Color Correction and Data Loss

Color correction is a subtractive process. The more adjustments you make, the more color data you lose. Although this is par for the course with any image adjustment, there are a few things you should avoid when it comes to tonal correction:

Never Apply Two Consecutive Edits That Counteract Each Other For example, never lighten the midtones in an image, and then proceed to do the exact opposite by darkening them. The only way to make these kinds of drastic edits and avoid posterization (creating stairstepping gray levels as opposed to smooth gradations), is to use nondestructive adjustment layers whenever possible.
Be Careful Not to “Stretch” Too Far When “stretching” the tonal range of an image, pixels that previously shared the same values can be made different by several levels. Stretching the tonal range too far can cause the image to display noticeable jumps in tone or color.
Don’t “Oversqueeze” When “squeezing” the tonal range of an image, pixels that previously contained different values become compressed to the same value. Compress too much, and you will lose valuable image detail.
Enhance Detail—Minimize Noise Adjacent pixels with different values are what make up detail in an image. In areas where the difference is slight, detail can be accentuated by differentiating the pixels even more. However, if the image contains noise, enhancing the detail also enhances the noise. The challenge is to enhance image detail just enough without overaccentuating any noise.

Now that you know what not to do when color-correcting images in Photoshop, let’s focus on what you can do to make the task a little easier:

Work with 16-Bit Images 16-bit-per-channel images may be larger in file size, but they offer you considerably more headroom than traditional 8-bit images when it comes to color correction. 8-bit images contain up to 16.8 million colors, but 16-bit images contain trillions of colors. Therefore, data loss is much less apparent when working with high-bit images. If you’re working with RAW files, remember that the Camera Raw plug-in allows you to open them in Photoshop as either 8-bit or 16-bit. Working in 16-bit mode can be useful if the corrections you expect to make are extreme, as they can cause posterization in 8-bit images due to data loss.
Use Adjustment Layers Whenever Possible With most image adjustment commands, such as Curves or Levels, you have two options: to access the dialog box from the Image -->Adjustment submenu and apply the change, or to use the adjustment layer equivalent. Choosing the menu command applies a permanent change to your image that can be undone only by using the Undo command (F/Ctrl+Z) or by selecting a previous state in the History palette. Edits cannot be made to the adjustment unless the command is undone first. Take one too many steps past your history limit, and you’re stuck with permanent data loss that cannot be retrieved. Adjustment layers, on the other hand, allow you to edit or undo the adjustment anytime—without any limitations or penalties.
Choose Edit-->Fade to Tone Down an Adjustment If you’ve chosen not to work with adjustment layers, you can always use the Edit-->Fade command (F/Ctrl+Shift+F) to reduce the opacity of the last applied adjustment. However, there is a limitation to using Edit-->Fade, as Photoshop allows you to apply the Fade command only immediately after applying the adjustment.