how many of you use aRGB?

[Craig Jones]

I can see you aren't in it for the discussion, only for the disrepect (despite what you say). Why did you enter this thread anyway?

 

[Craig Jones]

Full disclosure of how both conversions were performed is required before anything meaningful can be concluded. As we all know, there is no such thing as Lab values direct from raw and I'm surprised Iliah would suggest there were.

It wouldn't surprise me if the first conversion comes from one if Iliah's converters and the other from ACR through ProPhoto. If that's the case, this is nothing other than a comparison of color conversions between competing products and says nothing at all about ProPhoto.

I would also be interested in seeing a third conversion using Adobe RGB instead of ProPhoto but otherwise the same conversion technique. I suspect the results will be identical. The raw file and full descriptions of the conversion methods would also be nice.

 

[CM_laptop]

...behind an ad hominem attack (I will not respond to that)

Once again: Prove it.

 

[Craig Jones]

Hide? More like not waste my time. What have you contributed here? I mean, rather than insults.

 

[ejmartin]

Hide? More like not waste my time. What have you contributed here?
I mean, rather than insults.

Please ignore this sort of garbage. Responding to it only clogs otherwise useful and informative threads, which this has been up to now.

--
emil
--



http://theory.uchicago.edu/~ejm/pix/20d/

 

[voider]

Hi Roman,

good to hear that you think the same way. Do you use in Editing then aRGB or ProPhotoRGB?

Timur
--
Visit http://www.voider.net

 

[RomanJohnston]

I use BetaRGB.....

http://brucelindbloom.com/index.html?BetaRGB.html

Reccomended by Julia Borg as one of the better free RGB workingspaces.

Dont know if any better free ones have come to light since that reccomendation.

Roman
--

'Our deepest fear is not that we are inadequate. Our deepest fear is that we are powerful beyond measure. It is our light, not our darkness that most frightens us. We ask ourselves, who are we to be brilliant, gorgeous, talented, fabulous.

Actually, who are we not to be?'

--Marianne Williamson

http://www.pbase.com/romansphotos/

 

[voider]

Interesting. I never thought that this simple thread would give so many different replies. I wonder what the others say to BetaRGB.

Since when do you use BetaRGB and do you feel or know that the results are better since using it?

Timur

I use BetaRGB.....

http://brucelindbloom.com/index.html?BetaRGB.html

Reccomended by Julia Borg as one of the better free RGB workingspaces.

Dont know if any better free ones have come to light since that
reccomendation.

Roman
--
'Our deepest fear is not that we are inadequate. Our deepest fear is
that we are powerful beyond measure. It is our light, not our
darkness that most frightens us. We ask ourselves, who are we to be
brilliant, gorgeous, talented, fabulous.

Actually, who are we not to be?'

--Marianne Williamson

http://www.pbase.com/romansphotos/

--
Visit http://www.voider.net

 

[Kaj E]

I have my camera set to Adobe RGB and I shoot NEF.

I only use my camera LCD for checking the histogram (and some times to check that the arrangement of the subjects within the frame are as intended). By using Adobe RGB the histogram more closely reflects exposure accuracy due to the larger gamut of Adobe RGB. I by now roughly know how much blinking highlights or lost shadows I can accept in the histogram without having a problem in the RAW file. I typically try to expose to the right as much as I safely can.

I could not care less for the colors in the LCD display as they can be fixed in PP anyhow.

Adobe RGB is also my working color space.
--
Kind regards
Kaj
http://www.pbase.com/kaj_e
WSSA member

 

[Kaj E]

Sorry, post above intended to reply to OP.
--
Kind regards
Kaj
http://www.pbase.com/kaj_e
WSSA member

 

[RomanJohnston]

I try to focus more on photography than the geek side of the fence.....I trust the Borgs (both of them) to provide factual and highly detailed information.

Through my research and readings on the web from people who are far more intelligent on the subject than I....I do know one thing....you dont want a color space too large OR too small.....I would hazard a guess that BetaRGB is probably as close to "just right" as you can get for free.

If they reccomend somthing....as far as I am concerned...it is fact and I follow it (if possable). They arent infallable I am sure...but their batting average is insanely high on the correct side of the fence.

They are members of a very short list of people I trust that much.

Thom Hogan would be up there as well...

Roman

Since when do you use BetaRGB and do you feel or know that the
results are better since using it?

Timur

I use BetaRGB.....

http://brucelindbloom.com/index.html?BetaRGB.html

Reccomended by Julia Borg as one of the better free RGB workingspaces.

Dont know if any better free ones have come to light since that
reccomendation.

Roman
--
'Our deepest fear is not that we are inadequate. Our deepest fear is
that we are powerful beyond measure. It is our light, not our
darkness that most frightens us. We ask ourselves, who are we to be
brilliant, gorgeous, talented, fabulous.

Actually, who are we not to be?'

--Marianne Williamson

http://www.pbase.com/romansphotos/

--
Visit http://www.voider.net

--

'Our deepest fear is not that we are inadequate. Our deepest fear is that we are powerful beyond measure. It is our light, not our darkness that most frightens us. We ask ourselves, who are we to be brilliant, gorgeous, talented, fabulous.

Actually, who are we not to be?'

--Marianne Williamson

http://www.pbase.com/romansphotos/

 

[CM_laptop]

Emil, I asked him to prove a point HE made.

What's wrong or "garbage" with that?

Please put the blame squarely where it belongs......

OT to Craig: you still haven't proven your point.

 

[CM_laptop]

Hide? More like not waste my time. What have you contributed here?
I mean, rather than insults.

No, it's not just the title of a film.....

It was indeed you who resorted to ad hominem attacks, (like your inane "cheese remark"; just after you demonstared you were talking about issues you had no idea about).

Now that we've swept your insults aside; let's get back to the original issue:

Provide proof fo your arguments.

 

[Bill Janes]

I try to focus more on photography than the geek side of the
fence.....I trust the Borgs (both of them) to provide factual and
highly detailed information.

Through my research and readings on the web from people who are far
more intelligent on the subject than I....I do know one thing....you
dont want a color space too large OR too small.....I would hazard a
guess that BetaRGB is probably as close to "just right" as you can
get for free.

If they reccomend somthing....as far as I am concerned...it is fact
and I follow it (if possable). They arent infallable I am sure...but
their batting average is insanely high on the correct side of the
fence.

They are members of a very short list of people I trust that much.

Thom Hogan would be up there as well...

Roman

Certainly the Borgs and Bruce Lindbloom are right up there in terms of expertise. Mr. Lindbloom designed BetaRGB to be as small as possible yet encode all the important colors in his color set, which included colors of the MacBeth color checker, various film and photographic paper stocks, and FOGRA color sets, which I take to represent printing press colors.

Kodak and Prof. Gernot Hoffmann have researched real world surface colors, which represent colors occurring in nature from non-self-luminous reflectors (excludes neon lights, florescent colors, etc). Hoffmann states that these surface colors probably should be reproducible by photography, but that it is not important to record all colors in the L*a*b space. sRGB and aRGB do not include these surface colors, but ROMM RGB (basically ProPhotoRGB) does.

http://www.fho-emden.de/~hoffmann/gamuts08072002.pdf

Now that we have digital photography and wide-gamut ink jet printers, I'm not certain that Mr. Lindbloom's color set is the most appropriate, but it appears similar to Prof. Hoffmann's surface gamut.

If you use Adobe Camera Raw, it does not support BetaRGB or other custom color spaces, so it would be necessary to render into ProPhotoRGB and then convert to BetaRGB. It you use a 16 space, the size of the gamut is probably not as important as with 8 bit, which is not adequate for ProPhotoRGB. I don't know if 8 bits are sufficient for BetaRGB.

If you use Nikon CaptureNX or some other raw converters, I think you can render directly into BetaRGB. If you shoot in camera JPEGs, you are out of luck.

--
Bill Janes

 

[Iliah Borg]

Kodak and Prof. Gernot Hoffmann have researched real world surface
colors, which represent colors occurring in nature from
non-self-luminous reflectors

We are talking here not about the real world, but about what of that real world can be recorded by the sensor, and how it can be recorded. The idea of using a hypothetical space, basically designed to be a universum of colours, to a limited task is wrong. You waste encoding, the same way as you waste encoding not using gamut-mapped device links. Limited encoding is one of the key issues, the blotchy noise in shadows is one of the results of it.

Take an sRGB colour of (48,7,7) and of (47,7,7). Convert those to ProPhoto RGB. With any given encoding width you will have this issue. You won't have that much of it in floating, but still you will have it. The whole idea of colour conversion in raw converter other then for output (and even there it is better to output in corrected Lab space) is unnecessary complication, source of problems, the cure that is worse then the disease. Not to mention the disease by itself is imaginary: strive for one size fits all never turned out well.

--
http://www.libraw.org/

 

[Bill Janes]

Kodak and Prof. Gernot Hoffmann have researched real world surface
colors, which represent colors occurring in nature from
non-self-luminous reflectors

We are talking here not about the real world, but about what of that
real world can be recorded by the sensor, and how it can be recorded.
The idea of using a hypothetical space, basically designed to be a
universum of colours, to a limited task is wrong. You waste encoding,
the same way as you waste encoding not using gamut-mapped device
links. Limited encoding is one of the key issues, the blotchy noise
in shadows is one of the results of it.

While, technically speaking, digital cameras do not have a gamut, they can record nearly the entire visible spectrum from 380-750 nm as demonstrated by Christian Buil and others. See section 9 of Buil's paper:

http://astrosurf.com/buil/d70v10d/eval.htm

Wavelength corresponds to hue, and the two other parameters in the color definition are saturation and brightness. Brightness is related to the dynamic range of the sensor, which is well documented. I have no information on saturation, but it is well known that current sensors can capture colors well outside the gamut of sRGB and aRGB. I state without proof that if one wishes to capture all the colors in a digital camera capture, it would be necessary to render into a wide working space such as ProPhotoRGB.

If the selected output device can not reproduce the entire gamut of the capture or if the scene contains a limited range of colors, encoding space will be wasted in a wide gamut space as you mention. Output devices are continually being improved, and it may not be wise to limit the gamut of your working space to the gamut of the output device. You may get a better printer in the future. If you are processing a large number of images, it is not practical to select a working space matched to the gamut of the scene. For safety, you should render into a large working space at 16 bits per pixel.

How large should the working space be? BetaRGB, which you apparently endorse, was devised by Bruce Lindbloom to cover the output range of selected media, in contradiction to your dictum about hypothetical spaces.

Likewise, the real world surface colors described by Prof. Hoffmann limit the gamut to colors likely to be present in a real world scene, thus enabling a smaller space with better encoding. I see no real difference in these two approaches.

Take an sRGB colour of (48,7,7) and of (47,7,7). Convert those to
ProPhoto RGB. With any given encoding width you will have this issue.
You won't have that much of it in floating, but still you will have
it. The whole idea of colour conversion in raw converter other then
for output (and even there it is better to output in corrected Lab
space) is unnecessary complication, source of problems, the cure that
is worse then the disease. Not to mention the disease by itself is
imaginary: strive for one size fits all never turned out well.

Conversion of the working space rendered by the raw converter to another working space is an unnecessary complication, as you correctly point out. For most purposes, 16 bit integer encoding would give sufficient quality and 32 bit floating point may not be necessary. Perhaps you can demonstrate examples of the blotchiness you mention.

For my personal work, I use ProPhotoRGB and am pleased with the results. Bruce Fraser had a similar approach as does Adobe Lightroom.
--
Bill Janes

 

[Mark den Hartog]

What in your opinion fills the gaps between missing

colours then?

when you start editing or shifting colours in the image which you can easily do by changing something as simple as whitebalance!

basically it gives more room in the editing process.

 

[Iliah Borg]

While, technically speaking, digital cameras do not have a gamut,
they can record nearly the entire visible spectrum from 380-750 nm

Quite true, the input gamut is not limited. Output gamut is.

current
sensors can capture colors well outside the gamut of sRGB and aRGB.

Like IR, for example.

if one wishes to capture all the colors in a
digital camera capture, it would be necessary to render into a wide
working space such as ProPhotoRGB.

Can you reconstruct all real world colours from camera capture? If not, then huge colour spaces are extraneous for output.

Quantitative study of camera metamerism is necessary to see how capture works in reality.

It is a workflow question actually. For some reason in a lot of cases one step in the colour conversion pipeline is missing. Input is scene colours, recording is in camera output colours, output from the raw converter is in some working space which may be Lab, camera output space, or working space. Inside the converter conversion from camera output space to RGB working space is excessive, and creates conversion noise and artefacts.

Output devices are continually being improved, and it may not be wise
to limit the gamut of your working space to the gamut of the output
device.

Yes, but. Output devices like printers, projectors, and monitors have nothing to do with capture and raw conversion. It is a separate issue. What captured is captured, and output device can only use the captured colours directly or map them to something else. Intermediate colour space of raw converter has nothing to do with that. The choice of intermediate working space in Photoshop is a different issue. Clear separation between raw converter and Photoshop is essential when we address the colour issues.

For most purposes, 16 bit integer encoding would give sufficient quality

Not for DNG workflow. Do you know how many bits ACR/LR use internally? 20+?

Perhaps you can demonstrate examples of the blotchiness
you mention.

The example with 2 colours I suggested demonstrates exactly that and is in line with my practical experience with ACR, especially with higher ISO shots.

For my personal work, I use ProPhotoRGB and am pleased with the
results. Bruce Fraser had a similar approach as does Adobe Lightroom.

You have no other choice with Adobe converters and scenes with vivid colours, they use ProPhoto RGB chromaticities internally and clip the output if other output colour space is selected. However if clipping on the ACR histogram is not observed a smaller output colour space can be used, having more efficient coding of gradations. Images will only benefit from such a coding.

--
http://www.libraw.org/