Digital BW, The Print

Austin Franklin wrote:
 >>All other facts and misinformation aside, here is a bit of information
 >>which should give you pause:  8 bits per channel of RGB data gives you
 >>(only) 9 bits of lightness (black & white) information.
 >>
 >>-Jon Dubovsky ( entropy@... )
 >
 >
 > Jon,
 >
 > Please explain the basis for this claim.

In the best of worlds, three equally-weighted 8 bit channels give you,
at most, about 9.6 bits of lightness information.  ( ln(256*3) / ln(2) )
   It doesn't actually work out to that much because your eye does not
equally weight the three channels.

(On a lesser note, if you're actually trying to extract the lightness
info, the most common RGB to lightness conversions compute the L channel
by taking (max3(r,g,b) + min3(r,g,b)) / 2, which gives you one more bit
of resolution than the source per-channel space (in this case, it gives
you 9 bits).)

Mr. Wesley, I only bring this up because I often hear people say that 24
bit RGB space has a great deal more contrast capacity than 8 bit
black-and-white space.  While the hue information gives the human eye
some extra info for contrast, it turns out that the perceptual
difference is far less than most people realize.  Depending on whose
numbers you believe, you end up with about 9 or 10 bits of useful
contrast range... right up against the arguable limit of human perception.

Good luck, all, and good print-making.  :)
-Jon Dubovsky ( entropy@... )

> Austin Franklin wrote:
>  >>All other facts and misinformation aside, here is a bit of information
>  >>which should give you pause:  8 bits per channel of RGB data gives you
>  >>(only) 9 bits of lightness (black & white) information.
>  >>
>  >>-Jon Dubovsky ( entropy@... )
>  >
>  >
>  > Jon,
>  >
>  > Please explain the basis for this claim.
>
> In the best of worlds, three equally-weighted 8 bit channels give you,
> at most, about 9.6 bits of lightness information.  ( ln(256*3) / ln(2) )
>    It doesn't actually work out to that much because your eye does not
> equally weight the three channels.

I think you missed my question.  Where does your equation come from?  Also,
the channels aren't typically equally weighted.

Mind you, I'm not disagreeing with you, but just want to know the origin of
your claim.

> (On a lesser note, if you're actually trying to extract the lightness
> info, the most common RGB to lightness conversions compute the L channel
> by taking (max3(r,g,b) + min3(r,g,b)) / 2, which gives you one more bit
> of resolution than the source per-channel space (in this case, it gives
> you 9 bits).)
>
> Mr. Wesley, I only bring this up because I often hear people say that 24
> bit RGB space has a great deal more contrast capacity than 8 bit
> black-and-white space.

That I agree with, that it doesn't.  There is a difference between luminance
and chrominance, and what's significant to the human eye.

Austin