[Marc Sabatella]

I would contrarily argue that colour is defined by wavelength of the light reflected from an object, which is very much measurable.

Well, sure, but that's not ow we *perceive* color, and there's nothing at all metaphysical about it. It's why cameras have adjustable WB settings. If you want your camera to capture color exactly as a colorimeter would measure it, set your WB to flash and shoot for a while in a variety of conditions. See how closely that matches your perceptions.

In theory there is no problem with the idea of exactly duplicate real life colours with the right, calibrated tools.

True, and it would be possible to measure different RAW processing programs that way. But you've never done this - you've only compared to your own subjective impressions.

I can appreciate your idea that we might see colour differently, although I haven't seen any science that seem to confirm it.

My point about WB is common knowledge. See, for instance, Color vision - Wikipedia, the free encyclopedia, and in particular, Color vision - Wikipedia, the free encyclopedia.

The thing I don't agree with is that I and other I have asked to look at the difference shouldn't be able to see the difference in colours in between calibrated screen and real life.

You mean you've set up a still life right next to your computer and are directly comparing the still life against photogs of that that still life as interpreted by different RAW processing applications? Even if so, you're still seeing the colors of the still life through the "lens" of your brain's "wb" function, meaning it's subjective. More accurate would be to get a color-managed *print*, and view that print in the same light as the still life itself, so at least you're removing the WB from the equation. But that's still going to be misleading, because the range and distribution of values will differ between the actual scene and the print (as they also do between the scene and the monitor).

But I think we are getting a bit off topic here. To me this just show how different the priorities in what is needed in a RAW converter can be.

True, but it also suggest that unless you've really done extensive tests to remove the variables that influence your perceptions, what you see may not not really be what you get.

[wowtip]

You mean you've set up a still life right next to your computer and are directly comparing the still life against photogs of that that still life as interpreted by different RAW processing applications? Even if so, you're still seeing the colors of the still life through the "lens" of your brain's "wb" function, meaning it's subjective. More accurate would be to get a color-managed *print*, and view that print in the same light as the still life itself, so at least you're removing the WB from the equation. But that's still going to be misleading, because the range and distribution of values will differ between the actual scene and the print (as they also do between the scene and the monitor).

Yes, yes I know in theory there are many parameters that might affect what we (think we) see. But in my experience the difference in colour between RAW converters handling PEF files is not close to subtle.

I took some time compiling an example just to show what I mean, and why I find "Pentax" colours in other apps problematic.

Three images in the following order...

ACR5 default conversion - PentaxPhotoLab - ACR5 Tweaked conversion



I am as sure as can be that the colour of her pants in cloudy outdoor light is about the hue you see (on a calibrated screen) in the PPL version and the tweaked ACR5 conversion. They are pink with a hint of blue, not salmon coloured as in the ACR5 default conversion.

So, the ACR5 with tweaked colours looks pretty much like what I want, almost like the PPL conversion and ACR5 is much more user friendly. Why not use it?

The thing is, to get those colours with ACR5, I had to tweak the colour balance in the image for 15 - 20 minutes. To me that is too much time to spend when you have lots of files to process, with PPL they are acceptable straight out of the box. Most of the time you only need some level / contrast adjustment end then you are done. I know you can save colour profiles in ACR5, so in theory I should be able to save the adjustments I did on the image above and use on all images (set as default). Well, not in my experience, the profile used on the image above will cause severe colour artifacts if used on some other PEFs, with others it will work just fine.

But if someone know how to save a colour only profile or has a nice profile for ACR5 that give "Pentax colours" I would be happy as happy can be.

[grayboy]

wow and all this is just in the beginners corner - cant wait to get to the experienced users forum!!

[audiobomber]

Three images in the following order...

ACR5 default conversion - PentaxPhotoLab - ACR5 Tweaked conversion

Can't see the images.

[wowtip]

The images were there for a while, then they disappered? I tried to edit them back in, but that didn't work. Anyway, repost...

ACR5 default conversion - PentaxPhotoLab - ACR5 Tweaked conversion

Ok, for some reason the images disappear after I post them, they work when I post, then are gone on reload???

straight urls..

ACR5 default conversion
http://joj.smugmug.com/photos/953722196_8A345-L.jpg

PentaxPhotoLab
http://joj.smugmug.com/photos/953722751_o7V4z-L.jpg

ACR5 Tweaked conversion
http://joj.smugmug.com/photos/953721843_67CM2-L.jpg

[mattdm]

Well, sure, but that's not ow we *perceive* color, and there's nothing at all metaphysical about it.

I was just recommending my favorite author on the technical aspects photography in another thread, and with that in mind, I'd highly suggest Mastering Color Digital Photography to anyone interested in a very well-done overview of how color works in photography, from the technical to the physiological to the psychological.

Unfortunately this one is put out by Lark Press, who tends to do a sloppy job with editing, but very good overall even so.

[JeffJS]

The images were there for a while, then they disappered? I tried to edit them back in, but that didn't work. Anyway, repost...

ACR5 default conversion - PentaxPhotoLab - ACR5 Tweaked conversion

Ok, for some reason the images disappear after I post them, they work when I post, then are gone on reload???

straight urls..

ACR5 default conversion
http://joj.smugmug.com/photos/953722196_8A345-L.jpg

PentaxPhotoLab
http://joj.smugmug.com/photos/953722751_o7V4z-L.jpg

ACR5 Tweaked conversion
http://joj.smugmug.com/photos/953721843_67CM2-L.jpg

They don't show on your straight links either.

[wowtip]

They don't show on your straight links either.

Something is wonky with either PentaxForums or Smugmug. Posted them exactly like I always have, never any problem until now. Anyway, you should be able to see them by opening the link, deleting a single character in the URL, writing the same character back into the URL and then pressing enter.

So much trouble for posting an example...

And to get back on topic. If you need the best RAW editing software, and don't need "Pentax colours", then I would look at Silkypix, Capture One, Aperture or Lightroom. I would say these are the best, not necessarily cheap, but very good.

[JeffJS]

They are showing inline now. I have to say though that other than a slight difference in the pinkish shades on the bottom of her feet, I'm not seeing any difference between them.

[adwb]

well I can see them in every post and there is obvious colour shift in each one.Interesting point being raised with colour representation dependant on software. I will have to test this out as I stopped using pentax photo lab as I find it unwieldy to use and I have ended up using the raw converter in elements for no other reason than ease of use.
Alistair

[Marc Sabatella]

I see the images just fine, and I won't hesitate at all to call the differences subtle (the exposure curve differences are more noticeable than the color differences). And when I say our perception of color is subjective, I am talking about *much* larger variations than this. That same scene viewed under tungsten light, sunlight, and shade would be vastly different as measured by a colorimeter, but our eyes & brain do an amazing job of fooling us into think we are seeing something close to the same color. But whether that color we perceive is more like your image #1, #2, or #3 is going to be totally subjective - the actual color isn't likely to be anything close to any of those except under one specific color of light.

[adwb]

As Mark says although obvious the shades are in reality quite subtle and depending on the light source will or will not look like the original. perhaps the point to bear in mind is that you saw the original scene and in in all probability your viewer of you web or print image did not and will never ever know if the colours are not the same. Even on a skin tone of any race the same applies , try it , ask some one [other than a pixel peeper] if they think the colour is realistic and I bet they will say yes. You can waste hours of time trying to calibrate your screen to show what you saw but if the viewers screen is not calibrated then whats the point? you can also spend hours getting a print to come out the same as you calibrated screen but once in a gallery or shop how does the viewer know that shade of pink in the trousers is right or wrong?
I have for the exercise just taken a image in raw from my K10D and converted via adobe elements, Pentax Photo lab and Picassa,and Raw thereapee all with any adjustable settings on zero, to be really honest there is so little difference with the shades of colours that the choice should be dictated by which software you find the easiest to use rather than the colour saturation/hue output as a jepg.
Alistair

[wowtip]

I see the images just fine, and I won't hesitate at all to call the differences subtle (the exposure curve differences are more noticeable than the color differences). And when I say our perception of color is subjective, I am talking about *much* larger variations than this. That same scene viewed under tungsten light, sunlight, and shade would be vastly different as measured by a colorimeter, but our eyes & brain do an amazing job of fooling us into think we are seeing something close to the same color. But whether that color we perceive is more like your image #1, #2, or #3 is going to be totally subjective - the actual color isn't likely to be anything close to any of those except under one specific color of light.

Well, maybe the subjective difference is that me and Dan has less tolerance for colour deviations and see a very real difference that you just can't see? Like when older people can't her sounds above 15 kHz...

Seriously, the differences in the photos posted by both Dan and me is enough to make me consider a RAW tool that is much harder to work with, just to make my pictures look (subjectively) right. I am not posting this just to get you worked up.

To me these differences are major, obviously not so for you and probably not for others. But if Genshu has our sensitivity to colour differences it is a very valid point to bring up.

As Mark says although obvious the shades are in reality quite subtle and depending on the light source will or will not look like the original. perhaps the point to bear in mind is that you saw the original scene and in in all probability your viewer of you web or print image did not and will never ever know if the colours are not the same. Even on a skin tone of any race the same applies , try it , ask some one [other than a pixel peeper] if they think the colour is realistic and I bet they will say yes. You can waste hours of time trying to calibrate your screen to show what you saw but if the viewers screen is not calibrated then whats the point? you can also spend hours getting a print to come out the same as you calibrated screen but once in a gallery or shop how does the viewer know that shade of pink in the trousers is right or wrong?
I have for the exercise just taken a image in raw from my K10D and converted via adobe elements, Pentax Photo lab and Picassa,and Raw thereapee all with any adjustable settings on zero, to be really honest there is so little difference with the shades of colours that the choice should be dictated by which software you find the easiest to use rather than the colour saturation/hue output as a jepg.
Alistair

Yes, to be honest laziness might make me cave in, in the long run. But for now I will go the long way with my RAW images and process them with Pentax DCU. And version 4 isn't that bad at all, not as intuitive and filled with functions as other commercial solutions, but it is much better than earlier versions.

Not all images show a dramatic difference between converters in my experience. From what I have seen it is mostly pinkish hues that get really messed up. Unfortunately that includes many flowers and caucasian skin tones. It might not be obvious when you see them separately, as you noted, but when you see them side by side, the difference is a big deal to me. And as Dan said, sometimes the "wrong" colour from other converters even make subtle (but important) details in skin tone go missing in portraits.

I first noticed this when I realized the colours almost always looked better, more realistic (yes yes, subjectively) and showed more detail in the embedded JPEG in my RAW images than what a full RAW conversion with my converter of choice, CaptureOne 4, would give.

[mattdm]

I still do not see OS X as a support platform in the download section.

There are alpha-build quality binaries for Mac OS X released here:

The Digital Cave - RawTherapee Mac Binary

I'm not a Mac user, but I think it'd be worth trying them, since it's a great free program.

Well, maybe the subjective difference is that me and Dan has less tolerance for colour deviations and see a very real difference that you just can't see? Like when older people can't her sounds above 15 kHz...

I'm gonna have to say I see a pretty obvious difference in the different conversions too, both the flowers and the kid's pants. I can't say which is "correct", but there's definitely a difference in hue. My guess is that this is due to a more or less accurate camera profile between the different converter programs.

[Marc Sabatella]

Well, maybe the subjective difference is that me and Dan has less tolerance for colour deviations and see a very real difference that you just can't see?

No, you're completely missing the point. I can see the differences; what I'm saying is that they are much smaller than the actual variations in real life scenes caused by different colors of light, and our eyes trick us into not seeing the latter. Given the pictures you posted, there is *no objective sense* in which one can reliably say which is more "accurate". Take that same scene, set your WB to either "daylight" or "flash" (to get the most neutral color), then take three pictures: one using incandescent light, one using flourescent light, and one using light coming in the window on a cloudy day. You'll get three *vastly* different photos - that's reality - and yet your brian will have perceived all three scenes more or less the same way. This is the subjective element I am talking about, and it's *huge* compared to the measurably *small* variations in your posted pictures.

if Genshu has our sensitivity to colour differences it is a very valid point to bring up.

Perhaps, but just because he cares a lot about color doesn't mean he'll share your particular subjective impressions about which is "better" for any given image. He might prefer the one you don't like as much. And he may find his sensitivity to color is such that he's never completely satisfied with the defaults from any program, but always gets results he likes better by custom processing. The point being, it is just false to say say one program is objectively "better" than another for color.

And version 4 isn't that bad at all, not as intuitive and filled with functions as other commercial solutions, but it is much better than earlier versions.

Workflow-wise, though, it's still in the stone ages compared to modern programs.