Dual ISO - massive dynamic range improvement (dual_iso.mo)

[Armetz]

hello
I´m quite new to dual iso, maybe the issues I have here were discussed before  .. I have 2 problems with my dual iso shots:
1. some merged frames differ quite much in brightness but I can´t see that in the the original dual iso dngs. It appears for 20 or so frames in a sequence every few 100 frames. I uploaded exaple dngs here (original and merged)  https://www.dropbox.com/sh/dvwp3f2xkpkzlmh/AAAMo2z1bBiuxXjhvDHYMw2Ha?dl=0
2. the merged files appear to have a color temperature shift towards blue. That´s not really a big problem but maybe there´s an easy way to fix this..

thanks

[xorpi]

Hi! I have 2 questions:
1:
Is possible to combinate Canon Highlight Tone Priority mode with Dual ISO for max DR?

2:
And... is there any new improvments  with shadows/highlites/blacks without loosing details and informations? - It's really BIG FAIL

[DeafEyeJedi]

1. No (There are reasons why we've been instructed to have HTP off).

2. Not sure what ur asking but it depends on which ISO combo (100/800 or 100/1600) you use along with how you expose it and also depends on which camera you use

[Walter Schulz]

2. Seems like he wants to have Dual-ISO benefit without trade-off.

[xorpi]

Dual ISO with actual performance is not good for any types of hobby or professional photography... This "massive dr improvments" is like "massive miss of the best shots" I want the best performance as possible from my gear (EOS 700D) and with Dual-ISO is like lottery for few geeks. Not for hobby enthusiastic photographers or... professionals.

[dfort]

Guess that makes me a Geek that won the lottery. All my years as a professional photographer doesn't count?

Of course Dual ISO works. You just need to know how to use it and just as important, when it is better not to use it.


Sent from my iPhone using Tapatalk

[nikfreak]

+1 @dfort

[R]

Dual ISO with actual performance is not good for any types of hobby or professional photography... This "massive dr improvments" is like "massive miss of the best shots" I want the best performance as possible from my gear (EOS 700D) and with Dual-ISO is like lottery for few geeks. Not for hobby enthusiastic photographers or... professionals.

I have sold thousands of Dual ISO photos. I'd recommend you to experiment a lot, compare results, refine your exposure calculations, use zone system for metering areas, anticipate where your shadows will fall after running cr2hdr, and start again until you find your method. Once you know where to measure and in which zone the opposite side of the histogram will fall you'll be able to predict results.

Check my portfolio www.ramonclemente.com/portfolio, most of the photos are Dual ISO.

[DeafEyeJedi]

Spot on, @dfort and perhaps it takes a true geek to at least understand how to use (or not to use) such features properly, right? [emoji6]

Simple trial and errors goes a long way and for good reasons as @R said it perfectly!

Thanks for the friendly reminder @xorpi!

[Marsu42]

Dual ISO with actual performance is not good for any types of hobby or professional photography...

This is just FUD. I'm using dual_iso since it was released and use it on most outdoor shots nowadays. When in doubt, a little wasted dr and semi-clipping with dual_iso beats completely blown highlights and low-res blacks drowned in noise.

With recent Canon sensors ml supports (like 6d) you get excellent iq, 16bit raw files and can actually match the dynamic range of the scene and the image by selecting the correct iso spread (100/200, 100/400, ...) = optimal bit-res usage. Even on older sensors like the 18mp crop the shadow noise is ok up to 100/800, but with hugely improved dr.

As you need to back up your shots anyway, there's no additional file space requirement (back up the dual_iso dng, work on the cr2hdr-processed dng). The only hassle is the broken auto-wb, the striped preview and the time to process the files.

[Danne]

The only hassle is the broken auto-wb

Reenter CR2 white balance back into a dual iso dng can be done quite easily. Every CR2 file stores WB RGGB Levels As Shot. For instance:

WB RGGB Levels As Shot          : 1739 1024 1024 1931

1024/1739=0,588844163312248
1024/1931=0,528107271789582

Then inject the tag with exiftool.
exiftool "-AsShotNeutral=0,588844163312248 1 0,528107271789582" dualiso.dng

I don,t know how to do this in C but since the CR2 is always there in front of you I don,t see why this can,t be done inside cr2hdr.

Calculations coming from Chmee and a lot of trial and error from my part when building scripts.

[DeafEyeJedi]

Reenter CR2 white balance back into a dual iso dng can be done quite easily. Every CR2 file stores WB RGGB Levels As Shot.

I don,t see why this can,t be done inside cr2hdr.

Agreed. This would be nice. Should be doable.

[a1ex]

--wb=exif



See also http://www.magiclantern.fm/forum/index.php?topic=10265 (in particular, posts from ayshih and Marsu42)

[Danne]

Last time I went through --wb=exif this was happening .
http://www.magiclantern.fm/forum/index.php?topic=16024.msg155670#msg155670

I since deleted that particular CR2 but I could probably dig it out from some folder if necessary. In the thread is the solution coming from using dcraw -i -v multiplier method.

[a1ex]

Here, --wb=exif gives the same results as dcraw -i -v. I probably missed your post back then, so please re-upload the file (or other file that has the same issue).

[Danne]

I will see If I can find that particular file or if I can reproduce the issue otherwise. Meanwhile.
Tested just now three files like this.
The auto wb CR2 file might be worth checking. All three files except AWB was giving similar results as dcraw -i -v.
https://drive.google.com/file/d/0B4tCJMlOYfirLVVXZXIzNnQzOTA/view?usp=sharing

Checked in acr

Auto
CR2 6400 +35
DNG 6050 +27 (dcraw -i -v)
DNG 3850 +10 (--wb=exif)

2500 kelvin
CR2 2500 +3
DNG 2400 -1

10000 kelvin
CR2 10000 + 9
DNG 9300 +1



Auto (dcraw -i -v through MLP)
6050 +27


Auto (--wb=exif)
3850 +10

[Danne]

Here we go. Found the CR2. Issue reproduced.
Problematic file
https://drive.google.com/file/d/0B4tCJMlOYfirb1FqTzVNRXZzVkk/view?usp=sharing

dcraw i -v


--wb=exif

[a1ex]

Code Select Expand


exiftool DUAL2695.CR2 | grep RGGB
Measured RGGB                   : 144 1024 1024 1136
WB RGGB Levels As Shot          : 2036 1024 1024 1990
WB RGGB Levels Auto             : 2036 1024 1024 1990
WB RGGB Levels Measured         : 2036 1024 1024 1990
WB RGGB Levels Daylight         : 2001 1024 1024 1667
WB RGGB Levels Shade            : 2300 1024 1024 1421
WB RGGB Levels Cloudy           : 2153 1024 1024 1535
WB RGGB Levels Tungsten         : 1440 1024 1024 2485
WB RGGB Levels Fluorescent      : 1753 1024 1024 2389
WB RGGB Levels Kelvin           : 1324 1024 1024 2774
WB RGGB Levels Flash            : 2231 1024 1024 1511
Raw Measured RGGB               : 219 243 2608 1555
WB RGGB Levels                  : 2036 1024 1024 1990


cr2hdr --wb=exif:

Code Select Expand


White balance   : from RawMeasuredRGGB
AsShotNeutral   : 0.90 1 0.60


ayshih's comments in exiftool-bridge.c:

Code Select Expand


    //If WB mode is not Auto, use WB_RGGBLevelsAsShot values
    //If WB mode is Auto, read WB_RGGBLevelsMeasured values
    //  (not WB_RGGBLevelsAuto values because they can be temperature-shifted)
    //  Use WB_RGGBLevelsMeasured values if they have a significant difference between the G channels
    //  Otherwise, use RawMeasuredRGGB values


https://bitbucket.org/hudson/magic-lantern/pull-requests/453/white-balance-fix-for-dual-iso/diff

Looks like what works on one camera may cause trouble on others, so I think we need tricky test cases for more camera models in order to solve this puzzle.

[Danne]

What are we looking for? Checking auto wb cr2hdr seems to use dcraw auto wb when camera set to AWB(?) But viewing CR2 files you can match AWB output much closer with multipliers. The numbers aren,t exact but it seems close enough. To me .

Here is the same awb test from an eos M.
Original CR2 (5050 +29)


dcraw -i -v (4650 +26)


cr2hdr --wb=exif (3800 +18)

[a1ex]

I bet ayshih did this change for a reason, but he tested on 50D. Here's the exif info from his sample file:

http://www.magiclantern.fm/forum/index.php?topic=10265.msg107458#msg107458

[Danne]

Aha. A few samples from a 50D to start with then. Anybody?
Auto WB, 2500K and 10000K?

[josepvm]

50D samples:
---------------

WB auto, dual ISO enabled 100/800
WB auto, dual ISO disabled
WB 2500K, dual ISO enabled 100/800
WB 2500K, dual ISO disabled
WB 10000K, dual ISO enabled 100/800
WB 10000K, dual ISO disabled

Same scene, with an EOS M:
---------------------------------
WB auto, dual ISO enabled 100/800
WB auto, dual ISO disabled
WB 2500K, dual ISO enabled 100/800
WB 2500K, dual ISO disabled
WB 10000K, dual ISO enabled 100/800
WB 10000K, dual ISO disabled

[Danne]

Thanks josepvm. I,m on a limited time schedule atm but I managed to work your files. The relation is the same. Auto wb dcraw -i -v match much closer than --wb=exif.
CR2 5750 +35
DNG 5900 +34
DNG 4100 -7 --wb=exif

Conclusion over here is that reliable number to work with are the WB RGGB Levels As Shot values. The other stuff, at least, from my testing way back, I couldn,t make real sense of. So, what steps ahead? Getting back Ayshih in the discussion?

[Danne]

Read through some more interesting things and I wonder if this would be the solution? The missing matrices.
http://www.magiclantern.fm/forum/index.php?topic=10265.msg118799#msg118799

*Added all the recommended tags but it didn,t alter acr white balance settings.

[nikfreak]

The solution is to include the ColorMatrix2, CameraCalibration1, CameraCalibration2, CalibrationIlluminant2, ForwardMatrix1 and ForwardMatrix2 values for the camera in the DNG file. It also requires setting CalibrationIlluminant1 to 'Standard Light A' and CalibrationIlluminant2 to D65.

This?