> Two images with different color spaces appropriately processed for web
> display should not vary significantly in saturation or colour cast.
> Also when making any type of semi-critical colour comparison it really
> does help matters to have be using a similar or the same subject in
> the same light.
> 
> Firstly there is little point in displaying images on the web in other
> than sRGB colour space as at least 99% of web browsers (anyone can
> make up stats :-) are not colour space aware so will not render the
> images correctly. Secondly most images presented for web display have
> the colour tag information stripped so even a colour space aware
> Browser will generally assume sRGB rendering (depending its colour
> space management policies)
> 
> Just to highlight my point in the case of your two images only the
> AdobeRGB image has an embedded color space tag.
> 
> Viewers should only be able to potentially "see" a difference between
> (like) images when they are displayed using a media type that has
> colour gamut than can take advantage of the extra colour information.
> A typical video display (which typically approximates the sRGB gamut)
> isn't a good tool on which to make comparisons or draw conclusions.
> 
> All that said if you intend your final output to be to print and the
> printer can take advantage of the AdobeRGB colour space then using
> AdobeRGB probably isn't a bad idea but you need to really shouldn't be
> seeing a difference on screen if your editing work flow and image
> colour space handling is being managed correctly.
> 
        Well-said.  As a matter of fact, viewing two images (one in sRGB, 
one in aRGB) on an uncalibrated display will in general make the aRGB one 
look more saturated.  The smaller gamut available in the sRGB space will 
mean that for a given highly saturated absolute color in an image, it will 
be digitally encoded "wider" in the narrower colorspace.  

        Of course shooting RAW negates the requirement to make the choise 
at shoot-time.... it's all done in post-processing.  Strictly speaking, 
the white-point (and rest of the lighting spectrum) is required to 
accurately incorporate color management in the RAW conversion process, but 
the default ones are generally pretty close.  On the *export* to something 
else (like web, a color printer, sending off to a lab to have prints made, 
etc) the general concensus is that unless you're very clear about the 
remainder of the processing to be done, you should probably stick with 
sRGB.  Unless you calibrate your own printer, or use a lab that will give 
you a profile for their output or specifically states they use something 
wide (like ProRGB), chances are they use sRGB.

-Cory

-- 

*************************************************************************
* Cory Papenfuss, Ph.D., PPSEL-IA                                       *
* Electrical Engineering                                                *
* Virginia Polytechnic Institute and State University                   *
*************************************************************************


-- 
PDML Pentax-Discuss Mail List
[email protected]
http://pdml.net/mailman/listinfo/pdml_pdml.net

Reply via email to