Look at our sample image. Narcissus covers about 20% of overall area, inscriptions about 10% and approximately 70% of the picture are almost black. Why should we save all the image with equally high quality? Channel-based weighted optimization solves this problem.
1. Select Narcissus and/or the other key areas of the image with a selection tool.
2. In the Optimize palette, choose JPEG from the format menu.
3. Click the "Use a channel" button next to the Quality slider (movie).
4. In the dialog box, display the "Channel" menu and choose "Save Selection" item to save the selected area as alpha-channel (movie). Note on the status bar of the image window how the file size has changed.
5. Set the quality with "Maximum" and "Minimum" sliders. The selected channel area will be saved with maximum, the rest of the image - with minimum quality.
6. Estimate the result. Well... even 5 Kb without loss of quality is no bad.
Now let's test the weighted optimization for GIF format:
Now test the weighted optimization for GIF format.
7. In the Optimize palette, choose GIF in the format menu. Set "Colors" to 256, color reduction algorithm to "Selective".
8 . There are 3 "channel" buttons on the palette now (screenshot). They make saved channel to influence color reduction, dither and lossy settings. Use all these buttons in the same way as with JPEG optimization.