On Wed, 27 Oct 2004, Mark Roberts wrote: > >A change from one shade to the next, say from level #240 to #241 is > >very small in percent, but if you go from shade #1 to #2 it's a 100 > >percent increase. > > Only if you're using a linear scale.
Yes, the exact values I quoted are based on a linear scale, but the concept of quantisation noise exists regardless. Log or lin or something else; you will still get smaller resolution for small signals, and quantuum leaps for the very smallest. anders ------------------------- http://anders.hultman.nu/ med dagens bild och allt!
