Hello Nikolay, Correct. I search for soft proof functionality, which goes throuh three (or some times more) ICC profiles. The first profile A is the document colour space, the second one B is the to be simulated device space and the last one C is the proofing device, a monitor or a ink jet. .When you write about gamut mapping, I assume you mean each of the conversions from device A to the to be simulated device B and from B to C. My assumtion ist, that each separate ICC device link introduces new precission loss compared to a CMM computing in one go throuh all stages with a higher internat precission. Indeed I need DevA -> PCS -> SimulatedB -> PCS -> DevC. One expectedly loosy way is to separate into two device links and combine them to a final DL. kind regards Kai-Uwe Nikolay Pokhilchenko <nikolay_po@xxxxxxx> schrieb: 20 Sept. 2013, 20:59 +02:00 Kai-Uwe Behrmann wrote: I want to create a device link profile starting from colour space A converting into colour space C while simulating colour space B. One single device link profile shall be the result. Hellow Kai-Uwe. When You speaking about different color spaces, You're assuming that they are device spaces, not PCS. Right? As far as I know the device link represents the mapping between two DEVICE spaces, not PCS. collink allows to add a abstract profile to the conversion. Can a simulation profile of type abstract be created with argyll tools? I think Yes. But this may be not so straight and simple. If I find some free time I'll try to find the sequence. The desired conversion inside the abstract profile would be: PCS -> simulation space -> PCS In general, when the colors are in gamut and gamut mapping wasn't applied while conversion to the simulation space, such abstract profile is meaningless. It will be just 1:1 dummy conversion. So, may be You're speaking about the conversion lilke: DevA -> PCS -> Simulated_DevB -> PCS -> DevC ? If You really need PCS -> simulation space -> PCS then You should include in this patch the conversions that are applied to source image. The gamut mapping and (or) gamut clipping. Can You provide the files (ArgyllCMS compliant) that describes Your devices to experiment with? -=- Nikolay Pokhilchenko.