Don't be too concerned about this, I just wanted to make this point anyway.

Image processing has similarities to Audio processing. A point I tried to make here http://www.gnutellaforums.com/genera...tml#post377764.
For example, if you use an Equalizer (audio filters) upon the music file, it means some information will be shifted from one bandwidth to a nearby one and some other information will be thrown away. This also applies similarly to image manipulation when applied to levels, curves or any other brightness or color changes.

So let's say an image file has 200 dots of color and tonal information. If you adjust the color in the image, it will shift some tonal/color dots from one color bandwidth to another whilst simultaneously throwing a dot or few away.

Once destructive processing is used upon an image or audio file it will never be the same again. One reason it is best to work with layers (possible with some audio processing apps also) and also retaining backups of the original files and possibly also backups of work files at certain points of processing so you can always go back to a certain point in time of processing to fix things up.

Photoshop makes it quite easy to do this with layers. And sometimes you might even wish to duplicate an effects layer to make further changes but retain the previous one as a hidden layer just in case.