Re: 16 bits grayscale images

[Sergei Steshenko <sergstesh yahoo com>]

--- David Neèas (Yeti) <yeti physics muni cz> wrote:

> On Thu, Sep 27, 2007 at 04:45:10PM +0200, alex wrote:
> > I want to create a grayscale image from a CCD camera.
> > Data are from a function wich gives me an 16 bits array corresponding to
> > the CCD's pixels.
> > I take a look at the pixbuf class, but it seems to use a RGB structure.
> > How can I do my job ?
>
> If you don't expect the graphics system to be actually
> capable of displaying 16bpc images, it is easiest to just
> convert the data to 8bpp RGB (which is a trivial operation).
>
> Yeti
>
> --
> http://gwyddion.net/
> _______________________________________________
> gtk-list mailing list
> gtk-list gnome org
> http://mail.gnome.org/mailman/listinfo/gtk-list

Oh, I expect a lot of things. And I do not consider a lot of things
to be trivial.

First of all, modern displays have contrast ratios greater than 255:1, so
there is no physical reason not expect more than 8 bits of dynamic range,
and, actually, even with smaller contrast ratios the matter is brighness
resolution.

Second, 16 -> 8 bits conversion is not trivial - I am not talking about
taking just, say, upper 8 bits of a 16 bits word.

I am talking about brightness compression first. For example, I can decide
to use 12 bits of dynamic range of input image, then convert them to 8 bits
using something gamma-like curve, i.e. mapping function, changing the curve
steepness as needed for my image, for example, increasing it in lower values
range if I want to show shadow details.

...

Remember, KDE's KRITA uses ImageMagic for a reason; ImageMagic supports up
to floating point number per color per pixel IIRC.

Regards,
  Sergei.

Applications From Scratch: http://appsfromscratch.berlios.de/


       
____________________________________________________________________________________
Need a vacation? Get great deals
to amazing places on Yahoo! Travel.
http://travel.yahoo.com/