Re: [CinePaint] masks; detail 8 vs. 16+ bits

I've posted three images to these locations:

http://www.colorist.org/images/DSC_1180.tiff
    16-bit TIFF OpenEXR Half Float - Rgb - Wide Gamut D65
     (saved from Cinepaint, no adjustments)

http://www.colorist.org/images/DSC_1180_gimp.jpg
     8-bit jpg saved from The Gimp, nothing adjusted.

http://www.colorist.org/images/DSC_1180_lightroom.tiff
    16 bit (?) file saved from Adobe Lightroom, imported from Nikon
     camera same as above files.

I'm uploading them now, should be about 20 minutes from now they'll be  
available.

Acutance; we have had quite a few discussions about this on the TIG,  
as well as stochastic resonance, and these are fascinating subjects.    
Any feedback about what is happening in these images presented is very  
welcome.

Rob


On Mar 12, 2008, at 10:13 PM, Frank Peters wrote:

>>
>> Incidentally, yesterday I shot a still in Nikon's NEF (RAW) mode and
>> opened it in Cinepaint and in The Gimp.   The difference in fine
>> detail between The Gimp and Cinepaint was not small.   Since  
>> luminance
>> is the greatest contributor to fine detail, I was at first skeptical
>> that the improved detail in Cinepaint was due to bit depth, but then
>> upon further thinking I believe the luminance is derived from RGB, so
>> bit depth is just as important for luminance.  Am I right in this
>> assumption?
>>
>
> Fine detail depends on acutance, or the difference in contrast
> between closely spaced points.  Only the relative and not absolute
> luminosity would play a role here.  Post-processing can have an
> effect on acutance but it usually does not, so I would doubt
> that the difference in fine detail that you observe is due to
> a greater bit depth.  Also keep in mind that both 8-bit and 16-bit
> images are being viewed through an 8-bit monitor.
>
> Try this:  Open the RAW file in Cinepaint and then reduce the bit
> depth to 8 bits with the Image Menu.  Does the fine detail
> suddenly degrade or disappear?
>
> It would be useful to post the actual RAW file somewhere file so
> that other list members could experiment.  I could be all wrong, but
> without the actual RAW file only the most general of assumptions
> can be made.


--
Rob Lingelbach
[email protected]  http://www.colorist.org/robhome.html




-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/