[argyllcms] Re: targen problem when used with a pre-conditioning profile
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Wed, 08 Jun 2011 09:35:24 +1000
Dave Karipides wrote:
I am having difficulty with targen when used with a pre-conditioning
profile. The command I am using is as follows:
Depending on the value I use for the adaptation parameter (-A), I get
odd / incorrect output. In some cases the program seems to output
believable results. In other cases, it returns only a small number of
the requested 1380 patches or sometimes seems to spin in a loop.
I would appreciate any ideas on what may be happening here. If desired,
I can provide the pre-conditioning profile I am using. In the past, I
have used this initial/final profiling technique and have had excellent
results. I would like to understand what may be going on before I take
the time and effort to fine-tune the system.
Hi,
the algorithm used isn't actually guaranteed to work. That's the
nature of the geometry, particularly if the space is warped by the
pre-conditioning
profile. The best I can do is use heuristics to overcome the inherent problems.
To investigate any further, you need to send me the profile you were using
(please mail it to me or give me a URL I can download it from).
General things you can try are:
See how it behaves without the profile. Do you still have problems ?
Simplify the profile. You can do this by choosing and lower
grid resolution (quality setting), and/or increasing the smoothing
factors (-r parameter).
Graeme Gill.
Other related posts:
