Agreed, which is why I also noted there could be problems in the work flow. It's easy to confound the work flow with the profile's job of producing scene-referred images (that aren't going to look rather flat on a typical display). Delta E can tell us if the profile is performing OK. The rest depends on the capabilities of the software and if the person knows how to use the software effectively. ST. ________________________________ From: Iliah Borg <ib@xxxxxxxxxxx> To: argyllcms@xxxxxxxxxxxxx Sent: Sunday, 29 July 2012 11:43 PM Subject: [argyllcms] Re: help with camera profile > Visual evaluations are not satisfactory because there are dozens of variables > that affect whether or not the photo looks "good". You can review many > photos, some will be good and some will look wrong. Good quality profile does not make a photo to look good (most customers do not want to put a boring reproduction of Valley of Fire onto their walls, we need to make it more vivid, contrasty, and saturated to deliver the impression the market wants), it makes it a reproduction. If that is the goal, I can understand the struggle for perfect profile. Anything but reproduction goes through colour and tone editing anyway, and quite often it is beneficial to have correctional profiles (Kodak used to call those "looks") - like product, portrait, landscape; or film simulations. Such profiles make for a good starting point for further artistic work on the image. Speaking of camera profiles, another problem is gamut mapping. It is not uncommon to have sensors with chromaticities that exceed Lab and simultaneously have about 11 stops of dynamic range. Mapping say a Time Square night shot or even a Diner in the evening underexposed to preserve neon but still keeping nice details in shadows to some sane RGB or CMYK colour space needs a lot of manual intervention. Same about many sunrise/sunset photos, or generally combinations of direct and reflected light. Ironically, those are pretty straightforward when shot on film. -- Iliah Borg ib@xxxxxxxxxxx