[argyllcms] Re: Scanin Problem using -F option

The reason I manually added the fiducial line to the cht file was because I 
originally created the postscript file for the target and the cht file related 
to that target using V0.6 of argyll. I didn't realize that this was 
automatically created in V0.7Beta8, and since then I have created a new target 
that has the fiducials and run into the same problem.

I went ahead and attached two diagnostic images. I had to compress them quite a 
bit to fit the attachment size requirement.
1. diag_auto.jpg is the diagnostic image that is created when I allow argyll to 
automatically recognize the orientation of the target. The ti3 file that was 
created when I ran the scanin command to create this diagnostic image works 
fine and generates a working icc profile through the profile command.
2. diag_manual_fids.jpg is the diagnostic image that is created when I used the 
-F option with the scanin command. Obviously, this does not produce a working 
profile. However, it seems to have found the target patch locations correctly.

I also went ahead and attached the data file and cht file that I used to create 
the profile.

I figured I would outline the exact procedure that I use to create the profile 
so that you can tell me if I'm doing something wrong.
To create the postscript target and the cht file I issue the following two 
commands:
targen -v -d3 -f999 full
printtarg -v -iSS -s -pLetter full

I then send the postscript page to a printer and collect the data from the 
target using a Gretag-MacBeth Spectrolino. Once I have the data I issue the 
scanin and profile commands.

If I want argyll to automatically recognize the orientation of the target I use 
(this generates the diagnostic image diag_auto.jpg that I attached):
scanin -v -dipnoa full.tif full.cht full.data

If I want to manually specify the fiducial locations I issue the following 
command (this generates diag_manual_fids.jpg which is attached as well):
scanin -v -F 72,80,2464,72,2472,3236 -dipnoa full.tif full.cht full.data

If I have a working ti3 file i would create the profile using the profile 
command.

Thanks,
Andrew Asman
________________________________
From: argyllcms-bounce@xxxxxxxxxxxxx [argyllcms-bounce@xxxxxxxxxxxxx] On Behalf 
Of Graeme Gill [graeme@xxxxxxxxxxxxx]
Sent: Tuesday, June 10, 2008 11:54 PM
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Re: Scanin Problem using -F option

Asman, Andrew J wrote:

> In order to be able to use the -F option with the scanin command I added the 
> following line to
> the cht file immediately after the "BOXES 999" line:
>
> F _ _ _ _ 45.354331 38.834646 591.023622 38.834646 591.023622 773.007874

> Generating a profile using a ti3 file like the one above works flawlessly. 
> However, when I use
> the -F option the RGB_* and STDEV_* fields in the ti3 file are all set to 0, 
> like this:
>
> ... BEGIN_DATA_FORMAT SAMPLE_ID XYZ_X XYZ_Y XYZ_Z RGB_R RGB_G RGB_B STDEV_R 
> STDEV_G STDEV_B
> END_DATA_FORMAT
>
> NUMBER_OF_SETS 999 BEGIN_DATA AA37 14.390 20.860 6.6900 0.0000 0.0000 0.0000 
> 0.0000 0.0000 0.0000
>  AA36 38.240 33.610 5.0400 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 AA35 
> 3.9700 11.030 4.1500
> 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 AA34 4.2700 3.8500 4.5900 0.0000 
> 0.0000 0.0000 0.0000
> 0.0000 0.0000 AA33 10.940 22.410 7.5400 0.0000 0.0000 0.0000 0.0000 0.0000 
> 0.0000 ... END_DATA

The symptoms are consistent with the raster being completely blank (R=G=B=0).
If the patches are scanned completely off raster, then they will end up
with nonsense values (-2). It's a puzzle that you say the diagnostic looks ok.

Without having all the files to reproduce the problem, it's hard to say much 
more.

Is there a reason you manually added the fiducial line to the .cht file ?
printarg does this automatically in V0.7 Beta 8.

Graeme Gill.

Attachment: diag_auto.jpg
Description: diag_auto.jpg

Attachment: diag_manual_fids.jpg
Description: diag_manual_fids.jpg

Other related posts: