1. Channel Cutout

Channel Cutout What are the steps for channel cutout in PS?

What is a channel? I compare the channel to a searchlight. "Three lights, red, green, and blue, glow together at different brightnesses (RGB)" to form a colorful image. In the channel, the image is split into three lights: R, G, and B. , observe separately.

In terms of cutout mapping, the main function of the channel is to "delimit the selection." Please look at this button: Load the channel as a selection, its shortcut key is CTRL + left click. After executing this command, all pure white parts with high luminous intensity in the channel will be selected, and black will be ignored.

Therefore, the basic idea is: turn all the areas you want to cut out into white, and turn all the areas you don’t want to cut out into black. Let’s look at this picture: the left half of the hair is light, the right half of the hair is dark, and the edge hair is gray. The hair has 3 different luminous color levels, so you need to make the selection at least three times and then merge the selections together.

The basic route is: first pick out the complex hair, and then pick out other simple parts (body parts). The bright hair is A, the dark hair is B, the gray hair is C, and the body is D. First of all, the first step is to process A. Observe the three channels and find that the hair is the brightest under the red channel, so drag the red channel to the button in the lower right corner to make a copy and then select the part with a frame. I use the rectangular marquee tool. Then CTRL+L, adjust the color levels. The purpose of adjusting the color levels is to increase the black and white contrast between the hair and the background. The black is darker and the white is whiter. Drag the slider to adjust until there is an obvious black and white contrast between the hair and the background and then use the brush Paint the excess white on the background black, and then paint all the head white. After painting, CTRL+D cancels the selection. Your image should look like this. Next, process the upper part, select it again, and then CRTRL+L Adjust the color levels until the contrast is clear enough. It is still the old method. Use a brush to deal with the excess black and white. Be sure to apply it carefully and carefully, and then properly handle other parts of the hair.

At this point, part A is processed. Select it again, then CTRL+C to copy, or execute Edit-Copy and click on the lower right corner to create a new channel, create a new blank alpha1 channel and select the blank channel. CTRL+V Next, process B. Observe the hair on the right and find that all channels are very contrasting, so I just select a channel and make a copy (I still use the red channel), or use the method mentioned before to frame it. , then CTRL+L to adjust the color levels, and use a brush to blacken the head. Again, be careful when processing the edges. Next, CTRL+I to invert, and CTRL+L again to adjust the color levels. Continue to use the brush to blacken the excess white on the background. Part B is completed, and then CTRL+C is copied.

Select the alpha1 channel where part A was placed just now, press CTRL+V to paste it here. A and B are merged. Next, process C, with gray hair on the edge. Observe the three channels and find that the contrast is strongest under the green channel. Duplicate it, then use the lasso to select the hair above, use CTRL+L to adjust the color levels, and use the brush to process the hair like this. It is worth noting that the hair here has been blended with the background to a large extent. It is impossible to distinguish and cut out 100% of it. Generally, this part can be discarded appropriately. I used a 1-pixel brush to trace it out according to the rough texture. Then CTRL+SHIFT+I to invert the selection, set the foreground color to black, ALT+DELETE to fill in black, CTRL+D to cancel the selection, deal with the excess white, then press CTRL+left button to load the selection, then CTRL+C to copy, then CTRL+V Paste it into the alpha1 channel that holds parts A and B. Now it should look like this. Next, for the last bit of hair, use the lasso to select the rough shape along the hair, trying to avoid the background color. Next, you need to use a command: The purpose of calculation is to superimpose channels in a specific way, change the luminous intensity, and facilitate the creation of selections.

Perform image-calculation, change the blending mode to "Color Dodge", and leave other parameters unchanged. Then you will find that this part of the hair becomes brighter.

Then, still use the method mentioned before, CTRL+L color scale, and paint with a brush. Here you can do it according to your feeling and process it carefully.

Then CTRL+SHIFT+I to invert the selection, set the foreground color to black, ALT+DELETE to fill, CTRL+D to cancel the selection, deal with the excess white, then press CTRL+left button to load the selection, then CTRL+C to copy, and then CTRL+V to paste. At this point, the alpha1 channel that stores A B has been processed and merged together. At this step, you should be able to understand the meaning of the channel and selection.

The following is very simple, ctrl+left-click to load the selection and click on the RGB channel. At this point, many children with fast hands want CTRL+J, right? no! Make a copy of the original image and add a mask. Why use mask instead of direct CTRL+J extraction? Because the mask can be adjusted, if you are not satisfied with a certain part, you can redo it.

You can try adjusting the specific parameters until it feels natural. Finally, just cut out the body. Change the background to see the effect (since the body part is very easy to cut out, use quick selection tools, pens, lassoes, etc. Either way, I won’t go into details here) Finally, I want to say that the purpose of cutout is not to "cut out", but to integrate it. After integrating it into a new context, disharmony and rejection will inevitably occur. If you want to do When it reaches 100% integration, it must be discarded appropriately. It is impossible to cut out a single pixel without losing it. Source: Designer Colorful.

How to cut out pictures using channels in PS?

For example: use the channel function of PS to change the background image of a photo

1. The principle of channel clipping: channel There are: color channel, ALPHA channel, and spot color channel. The ALPHA channel is used to store the selection, and the ALPHA channel uses 8-bit grayscale between black and white to save the selection. Instead we can use the black and white contrast in the ALPHA channel to create the selection we need. (White is the selected area in the ALPHA channel) Levels can correct the tonal range and color balance of the image by adjusting the intensity levels of the dark tones, mid tones and highlights of the image. We can increase the black-and-white contrast of the image through color levels, and use this to determine the selection range. 1. Open the channel panel. We first select a color pair among the three channels of red, green, and blue to copy this channel. Here I selected the blue channel and right-clicked to copy the channel

2. Click Image→Adjustment→Levels, move the small black triangle to the right, adjust the dark tone level, and then add the white Move the small triangle to the left and adjust the color level of the bright tones to separate the tones of the characters and background in the image. You can also enter the color level above: 135/1.00/195.

3. The outline of the character has been basically separated from the background. Then we use the brush tool to paint the white inside the character into black, and the black outside the character into white.

4. Click Image → Adjust → Invert to turn the character part into white, which is the selected area.

5. Click "Load channel as selection" under the channel panel, then return to the layer panel and select the background layer.

6. The selection has been made, and then use the cut and paste command to select the characters.

How to use channel cutout in ps (complex background)

How to use channel cutout in ps:

1. Open the image material to be processed and copy the image The layer gets a "Copy of Layer 1".

2. Enter channel mode, select a channel with the greatest color contrast between the character and the background and copy it to get a "blue copy".

1. Use Ctrl+L to adjust the color levels of the "Blue Copy" layer to increase the contrast between the character's hair and the background. Hold down the Ctrl key and click on the "Blue Copy" layer to obtain a selection.

3. Shift+Ctrl+I inverts the "Blue Copy" selection and returns to "Layer" mode.

Ctrl+C copies the "Layer 1 Copy" selection, Ctrl+V pastes, and you get a hair layer like "Layer 2".

4. Hide the "Layer 2" cut out in the previous step, and use the "Quick Selection Tool" in the "Copy of Layer 1" to select the approximate area of ??the character. Ctrl+C copies the "Layer 1 Copy" selection and Ctrl+V pastes it to obtain the "Layer 3" body part layer.

5. Display "Layer 2" and "Layer 3", so that a character image with complex details is roughly cut out.

What is the principle of PS channel cutout?

If someone asks you, which knowledge point is the most difficult in Photoshop? I believe 80% of people will think it is a channel.

What you see today is the first article. We mainly learn the principle of the channel function. Later, I will publish the second article to explain the specific application of the channel, that is, how to use the channel to pick out the character's hair. . I hope you can read both articles and fully understand the usage of channels.

It is said that the channel is difficult. In fact, it is not that it is difficult to understand, but that in order to fully understand the channel, you need to understand multiple knowledge points, such as color principles, selection knowledge, and mask knowledge, the most critical of which It lies in the principle of color. I have published an article about the principle of color before. If you don’t understand the principle of color, you can learn about it first. I will not repeat the explanation in this article.

Channels are mainly divided into 2 categories: color channels and Alpha channels. Some friends may mention spot color channels. In fact, I personally think that spot color channels also belong to the category of color channels. In order to avoid confusion, I put them together.

Color channel Some people say that channels are used to cut out images. In fact, this understanding is very one-sided. The fundamental function of channels is to record color information. You may not feel the help of the channel when you are drawing, because it secretly helps the image display in the background.

It can be said that if there are no color channels, the image will not be visible at all. There are many color modes in PS software. In order not to confuse everyone, we will only talk about the RGB mode. The principles of other modes are the same.

If you read my article on color principles above, you should understand that no matter how colorful an image is on the screen, its essence is composed of the three primary colors of red, green, and blue. A combination of different color ratios. So now the question comes, how much red, how much green, and how much blue is used in an image? This is what the channel does, recording the proportions of the three primary colors in a picture.

I opened a picture and took a screenshot of its channels. As can be seen from the picture, a *** has three channels, namely the red channel, the green channel, and the blue channel. The top RGB channel is a composite channel, which is the effect of combining these three primary color channels, so it does not belong to a single channel.

Let me first tell you the principle of color channels, that is, the corresponding color ratio in each channel is displayed in black, white and gray. When the channel image is displayed as black, then the color ratio corresponding to this channel is 0, that is, there is no color of the current channel; when the channel image is displayed as white, then the color ratio corresponding to this channel is 100%; the ones in the middle are all Different levels of grayscale values ??correspond to different channel color proportions.

Maybe it’s not easy to understand. I’m only opening the red channel now, so you can see it and you’ll understand. As can be seen from the picture, the gray color of the clothes in the front is lighter and more inclined to white, indicating that the proportion of red in this area of ????the clothes is relatively high.

The background itself is green leaves, and the red component is low, so the gray is heavier, leaning more towards black. I believe most friends should have two questions.

First: The background behind is green leaves, but there is obviously no red at all. Why is it not pure black? This is because our naked eyes cannot distinguish it so accurately. In fact, there is a little red component in the green leaves in the background. Second: The hair and skin color are obviously not red, so why are they whiter? Is it more red? If you have the second question, it means you don’t understand that the RGB mode of PS is an additive color mode, that is, when the three 100% ratio colors of red, green, and blue are added together, it will become white.

(You can try this in the software yourself) Therefore, hair and skin color not only have a lot of red, but also a lot of green and blue, because its color tends to be white. . OK, then we will only open the green channel this time to confirm it again for everyone.

Obviously, the green color at the back is lighter this time because it has reached its exclusive channel.

OK, this is the essential principle of the color channel. I hope everyone can fully understand it, because this is the necessary knowledge to teach you how to use channel matting later.

Alpha channel Now let me tell you about the Alpha channel. Don’t be confused by the English. It feels like it is so high-end, but it is actually much simpler than the color channel. The principle of the color channel just now is that each gray value represents the proportion of the color of that channel.

The Alpha channel is also represented by black, white and gray, but each gray value represents the transparency of the selection. Still in the picture just now, I made a selection for the red clothes, and then created an Alpha channel. I only turn on this Alpha channel now. The white *** area is the selected area, and the black *** area is the non-selected area.

If you want to bring up the selection, just hold down the Ctrl key and click on the small thumbnail of the Alpha channel. Then a function of the Alpha channel appears, that is, it can store selections.

If you make a selection in this PSD and want to use it as a backup, you can create a new Alpha channel after making the selection. Even if you save the PSD file, the Alpha channel will still be there when you open it again.

Some friends may say that if it is so troublesome, it is better to just use a layer to store the selection. There are two major advantages to using channels to store selections. First, it does not take up space and does not increase the file capacity.

Second: Not only the PSD format can store Alpha channels, but also supports TIF and TGA formats. There is another question, that is, what happens if there is a gray *** domain.

I directly feather the selection, and you will understand. As you can see from the picture, the different gray values ??represent the opacity of the selection. White is a complete selection, black is no selection, and gray Just choose according to the transparency ratio. OK, the article is about to end here.

This article only explains the principles of channel knowledge. The next article will show you how powerful it is to cut out images based on the knowledge of channels. Thank you all for reading, please like and follow, thank you.

Source: Designer Colorful.

How to use ps channel to cut out pictures

We cut out the tree from a picture and put it on another background.

The trivial details of leaves are not easy to handle with ordinary cutout tools, so channels are used here. 1. Cut out the image. Prepare to open a file that needs to be processed as shown in the figure, or ctrl+o as shown in the figure: Here we have to cut out the tree from the picture. The branches and leaves of the tree are very sparse and there are many small places, which is difficult to use. If the single cutout method is comfortable, let's try using channels, and you will feel different.

Pay attention to the icon. Ctrl+J duplicates the background layer.

We will now use the PS channel to cut out the tree in the picture. 2. Enter the picture channel as shown in the figure: After entering the channel, you will see the red, green, and blue channels. This is the channel in the RGB color mode. All RGB pictures are made of red, green, and blue through a certain algorithm. Formation, for specific RGB imaging principles, please refer to the RGB entry on Baidu Encyclopedia.

3. Enter the red, green, and blue channels to check the image quality. Red channel: Green channel: Blue channel: When selecting color channels, be sure to select them one by one, as shown in the picture, do not click It turns blue and then displays red, so the color will not be black and white. Channel cutout means observing in the color channel. Whether the contrast between the image to be cut out and the surrounding images is clear. The clearer it is, the easier it is to cut out. This is the basic requirement for looking at the quality of the color channel. Sometimes the contrast may not be obvious. We also need to use tools such as color levels and masks to adjust the contrast between black and white channels.

4. Copy the channel and adjust the contrast. Here we compare and see that in the blue channel, the tree of the image to be extracted is black, and if there is a clear contrast between the other background and the tree, we will Select the blue channel as the cutout channel. Then you need to copy a layer of blue channel. Do not modify each RGB color channel, which will destroy the color of the original image. Copying the channel forms a color-specific channel, which will slightly change the color of the original image. Under normal circumstances, observe The color change of the original image cannot be seen, so we copy the blue channel.

As shown in the figure.

Right-click on the blue channel and copy: After the blue channel is copied successfully, select the copy of the blue channel and press Ctrl+l to adjust the levels to adjust the contrast between the tree and the background. You can also complete this step through the menu-Image-Calculate. , generally using color levels will adjust the contrast faster, but each has its own advantages, and using calculation will minimize noise.

Here we use color levels. Adjust the color levels: maximize the contrast between the tree and the background, then use the brush tool to paint the tree black and the other colors white. If you are not sure, you can use the pen tool to create a selection.

After whitening, it is as shown in the picture: 5. Channel detail correction We can see that the contact point between the tree and the grass is not well grasped, so let's enter the RGB color channel again to see the relationship between the tree and the grass. The contrast between the grass and the contrast in that passage is even more obvious. Relatively speaking, the contrast between the trees and the grass in the red channel is more obvious, so we adjust it through the color level, specifically the same as the blue channel, copy and adjust, as shown above.

1. Copy the red channel to correct the details of the unclear contrast between the trees and the grass. 2. Like the blue channel copy, process the red channel copy through color levels as shown in the picture above. Pay attention to the tree and background. The place of separation. Further processing with a brush.

After processing: See the red copy spot color channel as shown in the figure: 6. Create a blue copy selection of the tree. First use ctrl+left click to load the red copy into the selection, and then return to RGB color. model. Keep layer 1 selected, hide the background, and create a black mask on layer 1, as shown in the picture: Use a black brush to gently correct the blue unclean areas.

Or use the color range to select blue and replace it with green. It mainly depends on how you think.

As shown in the figure after processing: Go back to the color channel, create a selection for the blue copy channel in the same way, return to layer one, and keep the selection. Fill the layer 1 mask with black.

(When maintaining the selection, the selected color is white, and what we need is the opposite color of the selection - black, so the selected color of the blue copy must be painted out in the mask) as shown in the figure: 7 , observe the effect of cutting out, and perform micro-processing of details. We see that the image of the tree is not obvious, and the dense leaves are not highlighted. The edge of the tree is basically OK. We will copy the background layer again, and then load the blue channel separately into the background copy. Use the same method to load the blue channel into the selection, return to the RGB layer, create a black mask, and do it again.

As shown in the picture: After repeating it once, it is almost the same. For the part with noise, create a color range to load the noise into the selection and replace it with green. 8. It’s done. Let’s find a background to try out the effect. Let’s just fill in the background color to show the effect of our channel cutout.

9. Find a seaside scene, renderings. Seaside scene picture: move in the tree we cut out, and then replace the display picture: If you look at it, for example, leaves, hair, hair, transparent objects I still can't handle the cutout well. It is recommended to watch it a few more times to deepen your understanding of channel cutout. Of course, you can also contact me through Baidu for help.