Hi,

>Does a LMCS generated perceptual intent profile rescale to the DMin/DMax 
>of the reference file or to that of the target scan measurements?

Beta 3 of profiler does scale DMin to white point, and leave DMax alone. Next 
version will handle black point as well, probably with some sort of black point 
compensation. 

The difference can easily be seen, perceptual intent does look "nicer" at a small 
rise of dE cost, whilst relative colorimetric gives better numbers. 

>I have seen it recommended to scan an IT8 target at an exposure that 
>results in values of 250 for DMin. In my scans DMin is approx 238. Is 
>that relevant at all?

Right. If your DMin results in 255, then white is at gamut boundary and probably 
clipped. The idea is to get closer to maximum  without reaching it. 238 would give 
you some extra levels for highlights, so is not bad. 

Note that in such case difference between perceptual and relative colorimetric is
very evident. In perceptual, 238 will result on Lab = (100, 0, 0), on relative 
colorimetric will be the Lab in target reference (about L*=90).

>Another question: When using a scanner profile with either perceptual or 
>relative colorimetric intent, does it matter what rendering intent is 
>set in the application? My understanding is that the CMM will use the 
>intent of the profile unless the profile contains the tag for the intent 
>specified by the application.

In this particular case, doesn't matter because v2 input profiles can contain
just one intent. Only v4 input profiles can effectively hold several intents, so, 
since the profiles generated by lcms profiler are still v2, you need two profiles 
if you want perceptual *and* relative colorimetric. This was a limitation of ICC 
spec which I  must follow in order to make compatible profiles.

Regards,
Marti.


----- Original Message ----- 
From: "T E Schmitz" <[EMAIL PROTECTED]>
To: "Marti Maria" <[EMAIL PROTECTED]>; "lms mailing list" <[EMAIL PROTECTED]>
Sent: Wednesday, February 18, 2004 1:42 PM
Subject: Re: [Lcms-user] Do DMin/DMax limit the scanner's dynamic range?



Does a LMCS generated perceptual intent profile rescale to the DMin/DMax 
of the reference file or to that of the target scan measurements?

I have seen it recommended to scan an IT8 target at an exposure that 
results in values of 250 for DMin. In my scans DMin is approx 238. Is 
that relevant at all?

Another question: When using a scanner profile with either perceptual or 
relative colorimetric intent, does it matter what rendering intent is 
set in the application? My understanding is that the CMM will use the 
intent of the profile unless the profile contains the tag for the intent 
specified by the application.


Regards,

Tarlika Elisabeth Schmitz


Marti Maria wrote:
> Hi,
> 
> This is up to profile builder and the intent. For example, in the lcms profiler, 
> perceptual 
> intent does rescale dynamic range to DMin/Dmax. Relative colorimetric does not.
> 
> In scanners such rescaling would be desirable, but not in digital cameras .
> 
> Regards,
> Marti Maria
> The little cms project
> http://www.littlecms.com
> [EMAIL PROTECTED]
> 
> ----- Original Message ----- 
> From: "T E Schmitz" <[EMAIL PROTECTED]>
> To: "lms mailing list" <[EMAIL PROTECTED]>
> Sent: Tuesday, February 17, 2004 11:33 AM
> Subject: [Lcms-user] Do DMin/DMax limit the scanner's dynamic range?
> 
> 
> Hello,
> I have come across the following in various places: The luminance of 
> DMin and DMax on an IT8 target will cause any whiter white and blacker 
> black to be clipped. In other words: if using a profiled scanner, the 
> dynamic range will be limited to that of the IT8 target. Is this 
> correct? If so, does this apply to all profile types?



-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Lcms-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/lcms-user

Reply via email to