Re: Can DeviceLink conversions be better?
Re: Can DeviceLink conversions be better?
- Subject: Re: Can DeviceLink conversions be better?
- From: Klaus Karcher <email@hidden>
- Date: Tue, 02 Sep 2008 04:05:51 +0200
Iliah Borg wrote:
Dear Klaus,
I'm not sure what all this is about. I print with device link profiles
all the time, for many years, and I'm getting much better results then
when printing with regular profiles. More so when repurposing CMYKs
between USA and Euro, which I do routinely. But even RGB to RGB device
links are worth studying rather then simple waving out. I use those not
only with printers, but with raw conversion process when the client
wants RGB files.
Dear Marco,
ColorThink Pro is capable of generating device link profiles, though the
quality might be not as good as with DeviL.
Dear Iliah,
I don't want to belittle the advantages of device link profiles, they
are beyond question. The main advantages are improved smoothness and
precession (as they are not bound to the PCS grid) and the option to
account for the separation of the source image during the transformation.
Clever device link profiling tools will also benefit from the fact that
both the source and destination gamut are known at the time the device
link will be created: there is no need to assume a generic source gamut
(like the PRMG) -- the gamut mapping can be optimized to the specific
combination of source- and target gamuts.
The advantage of device profiles is greater flexibility: you can link
every device profile with any other, but you separate device link
profiles for every combination of two devices: to transform between n
colorspaces in any direction, you need only n device profiles, but
n*(n-1) device link profiles.
You will never get optimal results and unlimited flexibility at the same
time using device profiles with pre-calculated gamut mappings.
There is no principal difference in gamut mapping capabilities between
device- and device link profiles: Argyll's source- and image-dependent
gamut mapping for example works with both device- and device link
profiles (whereas device profiles forfeit their flexibility). As image-
or source-dependent gamut mapping can not preserve the flexibility of
device profiles anyway, there are few reasons to use them in this
context and to renounce the advantages device link profiles can offer.
What makes the main difference between the examples provided by Rolf is
IMHO the gamut mapping. I'm pretty sure that the device link profile's
gamut mapping is better, but this has nothing to do with device or
device link profiles in general. And to compare the results, one should
use appropriate methods.
Klaus
_______________________________________________
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