[Radiance-general] ICC profiles

[iebele]

Hi again,

I'm still thinking about the data provide by ICC files. Somehow I think 
it should be possible to calculate the top of the luminance range for 
the target output device (-u option of normtiff), given a set of 
photometric measurements which I believe an ICC profile provides. I 
tried to figure out _what_ precisely is measured in the "SpectroScan"  
section of the ICC file (attached), but even the support team of Epson 
could not answer that question. Maybe one of you does?

Also I wonder how to calculate the 'dynamic range for the target output 
device ' (-d option of normtiff) This is said to be 32 for CRT monitors 
in the manpage. I have no clue how to obatin this value for other output 
devices.

The only option I know for sure in normtiff is the gamma option, which 
it adjustable for Epson printers (1.5, 1.8, 2.0, 2.2 ).

Today I spent some time reading High Dynamic Range Imaging. Although 
most of the math is just magic for me, I understand the issue. I wonder 
however how much this issue is known by photo-labs and printing 
services. I wonder also how much of this knowledge is implemented in 
contemporary printing-technology. I don't know, if one of you does, 
please tell me. The funny thing is that operators say "O wow, 
32bit/channel images, great, no problem!" But in fact they convert the 
file to the printer's actual output, which  is 8 bit/channel. This is a 
kind of a communication problem, which I believe has to be solved at my 
side. I really like to be able to provide the photo-lab 'just' a 
standard  8bit/channel image. I understand such an image can not look as 
perfect in print as it does in RGB monitor color space. But can we come 
close to 'perfect' using proper calculations?

-Iebele

iebele schreef:
> Dear group,
>
> I have a tone-mapping problem with printing hdr files, which are 
> converted to 8-bit Photoshop-RGB colorspace.
> In the print some clipping occurs.  To solve the clipping I was 
> thinking to use normtiff  or pcond in combination with a Photoshop ICC 
> profile (.icm files), which I a received from the printer operator. 
> Regretably I don't understand the meaning of the parameters of 
> normtiff and pcond enough, not even enough to get started . Does 
> anybody have an idea if and how the data from ICC profiles can be used 
> by normtiff?
> The lines below are copy/paste from an .icm file, and i wonder if 
> these can be of any use in tonemapping. Are the the tabular organized 
> XYZ_X   XYZ_Y   XYZ_Z valuable for this kind of application?
>
> Any hints are very much appreciated,
>
> Iebele
>
>
>
> CREATED    "1/12/2007"  # Time: 16:17
> INSTRUMENTATION    "SpectroScan"
> MEASUREMENT_SOURCE    "Illumination=D50    ObserverAngle=2¡    
> WhiteBase=Abs    Filter=Unknown"
> KEYWORD    "SampleID"
> KEYWORD    "SAMPLE_NAME"
> NUMBER_OF_FIELDS    8
> BEGIN_DATA_FORMAT
> SampleID    SAMPLE_NAME    XYZ_X    XYZ_Y    XYZ_Z    LAB_L    
> LAB_A    LAB_B
> END_DATA_FORMAT
> NUMBER_OF_SETS    918
> BEGIN_DATA
> 1    A1    2.48    2.52    1.96    18.00    1.10    1.12
> 2    A2    3.04    3.14    2.53    20.61    0.15    0.53
> 3    A3    3.64    4.17    3.21    24.23    -5.73    1.61
> 4    A4    4.11    5.21    3.74    27.32    -12.11    3.35
> 5    A5    4.73    6.79    4.55    31.32    -20.99    5.45
> 6    A6    5.26    8.53    5.03    35.06    -30.46    9.31
> etc...
>
>
> _______________________________________________
> Radiance-general mailing list
> Radiance-general at radiance-online.org
> http://www.radiance-online.org/mailman/listinfo/radiance-general
>
>