Re: RV: Camera Profiles... was : Camera profiles and reference filecorrection
Re: RV: Camera Profiles... was : Camera profiles and reference filecorrection
- Subject: Re: RV: Camera Profiles... was : Camera profiles and reference filecorrection
- From: Clemens Beisch <email@hidden>
- Date: Tue, 6 Mar 2007 10:54:15 +0100
Not only a good idea - it works ;-)
we are just using the bayer function internally to convert camera
mosaic data
to an RGB image. The channels are multiplied so they are in balance.
There is no color matrix conversion or any other kind of profile
before reading
out the values for the profile creation.
Very important informations are stored in the meta data of the RAW
files.
For example exposure and aperture that we can use to calculate
the density of light - I hope you can understand my lousy English :-(
This is importent to set CIECAM02 parameters.
The next thing is the information of the illuminant.
Once again we use the stored information inside the meta data.
I know - this is neither perfect nor absolutly precise but it much
closer the
the default assumption of D50, D65 or what erver.
These are just a few examples of what's going on in our profiling
software.
So in a nutshell we have:
1. Image data without color manipulation
2. Reference data that best possible fitts the illuminant
3. a powerfull appearance model - CIECAM02 and last but not least
4. about 17 years colormanagement know how
To make the things as easy and confortable as possible for the user
our scientist and director of software development Gregory K. Remington
has develloped a very fast and precise pattern recognition to find
our color chart anywhere in the image. It dosn't matter if it is
rotated or
mirrored. We just need about 20x20 pixels per color patch (quality
reasons)
and this is not a problem with modern digital cameras. Also we
deliver our
charts in different sizes.
We are working very heavy to get a special solution so every body can
try it world-wide without having to buy it first.
And I can say - we are on a good way - but more later this year ;-)
Best regards,
Clemens Beisch
Am 06.03.2007 um 09:05 schrieb Xabier Urien:
Our ColorCharts are the optimal targets of profiling on location
because they have special
marks (fiducial points), that makes possible a fully automaticly
chart detection.
Our software ColorCruiser ist specialized for the computation of high
grade camera profiles.
it not only uses TIFF and JPEG as input, rather you can use directly
the RAW-files of most
common models.
Sorry Clemens... good idea to profile the Raw files directly.. but,
how can
you do that?? if im not wrong the raw file doesn't have any RGB
value, just
light brightness for each pixel (more or less)... then, how can you
profile
that???
Im really interested if it works, becouse I never trust in the camera
profiles.. until I know and correct me if Im wrong, you have to
profile a
profiled image.... I mean... when you use a TIFF or JPEG file these
images
are already "profiled"..
Regards and sorry my english...
Xabier Urien
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
40calibr8.com
This email sent to email@hidden
***********************************************************
Clemens Beisch - Calibr8 GmbH
Founder & Technical Director
email@hidden
www.calibr8.com
***********************************************************
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden