Why 2 degree observer measurements for ICC profiles?
Hi list members. Hoping some color science folk out there on the list could help answer a question I have.
From Konica Minoltas web page titled “Understanding standard observers in color measurement” “CIE standard observers help correlate instrumental color measurements to human visual assessments. The 1964 10 Degree Supplementary Standard Observer is considered to be more representative of how the human eye perceives color.”
So why do ICC profiles, for print production, still use 2degree observer measurements instead of 10 degree observer measurements? Given that the 10 degree observer “is considered to be more representative of how the human eye perceives color. (me: Than 2 degree observer)” Thanks Peter
Hi Peter, The difference between the 2° and 10° observer is basically the size of the area on your retina that is stimulated. For the 2° observer mainly the cones in the center of the retina are stimulated while for the 10° observer also the rods are stimulated. That leads to a difference in color perception. Here is a quite nice visualization: https://santhalakshminarayana.github.io/images/color-science/standard-observ... The type of application when our perception is like a 2° observer or a 10° observer respectively is closely related to the viewing distance. When you look at a printed product from a typical viewing distance of 50-60cm the field of view is close to 2°. That's why for ICC profiles that are mainly used for print applications the 2° observer is the standard. When you look at other products like let's say a painted wall the filed of view is close to 10°. The wording on our website is a little misleading I must admit. Most of our industrial products are used for applications where the viewing distance leads to a field of view where the 10° observer correlates better than the 2° observer. BTW, when you want to check if in a certain situation "you are" the 2° or the 10° observer there is a simple test. Stretch out your arm, thumb upwards and look at your thumb with one eye. The area that is covered corresponds to approximately 2°. If the area can be covered with the entire fist, this is approximately 10°. Best regards Claas Peter Miles via colorsync-users schrieb am 16.07.24 um 04:13:
Hi list members. Hoping some color science folk out there on the list could help answer a question I have.
From Konica Minoltas web page titled “Understanding standard observers in color measurement” “CIE standard observers help correlate instrumental color measurements to human visual assessments. The 1964 10 Degree Supplementary Standard Observer is considered to be more representative of how the human eye perceives color.”
So why do ICC profiles, for print production, still use 2degree observer measurements instead of 10 degree observer measurements? Given that the 10 degree observer “is considered to be more representative of how the human eye perceives color. (me: Than 2 degree observer)”
Thanks Peter _______________________________________________ Do not post admin requests to the list. They will be ignored. colorsync-users mailing list (colorsync-users@lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/colorsync-users/lists%40bickeboeller...
This email sent to lists@bickeboeller.name
Thanks for the explanation Claas! Peter From: Claas Bickeböller via colorsync-users <colorsync-users@lists.apple.com> Date: Tuesday, 16 July 2024 at 7:22 PM To: Colorsynclist <colorsync-users@lists.apple.com> Subject: Re: Why 2 degree observer measurements for ICC profiles? Hi Peter, The difference between the 2° and 10° observer is basically the size of the area on your retina that is stimulated. For the 2° observer mainly the cones in the center of the retina are stimulated while for the 10° observer also the rods are stimulated. That leads to a difference in color perception. Here is a quite nice visualization: https://santhalakshminarayana.github.io/images/color-science/standard-observ... The type of application when our perception is like a 2° observer or a 10° observer respectively is closely related to the viewing distance. When you look at a printed product from a typical viewing distance of 50-60cm the field of view is close to 2°. That's why for ICC profiles that are mainly used for print applications the 2° observer is the standard. When you look at other products like let's say a painted wall the filed of view is close to 10°. The wording on our website is a little misleading I must admit. Most of our industrial products are used for applications where the viewing distance leads to a field of view where the 10° observer correlates better than the 2° observer. BTW, when you want to check if in a certain situation "you are" the 2° or the 10° observer there is a simple test. Stretch out your arm, thumb upwards and look at your thumb with one eye. The area that is covered corresponds to approximately 2°. If the area can be covered with the entire fist, this is approximately 10°. Best regards Claas Peter Miles via colorsync-users schrieb am 16.07.24 um 04:13:
Hi list members. Hoping some color science folk out there on the list could help answer a question I have.
From Konica Minoltas web page titled “Understanding standard observers in color measurement” “CIE standard observers help correlate instrumental color measurements to human visual assessments. The 1964 10 Degree Supplementary Standard Observer is considered to be more representative of how the human eye perceives color.”
So why do ICC profiles, for print production, still use 2degree observer measurements instead of 10 degree observer measurements? Given that the 10 degree observer “is considered to be more representative of how the human eye perceives color. (me: Than 2 degree observer)”
Thanks Peter
Peter Miles via colorsync-users wrote:
So why do ICC profiles, for print production, still use 2degree observer measurements instead of 10 degree observer measurements?
A standard often has a momentum of its own. But in the case of color in media that is looked at in detail (such as images), 2 degree is appropriate. AFAIK, one of the reasons that the 10 degree observer is often recommended is that it was created with more observers and in a more refined way, and is technically a bit more accurate. In subsequent experience it has often turned out that the 10 degree standard matches people more accurately than the earlier 2 degree observer, even though it is not the appropriate viewing angle. Note that there are many subsequent more refined standards or proposed standards such as the 2012 2 and 10 degree observers. These are not in widespread use though, whereas every graphic arts instrument will return 1931 2 degree values. And of course there is nothing to stop you making ICC profiles using any observer you like :-) They may not be technically interchangeable with profiles using the 1931 2 degree observer, but that matters less in practice, and not at all if all the profiles you use use the same observer. Cheers, Graeme Gill.
Thanks Graeme, and everyone for your responses. A lot of food for thought. Reflik I had a looked at the “Individual Colorimetric Observer Model” paper, however I don’t have a training in color science to fully grasp the significance of what I am reading. Is there a way of qualifying what sort of size colour differences these observer variations are causing in the normally sighted population? If they were all looking at the same objects in a viewing booth, what sort of magnitude of potential colour differences would they be seeing? FYI: A few years ago I had a lens replaced in one of my eyes and so I now have two quite different ‘observers’ when it comes to assessing colour. Thanks Peter Miles From: Graeme Gill via colorsync-users <colorsync-users@lists.apple.com> Date: Friday, 19 July 2024 at 8:49 PM To: ColorSync <colorsync-users@lists.apple.com> Subject: Re: Why 2 degree observer measurements for ICC profiles? Peter Miles via colorsync-users wrote:
So why do ICC profiles, for print production, still use 2degree observer measurements instead of 10 degree observer measurements?
A standard often has a momentum of its own. But in the case of color in media that is looked at in detail (such as images), 2 degree is appropriate. AFAIK, one of the reasons that the 10 degree observer is often recommended is that it was created with more observers and in a more refined way, and is technically a bit more accurate. In subsequent experience it has often turned out that the 10 degree standard matches people more accurately than the earlier 2 degree observer, even though it is not the appropriate viewing angle. Note that there are many subsequent more refined standards or proposed standards such as the 2012 2 and 10 degree observers. These are not in widespread use though, whereas every graphic arts instrument will return 1931 2 degree values. And of course there is nothing to stop you making ICC profiles using any observer you like :-) They may not be technically interchangeable with profiles using the 1931 2 degree observer, but that matters less in practice, and not at all if all the profiles you use use the same observer. Cheers, Graeme Gill.
Peter Miles via colorsync-users wrote:
Is there a way of qualifying what sort of size colour differences these observer variations are causing in the normally sighted population? If they were all looking at the same objects in a viewing booth, what sort of magnitude of potential colour differences would they be seeing?
In some numerical simulations I've done, this depends heavily on how narrow band the colors are. Make them very narrow (such as lasers) and you can expect extreme differences in perceived color (I saw of the order of 30 dE for some observer pairs). Of course a print evaluation is the least provocative situation since typical colorants used on paper have relatively mild transitions vs. wavelength. This is why reflective and transmissive media that uses typical pigments or dyes has smaller gamuts than emissive displays or interference effects filters.
FYI: A few years ago I had a lens replaced in one of my eyes and so I now have two quite different ‘observers’ when it comes to assessing colour.
How different this will be in practice depends on how "bumpy" the difference in the transmission spectra of your two lenses are. A smooth difference will amount to a change in white point, but is less likely to change what you regard as a colorimetric match. More rapidly changing differences could have some effect, depending on the coincidence between those differences and the transition bands of the color stimulus. Cheers, Graeme Gill.
The difference between a natural lens, a replacement lens with UV absorption, and a clear replacement can be very large in daylight vs. other sources. I got UV absorbing lenses in both eyes, and during the time I had only one replacement, the difference was quite noticeable. A colleague who had one UV absorbing and one clear reported some extreme differences. I had a khaki colored parka that was mainly cotton, but synthetic knit for the collar. In daylight, I saw a slight mismatch, but he saw a strong purple cast to the collar. I wear photo gray glasses, and I can see an amber tint under artificial light and a much stronger amber tint under daylight. My natural lenses had browned to the point that I could not see the tint. -----Original Message----- From: Graeme Gill via colorsync-users <colorsync-users@lists.apple.com> Sent: Monday, July 22, 2024 5:29 PM To: ColorSync <colorsync-users@lists.apple.com> Cc: Graeme Gill <graeme2@argyllcms.com> Subject: Re: Why 2 degree observer measurements for ICC profiles? Peter Miles via colorsync-users wrote:
Is there a way of qualifying what sort of size colour differences these observer variations are causing in the normally sighted population? If they were all looking at the same objects in a viewing booth, what sort of magnitude of potential colour differences would they be seeing?
In some numerical simulations I've done, this depends heavily on how narrow band the colors are. Make them very narrow (such as lasers) and you can expect extreme differences in perceived color (I saw of the order of 30 dE for some observer pairs). Of course a print evaluation is the least provocative situation since typical colorants used on paper have relatively mild transitions vs. wavelength. This is why reflective and transmissive media that uses typical pigments or dyes has smaller gamuts than emissive displays or interference effects filters.
FYI: A few years ago I had a lens replaced in one of my eyes and so I now have two quite different 'observers' when it comes to assessing colour.
How different this will be in practice depends on how "bumpy" the difference in the transmission spectra of your two lenses are. A smooth difference will amount to a change in white point, but is less likely to change what you regard as a colorimetric match. More rapidly changing differences could have some effect, depending on the coincidence between those differences and the transition bands of the color stimulus. Cheers, Graeme Gill. _______________________________________________ Do not post admin requests to the list. They will be ignored. colorsync-users mailing list (colorsync-users@lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/colorsync-users/waynebretl%40cox.net This email sent to waynebretl@cox.net
Hi Wayne.I can relate to the increased short wavelength / UV sensitivity with a clear lens. My eye surgeon was very interested in my feedback around how the new lens affected my color perception, given my job.About a month after my surgery I emailed my surgeon color graded photo of an image in a GTI viewing booth, but in two views. The color as it appeared to each eye.I have attached a link to it below. I did wonder at the time of selecting a lens, if the eye surgery folk that engineer the spectral transmission of those lenses had heard of the two degree standard observer. Could be a great collaboration there for some folk on this list maybe? :-) Peterhttps://drive.google.com/file/d/1Xoqdyc2LuKywA7rkodHEn-zJ71heIvw9/view?usp=d... On Tuesday, 23 July 2024 at 12:41:49 pm NZST, Wayne E. Bretl via colorsync-users <colorsync-users@lists.apple.com> wrote: The difference between a natural lens, a replacement lens with UV absorption, and a clear replacement can be very large in daylight vs. other sources. I got UV absorbing lenses in both eyes, and during the time I had only one replacement, the difference was quite noticeable. A colleague who had one UV absorbing and one clear reported some extreme differences. I had a khaki colored parka that was mainly cotton, but synthetic knit for the collar. In daylight, I saw a slight mismatch, but he saw a strong purple cast to the collar. I wear photo gray glasses, and I can see an amber tint under artificial light and a much stronger amber tint under daylight. My natural lenses had browned to the point that I could not see the tint.
Hi Peter, There have been several attempts made by the scientific community to quantify observer metameric mismatch during the past couple of decades. Finally IEC (International Electrotechnical Commission) has published the technical specification on 2023-11-01 as “IEC TS 61966-13:2023”. (https://webstore.iec.ch/en/publication/67769) -------------------------- IEC Technical Specification 61966-13 IEC TS 61966-13:2023 Multimedia systems and equipment - Colour measurement and management - Part 13: Measurement method of display colour properties depending on observers IEC TS 61966-13:2023 defines an objective colour difference metric and a measurement method for observer metamerism caused by displays with different spectral power distributions. This document also specifies the measuring equipment, conditions and methods that are necessary to obtain the metric. This document applies to light-emitting or backlit transmitting colour displays measured under dark-room conditions. -------------------------- You can read more on the topic in a recent issue of the ID magazine published by The Society for Information Display (Volume 40, Issue 1, https://sid.onlinelibrary.wiley.com/toc/2637496x/2024/40/1) The above technical specification is mentioned in the article “Measuring Subjective Differences Objectively” by Johan Bergquist in the same issue. https://sid.onlinelibrary.wiley.com/doi/10.1002/msid.1453 Johan Bergquist happens to be one of the members of the technical committee that developed the specification. https://www.iec.ch/dyn/www/f?p=103:14:510370051608773::::FSP_ORG_ID,FSP_LANG... The January/February 2024 issue of the ID Magazine is dedicated to “Understanding Wide Color Gamut Displays”. At this point in time, I am not aware of the presence of any measurement device that is capable of measuring/reporting the OMMI (Observer Metameric Mismatch Index) out of the box. “If they were all looking at the same objects in a viewing booth, what sort of magnitude of potential colour differences would they be seeing?” Viewing real product samples, inkjet contract proofs or real prints in a viewing booth is somewhat a different ball game. Observer metameric mismatch is related with the interaction of the individual CMFs of the observers and the spiky nature of the SPDs of the pigments in the product/inkjet inks and printing inks. We just have to assume that the viewing booth has a full spectrum very high CRI light. Hence the visual mismatch between what the observers is seeing will be related to the spectral structures of the pigments. Visual mismatch takes place when the narrow band peaks of the incoming light hits the CMFs of different individual observers at different points. Best regards, - Refik PS. I have sent this message originally on 28 July 2024. However, the iCloud mailing system only sent the notification today that message has not been delivered. Hope it will reach its destination this time. -----Original Message----- From: Peter Miles via colorsync-users <colorsync-users@lists.apple.com <mailto:colorsync-users@lists.apple.com>> Reply-To: Peter Miles <P.Miles@massey.ac.nz <mailto:P.Miles@massey.ac.nz>> Date: 22 July 2024 Monday 09:07 To: ColorSync <colorsync-users@lists.apple.com <mailto:colorsync-users@lists.apple.com>> Subject: Re: Why 2 degree observer measurements for ICC profiles? Thanks Graeme, and everyone for your responses. A lot of food for thought. Refik I had a looked at the “Individual Colorimetric Observer Model” paper, however I don’t have a training in color science to fully grasp the significance of what I am reading. Is there a way of qualifying what sort of size colour differences these observer variations are causing in the normally sighted population? If they were all looking at the same objects in a viewing booth, what sort of magnitude of potential colour differences would they be seeing? FYI: A few years ago I had a lens replaced in one of my eyes and so I now have two quite different ‘observers’ when it comes to assessing colour. Thanks Peter Miles From: Graeme Gill via colorsync-users <colorsync-users@lists.apple.com <mailto:colorsync-users@lists.apple.com>> Date: Friday, 19 July 2024 at 8:49 PM To: ColorSync <colorsync-users@lists.apple.com <mailto:colorsync-users@lists.apple.com>> Subject: Re: Why 2 degree observer measurements for ICC profiles? Peter Miles via colorsync-users wrote:
So why do ICC profiles, for print production, still use 2degree observer measurements instead of 10 degree observer measurements?
A standard often has a momentum of its own. But in the case of color in media that is looked at in detail (such as images), 2 degree is appropriate. AFAIK, one of the reasons that the 10 degree observer is often recommended is that it was created with more observers and in a more refined way, and is technically a bit more accurate. In subsequent experience it has often turned out that the 10 degree standard matches people more accurately than the earlier 2 degree observer, even though it is not the appropriate viewing angle. Note that there are many subsequent more refined standards or proposed standards such as the 2012 2 and 10 degree observers. These are not in widespread use though, whereas every graphic arts instrument will return 1931 2 degree values. And of course there is nothing to stop you making ICC profiles using any observer you like :-) They may not be technically interchangeable with profiles using the 1931 2 degree observer, but that matters less in practice, and not at all if all the profiles you use the same observer. Cheers, Graeme Gill. _______________________________________________ Do not post admin requests to the list. They will be ignored. colorsync-users mailing list (colorsync-users@lists.apple.com <mailto:colorsync-users@lists.apple.com>) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/colorsync-users/rtelhan%40icloud.com <https://lists.apple.com/mailman/options/colorsync-users/rtelhan%40icloud.com> This email sent to rtelhan@icloud.com <mailto:rtelhan@icloud.com>
participants (6)
-
Claas Bickeböller
-
Graeme Gill
-
Peter Miles
-
Peter Miles
-
Refik Telhan
-
waynebretl＠cox.net