Pages: 1 [2] 3 4 ... 6   Go Down

Author Topic: Colour spaces  (Read 67582 times)

Panopeeper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1805
Colour spaces
« Reply #20 on: November 23, 2007, 09:26:44 pm »

Quote
If someone was to change the colour space of their image from AdobeRGB 98 to pro photo would the colour noise be more apparent and would this necessitate a stronger noise reduction in the chroma settings In a nut shell does a wider colour space mean a more noise in an image?TIA

The consequences of using Pro Photo in the ACR conversion are:

1. some colors, which are not contained in aRGB nor in sRGB are now representable, particularly in the deep blue and red,

2. the resulting gamut is much, much larger than what your monitor can reproduce (already aRGB is larger).

This means, that you don't see everything as it is in the image, consequently what you see is not only the noise in the image, but the noise due to the lack of gamut coverage of the monitor. If this is something perceivable, depends on the actual image.
Logged
Gabor

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Colour spaces
« Reply #21 on: November 24, 2007, 08:49:49 am »

Quote
You keep fishing for a raw file or DNG to have a "color space" as it is commonly associated with ICC style profiles and Photoshop working spaces. If you want to parse the words of the DNG spec and presume to attribute terms and their meanings, more power to ya, but it does a disservice to the industry to engage in a practice that will end up confusing people's understanding of raw linear captures and Photoshop working spaces and input profiles associated with scanners and cameras...

Raw captures do NOT have a "color space as it is defined by the ICC spec and associated with working color spaces in Photoshop". Raw captures are actually grayscale files whose color attributes are not yet assigned until the demosiacing process.

There, ya happy?

Ya know, sometimes you end up beating a dead horse bud...and the horse is beyond caring.
(and other people's eyes start to glaze over)
[{POST_SNAPBACK}][/a]

This thread reminds me the [a href=\"http://www.adobeforums.com/webx?14@@.3bc0d8df/0]thread[/url] on the Adobe Camera Raw forum that was set to read only by the forum administrator because of abusive ad hominem attacks on the participants. Schewe's characteristic shouting and bullying is not unlike the tone of that thread.

First of all, no one stated that the native camera space corresponded to an ICC profile, so arguments along that line of reasoning are invalid. Then the argument is that raw files are grayscale and couldn't have a color profile. As Bruce Fraser stated,  on p. 119 of Real World PSCS2, "In Photoshop, files saved in the RGB mode typically uses a set of three 8 bit grayscale files..." [bold added for emphasis]. Well, that argument is out.

Then it was stated that the camera data were not a space since there was no white point. Well, CIE XYZ has no white point either, but it is a color space even though it does not meet the ICC spec.

Chapter 6 of the Adobe DNG specification concerns converting from the Camera Color Space to the CIE XYZ Space. Adobe is pretty stupid to spend an entire chapter on something that does not exist. But then the mafia people know better, and they circle the wagons when challenged. If Schewe can't win his argument by reasoning, he resorts to bullying and shouting.

I'm sure that by this point no one is interested in pursuing this topic any further, but I just wanted to make a point. Furthermore, understanding of the camera color space helps us to understand the calibration procedure in ACR. We are modifying the 3 by 3 matrix values from the Adobe Camera Raw defaults to ones more descriptive of our own camera. These coefficients are used to convert the native camera color space to the internal color space of Camera Raw, and represent the Camera Raw Profile.
Logged

bobrobert

  • Guest
Colour spaces
« Reply #22 on: November 24, 2007, 10:03:03 am »

Any chance of someone answering my original question? In camera raw does setting Prophoto instead of AdobeRGB increase noise because there is a wider colour gamutTIA
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Colour spaces
« Reply #23 on: November 24, 2007, 10:09:29 am »

Quote
Any chance of someone answering my original question? In camera raw does setting Prophoto instead of AdobeRGB increase noise because there is a wider colour gamutTIA
[a href=\"index.php?act=findpost&pid=155464\"][{POST_SNAPBACK}][/a]

Quote
And, returning to the OP question;

No...the noise in an image is dictated by the ISO settings in combination with the expansion of darker tones to lighter tone (lightening the shadows) such as would occur when moving the Exposure setting (or Brightness) to the plus values...
[a href=\"index.php?act=findpost&pid=155308\"][{POST_SNAPBACK}][/a]

Bob, your question has been answered.
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Colour spaces
« Reply #24 on: November 24, 2007, 10:33:00 am »

Quote
Any chance of someone answering my original question? In camera raw does setting Prophoto instead of AdobeRGB increase noise because there is a wider colour gamutTIA
[a href=\"index.php?act=findpost&pid=155464\"][{POST_SNAPBACK}][/a]

No. It doesn't.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Jonathan Wienke

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5829
    • http://visual-vacations.com/
Colour spaces
« Reply #25 on: November 24, 2007, 10:56:20 am »

Quote
Any chance of someone answering my original question? In camera raw does setting Prophoto instead of AdobeRGB increase noise because there is a wider colour gamutTIA

I answered your question as well in post #6 of this thread. Color space has no effect at all on visible noise levels, unless you do something retarded like manually assigning the wrong profile after conversion, which will not only affect visible noise levels, but also cause over or undersaturation. Short answer: NO.
Logged

Panopeeper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1805
Colour spaces
« Reply #26 on: November 24, 2007, 11:27:10 am »

Quote
Any chance of someone answering my original question? In camera raw does setting Prophoto instead of AdobeRGB increase noise because there is a wider colour gamutTIA

The gamut of the camera is much larger, than sRGB or even aRGB.  Consequently, when mapping, the original pixel values get closer to each other, the originally existing noise diminishes.

The gamut of ProPhoto is perhaps somewhat larger than the camera's gamut, and anyway much larger than aRGB, so the mapping yields larger spacing between the original pixel values than with aRGB, i.e. more noise.

However, all this is present only in 16-bit TIFF. What you see is far from that; the more colors of the larger gamut "are not there", for the monitor supports only 8bit sRGB, and you probably see noise, which is not in the image; that's what I tried to explain above.

In order to see the "true noise", you have to inspect the raw image in the camera's color space, i.e. w/o converting the colors (but then the colors are not "true", for you are viewing them on an sRGB monitor).
Logged
Gabor

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Colour spaces
« Reply #27 on: November 24, 2007, 12:12:51 pm »

Quote
The gamut of the camera is much larger, than sRGB or even aRGB.  Consequently, when mapping, the original pixel values get closer to each other, the originally existing noise diminishes.

The gamut of ProPhoto is perhaps somewhat larger than the camera's gamut, and anyway much larger than aRGB, so the mapping yields larger spacing between the original pixel values than with aRGB, i.e. more noise.

However, all this is present only in 16-bit TIFF. What you see is far from that; the more colors of the larger gamut "are not there", for the monitor supports only 8bit sRGB, and you probably see noise, which is not in the image; that's what I tried to explain above.

In order to see the "true noise", you have to inspect the raw image in the camera's color space, i.e. w/o converting the colors (but then the colors are not "true", for you are viewing them on an sRGB monitor).
[a href=\"index.php?act=findpost&pid=155492\"][{POST_SNAPBACK}][/a]

In order to test the noise hypothesis, I rendered the same raw file into sRGB and ProPhotoRGB and analyzed the noise response with Imatest Stepchart. The overall luminance noise is the same (0.38% and 0.36% respectively), but the resulting TRCs are different and the distribution of the noise is different along the TRC. I think that this has to do with the gamma of the spaces (sRGB is 2.2 and ProPhotoRGB is 1.8). The actual gamma of the converted files are 1/1.58 for the ProPhoto and 1/1.68 for the sRGB. If one converts the ProPhotoRGB file to sRGB and reanalyzes, then the noise characteristics are exactly the same.

[attachment=4009:attachment]

[attachment=4010:attachment]

[attachment=4011:attachment]
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Colour spaces
« Reply #28 on: November 24, 2007, 12:33:39 pm »

Quote
The gamut of the camera is much larger, than sRGB or even aRGB.  Consequently, when mapping, the original pixel values get closer to each other, the originally existing noise diminishes.

Digital cameras don't have a color gamut, they have color mixing functions.

To even profile a camera, to get some idea of its so called gamut, you have to place a target of known color values in front of the sensor. That target does have a gamut, the gamut of the resulting profile can't be larger than the target, hence the problems here both profiling the camera and trying to define its so called gamut. Then we can discuss how this so called gamut is affected by the illuminant used to shoot the target as well as the dynamic range of whatever you now place in front of the camera or the role of the Raw converter in rendering the image. So, lets not even begin to try to discuss the gamut of a digital camera (because again, it doesn't have one).

Quote
The gamut of ProPhoto is perhaps somewhat larger than the camera's gamut, and anyway much larger than aRGB, so the mapping yields larger spacing between the original pixel values than with aRGB, i.e. more noise.

Again, that's simply speculation and I'll add, the gamut of ProPhoto RGB contains colors that fall outside human vision (although digital cameras can clearly capture stuff outside human vision). But, you're using such differing terms and processes that the two just don't sync up.

Quote
In order to see the "true noise", you have to inspect the raw image in the camera's color space, i.e. w/o converting the colors (but then the colors are not "true", for you are viewing them on an sRGB monitor).
[a href=\"index.php?act=findpost&pid=155492\"][{POST_SNAPBACK}][/a]

And that camera color space would be what? We're talking Raw data again (essentially Grayscale data to get back another post here that's up to debate).
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Jonathan Wienke

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5829
    • http://visual-vacations.com/
Colour spaces
« Reply #29 on: November 24, 2007, 12:37:14 pm »

Quote
The gamut of the camera is much larger, than sRGB or even aRGB.  Consequently, when mapping, the original pixel values get closer to each other, the originally existing noise diminishes.

This is completely false, at least in relation to how visible the noise is in the image, whether on screen or in a print. Converting from one color space to another alters how the noise is encoded, but as long as all image colors are in-gamut, converting from one color space will have no effect on how noisy the image looks on-screen or printed, as long as you do so in 16-bit mode so that quantization errors are not an issue. Minor differences in noise measurements of the sRGB file versus ProPhoto or whatever cancel out exactly once you correctly account for the differences in gamut and TRC defined by each color space.
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Colour spaces
« Reply #30 on: November 24, 2007, 02:14:38 pm »

duplicate post deleted by maker
« Last Edit: November 24, 2007, 02:19:29 pm by bjanes »
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Colour spaces
« Reply #31 on: November 24, 2007, 02:18:57 pm »

Quote
Digital cameras don't have a color gamut, they have color mixing functions.
[a href=\"index.php?act=findpost&pid=155510\"][{POST_SNAPBACK}][/a]

I was waiting for the mob to employ this tactic: falsis in unum, falsis in omnibus , which is used in legal arguments to discredit a witness. A digital camera does not have a gamut in the sense of a space with well defined boundaries. However, the dictionary definition is "an entire range or series <ran the gamut from praise to contempt>" and a digital camera does have a gamut in this general sense.

Rather than nit pick about peripheral issues (digital camera has no gamut, a raw file has no color space), it would be best to remain on topic and address the main issue rather than try to put down others and show one's own erudition. This is a forum where we are all trying to learn, not a court of law with an adversarial climate. The main thrust of the argument was false, as Jonathan pointed out.
Logged

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Colour spaces
« Reply #32 on: November 24, 2007, 04:50:38 pm »

Quote
This thread reminds me the thread on the Adobe Camera Raw forum that was set to read only by the forum administrator because of abusive ad hominem attacks on the participants. Schewe's characteristic shouting and bullying is not unlike the tone of that thread.
[a href=\"index.php?act=findpost&pid=155451\"][{POST_SNAPBACK}][/a]

Ironically, I never posted in that thread...so I guess Mr. Janes is trying to paint me with a brush of "similarity"...

"shouting and bullying"? Strikes me that THAT is an ad hominem attack...

Never the less, I thought I would show some images to help make my points. Here's a few images from RWCR CS3 showing some rarely seen real raw images:


The image above is a real raw image, processed but not demosiaced. It was processed through DNG Verifier to show what the image looks like when written to disk...


Here's the same image processed through Camera Raw.

So, what color space is the original grayscale raw image in? The actual raw image is indeed grayscale until the image is demosiaced...and not only is it grayscale, it's in linear gamma.

You see the tiny green rectangles? Here are some details from those areas...


The image above is taken from the original file processed through DNG Verifier then blown up in Photoshop to 3,200 % zoom. You can actually see the original Bayer array pixels. You'll note that the dark pixels in the area that is actually yellow shows the blue pixel photo sites.


This is the same area of pixels shown in the detail image above. The grayscale tones have now been interpolated into color information. It is this demosiacing interpolation that gives color information to the grayscale pixels.

So, again, in terms of the RAW file before demosiacing interpolation, what would be the "camera color space"? In actuality, it has no color space until the manner of demosiacing interpolation is determined and the tags regarding white point are provided and factored into the interpolation.

Even then those tags are subject to interpretation...what the original raw does have is a spectral response based on the properties of the Red, Green & Blue separation filters that are used to filter the white light to capture the RGB response properties, but that data is held in B&W until demosiacing.

So, ya still think that the original raw capture has a color space?
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Colour spaces
« Reply #33 on: November 24, 2007, 05:00:22 pm »

Quote
The image above is a real raw image, processed but not demosiaced. It was processed through DNG Verifier to show what the image looks like when written to disk...

Very cool Jeff. Where does one get DNG Verifier? This is a great way to illustrate a Raw processed by not yet demosiaced.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Colour spaces
« Reply #34 on: November 24, 2007, 05:02:44 pm »

Quote
Rather than nit pick about peripheral issues (digital camera has no gamut, a raw file has no color space), it would be best to remain on topic and address the main issue rather than try to put down others and show one's own erudition.
[a href=\"index.php?act=findpost&pid=155548\"][{POST_SNAPBACK}][/a]

No put down intended. The fact is, a digital camera doesn't have a color gamut and a Raw file doesn't have a color space. Some here don't consider this a nit pick.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Colour spaces
« Reply #35 on: November 24, 2007, 05:05:16 pm »

Quote
Very cool Jeff. Where does one get DNG Verifier? This is a great way to illustrate a Raw processed by not yet demosiaced.
[a href=\"index.php?act=findpost&pid=155590\"][{POST_SNAPBACK}][/a]

It's part of the DNG SDK...and intended for geeks (Zalman had to teach me how to run it because you run it via command line).

:~)
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Colour spaces
« Reply #36 on: November 24, 2007, 05:09:06 pm »

Quote
(Zalman had to teach me how to run it because you run it via command line).
[a href=\"index.php?act=findpost&pid=155593\"][{POST_SNAPBACK}][/a]

OK, never mind. I'm not about to mess around with command lines! But, a very interesting illustration of the processing none the less.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Panopeeper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1805
Colour spaces
« Reply #37 on: November 24, 2007, 06:02:58 pm »

Quote
Digital cameras don't have a color gamut, they have color mixing functions

The gamut is the amount (range) of reproducible colors. This has nothing to do with the question how it can be profiled.

To even profile a camera, to get some idea of its so called gamut, you have to place a target of known color values in front of the sensor

This is the description of how you would do it. Again, it has no role here.

Btw, the spectral responses of the filters define the gamut of the camera, but manufacturers don't publish this information. (Though one could measure it with suitable equipment.)

Quote
that camera color space would be what?

That camera color space would be that camera's specific color space. If you want a name to it, use the camera model, or in some cases the sensor "model" (sometimes the same sensor is used in different cameras).
Logged
Gabor

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Colour spaces
« Reply #38 on: November 24, 2007, 06:21:06 pm »

Quote
The gamut is the amount (range) of reproducible colors. This has nothing to do with the question how it can be profiled.

And you define this gamut, from a Raw file how?

To even profile a camera, to get some idea of its so called gamut, you have to place a target of known color values in front of the sensor

Quote
This is the description of how you would do it.

And YOU would profile it how?

In light of Jeff's example above, what gamut does the image have processed but not demosiaced?
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Panopeeper

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1805
Colour spaces
« Reply #39 on: November 24, 2007, 07:40:29 pm »

Quote
And you define this gamut, from a Raw file how?

I don't know, I am not that business. However, some others are obviously doing that (and the manufacturers, who know the spectral response of the filters, certainly know the gamut of their own equipment).

Micheal Reichmann shows the color space of the Canon 20D in here. He even thanks Bruce Fraser and Jeff Schewe for their feedback and suggestions; the latter must have forgotten it already.

Quote
In light of Jeff's example above, what gamut does the image have processed but not demosiaced?

You are mixing up the gamut of the camera and of the scenery/actual image. The subject is the former.
Logged
Gabor
Pages: 1 [2] 3 4 ... 6   Go Up