Luminous Landscape Forum
Raw & Post Processing, Printing => Digital Image Processing => Topic started by: Barry Pearson on May 24, 2007, 01:09:24 pm
-
A recent article posted here appears to assume that the purpose of calibrating ACR, (and by extension, Lightroom), using scripts such as the Thomas Fors script, is to cater for differences between cameras of the same model. I disagree!
The article said "they assume that the camera that was used to create the raw processor's profile and your particular camera of the same model have the exact same colour characteristics.... Or, it might be such that the generic profiling done by Phase One, Adobe, or someone else varies enough from what your particular camera does so as to make the creation of a custom profile (or as will be seen, an edited version of that profile) worth having".
http://www.luminous-landscape.com/reviews/...ib-profil.shtml (http://www.luminous-landscape.com/reviews/accessories/dig-calib-profil.shtml)
I believe that the primary purpose of calibrating ACR or Lightroom using such a script is that the default Adobe method of calibration gives results that are unacceptable to many people, while the use of such a script gives results that many of those people prefer. To be very simplistic indeed, these scripts correct the Adobe tendency to give reds that are too orange, and grass that is too yellow.
Have a look at the many results of using such scripts for many camera models that have been published in various forums. They nearly always have a significantly negative Red Hue, and a significantly positive Red Saturation. (The values in the article are respectively: -9; 20). This tendency is obviously systematic, and not to do with within-model variations which would give values either way. I suspect that using such a script on the ACTUAL cameras that Adobe used originally would also give such results!
Here are some results that have been published in various forums:
D200: Red Hue: -24; Red Saturation: 42
D2X: Red Hue: -20; Red Saturation: +20
K100D: Red Hue -27; Red Saturation 40
K10D: Red hue: -13; Red sat: 7 (mine)
*istD: Red Hue: -24; Red Saturation: 27 (mine)
*istDS: Red hue: -26; Red Sat: +35
LX1: red hue -20; red saturation +42
LX1: Red hue: -26; Red sat: +19
A2: Red hue: -26; Red sat: 0
G3: Red hue: -1; Red sat: 20
300D: Red hue: -5; Red sat: 20
20D: Red hue: -9; Red sat: 9
10D: Red hue: -5; Red sat: 30
5D: Red hue: -5; Red sat: 11
5D: Red hue: -11; Red sat: 4
These are not results from anti-Adobe people, or people with no knowledge of colour management. These are typically results from people who own a GretagMacbeth ColorChecker, who care enough about their colour balance to use the ColorChecker and a suitable script to try to get things right, and who then share their results so that others can benefit. We don't have all the answers, but surely we can raise serious questions!
-
I believe that the primary purpose of calibrating ACR or Lightroom using such a script is that the default Adobe method of calibration gives results that are unacceptable to many people, while the use of such a script gives results that many of those people prefer. To be very simplistic indeed, these scripts correct the Adobe tendency to give reds that are too orange, and grass that is too yellow.
[a href=\"index.php?act=findpost&pid=119407\"][{POST_SNAPBACK}][/a]
No...the primary reason for the calibrate tab is to account for variations between sensor lots of a given brand and model of camera. Bruce Fraser and I tested 3 Canon Rebels and found considerable variation in both ISO and color rendering. They required 3 different calibrate settings to bring them all into agreement. And, that's just 3 cameras from the same model-two of the serial numbers on the cameras were close (which means they were "probably" from the same sensor lot).
What this points to is I think there is a WIDE variation between sensors, even in the same lot and same model...perhaps the camera makers ARE doing sensor by sensor adjustments in their proprietary and undocumented file formats-we don't know since it's proprietary and undocumented...and perhaps this is why SOME people feel that the camera software does a more accurate rendering of color than 3rd party software. And this is something the camera makers may not want to admit. I don't know since THEY are staying mum on the whole issue.
This may change in the future...we can only hope. Kodak, when they first shipped a digital camera shipped a "Cal File" that was the calibration file for THAT camera and THAT sensor that was used to make sure the sensor was within spec and provided a method of finger printing that sensor's responce...I think something along these lines, embedded in the capture and metadata is going to be what is required eventually. But all this proprietary business seems to have gotten in the way of a standard method exchanging data on the sensor by sensor variations that we all know occur.
-
No...the primary reason for the calibrate tab is to account for variations between sensor lots of a given brand and model of camera. Bruce Fraser and I tested 3 Canon Rebels and found considerable variation in both ISO and color rendering. They required 3 different calibrate settings to bring them all into agreement. And, that's just 3 cameras from the same model-two of the serial numbers on the cameras were close (which means they were "probably" from the same sensor lot).
What this points to is I think there is a WIDE variation between sensors, even in the same lot and same model...
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=119583\")
There might well be considerable variation in sensors as Jeff asserts, but testing of 3 samples of one camera model does not constitute a statistically significant model for the universe of sensors. There may well be variation between CMOS and CCD or among other manufacturers such as Sony, Dalsa, and Kodak.
[a href=\"http://www.dpreview.com/reviews/nikond40/page17.asp]DPReview[/url] routinely tests the ISO sensitivity of cameras as a part of their testing procedure. The Nikons are usually spot on for ISO whereas Canon often gives a bit more sensitivity than nominal. Not a lot of ISO variation in these tests. Of course these are tests of the camera system and not of the sensor in isolation.
Bill
-
To answer the technical question posted in the title of the thread:
As the article notes, there are two profiles built-into ACR for each supported camera: one for illuminant D65, one for illuminant A. These are matrix-based profiles. Adjusting the ACR calibration knobs moves the actual RGB primaries used in the built-in ACR camera profiles.
-
There might well be considerable variation in sensors as Jeff asserts, but testing of 3 samples of one camera model does not constitute a statistically significant model for the universe of sensors.
[a href=\"index.php?act=findpost&pid=119597\"][{POST_SNAPBACK}][/a]
Obviously...but it does suggest that even with a single model and two sensors whose serial numbers were very close, one should assume there WILL be variations from camera to camera, which is why Thomas put the calibrate function in Camera Raw in the first place. And why Kodak shipped each of their cameras with cal files...I know of no other cameras (other than medium format backs) that are doing that openly.
-
No...the primary reason for the calibrate tab is to account for variations between sensor lots of a given brand and model of camera. [{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=119583\")
That doesn't answer my question, and doesn't respond to the point I made about the near-universal trends of Red Hue and Red Saturation.
Yes, I KNOW that the Calibration tab is intended to cater for differences between specific cameras. But its use has revealed something more significant.
Calibrating using the Fors script, or probably any other, IS NOT primarily catering for differences between specific cameras. If that is what they were doing, there would be a spread of Red Hue and Red Saturation from positive to negative. They would NOT be clustered with significant negative Red Hue and significant positive Red Saturation.
The best proof would be to test the ACTUAL cameras that Adobe used for generating the ACR profiles using the Fors script. My prediction is that they would consistently show a significant negative Red Hue and significant positive Red Saturation. But without access to those cameras, I can't test this for myself.
Many people in forums have complained that the default ACR settings give reds that are too orange, and grass (etc) that is too yellow. (In some quarters, this has become a joke!) That was my own problem until I used the Fors script to calibrate my first dSLR a couple of years ago, and then my latest more recently. These issues can be compensated for by a significant negative Red Hue and significant positive Red Saturation. And the Fors script will tell people to use those. Since using the Fors script, I no longer have that type of problem.
There is a SECONDARY affect of the scripts - catering for within-model variations. In the examples I published at the start of this thread, there were differences even for the same camera model. No one doubts that this is ALSO an issue. But the MAIN issue is that the standard Adobe method of profiling cameras give results that many people don't like, and the Fors script, and others, deliver better results for those people.
Anyone doubting what the scripts are doing needs an answer to the systematic biased results for Red Hue and Red Saturation. I posted some links to sources of those results at:
[a href=\"http://adobe.groupbrowser.com/t165599.html]http://adobe.groupbrowser.com/t165599.html[/url]
-
As the article notes, there are two profiles built-into ACR for each supported camera: one for illuminant D65, one for illuminant A. These are matrix-based profiles. Adjusting the ACR calibration knobs moves the actual RGB primaries used in the built-in ACR camera profiles.
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=119611\")
Yes, I am aware of that, since it is documented in the DNG specification which I know a lot about:
[a href=\"http://www.barrypearson.co.uk/articles/dng/]http://www.barrypearson.co.uk/articles/dng/[/url]
Also in the DNG specification are tags called CameraCalibration1 and CameraCalibration2, specifically for within-model variations. I would like to be able to convert to DNG and add those fields to my DNGs for my particular cameras, and so avoid the need for the Calibration tab. (This would, I hope, mean that the resultant DNGs would be equally applicable to all raw converters that accepted DNGs with their embedded profile).
So what I want to be able to do is calibrate my cameras and feed the results into the Adobe DNG Converter (or equivalent) to create values for those tags.
-
Obviously...but it does suggest that even with a single model and two sensors whose serial numbers were very close, one should assume there WILL be variations from camera to camera, which is why Thomas put the calibrate function in Camera Raw in the first place. And why Kodak shipped each of their cameras with cal files...I know of no other cameras (other than medium format backs) that are doing that openly.
[a href=\"index.php?act=findpost&pid=119613\"][{POST_SNAPBACK}][/a]
The ACR calibration appears to be matrix based. Most camera calibration profiles that I am aware of involve lookup table based profiles and may work well only under controlled conditions and can be quite large. What type of profiles are provided with the Kodak and medium format cameras?
-
What type of profiles are provided with the Kodak and medium format cameras?
[a href=\"index.php?act=findpost&pid=119657\"][{POST_SNAPBACK}][/a]
I don't know that the Kodak "cal files" were real profiles in a standard sense. Pretty sure they were just "reports" in text form of the results of a battery of tests-and not a whole lot of them either-just the basics. Leaf, Phase One and Imacon backs have embedded metadata calibration stats as well.
But if the camera makers could develop a "standard" (ooops, there's that concept again) ISO & color test array and write the results in metadata and document it, it would go a long way towards bringing about camera standards. But they seem more intent on their "LOOK".
Some of their attitudes may be changing though. If you look at the rate of progress and change in the industry, it's slowed way down. Neither Nikon nor Canon are churning out new models nearly as fast as they were. Part of that is the fact that there's not a whole lot left to stuff into a DSLR and part of it is the fact that pretty much everybody who needs a camera has one so they are competing not only against themselves (upgrades to existing cameras) but each other.
I don't know when...but at some point, some smart guy in Tokyo will realize that adopting a standard before the next guy might lead to some additional sales.
We can only hope...
:~)
-
DPReview (http://www.dpreview.com/reviews/nikond40/page17.asp) routinely tests the ISO sensitivity of cameras as a part of their testing procedure. The Nikons are usually spot on for ISO whereas Canon often gives a bit more sensitivity than nominal. Not a lot of ISO variation in these tests. Of course these are tests of the camera system and not of the sensor in isolation.
[a href=\"index.php?act=findpost&pid=119597\"][{POST_SNAPBACK}][/a]
They are tests of the metering, AFAIK. what a camera meters for, and what its sensitivity is (either saturation-based, or noise-based) may be different yet.
I bought a 30D that I returned for an exchange. It metered like my 20D, but it amplified the input going into the ADC 1/3 stop too high, losing 1/3 stop of potential DR in the highlights. Its histogram of a flat wall was at the 55% mark. My other Canons generally fall at the 45% mark on the histogram (green channel or luminance).
-
A plague on both their houses ! (Shak.)
Yes, Jeff, a "standard" is what Adobe says it is ? Adobe refuses to play nice with the ICC input profile standard, preferring something Thomas cooked up on vacation. We've already had plenty of people flame me on that so, yes, I have read, reread and probably even understand the color rendering parts of the DNG code. And it still fails to impress me. It's not that it's wrong, it's that it's not enough. And, yes, I've told Thomas that in person, and so have a few other people in the industry.
Adobe is trying to commoditize the cameras, so of course they want all cameras to look equal, like Microsoft wants all Windows hardware boxes to look equal. That way Adobe gets all the money, the camera makers do all the nasty physical work matching the physical things to what the software already expects. I'm not saying this is necessarily bad, but it does crimp innovation a bit.
The camera makers are going the other way, hoarding data on their hardware like the crown jewels making life unpleasant for anyone writing Raw conversion software. Also they doubtless find that careful calibration and linearization is more trouble than it is worth for consumer models like the Rebel and the 5D. From reports out there by users of my profiles there is at least half a stop variation in exposure alone in the 5D population. The high end models are usually a bit better matched.
Last not least, anyone who has measured Colorcheckers knows that matching charts is not what Gretag does best - I own Colorcheckers that are *immediately and visibly different*; as they told me one day "we have had production problems". This also accounts, no doubt, for some of the variation in the results of the calibration script.
Of course, every amateur who photographs a chart has perfect lighting and no reflections, yeah !
Edmund
I don't know when...but at some point, some smart guy in Tokyo will realize that adopting a standard before the next guy might lead to some additional sales.
:~)
[a href=\"index.php?act=findpost&pid=119723\"][{POST_SNAPBACK}][/a]
-
They are tests of the metering, AFAIK. what a camera meters for, and what its sensitivity is (either saturation-based, or noise-based) may be different yet.
I bought a 30D that I returned for an exchange. It metered like my 20D, but it amplified the input going into the ADC 1/3 stop too high, losing 1/3 stop of potential DR in the highlights. Its histogram of a flat wall was at the 55% mark. My other Canons generally fall at the 45% mark on the histogram (green channel or luminance).
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=119767\")
DPreview does not describe their testing process in detail, but in addition to exposing according to the camera light meter, they expose a gray card according to a reading taken by their Sekonic L-358 hand held meter. If a gray card is exposed under specified conditions, the ISO 12232 saturation speed can be determined as
[a href=\"http://www.normankoren.com/digital_cameras.html]Norman Koren on Digital ISO[/url] explains. With exposure under these conditions an "18% gray card has a voltage or pixel level 18/106 of full scale (e.g., a pixel level of 696 at the output of a 12-bit A-to-D converter, which can represent 4096 levels). In an 8-bit color space encoded for display at gamma = 2.2 (sRGB, Adobe RGB, etc.), the corresponding pixel level is 114 (from the formula, 255*(18/106)(1/2.2) )."
This exposure under specified conditions does not involve the camera's light meter and determines the effective ISO of the camera system, which includes the sensor as well as the amplifier and A-to-D converter, as I mentioned in my previous post. If the pixel level in a gamma 2.2 space with the specified exposure is 114, then the camera system ISO is on spec. If the results of the standardized exposure do not match the results of an exposure using the camera's meter, then the meter is defective.
Bill
-
A plague on both their houses ! (Shak.)
Yes, Jeff, a "standard" is what Adobe says it is ? Adobe refuses to play nice with the ICC input profile standard, preferring something Thomas cooked up on vacation. We've already had plenty of people flame me on that so, yes, I have read, reread and probably even understand the color rendering parts of the DNG code. And it still fails to impress me. It's not that it's wrong, it's that it's not enough. And, yes, I've told Thomas that in person, and so have a few other people in the industry.
[a href=\"index.php?act=findpost&pid=119800\"][{POST_SNAPBACK}][/a]
Edmund,
Adobe does not use a standard ICC profile in their calibration as is done by Capture One and some other raw converters. This has been discussed at length previously by such experts as Bruce Fraser, and the consensus is that such profiling may be useful under controlled studio conditions, but does not work well in general photography.
If you use a lookup based profile under controlled conditions, it is possible to match a target such as the ColorChecker SG reasonably well. If you use a matrix based profile, then you are determining the coefficients for converting from camera CIE XYZ to working space CIE XYZ, and accurate color matching depends on the linearity of the color filters in the camera sensor. If the filters were perfectly linear, then an exact match could be obtained according to Thomas Knoll. Since the filters are not perfect, then a color match may not be attainable for some colors.
Thomas has chosen a calibration method that works well for most users, but if it does not meet your needs, why don't you use Capture One or some other converter that does use ICC profiles?
Bill
-
In my original post for this thread, I pointed out that using the Thomas Fors script appears to result in negative Red Hues, and positive Red Saturations. (In fact, I have never seen a positive Red Hue, or a negative Red Saturation, in many results - more than just the ones in the original post - I've seen posted to forums after using this script!)
Furthermore, and obviously important, people appear to prefer the result of using the Calibrate tab values identified by running the script. It typically improves a problem that a number of people have complained about in forums, that the default Adobe profiles give reds that are often too orange, and greens that are often too yellow.
If true, then this offers the chance of identifying generic Calibrate tab values per camera model, which would help people without a GretagMacbeth ColorChecker, and/or people who use Lightroom but don't have ACR+Photoshop. We could have camera-model presets for ACR or Lightroom. (Or an evolution of Adobe's profiling).
I've just started a thread in Thomas Fors' forum for the AcrCalibrate script:
http://groups.google.com/group/acrcalibrat...5020c77c83e07dd (http://groups.google.com/group/acrcalibrator/browse_frm/thread/45020c77c83e07dd)
In it, I asked that people post their results from running this script, and say whether they prefer the results to the default Adobe value. I repeat that request here:
Please post (here or there) your results from running the Thomas Fors script (or the Rags Gardner one, etc). Let's gather more data to see whether this effect is real, and to draw some conclusions.
-
Adobe is trying to commoditize the cameras, so of course they want all cameras to look equal, like Microsoft wants all Windows hardware boxes to look equal. That way Adobe gets all the money, the camera makers do all the nasty physical work matching the physical things to what the software already expects. I'm not saying this is necessarily bad, but it does crimp innovation a bit.
[a href=\"index.php?act=findpost&pid=119800\"][{POST_SNAPBACK}][/a]
Ya know, you "think" you know what Adobe is trying to do, but you're not even sneaking up on having half a clue doode.
Adobe is doing DNG because Thomas Knoll thinks it's the right thing for the industry. There is no money in developing DNG. They aren't making any money on DNG now and they don't expect to in the future. There is no money changing hands when a camera vendor uses DNG nor when a user converts to DNG. You have it in your head that's DNG is an attempt by Adobe to engage in monopolistic practice? Well, that's YOUR baggage you're bringing to the table.
There was exactly one guy-Thomas Knoll-that came up with the concept and idea of producing a raw file format standard. Not to make money...Thomas doesn't need to make money in that way. He thought it was the right thing to do for the photo industry. I happen to believe he's correct.
Other than Dave Coffin and perhaps Eric Bibble, there's nobody else on the planet that has as much knowledge and experience in dealing with raw file formats. He's already worked with Coffin, and Eric is a story "not for public consumption".
There is one other guy that also has a lot of experience dealing with raw file formats-Michael Jonnson-formerly of Phase One then pixmantic. He's also now working on Camera Raw/DNG along with Zalman Stern (who used to work on Photoshop then spent a decade or so at Macromedia).
You keep talking like there's this big attempt in the boardrooms at Adobe to plot to "commoditize" camera maker's file formats...uh, no. You simply couldn't be further than the truth.
Adobe is just trying to do the right thing when it comes to the long term conservation and preservation of digital files and that includes PDF-A as well as DNG and the old and venerable (although some say outdated) Tiff.
If you want to see an example of Adobe's true motives, look at what Adobe did for the ISO regarding Tiff-EP. They granted the ISO the right (and waived or set aside their patent rights) to incorporate Tiff-EP as a public standard-which was turned around and adopted by Nikon and Canon for their proprietary and undocumented file formats-see recent NEFs and CR2s are so close to being DNG by virtue of being rather fully formed Tiff-EP files that the big joke is that NEFs and CR2s are already DNGs except for the standardized location of metadata and meaning of the metadata.
You may think you know what Adobe is trying to do-but you don't. And, it would be real useful if you did get a clue pretty quick cause misstating Adobe's motive simply gives aid and comfort to the enemy...undocumented and proprietary file formats.
The sooner the photo industry realizes just how dangerous proprietary and undocumented file formats are to the future of the photo industry, the sooner the camera companies will be forced to address the issue.
-
Bill,
Disagreeing with a consensus is sometimes a useful strategy, and one with an implicit attraction to a contrarian
It's an interesting academic question whether it's the separability of the color filters or the non-linearities of the sensor electronics which creates the most practical issues with matrix models. I have heard it said that the high sensitivity of one major vendor's dSLR sensors is incurred at the expense of separability.
Regardless: Real engineering will evidence non-linearities, and I believe Thomas relies excessively on the camera makers exposing utopically perfect data. Forcing camera makers to market externally linear devices is a good political strategy for Adobe, but not so useful to present day users.
I expect that realism and a transition to a more complete model will intervene soon. Of course, Adobe won't be the last company to market a new version as solving some inadequacies that a previous version introduced as a triumphant breakthrough.
As for your suggestion to use C1 - I find it so good that I have been distributing C1 profiles for the M8 for a few months, and these are in widespread use in the Leica community.
I have even given a talk about this to Xrite at Regensdorf, who have -with great generosity and a great sense of humor- supported my efforts. The upshot is that as you say chart-based profiling often has issues, but my profiles seem to be useful to many users. We don't know exactly why I happen to be successful practicing an approach that has, as you indicate, often disappointed.
Edmund
Edmund,
Adobe does not use a standard ICC profile in their calibration as is done by Capture One and some other raw converters. This has been discussed at length previously by such experts as Bruce Fraser, and the consensus is that such profiling may be useful under controlled studio conditions, but does not work well in general photography.
If you use a lookup based profile under controlled conditions, it is possible to match a target such as the ColorChecker SG reasonably well. If you use a matrix based profile, then you are determining the coefficients for converting from camera CIE XYZ to working space CIE XYZ, and accurate color matching depends on the linearity of the color filters in the camera sensor. If the filters were perfectly linear, then an exact match could be obtained according to Thomas Knoll. Since the filters are not perfect, then a color match may not be attainable for some colors.
Thomas has chosen a calibration method that works well for most users, but if it does not meet your needs, why don't you use Capture One or some other converter that does use ICC profiles?
Bill
[a href=\"index.php?act=findpost&pid=119811\"][{POST_SNAPBACK}][/a]
-
You may think you know what Adobe is trying to do-but you don't. And, it would be real useful if you did get a clue pretty quick cause misstating Adobe's motive simply gives aid and comfort to the enemy...undocumented and proprietary file formats.
The sooner the photo industry realizes just how dangerous proprietary and undocumented file formats are to the future of the photo industry, the sooner the camera companies will be forced to address the issue.
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=119859\")
Jeff,
I certainly don't know much about what is happening at Adobe. Indeed, Adobe is to praise for releasing documentation and source code, and freely licensing DNG. I have publicly apologized in the past because I did not realize initially that DNG is completely freely licensed and some converter code provided.
However, regarding color models I do think that isssues of digital color have an existing venue for their discussion that happens to be the ICC, which has made a practice of releasing reasonably well documented standards and even sample code. ICC profiles are not closed technology as far as I can see.
All of us abhor the undocumented file formats. But forcing an incomplete color model on the rest of the industry sounds a bit like burning the village to save it.
I have written an essay entitled [a href=\"http://photofeedback.blogspot.com/2007_02_01_archive.html]Raw: The Adobe Way, The Microsoft Way and the Open Way[/url] which may prove entertaining to some here.
Edmund
-
Adobe is just trying to do the right thing when it comes to the long term conservation and preservation of digital files and that includes PDF-A as well as DNG and the old and venerable (although some say outdated) Tiff.
If you want to see an example of Adobe's true motives, look at what Adobe did for the ISO regarding Tiff-EP. They granted the ISO the right (and waived or set aside their patent rights) to incorporate Tiff-EP as a public standard-which was turned around and adopted by Nikon and Canon for their proprietary and undocumented file formats-see recent NEFs and CR2s are so close to being DNG by virtue of being rather fully formed Tiff-EP files that the big joke is that NEFs and CR2s are already DNGs except for the standardized location of metadata and meaning of the metadata.
[a href=\"index.php?act=findpost&pid=119859\"][{POST_SNAPBACK}][/a]
I agree with your view there, but I'll clarify and qualify it a bit.
Adobe granted ISO the right to incorporate TIFF 6.0 (which Adobe owns) into a public standard, which became ISO 12234-2 (TIFF/EP).
DNG, and apparently NEF, CR2, and perhaps others, were then based on TIFF/EP. Examine a NEF with "dng_analyse" from the DNG SDK, and it shows its tags and values. I suspect, as you imply, that converting a NEF to a DNG is little more than re-arranging the data and adding the unique DNG camera profile metadata. (For interest, DNG and CR2 both use JPEG lossless compression for the raw image data). Contrary to statements that using DNG would inhibit the innovation of Canon and Nikon, it would probably be pretty trivial for them and be no inhibition at all.
A recent email from the ISO working group revising ISO 12234-2 (TIFF/EP) says: "The Adobe DNG format was derived from this standard and the group has Adobe's permission to incorporate modifications and developments made for DNG in the new standard". Perhaps the next version of ISO 12234-2 will be much closer to DNG?
DNG is, as far as I know, the ONLY raw file format specifically designed to be suitable for archiving. It would be helpful if ISO made that clear, and exploited it, because there are lots of national libraries and archives that would like to adopt an ISO standard archival raw file format. They would nominate DNG immediately if it had ISO's blessing as an archival format. (As the US Library of Congress already has).
-
I certainly don't know much about what is happening at Adobe. Indeed, Adobe is to praise for releasing documentation and source code, and freely licensing DNG. I have publicly apologized in the past because I did not realize initially that DNG is completely freely licensed and some converter code provided.
[a href=\"index.php?act=findpost&pid=119878\"][{POST_SNAPBACK}][/a]
So, why do you insist in characterizing Adobe's reasons for DNG as something south of noble?
Adobe is trying to commoditize the cameras, so of course they want all cameras to look equal, like Microsoft wants all Windows hardware boxes to look equal. That way Adobe gets all the money, the camera makers do all the nasty physical work matching the physical things to what the software already expects. I'm not saying this is necessarily bad, but it does crimp innovation a bit.
And that is what you said earlier in the thread...this is a halfbaked economic model you are STILL trying to paint Adobe with-even though you now admit Adobe is giving DNG away. You still don't get it.
Unfortuantely, to even answer you tends to give credence to your crackpot ideas-which is why, in the past, I've let you say what you will, knowing full well you didn't have a clue. But you keep coming around and totally misrepresenting what Adobe (ie: Thomas Knoll) is trying to do and why he's trying to do it.
And your "post" regarding Coffin's Dcraw also is wrong. Have you asked Davis what he thinks of DNG? Dcraw actually benefits since he now has color processing data on other cameras where he hasn't had to do the work-Thomas has.
You also pretty much mis-characterize MSFT's efforts at providing raw access via a raw codec. This was MSFT's punt...they bowed to the pressures of the camera makers by letting them keep their stuff under wraps while exposing the default rendered raw file for use in Windows....and only Windows. This isn't some sort of nice thing for the photo industry, this is MSFTs dodge around the whole issue. Apple on the other hand went a different direction by basically using Coffin's Dcraw for both the OS as well as basic raw support of Aperture.
Unfortunately, you write about this stuff as though you know what's going on in the trenches...and you don't. As a result, you end up being part of the problem, not the solution. You are spreading FUD. And, I can't quite figure out why...cause if you THINK you are doing the industry a service, you are wrong-so I can only guess that you have some other agenda...
-
DNG, and apparently NEF, CR2, and perhaps others, were then based on TIFF/EP. Examine a NEF with "dng_analyse" from the DNG SDK, and it shows its tags and values. I suspect, as you imply, that converting a NEF to a DNG is little more than re-arranging the data and adding the unique DNG camera profile metadata. (For interest, DNG and CR2 both use JPEG lossless compression for the raw image data). Contrary to statements that using DNG would inhibit the innovation of Canon and Nikon, it would probably be pretty trivial for them and be no inhibition at all.
[a href=\"index.php?act=findpost&pid=119882\"][{POST_SNAPBACK}][/a]
Absolutely...and I would argue that the reason (at least one big reason) that recent NEF and CR2 formats from newer Nikon and Canon camera _ARE_ better is because of DNG.
I know for a fact that both Nikon and Canon have studied DNG very closely. And while they have yet to bend to pressure to adopt DNG, they HAVE been influenced by DNG if for no other reason to improve their own raw file formats-even if they remain propriatary and undocumented. In other words, DNG has taught them a few things...
I'm not aware of what's going on at the ISO with regards to a TIFF-EP update-other than there's something going on-finally. Which is good news I suppose.
-
All of us abhor the undocumented file formats. But forcing an incomplete color model on the rest of the industry sounds a bit like burning the village to save it.
I have written an essay entitled Raw: The Adobe Way, The Microsoft Way and the Open Way which may prove entertaining to some here.[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=119878\")
Your article "Raw: The Adobe way, the Microsoft way and the Open way" advocates using dcraw:
"I suggest the Open Source community and maybe even the OpenRaw guys write or commission dcraw plugins for Microsoft Windows and Adobe's software, so that dcraw becomes an additional input module to all existing and future software, thereby future-proofing it. In effect, dcraw would become the ACR of the open source community".
dcraw is based on the Adobe colour model!
Dave Coffin: "Adobe Digital Negative (DNG) is a great format -- I totally redesigned dcraw for maximum DNG compatibility".
[a href=\"http://www.dpreview.com/news/0504/05042701davecoffininterview.asp]http://www.dpreview.com/news/0504/05042701...ininterview.asp[/url]
Dave Coffin: "Not only is Adobe DNG now supported, the entire codepath has been redesigned for it. Adobe's XYZ->CAM matrices allow color science to replace black magic, whether decoding DNG or the original raw files".
http://forums.dpreview.com/forums/read.asp...essage=20585662 (http://forums.dpreview.com/forums/read.asp?forum=1018&message=20585662)
dcraw appears to build-in inverses of the Adobe colour matrices as found in DNG files. Look for "void CLASS adobe_coeff".
ps: OpenRAW (which you mention) appears to be in terminal decline. They made themselves irrelevant by their hostility towards DNG.
-
I'm not aware of what's going on at the ISO with regards to a TIFF-EP update-other than there's something going on-finally. Which is good news I suppose.
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=119887\")
You and me both!
ISO produced a draft for TIFF/EP in 1998. It was published as a standard (ISO 12234-2) in 2001. (Which costs a significant amount of money). It began its 5-year revision in 2006, and that is still ongoing. Perhaps the revision will be published within a year or two.
Anyone who believes that ISO rather than Adobe is the place to develop a format such as DNG should contemplate those dates.
[a href=\"http://www.barrypearson.co.uk/articles/dng/history.htm]http://www.barrypearson.co.uk/articles/dng/history.htm[/url]
Peter Krogh said: "As to the standard being owned by Adobe, I checked into that as well. The response I got was that the standards bodies don't want it until it's done. If you know someone else out there who is willing to fund and manage the effort, I'm sure the folks at Adobe would be interested in a discussion. At this point nobody is interested in doing that work".
-
ps: OpenRAW (which you mention) appears to be in terminal decline. They made themselves irrelevant by their hostility towards DNG.
[a href=\"index.php?act=findpost&pid=119888\"][{POST_SNAPBACK}][/a]
Can't keep the troops from fragging their officers these days. Terrible world we live in
Edmund
-
All this discussion about what ACR is doing, is technical and interesting, but what about the visual psych? I fight a constant battle to leave the color temp warm enough. It has taken me a year to realize the Canon's "as shot" generally does a better (warmer) job than I do. There is something appealing that happens to skin when when you take the "as shot" of 4900 and move it down to 4600. And even though it is often very appealing, I don't think the lower setting is as correct as it often seems. But why does it look good? It's a trap. It's quicksand. It recenters your color judgement way off track. It leads to that prutrid green that was discussed in this thread. The mind is incredibly difficult to keep objective in this adjustment.
Maybe we are pushed to this because we are compensating for a camera that has not been calibrated. The way Adobe processing things? Hugh? Couldn't that be verified by carefully comparing ACR with other RAW processors? What about just a tendency in cameras because of some kind of technical limitations?
Neal
-
However, regarding color models I do think that isssues of digital color have an existing venue for their discussion that happens to be the ICC, which has made a practice of releasing reasonably well documented standards and even sample code. ICC profiles are not closed technology as far as I can see.
Edmund
[a href=\"index.php?act=findpost&pid=119878\"][{POST_SNAPBACK}][/a]
Edmund, have you ever sat in on any ICC meetings about digital camera technology? I have, Thomas, Eric (from Bibble) are both regulars. The ICC has their hearts in the right place. But if a high tech company like Adobe moved at this pace, we'd be on about Photoshop 3 these days.
Look at what's up with the V4 spec (how long has that been around) and what its supposed to do to reduce ambiguity among profile vendors. They recommend using a specific colorspace for communication with the PCS now called PRMG but the vendor doesn't have to use it (they can do what they please). Tight standards are not something you're going to see from this body.
-
Andrew,
I realize the ICC moves at the speed of a beard growing on a woman's face
But looking at Adobe sidelining them on camera color with ACR while Microsoft does an end run around them for screen and print (Windows Color System) should shake them up a bit, one would think.
Edmund
Edmund, have you ever sat in on any ICC meetings about digital camera technology? I have, Thomas, Eric (from Bibble) are both regulars. The ICC has their hearts in the right place. But if a high tech company like Adobe moved at this pace, we'd be on about Photoshop 3 these days.
Look at what's up with the V4 spec (how long has that been around) and what its supposed to do to reduce ambiguity among profile vendors. They recommend using a specific colorspace for communication with the PCS now called PRMG but the vendor doesn't have to use it (they can do what they please). Tight standards are not something you're going to see from this body.
[a href=\"index.php?act=findpost&pid=120155\"][{POST_SNAPBACK}][/a]
-
In my original post in this thread, I identified that ACR Calibration using the THomas Fors script wasn't, as typically claimed, primarily concerned with variations between different cameras of the same model.
Instead, it appears to provide a systematic change to the colour balance compared with the Adobe default balance - typically reducing the tendency for reds to be too orange, and greens too yellow. (Camera variations are a secondary feature).
I have now obtained more evidence of a consistent bias in the settings resulting from running the Thomas Fors script. I often download raw files that people have posted so that I can examine them. Some of those files are of images of a GertagMacbeth ColorChecker. So I have just run the Thomas Fors script (version 1.0) against those raw files, using ACR 4.0. Here are just the results for the Red sliders:
Canon 350D: Red Hue: -18; Red Sat: 26
Canon 1Ds II: Red Hue: -16; Red Sat: 5
Leaf Aptus 17: Red Hue: -37; Red Sat: 100
Nikon Coolpix 5400: Red Hue: -26; Red Sat: 40
Nikon D70: Red Hue: -15; Red Sat: 8
Nikon D70s: Red Hue: -15; Red Sat: 8
All of those show a negative Red Hue and positive Red Sat, which is a typical result of running such a script. But there was a very interesting exception:
Sigma SD10: Red Hue: 51; Red Sat: -29
That is the first time I've seen a positive Red Hue or a negative Red Sat from one of these scripts. A characteristic of the Sigma X3F raw files is that Adobe use Sigma/Foveon code to do the first stage of colour handling. In other words, when code other than Adobe's is involved, the scripts no longer have the same consistent trend.
In fact, another exception came from the Hasselblad-Imacon H2D camera, which uses DNG as its raw file format, and so has colour matrices supplied by H-I not Adobe:
Hasselblad-Imacon H2: Red Hue: 2; Red Sat: -14
Once Adobe do not control the colours directly, either because Sigma/Foveon do so, or Hasselblad-Imacon do, different results appear.
I ask people here to post their results from doing an ACR Calibration using the Thomas Fors script or similar. Let's try to see what is happening.
-
But looking at Adobe sidelining them on camera color with ACR while Microsoft does an end run around them for screen and print (Windows Color System) should shake them up a bit, one would think.
Edmund
[a href=\"index.php?act=findpost&pid=120157\"][{POST_SNAPBACK}][/a]
What exactly has Microsoft done that anyone can use?
-
What exactly has Microsoft done that anyone can use?
[a href=\"index.php?act=findpost&pid=120162\"][{POST_SNAPBACK}][/a]
Nothing that I know of - but they are certainly threatening to do something.
Old joke: Two chessmasters at a tournament, smoking has been agreed is forbidden. One GM pulls out a huge cigar and places it in his mouth, moves it from mouth to hand etc without lighting it. Second GM appeals to referee - who replies "But, he isn't smoking" And the GM, trembling "But he is threatening to smoke".
Edmund
-
Nothing that I know of - but they are certainly threatening to do something.
[a href=\"index.php?act=findpost&pid=120164\"][{POST_SNAPBACK}][/a]
Great. When you see them actually produce something that's not too messy, you let me know.
-
Great. When you see them actually produce something that's not too messy, you let me know.
[a href=\"index.php?act=findpost&pid=120165\"][{POST_SNAPBACK}][/a]
Non-messy Microsoft software ? hahaha.
Edmund
-
A recent article posted here appears to assume that the purpose of calibrating ACR, (and by extension, Lightroom), using scripts such as the Thomas Fors script, is to cater for differences between cameras of the same model. I disagree!
I believe that the primary purpose of calibrating ACR or Lightroom using such a script is that the default Adobe method of calibration gives results that are unacceptable to many people, while the use of such a script gives results that many of those people prefer. To be very simplistic indeed, these scripts correct the Adobe tendency to give reds that are too orange, and grass that is too yellow.
Have a look at the many results of using such scripts for many camera models that have been published in various forums. They nearly always have a significantly negative Red Hue, and a significantly positive Red Saturation. (The values in the ACTUAL cameras that Adobe used originally would also give such results!article are respectively: -9; 20). This tendency is obviously systematic, and not to do with within-model variations which would give values either way. I suspect that using such a script on the
These are not results from anti-Adobe people, or people with no knowledge of colour management. These are typically results from people who own a GretagMacbeth ColorChecker, who care enough about their colour balance to use the ColorChecker and a suitable script to try to get things right, and who then share their results so that others can benefit. We don't have all the answers, but surely we can raise serious questions!
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=119407\")
Most of us agree that the purpose of calibration with the Fors script or using Bruce Fraser's manual method is to cause the pixel values of the patches to be as close to the published nominal values as possible. The above methods adjust for only the primary colors of the red, blue, and green patches, and one hopes that the values for the other patches will also fall into line.
I used the most recent Fors script (ver 1.0) and ACR 4.0 to calibrate my Nikon D200 and got the following values: tint -1, red hue -20, red sat 21, green hue -4, green sat 8, blue hue 6, and blue sat 2. The red values are in the range Barry has documented. If the purpose of the calibration is merely to compensate for variations from camera to camera in the same model, the values of many calibrations with different cameras should have a bell shaped distribution with a mean of zero. Pending further data, it would seem to me that Barry is on to something.
The standard way to evaluate color error is with ΔE*ab, which is the Euclidian distance between the measured and ideal (reference) values in the a*b* plane of the CIELAB color space as explained by [a href=\"http://www.imatest.com/docs/tour_colorcheck.html]Norman Koren[/url] on his Imatest web site. His Imatest program can plot the ΔEs for the Macbeth Color Checker. Results for many cameras are given on the Imaging Resource (http://www.imaging-resource.com/PRODS/E5D/E5DIMATEST.HTM) web site, and a link to results for the Canon EOS 5D is shown. Readers should refer to the Imatest web site and the Imaging Resource web site for information on the interpretation of these plots. Color error can be with Chroma (saturation), in which case ΔE lies on a radial line extending from the white point, or with hue, where ΔE is not on this radial line. Increased saturation (positive chroma error) is not necessarily bad: many people prefer saturated blue skies and green grass. However, hue error in the form of purple skies or yellowish grass is usually not desired.
Here are plots of my recent calibration for discussion. Exposure was in direct sunlight with the checker mounted on a black background and care was exercised to that no colored objects that could reflect colored light were near the target and exposure was made according to the Imatest suggestions. The red, blue, and green patches are 13, 14, and 15 on the plots.
ACR default. Note the reds are in the yellow direction as Barry noted.
[attachment=2573:attachment]
ACR calibration with brightness 36 and contrast -34 as determined by the script for the low contrast target:
[attachment=2575:attachment]
ACR calibration with brightness and contrast at their default ACR values of +50 and +25:
[attachment=2576:attachment]
Nikon Capture NX for comparison:
[attachment=2577:attachment]
The Nikon Capture rendering is not that much different from that of the default ACR. The script calibrated results with the contrast and brightness as set by the script gave a very accurate result, but one does not use these contrast and brightness settings for normal photography. With the ACR default contrast and brightness but with the script calibration settings, the overall saturation is increased as indicated by the chroma of 134% as reported by Imatest. However, the chroma corrected color difference (ΔC) is quite small. The increased saturation could be handled easily in ACR or Photoshop with the saturation control, and the calibration was quite successful.
Comments and additional calibration results are welcome.
Bill
-
Hi Bill,
Thanks for the data. BTW, the delta-E 1976 metric (the Euclidean metric you mention) is a bit dated since many recent studies have shown that CIE L*a*b* is not perceptually uniform. (I'm sure you've seen Bruce Lindbloom's article on this.) Does Imatest let you use some of the newer metrics such as delta E CMC 2:1?
Eric
-
Hi Bill,
Thanks for the data. BTW, the delta-E 1976 metric (the Euclidean metric you mention) is a bit dated since many recent studies have shown that CIE L*a*b* is not perceptually uniform. (I'm sure you've seen Bruce Lindbloom's article on this.) Does Imatest let you use some of the newer metrics such as delta E CMC 2:1?
Eric
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=120529\")
Eric,
Actually, I used the delta-E 1996, which is reportedly a bit better. Imatest also has a CMC metric, but I am not familiar with its specifics. Here is the [a href=\"http://www.imatest.com/docs/colorcheck_ref.html#colorerr]Imatest[/url] reference.
Bill
-
Most of us agree that the purpose of calibration with the Fors script or using Bruce Fraser's manual method is to cause the pixel values of the patches to be as close to the published nominal values as possible. The above methods adjust for only the primary colors of the red, blue, and green patches, and one hopes that the values for the other patches will also fall into line.
Out of interest, and with reference to your results: how many degree of HSB-hue does the Light-skin patch no2 become worse due to calibration? Could it be that the hue-accuracy of this memory color is sacrificed to get dead-on Red... At least (and as expected) this trend seems to be obvious from the imatest plots
Btw, great analysis.
Peter
--
-
Out of interest, and with reference to your results: how many degree of HSB-hue does the Light-skin patch no2 become worse due to calibration? Could it be that the hue-accuracy of this memory color is sacrificed to get dead-on Red... At least (and as expected) this trend seems to be obvious from the imatest plots
Btw, great analysis.
Peter
--
[a href=\"index.php?act=findpost&pid=120657\"][{POST_SNAPBACK}][/a]
Peter,
Thanks for the kind words.
According to the HSB of the Photoshop color picker and the Imatest results output to a CSV file for analysis, the ideal HSB values for the skin patch are 32, 28, 61. The HSB values for ACR without calibration are 25, 24, 70 and with the calibration they are 20, 22, 76. So the calibration worsens the skin color by 7 degrees of hue. So your observation is a good one.
Since skin tones are a crucial memory color, the results of the calibration are not all good.
Bill
-
Hi,
Thanks for the OP and others who have contributed to this interesting topic.
Skin tone reproduction sounds like a very probable and obvious reason for overall behavior of the profiles built in ACR, especially if we think that the majority of photographs shot in the world represent people (or their cats/dogs ). The interesting Imatest plots posted by Bjanes show what many have observed, that is a hue shift of reds and greens towards yellow. In commercial printed material (at least in fashion related advertising) a trend can be seen towards a quite yellow or "golden" tone in skin color that many find very pleasing although it is rarely faithful to "reality". As Andrew Rodney has many times here pointed out most people do neither want nor need scene referred color representation or accuracy. Thus it could be the case that the default color rendering in ACR is directed to make it relatively easy for the "average" user (or a professional digital retoucher) to get at least decent skin color in most photographic situations. Also it would seem logical that the camera manufacturers would tweak the color response of their imaging sensors towards similar direction.
Although in most cases we indeed want pleasing not accurate color, there is a huge number of dslr-cameras used for both commercial product and fashion photography today. In many cases there are very different needs in color reproduction between these (and other) photographic pursuits. Especially in product photography the way to get relatively easily a good perceived match between the objects color and the color in both the file and printed material would be a highly desired situation. Some very succesful product photographers insist on using camera calibration and icc-profiles as the only way to create this match (although proper camera calibration is certainly not an easy or trivial task. Some people also claim that a well produced and edited daylight balanced camera profile may be used in all normal shooting situations to produce also pleasing color if carefully gray-balanced for ambient lighting conditions, ie. not only in studio).
Thus it might be useful to have a set of different "profiles" for different purposes available in ACR. This could be done by making it possible to use icc-profiles or some other system to have "presets" for different image content for different camera models. Although any particular way of dealing with color may not be perfect for all possible situations, there are more people doing some sort of product or repro photography these days than most people think. Similarly the need for relatively accurate (scene referred) color reproduction without the emphasis on skin tone reproduction is also in great demand. Although one might say that scanning backs and multi exposure MFDBs are the only correct tools for that kind of work, there are many situations where the IQ of the best dslr's is certainly good enough if the color reproduction works well enough.
Personally I would like to hear real world experiences when the Gretag/ACR-calibration script method has worked well (especially for demanding product photography). Also it might be useful to hear comments if the ACR-calibration has failed.
I haven't been too succesful with this aproach (luckily I haven't really needed super accuracy yet...) and I have been wondering if the reason is in my colorchart (I haven't had it measured yet) or in something else I have done. I have tried 10D, 1Ds, 1Ds2, studio flash, sunlight (photographed on a top of a hill without any color reflecting objects near), color corrected daylight fluorescent, tungsten halogen, evennes of lighting within 0-2 rgb levels in the whole chart area in all cases, have tried different lighting directions and types of light (diffused vs harsh). The bottom line is that although the calibration settings are allways towards the same direction (and similar to the settings discussed in earlier posts) and the overal accuracy has been better after running the script, some colors and most notably the red and orange patches are allways off and usually visually worse after calibration. There is also a notable difference between different camera model's color response when photographed in the exact same lighting and the same lens, at least between 10D and 1Ds2 the latter being more neutral as a default. Usually after calibration the overal delta E has been around 2 according to the script.
So my question is if there is a way to edit the target values in the script if after measuring the particular chart in use there happens to be a meaningfull difference between the measured values and the script's default values?
Thanks for any comments.
Cheers,
Juicy
-
What this points to is I think there is a WIDE variation between sensors, even in the same lot and same model...[a href=\"index.php?act=findpost&pid=119583\"][{POST_SNAPBACK}][/a]
Maybe, but... I daylight calibrated my first 5D, then 6 months later got a second. I applied the original calibration to the new camera to test it and avoid having to repeat the calibration proceedure. The converted files are indistinguishable using the same lens on either camera. The worst color difference was 2 points in the R channel and most other patches were either equal or only one point off in one or two channels. Expoure in Av Matrix was barely measurable at less than 1/10th stop different... Moreover, I supplied that calibration to four of my shooting buddies who also shoot 5D's and they all claim it is "excellent" with their cameras.
So I supect some INTERNAL profiling may be happening with some cameras before they leave the factory -- the performance between this many different 5D's is too consitent to believe otherwise.
Cheers,
-
Hi Bill, thought I'd add some grist to the mill. The calibration results are about 6-8 weeks old, and as you will see not so far away from yours. The chart was inside in bright, naturally lit room generally free of reflection.
For what it's worth I've added the first few digits of the D200 serial.
D200 8022xxx
AcrCalibrator V1.0 (x89)
ACR 4.0
tint, -2
red hue, -20
red sat, 29
green hue, -2
green sat, 13
blue hue, 8
blue sat, 3
-
Hi Bill, thought I'd add some grist to the mill. The calibration results are about 6-8 weeks old, and as you will see not so far away from yours. The chart was inside in bright, naturally lit room generally free of reflection.
For what it's worth I've added the first few digits of the D200 serial.
D200 8022xxx
AcrCalibrator V1.0 (x89)
ACR 4.0
tint, -2
red hue, -20
red sat, 29
green hue, -2
green sat, 13
blue hue, 8
blue sat, 3
[a href=\"index.php?act=findpost&pid=120848\"][{POST_SNAPBACK}][/a]
Greg,
Thanks for the feedback. The results are pretty similar and show a strong bias in the red hue and saturation as the OP noted. FWIW, the SN on my camera is 3005xxx. I wouldn't want to over interpret the data, but if we got more data, we could calculate a standard deviation and confidence interval.
Bill
-
Peter,
Thanks for the kind words.
According to the HSB of the Photoshop color picker and the Imatest results output to a CSV file for analysis, the ideal HSB values for the skin patch are 32, 28, 61. The HSB values for ACR without calibration are 25, 24, 70 and with the calibration they are 20, 22, 76. So the calibration worsens the skin color by 7 degrees of hue. So your observation is a good one.
Since skin tones are a crucial memory color, the results of the calibration are not all good.
Bill,
Many thanks for the additional data. I have to agree that even for a state-of-the art camera (sensor), the combined hue error: /dH-Red/ + /dH-Lightskin/ can easily be 10 deg. or more. So the Fors script may have its value e.g. for Ferraristi (people who like these fast red cars) ...
Best regards, Peter
--
-
"Could it be that the hue-accuracy of this memory color is sacrificed to get dead-on Red..."
Peter might be on to something... according to this: http://photoshopnews.com/2005/05/01/photos...e-real-history/ (http://photoshopnews.com/2005/05/01/photoshop-widows-club-part-two%e2%80%93camera-raw-the-real-history/) , we can maybe say the opposite, that the pure reds have maybe been sacrified for the (ears) skintones in ACR?
J/K
-
FWIW, here are the ACR calibration values I've gotten for red hue and saturation from the Fors script, by camera model:
Olympus SP-350:
-19, -3
Canon 10D:
-6, +16
Canon 1D-MkII:
-20, +27
Canon 1Ds:
-12, +70
I'm not sure I buy into the Adobe conspiracy theory though. I have no opinion regarding why the red hue/sat values aren't distributed arouind zero, but I don't really care. There's a lot to be said for the notion of going to a job with several cameras and getting closely similar and acceptaby accurate color from all of them straight from the RAW converter. For me, the practical benefits of getting reasonably consistent and accurate color regardless of whether it was shot by an assistant with a P&S or myself with a DSLR are high, and making conspiratorial accusations about the people & companies who made it possible seems a bit counterproductive.
What I'd like to see discussed instead is ways to improve the process, like by integrating the calibration/profiling process directly into ACR, so that one can shoot a RAW of a Color Checker or other known reference target, and be able to either cook up new calibration values, or create a custom profile to replace the internal ACR ones on a by-camera basis straight from the Calibration tab. I'm sure that would speed up the calibration/profiling process at least 20-fold, and probably improve the color accuracy greatly as well. Have a list of definition files for the standard and enhanced Color Checkers (as well as other common profiling reference targets) in a documented format, and be able to add new ones at will simply by dumping them in the right folder. Hell, make it a cooperative effort between Adobe and the profiling target makers, so that each target comes with its own definition file containing actual measured patch values to eliminate the problem of variations between batches. I'd gladly pay to upgrade to CS4 for that feature alone.
-
... I'm not sure I buy into the Adobe conspiracy theory though. I have no opinion regarding why the red hue/sat values aren't distributed arouind zero, but I don't really care. ...
[a href=\"index.php?act=findpost&pid=140743\"][{POST_SNAPBACK}][/a]
It's not just that the red hue/sat values aren't distributed arouind zero. For all the Nikon and Canon results posted in this thread, the red hue/sat corrections are in the same direction (my 1DII is -23/+48).
DPP 3.0.2.6 is doing a better job with regard to the ColorChecker red square (#15) than ACR 4.2, before or after running AcrCalibrator 1.1.
Ken
-
It's not just that the red hue/sat values aren't distributed arouind zero. For all the Nikon and Canon results posted in this thread, the red hue/sat corrections are in the same direction (my 1DII is -23/+48).
Statistically speaking, those statements are exactly the same. If the distribution of values is centered above or below zero, then the probability of negative values is going to be higher than positive ones, or vice versa. In this situation, the center point of the distribution of values is significantly away from zero, so that most of the values are have the same sign. One requires the other mathematically.
But before breaking out the tinfoil hats over this, comparing a decent sample size of Olympus, Konica-Minolta, Pentax, Fuji, Sigma and other camera brands would probably be a good idea.
-
Statistically speaking, those statements are exactly the same. If the distribution of values is centered above or below zero, then the probability of negative values is going to be higher than positive ones, or vice versa. In this situation, the center point of the distribution of values is significantly away from zero, so that most of the values are have the same sign. One requires the other mathematically. ...
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=140780\")
It appears to me that as far as ACR and AcrCalibrator are concerned, the reds in all the Canons tested so far,and all the Nikons (with a different raw format) are too orange and need correction in the magenta direction. IOW, Canon and Nikon both got it wrong the same way. DPP seems to indicate otherwise.
There have been various comparisons of the accuracy of raw converters but I haven't seen an up-to-date one. Some years ago, Don Lashier compared ACR, C1, and DPP; and ACR didn't fare too well.
[a href=\"http://www.lashier.com/home.cfm?dir_cat=20503&gal_col=2]http://www.lashier.com/home.cfm?dir_cat=20503&gal_col=2[/url]
Ken
-
Have read this thread even though it's technical information is WAAAAY beyond my scope (I've only been shooting digital for less than a year).
Not sure if this is of any help, but I have heard it said that Nikon and Canons are designed to create pleasing colors for Asian skin tones.
Does this help explain the aforementioned results?...or is that just a load of racist hogwash with no basis in reality?
I'm just throwing it out there in the hope that it could explain the color differences with Canon and Nikon cameras. Please don't interpret it as an admission of belief in the rumour about how these companies try to create pleasing Asian skin tones. I don't know if that's true, it's just something that I as a noob have heard mentioned other forums and took with a grain of salt.
-
Not sure if this is of any help, but I have heard it said that Nikon and Canons are designed to create pleasing colors for Asian skin tones.
Does this help explain the aforementioned results?...or is that just a load of racist hogwash with no basis in reality?
It's hogwash, but not necessarily racist. The internal profiles used by ACR are created by Adobe, not Canon or Nikon.
-
"Could it be that the hue-accuracy of this memory color is sacrificed to get dead-on Red..."
Peter might be on to something... according to this: http://photoshopnews.com/2005/05/01/photos...e-real-history/ (http://photoshopnews.com/2005/05/01/photoshop-widows-club-part-two%e2%80%93camera-raw-the-real-history/) , we can maybe say the opposite, that the pure reds have maybe been sacrified for the (ears) skintones in ACR?
J/K
Sense & Simplicity of ACR Calibration … was the title of a post which I placed some long time ago in the Adobe forums. Purpose was to share a quite simple, manually operated procedure for ACR Calibration. Some points may still be of interest:
Guiding principles:
1.) In the course of ACR Calibration, it is hardly possible to compute a best-fit matrix space by just nailing the three primary patches i.e. red, green and blue. Background is as follows: When the spectral response of a real-world sensor deviates from an ideal linear-combination of CIE XYZ weighting function, colors and RGB data released from Bayer interpolation deviate from their ideal position (within a matrix space as a subset of CIE XYZ) in a quite idiosyncratic way. There are no obvious systematic shifts in terms of hue and saturation. For example, get the red patch right, and the hue of skin tones becomes worse… (see above).
The latter is one problem with the Fors Script which just analyzes the three primary patches. This may make sense if your camera is considerably different from the one profiled by Thomas Knoll (device-to-device deviations). However, if your camera is somewhat close to the camera which was used by Adobe, the Fors Script will just violate Mr. Knoll’s profiling efforts. At least that’s obvious for me from quite a bit numerical analysis.
Anyway, it’s of course possible to compute a best-fit matrix by minimizing the color error for many colors. But then, we are having memory colors which are more important for us than others - and that’s an individual thing. So any procedure for ACR Calibration which doesn’t call for an input by the user about preferred memory colors is prone to fail in practice.
2.) There is no need to linearize Camera Raw i.e. to match the grayscale first, for the purpose of subsequent ACR Calibration. This is not only unnecessary, at the end it leads to a high-sat look due to the RGB tone curve (and it’s side effect on color saturation) which in practice is applied on the top – mainly via Brightness and Contrast settings.
That’s the second problem with the Fors script and again there is evidence that the in-built Thomas profiles already absorb this more or less unavoidable boost of saturation from such sigmoidal RGB curve which is needed for a pleasing tonality.
Anyway, the Calibrate Tab sliders actually do what the labels suggest. They adjust the red, green and blue Hue & Saturation for these primary colors with decreasing influence on secondary colors, respectively. While this is done behind the scenes on a CIE xy basis, there are enough similarities with the HSB color model to trace respective changes via HSB readings.
In other words, given that ACR calibration is 2D per primary color, why bother with a 3D fit.
----
Proposed procedure:
/> Reset the Main Adjust Tab and in particular all tonal controls to ACR 2.x defaults i.e. Shadows (Blacks) 5, Brightness 50 and Contrast 25. Everything else zero and Curve Tab linear.
/> Nail the second gray i.e. patch #20 of Gretag’s ColorChecker to RGB= 190 by means of click-whitebalance and the Exposure slider (All numbers referring to ProPhoto RGB).
/> Address the color patches in pairs of one primary color and one corresponding memory color. My choice is: Red #15 + Light skin #2, Green #14 + Yellow-green #11 and Blue #13 + Blue sky #3.
Adjust the Red Hue slider of the Calibrate Tab to get the HSB-hue somewhat right for both, the Red patch #15 and Light skin patch #2. In all probability this will require to make a compromise. It’s a balance. Continue with the Green Hue slider of the Calibrate Tab to get the HSB-hue somewhat right for both, the Green patch #14 and the Yellow-green patch #11. Same with the Blue Hue slider and the respective patches Blue #13 and Blue sky #3.
Same principle applies to the adjustment of saturation, at least as far as the Red Saturation is concerned. With the Green Saturation my preference is to stay a bit under-saturated in order to prevent fresh green grass from getting too vivid. With the Blue Saturation my preference is to have it a bit over-saturated in order to get a deep blue sky. However, that’s not really essential. You can also strive for a somewhat accurate calibration of saturation while making such tweaks later on via the HSL Tab.
Finally, re-exercise through all sliders in order to eliminate interactions. My experience is that finally only more or less small adjustments are needed. A kind of minimum invasion which tells me that this approach is quite close to the logic of Thomas Knoll’s profiling software.
As for the target values for Gretag’s ColorChecker, the following numbers given in terms of HSB hue & saturation were derived by averaging the simulated charts as offered by ColorRemedies as well by BabelColor. Perhaps you have better ones, if not, the following ones should not be so bad:
red patch # 15: H= 10°, S= 62%
# 2 light skin: H= 30°, S= 28%
green patch #14: H= 101°, S= 45%
# 11 yellow green: H= 75°, S= 56%
blue patch # 13: H= 248°, S= 60%
# 3 blue sky: H= 230°, S= 29%
Again, the numbers are referring to HSB readings in ProPhoto RGB. In the past it was necessary to convert from ACR to Photoshop in order to access HSB readings. So in the course of above described procedure it was necessary to do repeated RAW conversions while proceeding step by step. However, while I have not yet arrived with latest version of Camera Raw I can not say if HSB readings are now accessible in ACR. Would be nice.
That’s it - really not as complicate as it may sound. It’s however nice to see from above quoted article that Mr Knoll also had to go through a learning curve.
Peter
--
-
Yikes - this is just a hobby for me....maybe it's time to explore B&W more fully
-
Sense & Simplicity of ACR Calibration … was the title of a post which I placed some long time ago in the Adobe forums. Purpose was to share a quite simple, manually operated procedure for ACR Calibration. Some points may still be of interest:
Guiding principles:
2.) There is no need to linearize Camera Raw i.e. to match the grayscale first, for the purpose of subsequent ACR Calibration. This is not only unnecessary, at the end it leads to a high-sat look due to the RGB tone curve (and it’s side effect on color saturation) which in practice is applied on the top – mainly via Brightness and Contrast settings.
[a href=\"index.php?act=findpost&pid=142876\"][{POST_SNAPBACK}][/a]
Peter,
In his ACR for PSCS2 book, Bruce Fraser suggested that if the grayscale adjustments are not made as per his calibration method or the Fors script automation, the colors will be wildly off and the results will be unpredictable. His suggested brightness and contrast values yield a very flat image. When one does the calibration, the Color Checker patches match reasonably well, but when one restores brightness and contrast to their normal default values, saturation is markedly increased as you suggested. Experts can and do vary in their opinions.
This is shown by some Imatest plots that illustrate your findings. Readers should refer to the Imatest web site for information on how to interpret the charts. However, values radiating outward from the white point have equal a* and b* color values (no tone shift), but increased saturation (chroma).
ACR defaults:
(http://bjanes.smugmug.com/photos/164350910-O.png)
After calibration with Fors script, using script grayscale values:
(http://bjanes.smugmug.com/photos/163188035-O.png)
With ACR defaults restored, but using the calibration:
(http://bjanes.smugmug.com/photos/163194057-O.png)
One could decrease saturation when using the calibration (easier than adjusting color balance), but I have never been comfortable with this. When I get the time, I will try to implement your suggestions. Is your essay still on the Adobe site and can you give a link if so?
Bill
-
... Experts can and do vary in their opinions.
This is shown by some Imatest plots that illustrate your findings....
One could decrease saturation when using the calibration (easier than adjusting color balance), but I have never been comfortable with this. When I get the time, I will try to implement your suggestions. Is your essay still on the Adobe site and can you give a link if so?
Bill,
You’re right that there are different opinions and options. While many fundamental threads on this subject are hardly accessible any more, let me take the opportunity to re-compile some aspects I kept in mind. Referring to ACR calibration or profiling in general, two main strategies are as follows:
a.) to calibrate a lineareized state and to accept a subsequent boost of color saturation which is introduced as a side effect from the pleasing (RGB) tone curve needed to compensate for dynamic range compression.
b.) to strive for somewhat accurate HSB hue & saturation in conjunction with a realistic tone curve being already in place. Obviously, this requires to ignore the dimension of brightness in the course of calibration, which is quite easy because the Calibrate Tab hasn’t brightness sliders anyway. That’s the nature of matrix primaries. As mentioned above, I'm more in this second camp.
Anyway, why HSB? HSB hue & saturation, both together, are representing the intensity ratios of R:G:B. That’s actually the only attribute which is left from the spectral intensity ratios of the original color. Exposure which operates in terms of linear scaling just changes HSB brightness. Multiplier do not change ratios [aR:aG:aB = R:G:B]. As a matter of principle I would not discard this information i.e. accurate HSB hue & sat. too early. Whereas Brightness is subject to many changes anyway – i.e. the almost unavoidable tone curve and finally the monitor’s luminance not matching the original scene.
From a perceptual point of view, option b.) gives a more neutral starting point which I prefer for portrait and skin tones, while landscape often requires a push of color saturation. This is then a decision by purpose, using a dedicated saturation slider, and often the specific needs are different for different colors, tonal sections or areas in an image. However, it’s of course possible to proceed the other way round starting from option a.). So let’s look for possible references:
As far as I can tell, the in-built ‘Thomas profiles’ are closer to option b.) than to a.). Means that the uncalibrated state should not look as “exploded” as with above diagram showing the calibrated state after the Fors script was applied, plus the default tone curve. I think this is basically in line with with the before/after Imatest plots offered above. Accordingly, option b.) requires less strong settings in the Calibrate Tab compared to option a.). So with option b.) it should be more a kind of minimum invasion.
I'm sure you know, however, as a side note it may be worth to mention that ACR can be linearized by setting the tonal controls to zero and selecting a linear tone curve. IIRC, this was even confirmed by Thomas Knoll in the meantime. Whereas the implementation of linearization with the Fors script is quite worrisome – i.e. trying to adjust 4 interacting parameters based on only 6 measuring points. This can be seen as the reason why we find such exaggerated tonal adjustments which then effect color again. Some times ago I experimented with the Fors script: If the respective section of the script was not completely disabled the script will produce weird tonal settings even if the given tonal values are defined as the target and the given tonal settings are defined as the starting point. Oops.
Another reference FWIW: I’ve recently analyzed the colors of my digicam, yes just a digicam, however equipped with Canon’s digic IIII processor. Referring to JPEGs’ from in-camera conversion, colors were found to be surprisingly accurate in term of HSB hue & saturation. Not kidding. Compared to earlier generations of the digic processor they have apparently dropped over-saturation – at “normal” shooting mode. It’s now accessible via “My Colors” or “Positive Film” if the user wants so. Emulation of positive film boosts saturation, however, even in this case skin tones seem to be masked. This may require confirmation with the 40D (if I’m going to order; not yet sure), though Micheal Reichmann’s comment on the 40D seems to point in the same direction.
My original post on this subject obviously already dropped from the Adobe forum archive. However, this discussion here is more detailed anyway. For the theoretical part I recommend Simon Tindemann’s website (see below link). It may seem that he strives for option a.) but his special Luminosity curves leave HSB hue and saturation unchanged. Another, third option and long story. Option b.) is offered with the “use XMP” option of his script – not sure if it works with latest version of Camera Raw. One point I wanted to make with above post is that a manually operated procedure can still be up to the task. Hope the description is understandable. For me, this is more in the tradition of Bruce Fraser’s original procedure, which definitively was a milestone for systematic use of Calibrate Tab, then the Fors script which even does not precisely represent Bruce’s method. Another long story, and many good advice – by guys like Allen Pacheco or Ethan Hansen – was ignored by Mr. Fors at times when he struggled with scripting. Anyway.
Take care
& best regards, Peter
--
http://www.xs4all.nl/~tindeman/raw/color_reproduction.html (http://www.xs4all.nl/~tindeman/raw/color_reproduction.html)
-
The latter is one problem with the Fors Script which just analyzes the three primary patches. This may make sense if your camera is considerably different from the one profiled by Thomas Knoll (device-to-device deviations). However, if your camera is somewhat close to the camera which was used by Adobe, the Fors Script will just violate Mr. Knoll’s profiling efforts. At least that’s obvious for me from quite a bit numerical analysis.
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=142876\")
Peter,
The OP wanted to determine if the calibration adjusted mainly for camera to camera variation in color response, or if there was a systematic error in the original calibration. The approach taken by Bruce Fraser and Jeff Schewe was asssumed the former. However, the OP's hypothesis was that most calibrations resulted in similar results, suggesting that there was not that much variation in camera to camera. Subsequent data provided by users seemed consistent with this hypothesis. In this case, according to your analysis, the calibration would violate Mr. Knoll's profiling efforts.
The issue was never resolved satisfactorily.
Anyway, it’s of course possible to compute a best-fit matrix by minimizing the color error for many colors. But then, we are having memory colors which are more important for us than others - and that’s an individual thing. So any procedure for ACR Calibration which doesn’t call for an input by the user about preferred memory colors is prone to fail in practice.
[a href=\"index.php?act=findpost&pid=142876\"][{POST_SNAPBACK}][/a]
Skin tone, blue sky, and green foliage are important memory colors. Increased saturation is often desired in landscape work (Velvia look), whereas these attributes are not desirable for portraiture. It might be that one profile can not cover both these situations. Perhaps Mr. Knoll's profiles are a reasonable compromise.
Again, the numbers are referring to HSB readings in ProPhoto RGB. In the past it was necessary to convert from ACR to Photoshop in order to access HSB readings. So in the course of above described procedure it was necessary to do repeated RAW conversions while proceeding step by step. However, while I have not yet arrived with latest version of Camera Raw I can not say if HSB readings are now accessible in ACR. Would be nice.
[a href=\"index.php?act=findpost&pid=142876\"][{POST_SNAPBACK}][/a]
The latest ACR does not support HSB, but there is a useful online calculator by [a href=\"http://www.rags-int-inc.com/PhotoTechStuff/AcrCalibration/RGB2HSB.html]Rags Gardner[/url] that will give the HSB values for given RGB values of the working space. However, real time feedback is not available. The points you raise here and in your subsequent post are worthy of further investigation and discussion.
Finally, if one's camera does not differ that much from the one Mr. Knoll used to build his original profile, the calibration may not be advantageous in some situations.
I am still absorbing the points of your reply to my post. I certainly appreciate your taking the time to post such a sophisticated analysis.
Bill
-
Interesting thread.
I don't use the Fors method anymore with my 5D, but a 'long hand' method that suits the type of work I do - mostly fine art catalogues.
In any given copy setup, I shoot one frame that includes a basICColor grey card along with a Gretag CC.
The basICColor card is used for the neutral balance - far better than using any of the patches on the Gretag CC since none of them is strictly neutral.
After adjusting the Gretag CC greyscale a la Fraser, I then adjust the eight colours on Gretag CC that have equivalents in ACR4's HSL tab.
The difference is noticable and better.
Since I can now do it in under five minutes, I do it every time rather than saving the settings.
Just my tuppence worth.
D.
-
Well here's another data point, and it lines up with all the others Barry has gathered. I calibrated my 20D, shooting the CC chart in sunlight where I do most of my shooting. I was put out with the orange-ish reds that ACR put out. Using the Rags Gardner script, I came up with the calibration parameters
tint: 0
red hue: -16
red sat: +28
green hue: 0
green sat: +8
blue hue: +6
blue sat: 0
I mean really, the default rendering of reds on ACR is a joke. Just compare them to ANY other raw converter -- the others I have tried are C1, DPP, Silkypix, Bibble. The reds are simply better from any other raw converter. Adobe is so pig-headed about this; it's like Apple stubbornly maintaining that the one-button mouse was superior (though at least they finally saw the light).
-
Peter, thanks for the descriptive answer to my attempt of humour (I do know that the ear was used at the beginning of ACR's birth was not used for color calibration)
Your posts reminded me of Kodak's raw development tool, where we could use a product look (logos, thus often primary colors) or a portrait look.
Getting a "one size fits all" Camera Profile does not seem a good idea in that regard, (different color temperatures, body to body variation, subjet captured) and Thomas Knoll was once again very wise by letting the user calibrate for the given situation, or look that he/she is trying to acheive.
I wonder if the "Magne Profiles" take this in consideration (but one would need many of them: product under tungsten, portrait under sunlight, etc.)