RawTherapee 5.8 WB CTE problem?

Oxygenum · 11-16-2020, 10:07 AM

RAW from Pentax K-70. In camera settings I set WB balance CTE.
In ZIP are DNG and output from DCU (Soft from Pentax) and RT 5.8
It is totally different on neutral settings ?!

RT 5.8 WB CTE bug ?! - RawTherapee - discuss.pixls.us

Only Pentax soft work ok with CTE WB?

photoptimist · 11-16-2020, 11:26 AM

Two key points:

1) CTE (Color Temperature Enhancement) is not a neutral setting, quite the opposite. It is intended to increase the effects of a non-white, non-neutral lighting source. It's almost the opposite of the white-balance correction concept.

2) In theory, RAW files are not directly affected by any of the WB settings or other JPEG-related settings. That statement is true for the main RAW image data file. However (and it's an important however) the thumbnail image for the RAW file and many photographic software programs use these JPEG settings from the EXIF and apply them when converting the RAW file into a viewable image.

It's possible that RT 5.8 does not know about the CTE setting (and ignores it) or has a different algorithm for implementing it than the Pentax version.

The Squirrel Mafia · 11-16-2020, 11:51 AM

RawTherapee doesn't recognize certain manufacturer specific camera settings like CTE WB, D-range settings, & things like that. It can't read color tones like Bright, Neutral, Vivid, & similar, unless you have a .dcp file from Adobe or one that you've made yourself.

Funny thing is that I don't even think that the Pentax Digital Camera Utility 5 recognizes CTE WB as well, but I could be wrong on that one. It has the typical WB settings, but I don't see one for CTE. Maybe CTE is recognized through the Camera setting WB option.

pschlute · 11-16-2020, 12:35 PM - **1 Like**

If you are shooting raw, the WB is for you to assign when processing. Use a grey card in the scene and work from there.

stevebrot · 11-16-2020, 04:44 PM

Originally posted by Oxygenum

RawTherapee 5.8 WB CTE problem?

White balance is sort of a sticky subject. As noted above, the DNG data has no notion of white balance. That said, the makernotes section of the EXIF metadata does include the camera's white balance settings as a "hint" to software such as Lightroom or RawTherapee.* The tag name used by ExifTool is WhitePoint and the value is four integer numbers, representing factors to be applied to each of the four color channels (R, G1, G2, B) for processing. The numbers for IMGP6092.DNG are:

21024 8192 8192 12640

Strangely enough, that is all there is. White balance is a surprisingly course adjustment. Looking at IMGP6092.JPG, the same four numbers are present as documentation of the camera's setting, that actual factors having already been applied.** IMGP6092neutralWBRT.JPG has no manufacturer's metadata, being that it was made by RawTherapee.

I don't know what the "neutral" setting is, but I did notice something interesting in comparing the DCU JPEG with the RT JPEG. The DCU version has sRGB as declared color space and also has that standard profile embedded in the file. The RT version on the other hand declares RTv4_sRGB and embedded a quite different profile matrix in that file. Depending on the display software/device, that in itself might be enough to account for the difference in colors.

Beyond that, the RT JPEG appears to have been processed darker by quite a bit with less contrast.

Addendum: I processed in Lightroom using default import and Adobe Standard and got results almost identical to that from DCU. Lightroom evaluated the color temperature at 6900.

Steve

* Lightroom will also apply a tint value, but that is not pulled from the EXIF.

** In the case where a RAW converter is not able to pull WB factors from the EXIF, it uses a standard set of values for 5200K and a neutral gray subject.

stevebrot · 11-16-2020, 05:29 PM

Originally posted by The Squirrel Mafia

Funny thing is that I don't even think that the Pentax Digital Camera Utility 5 recognizes CTE WB as well, but I could be wrong on that one.

v5.8.6 definitely does in the Laboratory module, White Balance tab, choosing override, followed by CTE.

Steve

stevebrot · 11-16-2020, 05:39 PM

Originally posted by stevebrot

Looking at IMGP6092.JPG, the same four numbers are present as documentation of the camera's setting, that actual factors having already been applied.

OK...I opened IMGP6092.dng in DCU and changed the WB type to "auto" and saved to JPEG. The factors recorded for that method were:

16849, 8192, 8192, 14228

...quite different from the CTE method used for IMGP6092.jpg.

Near as I can tell, RT probably used the WB factors as suggested by the camera, but did some other things along the way.

Steve

stevebrot · 11-16-2020, 07:01 PM

Originally posted by The Squirrel Mafia

RawTherapee doesn't recognize certain manufacturer specific camera settings like CTE WB, D-range settings, & things like that. It can't read color tones like Bright, Neutral, Vivid, & similar, unless you have a .dcp file from Adobe or one that you've made yourself.

Near as I can tell, CTE is just the usual set of WB factors and not a custom processing feature. Have you looked at the PP3 file provided by the OP in the ZIP?

Steve

The Squirrel Mafia · 11-16-2020, 09:30 PM

^ Finally noticed that CTE setting in PDCU5.

I extracted the embedded jpeg with PDCU5. The jpeg is set to the Bright color tone & has CTE white balance. I tried to do something similar in RawTherapee.

RawTherapee.zip - Google Drive

^ That will give you an idea.

I think the D-range settings are on. RawTherapee can't really read them & when using the auto-match tone curve, RawTherapee can't quite match it either.

stevebrot · 11-17-2020, 09:15 AM

Originally posted by The Squirrel Mafia

^ Finally noticed that CTE setting in PDCU5.

I extracted the embedded jpeg with PDCU5. The jpeg is set to the Bright color tone & has CTE white balance. I tried to do something similar in RawTherapee.

RawTherapee.zip - Google Drive

^ That will give you an idea.

I think the D-range settings are on. RawTherapee can't really read them & when using the auto-match tone curve, RawTherapee can't quite match it either.

I played a bit more with the DNG this morning and processed through dcraw to 16-bit TIFF. The results were very similar to the OP's results from RT, though a bit darker with higher contrast. I bumped the contrast on the RT JPEG a bit in LR and it matched the dcraw. Actual olor balance is close between all three RAW convertors.

My conclusion is that no proprietary magic is required to honor the CTE and that the differences between the DCU/in-camera and the OP's RT work is mostly the "Bright" profile and whatever curve tweaking was applied by DCU/in-camera.

Steve

The Squirrel Mafia · 11-17-2020, 10:16 AM - **1 Like**

Yeah. PDCU5 is probably the only piece of software that can read just about every single setting available in the camera like D-range settings & it will apply the same settings to the RAW file that results in a jpeg that looks nearly identical the OOC jpeg, but with better fine detail & color retention. That's PDCU5 shines, but compared to other software out there, it's very limited in features.

I have my K-50 set up only to take RAW images only, but I have turned off all D-range settings, lens correction settings, noise reduction settings, & other similar jpeg corrections. This way I expose the shot more properly, instead of relying on the camera to "fix" it for me. The end result is getting nice looking jpegs out of RawTherapee that have very similar contrast & brightness of the embedded OOC jpegs instead of getting a very dark or bright jpeg out of RawTherapee that is completely different from the OOC jpeg. The whole challenge manually of trying to match the contrast & brightness gets annoying. Programs like DxO & similar also can't read Pentax specific jpeg corrections, so it's a lot easier post processing RAW files.

11-16-2020, 10:07 AM	#1
Oxygenum Forum Member Join Date: Feb 2014 Posts: 70	RawTherapee 5.8 WB CTE problem? RAW from Pentax K-70. In camera settings I set WB balance CTE. In ZIP are DNG and output from DCU (Soft from Pentax) and RT 5.8 It is totally different on neutral settings ?! RT 5.8 WB CTE bug ?! - RawTherapee - discuss.pixls.us Only Pentax soft work ok with CTE WB?

11-16-2020, 04:44 PM	#5
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by Oxygenum RawTherapee 5.8 WB CTE problem? White balance is sort of a sticky subject. As noted above, the DNG data has no notion of white balance. That said, the makernotes section of the EXIF metadata does include the camera's white balance settings as a "hint" to software such as Lightroom or RawTherapee.* The tag name used by ExifTool is WhitePoint and the value is four integer numbers, representing factors to be applied to each of the four color channels (R, G1, G2, B) for processing. The numbers for IMGP6092.DNG are: 21024 8192 8192 12640 Strangely enough, that is all there is. White balance is a surprisingly course adjustment. Looking at IMGP6092.JPG, the same four numbers are present as documentation of the camera's setting, that actual factors having already been applied. IMGP6092neutralWBRT.JPG has no manufacturer's metadata, being that it was made by RawTherapee. I don't know what the "neutral" setting is, but I did notice something interesting in comparing the DCU JPEG with the RT JPEG. The DCU version has sRGB as declared color space and also has that standard profile embedded in the file. The RT version on the other hand declares RTv4_sRGB and embedded a quite different profile matrix in that file. Depending on the display software/device, that in itself might be enough to account for the difference in colors. Beyond that, the RT JPEG appears to have been processed darker by quite a bit with less contrast. Addendum:** I processed in Lightroom using default import and Adobe Standard and got results almost identical to that from DCU. Lightroom evaluated the color temperature at 6900. Steve * Lightroom will also apply a tint value, but that is not pulled from the EXIF. ** In the case where a RAW converter is not able to pull WB factors from the EXIF, it uses a standard set of values for 5200K and a neutral gray subject. Last edited by stevebrot; 11-16-2020 at 06:55 PM.

11-16-2020, 05:29 PM	#6
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by The Squirrel Mafia Funny thing is that I don't even think that the Pentax Digital Camera Utility 5 recognizes CTE WB as well, but I could be wrong on that one. v5.8.6 definitely does in the Laboratory module, White Balance tab, choosing override, followed by CTE. Steve Last edited by stevebrot; 11-16-2020 at 05:40 PM.

11-16-2020, 05:39 PM	#7
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by stevebrot Looking at IMGP6092.JPG, the same four numbers are present as documentation of the camera's setting, that actual factors having already been applied. OK...I opened IMGP6092.dng in DCU and changed the WB type to "auto" and saved to JPEG. The factors recorded for that method were: 16849, 8192, 8192, 14228 ...quite different from the CTE method used for IMGP6092.jpg. Near as I can tell, RT probably used the WB factors as suggested by the camera, but did some other things along the way. Steve

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
A stunning number of NFL players who suffer from CTE.	Dewman	General Talk	9	03-22-2016 12:26 PM
What does CTE do and when to use it?	uday029	Pentax DSLR Discussion	6	12-15-2014 06:21 AM
K-x tungsten WB is not a fixed WB	kxr4trids	Pentax DSLR Discussion	3	05-18-2010 12:43 AM
K-7's auto-WB vs. Lightroom's auto-WB	inferno10	Post Your Photos!	13	07-16-2009 08:54 AM
CTE WB on K20D?	yipchunyu	Pentax DSLR Discussion	0	05-26-2009 05:59 AM

11-16-2020, 11:26 AM	#2
photoptimist Pentaxian Join Date: Jul 2016 Photos: Albums Posts: 5,128	Two key points: 1) CTE (Color Temperature Enhancement) is not a neutral setting, quite the opposite. It is intended to increase the effects of a non-white, non-neutral lighting source. It's almost the opposite of the white-balance correction concept. 2) In theory, RAW files are not directly affected by any of the WB settings or other JPEG-related settings. That statement is true for the main RAW image data file. However (and it's an important however) the thumbnail image for the RAW file and many photographic software programs use these JPEG settings from the EXIF and apply them when converting the RAW file into a viewable image. It's possible that RT 5.8 does not know about the CTE setting (and ignores it) or has a different algorithm for implementing it than the Pentax version.

11-16-2020, 11:51 AM	#3
The Squirrel Mafia Pentaxian Join Date: Jan 2015 Posts: 3,056	RawTherapee doesn't recognize certain manufacturer specific camera settings like CTE WB, D-range settings, & things like that. It can't read color tones like Bright, Neutral, Vivid, & similar, unless you have a .dcp file from Adobe or one that you've made yourself. Funny thing is that I don't even think that the Pentax Digital Camera Utility 5 recognizes CTE WB as well, but I could be wrong on that one. It has the typical WB settings, but I don't see one for CTE. Maybe CTE is recognized through the Camera setting WB option.

11-16-2020, 12:35 PM - 1 Like	#4
pschlute Moderator Loyal Site Supporter Join Date: Mar 2007 Location: Surrey, UK Photos: Gallery Posts: 8,214	If you are shooting raw, the WB is for you to assign when processing. Use a grey card in the scene and work from there.

11-16-2020, 07:01 PM	#8
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by The Squirrel Mafia RawTherapee doesn't recognize certain manufacturer specific camera settings like CTE WB, D-range settings, & things like that. It can't read color tones like Bright, Neutral, Vivid, & similar, unless you have a .dcp file from Adobe or one that you've made yourself. Near as I can tell, CTE is just the usual set of WB factors and not a custom processing feature. Have you looked at the PP3 file provided by the OP in the ZIP? Steve

11-16-2020, 09:30 PM	#9
The Squirrel Mafia Pentaxian Join Date: Jan 2015 Posts: 3,056	^ Finally noticed that CTE setting in PDCU5. I extracted the embedded jpeg with PDCU5. The jpeg is set to the Bright color tone & has CTE white balance. I tried to do something similar in RawTherapee. RawTherapee.zip - Google Drive ^ That will give you an idea. I think the D-range settings are on. RawTherapee can't really read them & when using the auto-match tone curve, RawTherapee can't quite match it either.

11-17-2020, 09:15 AM	#10
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by The Squirrel Mafia ^ Finally noticed that CTE setting in PDCU5. I extracted the embedded jpeg with PDCU5. The jpeg is set to the Bright color tone & has CTE white balance. I tried to do something similar in RawTherapee. RawTherapee.zip - Google Drive ^ That will give you an idea. I think the D-range settings are on. RawTherapee can't really read them & when using the auto-match tone curve, RawTherapee can't quite match it either. I played a bit more with the DNG this morning and processed through dcraw to 16-bit TIFF. The results were very similar to the OP's results from RT, though a bit darker with higher contrast. I bumped the contrast on the RT JPEG a bit in LR and it matched the dcraw. Actual olor balance is close between all three RAW convertors. My conclusion is that no proprietary magic is required to honor the CTE and that the differences between the DCU/in-camera and the OP's RT work is mostly the "Bright" profile and whatever curve tweaking was applied by DCU/in-camera. Steve

11-17-2020, 10:16 AM - 1 Like	#11
The Squirrel Mafia Pentaxian Join Date: Jan 2015 Posts: 3,056	Yeah. PDCU5 is probably the only piece of software that can read just about every single setting available in the camera like D-range settings & it will apply the same settings to the RAW file that results in a jpeg that looks nearly identical the OOC jpeg, but with better fine detail & color retention. That's PDCU5 shines, but compared to other software out there, it's very limited in features. I have my K-50 set up only to take RAW images only, but I have turned off all D-range settings, lens correction settings, noise reduction settings, & other similar jpeg corrections. This way I expose the shot more properly, instead of relying on the camera to "fix" it for me. The end result is getting nice looking jpegs out of RawTherapee that have very similar contrast & brightness of the embedded OOC jpegs instead of getting a very dark or bright jpeg out of RawTherapee that is completely different from the OOC jpeg. The whole challenge manually of trying to match the contrast & brightness gets annoying. Programs like DxO & similar also can't read Pentax specific jpeg corrections, so it's a lot easier post processing RAW files.