New X-Rite Software announced - and upgrade eligibility (or lack thereof)

[Iliah]

> But in no way are the older profile engines in any way responsible for the dark print issue which is mostly user error in terms of display calibration and/or print viewing conditions.

Here is the question I'm asked many times: why I can set my monitor to 1:200 contrast ratio and make a good representation of the 1:1000 scene I shot; but to match a 1:120 print viewed in a Normlicht booth I still need to set my monitor to 1:120 or even lower. I'm told by far too many seasoned photographers that current printer profiling does not match user expectations, and those expectations are not baseless.

[BartvanderWolf]

Here is the question I'm asked many times: why I can set my monitor to 1:200 contrast ratio and make a good representation of the 1:1000 scene I shot;

Hi Iliah,

Good question, but I guess it comes down to "a good repersentation". How is that defined? It's probably largely subjective.

[QUOTE... but to match a 1:120 print viewed in a Normlicht booth I still need to set my monitor to 1:120 or even lower.[/QUOTE]

Well, to me that makes sense. One needs to set the monitor to something similar as the goal to allow a predictable preview. Therefore, I think the issue lies in the appreciation of "good" in the high contrast case. That's where the human visual system comes in. When viewed on a large enough screen, our eyes will automatically compensate for local contrast differences. When the differences are large, then there is more compensation. However, when the differences are small, it's going to remain pretty much as it is.

I'm told by far too many seasoned photographers that current printer profiling does not match user expectations, and those expectations are not baseless.

When the mismatch is not based on physics (as measured with a spectro, and corrected with a profile, and viewed in a calibrated environment), it must be psychological/vison related.

Cheers,
Bart

[Iliah]

Good presentation is when the paying client considers it good. Let's talk b/w film for a moment. Film in the camera captures a routine landscape with 10 stops contrast ratio, in a more or less (log-log) linear manner. Next, it is printed on a photopaper, with its 6 stops contrast ratio, less than for of which are linear while the other part is toe and shoulder. The system is balanced in the way that it is easy to use. All necessary perceptual compressions are in place. Profiles are concerned not with the physiological, but with colorimetric presentation. That's where colour management fails in the eyes of many end-users.

[BartvanderWolf]

The system is balanced in the way that it is easy to use. All necessary perceptual compressions are in place. Profiles are concerned not with the physiological, but with colorimetric presentation. That's where colour management fails in the eyes of many end-users.

Hi Iliah,

I agree with the assessment, but not with the conclusion we should draw from that. It may fail in the eyes of many end-users, but perhaps they are blaming the wrong party ... It's not the tool, it's the photographer!

I'll give an example, imagine a scene with a huge dynamic range, and a capture medium that's capable of recording it without compromises. Now, we output that huge range, but with a simple level adjustment to squeeze the huge dynamic range into what's possible on the output medium while preserving shadow and highlight info. We end up with a dull low contrast image. It's not the profile that's to blame. The reason it doesn't look good is because we ignored the perceptual aspects. What should have been done is proper tonemapping (which is scene dependent and cannot be caught in a standard physical recipe).

Another example. We know that on non-glossy paper, we need to increase shadow contrast, yet the measurable physical correction by a profile is not adequate. That's where the photographer/printer can make a difference, IMHO of course.

Cheers,
Bart

[Iliah]

Same photographer who can successfully use a traditional (legacy, as some try to put it) darkroom, who understands grades, compression, sensitometry lacks those convenient and proven tools assuring a systematic approach in a digital darkroom. Have you ever compared a characteristic curve of photopaper and, say, inkjet paper? Does any profiling tool really takes care of the issue? Why it is so difficult to realise users have certain expectations that need to be met? Why the regress?

In regards to the first example you suggested - I'm afraid I do not know what level move you mean, but in photography tonal compression starts with contrast, or grade. Based on the particular neg density range that constitutes the scene to be printed and midtone placement we decide on the paper grade and exposure. It is a standard recipe.

Same for the second example - knowing paper characteristic curve and density range / midtone placement on the negative all things are banal.

We had a system that was developed for perception. Now the system is broken.

[Wayne Fox]

I don't think color management was designed nor intended to address the problems of compressing real world data to be pleasing to the eye, handling compression of toe/shoulder, etc.  That's the job of other parts of the system (and yes camera profiles can do this do a degree).  Color management is intended to make device translations to achieve similar results, no matter if that information is good or bad.  While profile's can be used in a way to handle this, it's not it's purpose and this makes the profiles extremely circumstance specific.

While there are certainly many aspects that could be improved (like a standard that could be implemented directly in all hardware without user intervention), it's not really that broken.

[Iliah]

> I don't think color management was designed nor intended to address the problems of compressing real world data  to be pleasing to the eye

No, not to be pleasing to the eye, but to be printable and displayable. And yes, it was not designed to do it - so, it needs redesign.

> yes camera profiles can do this do a degree

Absolutely not. To do it (even to a degree) the characteristic curves of both input and output must be taken into account.

> Color management is intended to make device translations to achieve similar results

Exactly. And same perception between two devices is about similar results.