Re: LAB as working/interchange space
Re: LAB as working/interchange space
- Subject: Re: LAB as working/interchange space
- From: "shAf" <email@hidden>
- Date: Wed, 23 May 2001 20:39:28 -0700
Tom writes ...
>
on 24/5/01 5:08 AM, Chris Cox at email@hidden wrote:
>
>
> Look at the encoding. 16 million colors in 8 bit/channel
>
> sRGB will transform to only 2.47
>
> million colors in 8 bit/channel LAB (without dithering).
>
>
Huh? RGB has 3 channels of 8 bits each. So does Lab.
>
Which means that each has 24 bits each. Doesn't that
>
mean the same number of colors?
If you open "ole Moire", and image which should contain a full
representation of color and fill any gamut, and convert to Lab ... and
then use the 'adjust|levels', we all realize ... yes, ... 0-255 value
available, ... full 8bit precision. However ... take a good look at
the 'a' and 'b' histograms. 'a', for this supposedly "full gamut"
image takes up values ~50 thru ~210. This would imply, that
effectively, you are adjusting what everyone considers a wide gamut
with only ~160 values or ~7bit precision. On top of this ... is the
confusion caused by 8bit values being available, but which send your
image beyond perception. This behavior underlines what many describe
Lab as ... "unintuitive" and "difficult".
shAf :o)