Re: Untagged RGB to Lab conversion
Re: Untagged RGB to Lab conversion
- Subject: Re: Untagged RGB to Lab conversion
- From: Steve Upton <email@hidden>
- Date: Wed, 19 Dec 2001 00:47:20 -0800
At 9:35 PM -0500 12/17/01, David Remington wrote:
This topic has been beat around here a few times but I still have
questions. I'm going to try re-phrasing it a few different ways to
see what the group can come up with. I'm currently evaluating
PhaseOne and Better Light scanning back cameras for an upcoming
purchase. As a part of this process I would like to compare capture
values of various targets (Macbeth regular and DC, Kodak
Q-60...etc) to target definition file measured values. Measured
values are ICC Lab. Captured values are RGB. How can these values be
meaningfully compared?
without a comparison table, they cannot. One set is numbers applied
to human vision, the other is numbers applied to machine vision.
For that matter, and this is a question no doubt best directed to
PhaseOne and Better Light, how are RAW RGB values determined? Sensor
R,G, and B filters as primaries plus secret sauce?
probably. The raw stuff you get off a chip is probably not very
useful and needs to be expanded in to the 0-255 range we use for RGB
Is there any way to convert from RGB to Lab without defining the
RGB colorspace?
nope, that would be, by definition, a profile. So yes... I guess....
using a profile.
Having said this I should add that I'm fully capable of seeing the
circular logic involved. If you could make an ideal conversion to
Lab from your raw camera data you would have no need of a camera
profile.
well..... I would say that if you could make an ideal conversion to
Lab from the raw camera data you would have the ideal camera profile.
This is not a matter of converting from English to English it is
converting from binary to English (for example). They are not the
same thing, really, the comparison table between them defines the
relationship completely.
What I would like to investigate is just where that raw data is
starting from. Is it starting anywhere? Is RGB data completely
defined by the box you draw around it?
I think it is fair to say that the RGB data in the camera is defined
by how the camera manufacturer chooses to map it out to RGB values.
The box that results numerically is defined by the 0-255,0-255,0-255
range of values and the box that results colorimetrically is defined
by the profile.
These are good questions and tough to get one's head around I think....
Regards,
Steve
________________________________________________________________________
o Steve Upton CHROMiX www.chromix.com
o (hueman) 866.CHROMiX
o email@hidden 206.985.6837
________________________________________________________________________
--