High-level editing in Photoshop 6.0
Newborn High Chairs,Baby High Chair With Wheels,Baby High Chair 3 In 1,Children Dining Table Zhejiang Lamon Technology Inc. , https://www.baby-chair.com
Photoshop captures and edits more than 8 bits of image in each color channel, a feature that many designers are fascinated by. The “high position†in the title refers to a color channel that exceeds 8 bits of data. Photoshop treats images with more than 8 bits per channel as a 16-bit algorithm, but most image capture devices provide bits. The depth is relatively low, usually 10, 12 or 14 bits per channel. The "high position" covers all these situations.
There are two main reasons for working in high-level mode. First of all, when working with 8 bits per channel, each channel contains only 256 possible values ​​(0-255). This amount of data is enough for our human eye to recognize the smooth tone and color, but actually When you edit an 8-bit image per channel, you will find that the available data is sometimes much less than 256, which is mainly related to the accuracy of the edit. For example, when applying a 1.2 gamma value (teaching is softer) to adjust the midtones, the dark tone value is pulled off, no longer a continuous value, there will be some interruptions in the middle. At the same time, the high light values ​​are squeezed together so that some different pixels now become the same. The spacing will cause multiple tones or ribbons, and the loss of different pixels will cause loss of detail.
Figure 1 shows a histogram of a 24-bit RGB spectral gradient image, which is a histogram before and after gamma correction. The corrected histogram has a clear pitch in the dark tone area, and some spikes appear in the high light area, indicating that some pixels have been squeezed together.
When using the high-level file for the same edit, each channel contains 65,536 values, and the amount of data is very large, although at this time if you compress the image will lose some pixel values, but there will not be multi-tone or ribbon phenomenon, The main reason is that when the image is down converted to 8 bits per channel for output, there will be a full 256 steps per channel.
FIG. 2 shows a histogram of a 48-bit RGB spectral gradient image, which is also a histogram before and after gamma correction, the same as the method of FIG. The corrected histogram is completely free of pitch and compression.
Second, most scanners and professional digital cameras can capture more than 8 bits of image per channel. Although the data they provide to Photoshop is not always ideal, most of them support the editing of pre-scanned high-order data in the size of a postage stamp. , and send the scanned RGB values ​​directly to the monitor. When you open an image in a workspace in Photoshop, the image you find and prescan is not at all like that. An original, high-level scan allows you to get what the scanner can see. This scan has been corrected in the scan driver and can be edited using tools in Photoshop. You can even make a description file for the original scan map, apply it to Photoshop, you can save a lot of labor.
At first glance, the set of tools for editing seems to have little functionality and can only accomplish the overall gradation and color correction. The only tools available for direct manipulation of high-level images are Marquee, Lasso, and Polygon Lasso. However, Magnetic Lasso cannot be used. Magic Wand, Color Range commands, and Quick Mask modes cannot be used when working with high-level images; in addition, layer adjustments and brushes cannot be used. Most of the filter drop-down menus cannot be used except for a few filters that can be applied to high-level images. Unsharp Mask is one of the most important filters applied to high-level images. Most of the tools in the toolbox cannot be applied to high-level images.
In short, the situation is very unoptimistic, but when we see the Load Selection command and the History Brush, this view changes. With these tools, you can choose whatever you want to select, and you can also get the effect of any kind of brush.
The low-level technique for selection is to copy the high-level image and down-sample it into 8-bit images for each channel so that all the tools in Photshop can be used. This extremely simple concept is very helpful for obtaining powerful technology. Here is an extreme example to illustrate this technique.
This is a very bright and dark image. The bright part is too bright and the dark part is too dark. Obviously the overall editing is not suitable. At this time we have to select part of the image.
First, copy the image (Duplicate command under the Image menu) and select the 8bit/Channel command in Mode under the Image menu so that you can downsample to make a The image is represented by 24 bits. Now that all of the selection tools in Photoshop are available, use the Marquee tool to select bright areas and then use the Color Range command to select them.
At this time, we can save the selected area as a channel, or we can make an adjustable layer when the selected area is activated. It will automatically treat the selected area as a layer mask. Both of these methods can be used by us, and the adjustable layers are more efficient. If you are concerned about losing the complex selection, you can save it directly as a channel, take a suitable name for later identification, and experience shows that if you complete a selected area for more than 3 seconds, it is best to save it. .
When the selected area is activated, the selected area can be treated as a layer mask, and the area can then be adjusted using the Image-Adjust-Curve command.
After the adjustment, we can load the selection and curve into the high-order file. Now let's look at several ways of transferring these edit operations to the high-level file.
There are two ways to load the selection into the high-level file:
You can use the adjustment layer as a target in the replica, then convert to the high-level file, and select the Load Selection command from the Select menu to bring up the Load Selection dialog box. Select the copied file in the document menu, select Curves1Mask from the Channels menu, and click OK.
Note that Curves1Mask is represented in italics, indicating that it is a layer mask rather than a stored channel. This dialog appears only if the curve adjustment layer is positioned in the replica.
When the Curve layer is positioned in the replica, enter the Channel panel. At this point you can see that the layer mask appears as a channel, but only if the Curve layer is positioned. The function of saving the selected area as a channel is always available. Select the Duplicate Channel from the pop-up menu in the Channel panel. Name it as A. Save it. A new channel will appear on the Channel. On the panel. When you convert to a high-order file and select Load Selection, A appears in the Channel menu options in the Load Selection dialog box.
There are also two ways to load curves:
In the copied image, double-click the thumbnail of the adjustment layer in the Layers panel, open the Curves dialog box for the curve to be loaded, and click OK. You may think that nothing has been done, but what it actually remembers is to use the curve for the last time. Switch to the high-order file, make sure that the selection is loaded, press Curve/ctrl-opt/alt-M to open the Curves, and you can see the settings of the last-used curve.
Activate the copied image, double-click on the thumbnail of the adjustment layer in the Layers panel, open the Curves dialog for the curve to be loaded, and click Save, name the curve, and save the curve to Disk. Switch to the upper file, open the Curve, and click the Load button in the dialog. Browse to the curve saved to disk and load it.
Each method can be used, but the first method is found to be more convenient in practice because it uses fewer keys and generates fewer files. It saves all the necessary information in a 24-bit copy image. The content, in an optimized workflow, can now be edited on a 24-bit copy without considering the high-level image. You will find that you need to continuously go back and perform fine-tuning during continuous editing, so make a complete set of edits in the copy, allowing the operator to complete all fine-tuning before applying these operations to the high-level image.
The next step is to deal with the background (dark tone area), reduce the contrast of the entire image, and remove the bias (brighter blue) of the dark tone. The selected area is the opposite of what was previously described. Hold down cmd/ctrl and click the thumbnail of the curve layer mask in the Layers panel. Then press cmd/ctrl-shift-I. Invert Selection (shortcut key) under menu). In this way, you can directly select the background. Now you can make a second curve adjustment layer. It will use the new selection as a layer mask.
Use curves to reduce contrast and remove blue shadows. Next, apply this step to the high-level file, and we can always work on 24-bit replicas. After such processing, the quality of the image has greatly improved