Hi Graeme,
thanks for your clarification.
So everything is fine. The results look really good. I was just wondering...
Hermann-Josef
--- Begin Message ---Hermann-Josef Röser wrote:
- From: "Graeme Gill" <graeme@xxxxxxxxxxxxx>
- To: <argyllcms@xxxxxxxxxxxxx>
- Date: Thu, 11 Jun 2020 15:57:58 +0200
Sorry, I forgot to say that this was the scanner target for Ektachrome Ihad
create with Argyll as a shaper/matrix profile.
Right. Given the nature of input devices, this is not at all unusual.
Consider the standard observer - it has primaries well outside
the spectrum locus.
An input device can be arranged so that real world colors
correspond to almost any reasonable device value, and typically
these device values are not at the limits of the device range.
So extrapolating device limit values back into XYZ space
will mostly land outside the spectrum locus.
Or to put it another way, if an input device is not to be artificially
limited in its gamut, then by design its primary triangle has to
lie outside the spectrum locus in order to encompass it.
Cheers,
Graeme Gill.
--- End Message ---