255, it will still cause calculation errors in the transition part of the color scale. This is because of the merging of pixels at the left and right endpoints of the histogram, as we learned earlier.

So, will narrowing the color scale range not cause gaps? Yes, the original continuous color levels still maintain continuity after being reduced. Reducing the calculation will only produce overlapping levels, but not discontinuities. We can know through calculation:
114÷1.3795＝82.6386
115÷1.3795＝83.3635
116÷1.3795＝84.0884
After rounding, it is 83, 83, 84. It can be seen that the original 114 and 115 color levels have been merged. So wouldn’t color level 82 be idle again? No, you will understand if you use 113÷1.3795.

How to avoid faults?
Unless you don’t expand the range of color levels, gaps are inevitable.

Will faults affect the image?
It has both impact and no impact. The impact is absolute. Since the color level is reduced, it must be an impact. No impact is relative. As long as it is visible to people on earth, it makes no difference whether the fault exists or not. Can your eyes distinguish color gradients?

How to repair faults?
It’s simple. If you change the total number of pixels in the image, or change the color mode, Photoshop will refill the color range.
It should be noted that changing the total amount of image pixels refers to “global adjustment” operations for all pixels such as [Image>Image Size] or free transformation [CTRL T]. However, [Image>Canvas Size] and Cropping Tool [C] belong to “area adjustment”, so they are invalid. We will introduce the free transformation and cropping tools later.
Finally, Photoshop’s internal operations are far more complex than our above example. Here we have made an easy-to-understand metaphor to facilitate everyone’s understanding.

<!–Collectio