DNG profile editor - truncated RGB Lab HSL values

It seems that output is truncated to last two digits in all fields.
When I point the mouse to a blown highlight area, I got:
R=55 G=55 B=55
L=00 a=0 b=0
H=0 S=0 L=00
Windows platform, desktop size 1280x1024. Tried on 2 different computers

For clarification -- does this affect the profile created?
I ask because PE gags on one of my DNG images and reports that the 'white' patch has a "significant" color cast. I'll allow for the possibility that it may have a color cast, but it isn't "significant". When I roll over the patch I can occasionally see the 'R' value fall to '0', while 'G' & 'B' appear fine.
I see other inconsistencies: the approximate average RGB values for the neutral patches (white - black) are reported as: 90, 11, 63, 34, 15, 4.
Olympus E-3

Similar Messages

Integrate DNG Profile Editor with Lightroom

Worthy of an official feature request, me-thinks...
Its a great tool that's under-utilized - why not integrate it with Lightroom?
Some ideas:
- Dis-associate it from the DNG file format, so it does not need an intermediate DNG file to transfer a profile to a proprietary raw. (for those of you who dont know - you need a DNG file for initial profile creation, but then Lightroom can extract it and apply it to proprietary (non-DNG) raws.
- Make options in Profile drop-down menu in Lightroom:
- Edit Profile
- New Profile
- Delete Profile
I think these ideas would be relatively easy to implement. Maybe Lr4.
Other ideas that would probably be harder to implement, but would also be worthwhile in my opinion:
- Make a color editor based on the DNG Profile Editor technology, that allows one to fine tune color without a profile and copy/paste/sync it like you can do now with HSL adjustments - including saving with presets. Maybe Lr5 or 6...
Rob

Hi, I converted a Canon CR2 from a Rebel xti (or 400d) to DNG.
Good news: I've copied the same DNG file and the same copy of DNG Profile Editor onto a computer running win vista sp2 and it works like a charm! I got the message "Canon EOS 400D 100iso_day_curve2.dcp was exported successfully". The profile was saved on a wrong folder: somewhre in the adobe tree but not in the camera profile folder (in this computer is C:\Users\Fernando\AppData\Roaming\Adobe\CameraRaw\CameraProfiles\).
So I think it is a win xp related problem.
I've noticed another problem in both machines (xp and vista): In the "Color Tables" tab the RGB, Lab and HSL number displays only two digits. The biggest rgb shows 55, instead of 255.
Good luck

Questions about DNG Profile Editor recipe code

I have been playing around with DNG Profile Editor. With a text editor I made the following recipe:
I have some questions:
1. Right now I have my control points at 60 saturation. Will using two points for each color (say 70 and 40) make my changes more consistent across darker and lighter shades of the specific color?
2. Does DNG Profile editor respect HueLow and HueHigh, are they just placeholder numbers, or are they ignored if two points are close to each other?
3. What does FeatherAdjust do? I'm guessing it controls the rate of drop off of the corrections. Is the value respected by DNG Profile editor when it creates a profile?
Thanks for your time and attention, -Bruce.

1. For now DNG-only. Here's why:
http://labs.adobe.com/wiki/index.php/DNG_Profiles_FAQ#PEOnlyDNG
http://labs.adobe.com/wiki/index.php/DNG_Profiles_FAQ#WhyNameDNGPE
http://labs.adobe.com/wiki/index.php/DNG_Profiles_FAQ#WhyPESeparate
2. Standalone for now. See above links for why.
3. Use the 'Preview Color Changes' option in the Options menu.
4. You can use 'Apply Raw Adjustments' in the Options menu to see your raw adjustments. This is __not__ recommended for building a general-purpose profile because you are then optimizing a profile with specific image adjustments in mind, rather than building a profile based more on the inherent camera characteristics.
5. True, that is a limitation of this implementation.
6. I am not sure what you mean. The Chart Wizard automatically optimizes the color patches in a test shot based on reference values for many physical charts, which is more practically useful than the numbers printed on the reference card that comes with the chart.
7. Make sure you avoid color casts in the bottom row. The PE is picky about making sure your gray patches are relatively neutral. It is an attempt to help you get a better profile.
8. Use 'Show Affected Colors' from the Options menu.
You may wish to read this page carefully and thoroughly:
documentation
It is the reference online documentation for the DNG Profile Editor and a few of the things you wish to do, such as preview all color changes, apply raw adjustments, and visualize the extent of each color adjustment, are all documented there.

DNG profile editor

A couple of questions about DNG Profile Editor:
1-I know Lab values on my ColorChecker (measured with an EyeOne 2° D50). I shooted it with a Nikon D80 and I want to tweak one of the new Camera Raw profiles to match those values or to go closer. I see that changing the base profile (popup menu in Color Tables Pane) the image appearence changes a lot but the Lab numbers readout doesn't. I followed the tutorial on the Adobe site but I can't figure out how to make the camera calibration without having a numeric value, before and after, to look at. The tutorial tells that you may 'adjust the selected color via the Hue, Saturation and Lightness slider and you will see the preview in real-time'. That's true but does it means that is a sort of 'visual calibration'? Is there a way to tweak colors by the Lab numbers in order to have the best match from original Lab values in input (on the target) and Lab values in output (in a ProPhoto rendered image)?
2- When I create a Color Table from my ColorChecker it appears that in the Color Tables colors are already (and automatically) changed. Does it work like the ACR Calibration Scripts that way? And is it possible to change the reference Lab values (ColorCheckers are different in color values) as it was with a simple ACR script's editing?
Giuseppe Andretta

Eric,
Before starting, I want to say that I had cataract surgery last month. The vision improvement was almost instantaneous and dramatic. I can now state that color differences in individuals can be much greater than expected. It is hard to express how big the difference is in my own vision, let alone between two other individuals. Any color assessment tools must accommodate numeric as well as visual comparisons. That said, the healing process has also affected how much time I can spend comfortably in front of my screen. I hope my comments will be taken constructively.
As others have commented, I also feel that the CC24 Lab target values being used should be documented. Published values from Gretag, Lindbloom, Babbage, and others disagree. Without the target numbers it is difficult to determine accurately the calibration result.
My first attempt seemed to go smoothly, but when I verified the results with ReadColors.jsx (my script) the numbers degraded slightly. I was using the Gretag target values, so these may not be what Adobe is using. Also, I had already calibrated ACR for my D3. The doc seemed to imply that the ACR tabs would all default to zero and that was what displayed in the Color Matrices panel. But that was not what showed in the ACR panel. So I tried again, setting all ACR sliders to zero. This calibration attempt was a disaster, numerically and visually. Now I am confused. What is the relationship between these and what are the recommended user actions? In each case, I exported the DNG profile and selected it subsequently in ACR to re-open the image.
Next, I played with the Color Tables panel. I set a watchpoint on the blue patch. As long as I hold the eyedropper over the blue patch the Lab (and other numbers) track. But as soon as I move the mouse to the sliders, the numbers disappear. If I move the hue and saturation sliders the image and the patch sample both change. And, yes a little arrow shows in the color wheel. But after moving the eyedropper back, none of the numbers have changed. The lightness slider is even more bizarre. Obviously no arrow (z-axis) but the image brightens and darkens (OK), without any corresponding change in the patch sample (watchpoint) on the right (not OK). And again, no change in the numbers at all. It would be very helpful if the target values were shown here and if the currently selected watchpoint that would track the numeric changes. It would also be nice if the watchpoint could show the target color as well as the initial and changed image patch colors.
I dont understand the purpose of the Tone Curves. I understand gamma and Adobe linear, but these do not correlate to the ACR tone curves. And again the image changes but the numbers do not. Since we cannot set watchpoints in the neutral patches it is very difficult to assess any changes the user might make.
Next, I used the Chart panel to Create Color Table. When I return to the Color Tables panel, there are the 18 color (no neutral) watchpoints. I can see that blue has changed in the watch point and the arrow in the color wheel. But the numbers are still the same as before. Scrolling through the watch points I can also see that the hue and saturation sliders have changed. But the lightness slider did not change in any of the color patches. I anticipated that this would be the major improvement over the ACR sliders. Is this just not in the calibration algorithms yet?
This is a good start, but until I know what the target values are and what I should be doing about the existing ACR slider settings, I am at a stopping point.
Cheers, Rags :-)

DNG Profile Editor can't read ColorChecker shot on slide film

I use my DSLR (Olympus E-5, 50mm macro) to digitize old Kodachrome and Ektachrome slides. As it happens, I have reference shots of a Macbeth Color Checker I shot back then. I thought I would be able to create custom DNG profiles to get accurate digital conversions of the slides. But when I shoot the slide of the color checker, convert to DNG and try to build a profile in the DNG Profile Editor, I get an error when I hit the Create Color Table button, specifically, "Non-neutral grey patches. The gray patch in row 4, column3, has a significant color cast. Please reshoot the chart carefully to avoid color casts and try again." I've had the same problem with Kodachrome 64, and Ektachrome 200.
If I shoot a ColorChecker directly using the same illumination (I've tried both strobe and incandescent for the slides, neither works) the profile is generated OK.
BTW, the RGB values for neutral color checker patches on the short of the slide (strobe illumination) when I open it in the profile editor are
White 123, 127, 92
LLG 91, 94, 78
LG 61,60,60
DG 27,26,31
DDG 7, 6, 9
Black 0, 0 ,0
I tried reshooting the slide gelling the strobe with 30Y, so I got 89,89,88 for what I think is row 4 column 3, but I got the same error message.
So you can see that there is not a consistent color shift across the brightness range (skews from -B at the light end to +B at the dark end, while R and B are relatively constant)
OTOH, the values for the Color Checker shot straight with the camera are consistent in the color shift across the brightness range:
White 104, 104, 107
LLG 71, 70, 75
LG 46, 45, 48
DG 24, 23, 26
DDG 10, 10, 11
Black 2, 2, 2
Any way to get around this or am I just screwed.
Thanks

Hi!
I tried YouSendIt. The Express app did not install, and Safari reported an error, bad server response or something. I have earlier uploaded files using SendThisFile, but that required a URL, not an e-mail address. so if you are still interested in the files, you might mail me an upload URL. - Fine anyway that you already fixed the bug.
-BTW, the "Subscribe to this discussion by email"-button does not work for me, it keeps deselecting itself.
Kind regards - Hening.

DNG Profile Editor - questions about

0) Thank to Eric Chan for your work on this. Until now I used the Camera Calibration process through Photoshop and Camera Calibration table. But I see this as a big progress.
1) Will it stay only DNG editor or it will be also for raw pictures without a necessity to convert them to dng?
2) Will it be only stand alone aplication or as well a part of the LR?
3) Now I can switched off and on one color change (by the left mark in the Color List Box). Could it be possible to switch off and on the whole Color List Box to see all done changes?
4) I have red your comment that this tool doesn't work with lightness, only hue and saturation. But if I want to work on a specific picture (i.e. a man wearing a red sweater), it will be fine to have it in the brightness close to an end picture. Will it be possible to open a picture not only with WB, but also with exposure, blacks and other basic development setting?
5) Under the eye drop tool, I can see numbers of "initial" or "starting" color. But I can't see the numbers of changed color. It would be fine if I know the new numbers to see how close I am to them. It would be fine if the number over the color List Box will be also for starting color and also for end colors.
And one little proposal: if the color will not change, show in Color List Box only the starting (left) half of a color. Than I can simply see that I didn't set a new color or the color stay exactly the same (it is only a question if exactly the same or very close to the same).
6) It could be, that I will not use the exact color checker table for what you did process on the tab Chart. But many of color checker are delivered with exact numbers and picture on a cd, so it will be fine if I could open this "master" picture of a checker producer in a right panel under the Color List Box and simply take from it the end colors. First go to the main picture on left, choose the "initial color" and then simply go to this "master" picture in right and take the "end color". What do you think about this?
7) If I work on the tab Chart, I get the error window "not neutral gray patches, row 4 culumn 1". But if I move the circle to row 4 culumn 2 it works fine. Is there any difference or problem?
8) And last: how far is acting the shift of color. It will not work only for the one color but close surroundings will be also efeccted. How smal or large is the surroundings? E.g. if we take 360 degres for hue.
Tom

1. For now DNG-only. Here's why:
http://labs.adobe.com/wiki/index.php/DNG_Profiles_FAQ#PEOnlyDNG
http://labs.adobe.com/wiki/index.php/DNG_Profiles_FAQ#WhyNameDNGPE
http://labs.adobe.com/wiki/index.php/DNG_Profiles_FAQ#WhyPESeparate
2. Standalone for now. See above links for why.
3. Use the 'Preview Color Changes' option in the Options menu.
4. You can use 'Apply Raw Adjustments' in the Options menu to see your raw adjustments. This is __not__ recommended for building a general-purpose profile because you are then optimizing a profile with specific image adjustments in mind, rather than building a profile based more on the inherent camera characteristics.
5. True, that is a limitation of this implementation.
6. I am not sure what you mean. The Chart Wizard automatically optimizes the color patches in a test shot based on reference values for many physical charts, which is more practically useful than the numbers printed on the reference card that comes with the chart.
7. Make sure you avoid color casts in the bottom row. The PE is picky about making sure your gray patches are relatively neutral. It is an attempt to help you get a better profile.
8. Use 'Show Affected Colors' from the Options menu.
You may wish to read this page carefully and thoroughly:
documentation
It is the reference online documentation for the DNG Profile Editor and a few of the things you wish to do, such as preview all color changes, apply raw adjustments, and visualize the extent of each color adjustment, are all documented there.

About to use DNG Profile editor

I've had my Colorchecker for a few weeks, and I finally have time to set up my profiles for LR 3.4. In reading the DNG Profile editor tutorial page (http://labs.adobe.com/wiki/index.php/DNG_Profiles:Editor), I have a few questions:
1) If I use Tutorial 5, I am guessing I will have to create a profile for each lighting/WB condition, i.e., one for Tungsten, one for Fluorescent, one for sunny, one for overcast, one for flash, etc... Is that correct? Is there an advantage to doing this vs. going for the dual-illuminant approach explained in Tutorial 6?
2) In Tutorial 6, for the dual illuminant profiles, it states that one of the reference shots should be taken at 6500K. I'm not quite sure how to achieve this, since daylight and flash, to semi-constant WB sources I can easily achieve, are around 5000-5600K.
3) Should I (in LR) White-balance the Colorchecker reference shots, then convert to DNG, prior to bringing the DNG file into DNG Profile editor? Or is it preferable to WB as explained in Tutorial 1, step 3?
4) Finally, it's somewhat unclear how I bring the profiles into LR for use during the Import or Develop phases. Tutorial 1, step 9 mentions a "CameraProfiles directory" for ACR and LR, so I'm guessing I need to go digging through the Adobe folders to locate this directory and place/save my profiles there?

eswrite wrote:
1) If I use Tutorial 5, I am guessing I will have to create a profile for each lighting/WB condition, i.e., one for Tungsten, one for Fluorescent, one for sunny, one for overcast, one for flash, etc... Is that correct?
If you want a broad purpose DNG profile then yes, doing a dual illuminate profile is useful. Shoot a target under tungsten and under daylight and make the dual illuminate profile. There's no real reason you need to spawn off a bunch of other profiles for only slightly different light. The dual illuminate will handle cloudy or overcast just fine. If you also shoot with special lighting sources like fluorescent (which doesn't have a complete spectrum) or other non-standard lighting sources then do a profile for those special conditions. If you are creating a profile specifically for studio flash, you can get by doing only a single illuminate profile for the strobes...but in that case, the profile won't be as accurate if you also try to use it under tungsten–which would be the reason to do a dual illuminate profile.
As far as the D65 color temp, the closer you can get to it the better, but D55 should work fine. The key is to make sure it's evenly lit. The big difference between D65/D50/D55 is the relative amount of the blue light components-all three will be fairly close. Tungsten however has vastly less blue which can impact the sensor's spectral response–which is why the dual illuminate is suggested.
Once you do the CC shot, don't bother with tone/color corrections in ACR before converting to DNG. They won't have a material impact on the profile creation. The CC shot MUST be evenly lit and of an optimal exposure...
As far as where the profile goes, it depends on the system. Once you've created the profile, PE should default to the correct location, otherwise state your system and we can tell you where to put them.

Creative use of CC calibraton in the DNG Profile Editor

The DNG Profile Editor can be used to create profiles that mimic film looks.
I think right now it is manual trial & error process involving changes in hue, saturation and lightness, but it could be theoretically achieved with Color Checker calibration working in reversed order.
- Photograph the CC using your favorite film
- If B&W, develop in your favorite developer
- Scan the film
- Use the scan in the DNG Profile Editor as target values
The current DNG Profile Editor obviously does not allow that, it would require adding a new tab and clone the CC calibration feature and alter it to use the photographed values as target values as opposed to internal values.
I don't know if it would be worth the effort, but I thing the idea is worth of sharing.
By the way, unless I'm doing something wrong, the saturation slides don't seem to go 0%, so creation of B&W profiles is not possible.

Good idea. Yes, I've posted an example on my Flickr.
I chose a picture with good reds and yellows. There are four images in the set:
1. The original, untouched image. This is using the new Adobe Standard beta profile, brought into Lightroom with all default settings, including linear tone curve, no exposure or color adjustments, white balance "as shot", etc. For reference.
2. The image with my test profile applied, but no further adjustment.
3. My original attempt at the image, using the "old fashioned" pre-DNG-profiles raw processing method, for reference. The yellows are orangeish, which isn't that objectionable but also isn't accurate. The reds are weaker than I would like.
4. The "final" image, with my test profile plus additional tweaks -- white balance, tone curve for contrast, crushing out the blacks a bit more, etc. Two or three minutes of total processing, versus quite a bit more for the previous attempt with no "special" profile. The reds and yellows are perfect.
The Flickr set:
http://flickr.com/photos/100mph/sets/72157606547890574/
This was just a first, pretty quick attempt at playing with the DNG Profile Editor. I'm sure it can be tweaked more and even more can be accomplished.
[Edit: note: I changed the "upload date" on the test images so they wouldn't stay at the top of my Flickr page. Ignore the 2007 date, it doesn't mean anything.]

Camera Profiles and DNG Profile Editor

I saw the lightroom 2.0 eseminar and the presenter mentioned Camera Profiles and the DNG Profile Editor. All I see is ACR4.4 and 4.3. The the FAQ page says I need ACR4.5 and I can't find it and haven't gotten any update notice. Should I just wait for an update or forget it?
Don

>I have CS4. Will this overwrite ACR 5.1 in CS4?
If you have CS4, you should update to ACR 5.2. If you install 4.6, you will break your CS4 install. Either just run the Adobe updater app, or go to http://www.adobe.com/support/downloads/new.jsp to download it. DNG converter is a separate download. Both camera RAW and NDG converter contain the final release version of the new profiles. DNG profile editor can still be downloaded from Adobe Labs: http://labs.adobe.com/wiki/index.php/DNG_Profiles. There is also a profiles download there that you don't need if you install DNG converter 5.2.

Camera Profiles and DNG Profile Editor beta 2 now available

Hi everyone,
Beta 2 of the camera profiles and DNG Profile Editor are now available. Please visit here and enjoy:
http://labs.adobe.com/wiki/index.php/DNG_Profiles
I hope to have more detailed release notes/changes for you soon. You are welcome to ask questions, but please note that it may take a while for me to respond.
Eric

Eric,
I followed very carefully your instructions for installing the new beta2 profiles and deleting the beta1 profiles, but have the same problems as many others. I'd like to provide a bit more information. I'm running Windows Vista Home Premium and have installed Photoshop CS4, updated with ACR5.1 and Lightroom 2.1 final release. As others have described, the default for all of my images was set to one of the Camera profiles (Canon faithful beta1), but now the profile name in Lightroom is blank. I tried looking at the other profiles and it does indeed appear that the beta2 profile is being used, but if I then reset the image it now shows ACR4.4, whereas the image had been imported with the camera profile as default. Also, if I open an image in ACR5.1 that was specified to use the same camera profile, it now shows ACR4.4 and it is not using the beta2 profile of what I had been using, I can see this by selecting the other profiles. Going back to Lightoom, I guess I could select all my images and select the Canon faithful beta2 profile, but then the mark shows up bottom right of the image showing that they have all been edited/modified, is there anyway to get all my images using the profile I was using but the beta2 version, without this happening and how about new imports?
Thanks, David.

Converting LUTs to Camera Profiles in the DNG Profile Editor

Hi,
I've just started using the DNG Profile Editor, and can't seem to get the process right.
I have some film LUTs that I'd like to convert to camera profiles for my 5D Mk II.
Here's how I'm trying to do it, maybe someone can help me figure out what I'm doing wrong.
This is what I tried first:
1. Take a Color Chart into Photoshop, apply the LUT via Color Lookup adjustment layer.
2. Save out as 16bit TIFF
3. Open as Camera Raw, then save out as DNG
4. Open the adjusted color chart into DNG Profile Editor, run it through the color chart section to create a recipe of color adjustments.
5. Export out as Camera Profile
But when I open up a 5D image in Lightroom, the profile doesn't show up in the list, so I realised maybe it needs to be saved specifically as a 5D profile, but I can't figure out how.
If I:
1. Do steps 1-4 as above
2. Open a 5D DNG file so that I can set a 5D profile as the base profile
3. Load the recipe
The color adjustments appear on the color wheel, but nothing happens to the 5D image, no colour change at all.
Exporting it (which does say 'Export as 5D profile') does then give me a profile that shows up in Lightroom, but it doesn't do anything to the image...
What's the correct way of doing this?
Thanks!
S

I don’t know claim to know anything about the internals of DNG Camera Profiles, but I’m pretty sure you cannot take a RGB-file (TIF/JPG) DNG and make a raw-file DNG Profile for use with your camera raws, which sounds like what you’re trying to do.
Despite looking like it might work in DNGPE, starting with an RGB-file recipe (rather than a RAW-file recipe) will only lock the specific colors on the color wheel, rather than create a raw-capable camera profile.

Adobe DNG Profile Editor

Has there been any update to this beyond that in 2008 found on Adobe Labs? (I want to try using it to calibrate ACR for my camera with a shot of a MacBeth Color Chart.)
To be more precise about this, I currently use ACR Calibrator to calibrate ACR for my Canon 1Ds II. I now see there is beta 3 of Adobe DNG Profile Editor but note that it utilises a small subset of the GMB Digital Colorchecker SG target chart.
ACR Calibrator seems to produce quite different results from Adobe DNG Profile Editor. I have followed the steps below when using the latter.
Download DNG Profile Editor and drag to Applications
Launch Photoshop and open the raw file of your macbeth target.
In camera raw, place all settings to zero, curve to linear and turn off any sharpening etc. Click Save Image, choose dng as the file format and save it.
Launch DNG Profile Editor.
File / Open DNG File and choose the dng file you saved previously.
Select the chart tab in DNG Profile Editor.
Move the colored circles to the center of the matching colored corner square in your raw file
Click Create Color Table.
Choose File/Export <Camera> profile... Give it a name and save.
Exit DNG Profile Editor.
Re-launch photoshop and open a raw file. In Camera Raw, on the calibration tab, you should now see the profile you just saved.

ssprengel wrote:
I don't know of any other references or CC Passport numbers other than what Google can come up with. Babelcolor has some standard-deviation of spectrum numbers for a sampling of 20 charts done in 2006 and also some worst-color vs average-color visual comparisons to get an idea of how variable charts or at least their measurements can be. Somewhere it says the me4asurements were done of the standard and also the minichart, mixed together, I assume. I don't know if there is a list of the raw data or not.
To understand the source of the variation you really need to see multiple measurements with the same instrument, with different instruments, and of different charts from different manufacturing runs and different ages and measured in different temperatures and humidities, otherwise it's not easy to say whether the variations we're seeing between the CC standard and the CC passport in the article are within the normal variations or not or if they represent a significant difference not explained by normal variations.
Yes, I know the Babelcolor data and I have already included them in the study.
Obviously I do not have the opportunity to have the information to get an overview of the real things, but by analyzing various measurements I saw that Myers had to be the closest to official data of the CC Passport and instead is the more distant. Hence my doubts about the measurement.
Thanks anyway.
Marco

Documentation for DNG Profile Editor??

Hi. I am looking for the tutorial and documentation on Adobe DNG Profile Editor (v 1.0.0.39 beta 2). I assume this is the latest release, but if not, I would appreciate a link to the latest. I tried downloading documentation and tutorial from the following link, but the site never comes up.
http://labs.adobe.com/wiki/index.php/DNG_Profiles
I'd like to build some profiles for my 5DmkII and studio lighting situations, specifically for portrait / skin tone type work. Suggestions welcome.
Thanks,
Lou Dina

In the past I did some research in these fields and I want to share the fruit of that work (but English language is a problem).
> There can be Variations Among cameras.
I have analyzed two sets of cameras, two Canon 40D and two 400D illuminated by Solux Daylight 4700K bulbs.
The first test was done with a 400mm. Unfortunately, very little movement of the lens at f/5.6 (a mistake) has darkened the upper rows of ColorChecker and I had to correct them with the brush of ACR Exposure (the worst intervention reaches 0.17). After this intervention the cameras differ by an average error of delta E 2000 0.30 and a maximum of 0.75.
Link: http://www.boscarol.com/forum/viewtopic.php?f=58&t=321
The second test was done with a Canon 400D and a Canon Rebel XTi. In theory, the same model but for different markets.
Here the things were particularly interesting because:
- the cameras could be build from different processes due to different market areas
- the cameras were put on the market in 2006, and were produced with the technology of almost five years ago
- the cameras belonged to the entry level market segment, the techniques of production could be consistent with this low target
All these factors could lead really to differences in behavior. The results were an average error of 0.65 and a maximum of 1.26.
Link: http://www.gialandra.it/blog/files/343e438a6021337b5fdffb2cfbe0f08b-0.html
And probably if these tests had been made in laboratory the differences were even smaller.
I concluded that the differences in the behavior are all to prove.
On the Internet I found nothing of scientific.
> There are subtle Also Variations Among Conditions lighting, as well as chart shooting conditions (Exactly how the light is falling on the chart, the angle at Which you photograph the chart, etc.)..
These are the most important points. If the chart shooting conditions are a bit easier to control with geometry, get a light with a spectral power distribution right it is very difficult (and I suspect that even Adobe has not a decent D65 simulator). To obtain a light close to illuminant D50 I use 4700K Daylight Solux bulbs without glass diffuser. In front to the lamps I mounted a shape ring protection with a hole less than two centimeter (to avoid protection I have tried the black back version of the Solux bulbs but they do not have the correct spectral power distribution). Bulbs are powered by an adjustable power supply. Using the spectrometer as a feedback I can reach a CRI of 97.
It seems to me that even the not winter sunlight when the atmosphere is really clear is close to the D50, but I've never had a chance to measure it. In this condition the problems of non-uniform illumination do not exist, but be careful with the camera and target angles because the light is even more violent, the target should be perfectly at 45 degrees and the sensor should be parallel to the target.
Finally to have a more homogeneous illumination and all other problems I use only the mini format of the ColorChecker.
> There Are Also Among Variations lenses (eg, spectral transmission of Canon 50 f/1.4 is Not the same as Canon 180 f/3.5 macro, But They Are close).
I have studied the effect of five very different lenses in type, brand, cost, etc.. The spectral transmittance of the lens was so different to cause a shift of the white balance of 450K. On films this would lead to very different results, on digital system instead the human visual system is not able to appreciate differencies (worst deltaE 2000 is 1)
Link: http://www.photoactivity.com/Pagine/Articoli/052ResaCromaticaObiettivi/Resa_cromatica_degl i_obiettivi_ENG.asp
For my studies and my experience at this moment I think the best way to find the faithfull to the scene with a safe semi-universal profile is using the calibration sliders in D50 condition. From the linked experience below this gives faithfull from tungsten light (2600K) to above 10000K, of course this last only in daylight.
Link: http://www.photoactivity.com/Pagine/Articoli/023%20Calibratori%20al%20sole/Calibratori%20a l%20sole_en.asp
In this context I still use the script of Tindemans but it is time (many years now) that Adobe produces his matrices tweaker (XRite Passport stresses the matrices as far as possible before creating some sort of table)
I hope that I'm not boring
Ciao
Marco

Colorchecker passport and the DNG profile editor

I finally got around to getting a Colorchecker passport to use with the DNG profile editor.
I was secretly hoping I could save time by shooting the chart on location and make an instant profile out of that - then the colors would be at least "close ballpark" so I would just have to apply white balance and then move on to more specific adjustments. But of course it wasn't that easy, it still takes a fair bit of color tweaking to get right in most circumstances.
So I'm wondering if my energy would be better spent trying to perfect one good dual illuminant general profile. For this I would use 3000 K tungsten at one end and 5500 K studio flash at the other. Or would it be better to use overcast daylight? In either case it's obviously important to ensure the chart is absolutely evenly lit.
And then I should probably have a separate profile for fluorescent tubes.
I'm curious what other people's experiences with the colorchecker/DNG profile editor are? How are you using it?

The color profile changes with the lighting, so a profile that was computed for daylight won’t be right for tungsten or fluorescent. Keep in mind that a custom camera profile is the set of corrections on top of an existing Adobe profile that the DNG has assigned to it, perhaps Adobe Standard, and all Adobe profiles are dual-illuminant profiles, so it makes sense you’d want to also create a dual-illuminant profile for general purpose use that has the same two lighting scenarios as Adobe used: 2850K and 6500K. The 6500K is the tricky one because full sun is warmer, closer to 5000K, and you need the right amount of haze to create 6500K sunlight, unless you are using standard D65 bulbs, indoors, which is likely what Adobe does.
I use a dual-illuminant profile for general purpose use, but since the WB-Tint values of both the tungsten and daylight ends are close to zero, lighting that has a tint not close to zero needs a separate profile. The common example is the greenish light from fluorescents, and I have several single-illuminant profiles for various artificial lighting scenarios with WB-Tint values that are relatively non-zero. I also have ones for very red sunsets and very blue twilight. If I had studio lighting I’d make a profile for that. Sometimes I make a custom profile for a church or other venue that has significant non-neutral walls or ceilings, or where sunlight through the stained-glass windows shining on the walls are giving a significant color-cast including a non-zero tint value to the lighting. If you know you’ll be doing some shots near a large amount of tall, green vegetation then a custom color profile can correct for the green tinge to the ambient lighting coming down from above mixing in with the sunlight.
Adding to the suggestion about wearing neutral clothing, I’d argue for dark neutral clothing because something lighter clothing will reflect environmental colors (sunlit vegetation or brick walls, etc, outdoors) and not be neutral.
Besides non-neutral clothing, try to avoid areas that have green grass or tree leaves or brick buildings that also color the light and reflect off the colorchecker. For the sunlit shots, I put my colorchecker face up on the sidewalk or parking lot, to avoid coloration from the grass and far away from trees or buildings especially in the direction I’m pointing the camera, so the only lighting is direct sun and ambient blue-sky, possibly diffused by hazy clouds. I point the camera down at the colorchecker, at a 45-degree angle and at a compass direction and not directly toward the sun, nor directly away from the sun. I try to shoot the target that is face up on the ground perhaps at a 90-degree angle from the sun to minimize any residual glare from the colorchecker’s slightly non-matte surface. Obviously directly away from the sun, the color-checker will have my shadow on it or I will be blocking the light from the sky near the sun.
I use the same sort of process for the tungsten end, choosing a room with neutral walls and put the target relatively flat on the floor facing up, perhaps tipped up somewhat toward the light, but not so much as to have any glare from the lighting, either, in other words, not with the lighting directly behind the camera, but somewhat over to the side.

Do DNG HueSatDeltas correspond to DNG profile editor points?

I understand that the HueSatDeltas table in a DNG profile has 90 hue divisions & 25 saturation divisions (2250 total, excluding the value positions)... and these in turn record the hue shifts and saturation scales from these points.
Does anyone know if the dropper points on the color wheel in the DNG profile editor correspond directly to these 90 x 25 positions... Or are they simply arbitrary positions in the editor from which the resultant surrounding hue shifts and saturation scales of the adjacent points in the 2250 DNG profile positions are interpolated/calculated?
I had hoped that I could find the values corresponding with the 24 patches on a color checker and determine the hue shift/saturation scale recorded from these by reading the DNG profile... Or will that simply not work?
Thanks!

I am also uncomfortable with the 1.0.0.46 release with a D800.
Using Lightroom 4.2 I looked at the color checker chart that I used to generate my profiles using the DNG Profile editor. The blue change with this release was very significant.
Hovering over the blue panel of the chart after changing just the camera profile I got the following general readings:
Adobe Std.         R0    G22 B60
Beta 3                   R4    G18 B59.5
Beta .46                R17 G22 B59
Without an explanation, such a deviation from Adobe Standard is disconcerting as my base profile in the DNG Profile editor, both releases, was color checker.

DNG profile editor - truncated RGB Lab HSL values

Similar Messages

Maybe you are looking for