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

Started by a1ex, July 16, 2013, 06:33:50 PM

Previous topic - Next topic

0 Members and 3 Guests are viewing this topic.

Alex Roman


Canon eos m

Folks, any idea how to recover the exif data of a photo taken using dual_iso. Want to see what settings was the camera using when I took the shot.

You may have noticed the filmy grain look (as opposed to noise on a DSLR) when shooting with dual_iso. There are situations that don't need the dual functionality since the DR is not too high but I still cannot get myself to not using dial_iso. I just love the results. ACR 8.1 handles RAW so much better when shot with multiple iso.
Canon 5D Mark III, Gopro Hero Blacks with 3D Casing, A Few Lenses, Adobe CC 2014, MacBook Pro, Windows 8 PC, Lots of Video Rig!

Started Nuke. Loved it but then the 15 day trial ran out. Back to After Effects and loving it :-)

sletts02

Any update on if you can shoot 'auto' video exposure and dual iso raw?

Magic 7D

Quote from: Danne on July 26, 2013, 07:19:02 PM
check out my picture of a 100-6400 some posts up. Worked rather well.

First post.. So please bear with me if I seem dumb.
But Danne I'm very impressed with your pictures! Would you like to share how managed to render them? I've been trying for a few days now almost full time to get some good pics out of this Dual ISO function, but always get the horrid pink magenta highlights from cr2hdr and even some green stuff when under exposed, don't know what I'm doing wrong.. Did you use cr2hdr? Or how did you process your images can't wait to get a convenient workflow for this amazing function. And can't thank everyone working with it enough!!
A MAGIC 7D and 5D III + Lenses in the range of 15 - 400mm. Man it's shining.. it's a Lantern oh it's Magic, Aha it's Magic Lantern! ;)

Danne

Yup, used cr2hdr. In the nightpictures there is some green and red moiré or something. Only minor though. In the daylight pictures almost none. When filming a lot more green/red issues appear. Mostly 100-1600 or 100-3200 used in the pics. Clean the pictures in lightroom to get rid of purple fringing

a1ex

Quote from: Pelican on July 26, 2013, 11:52:58 AM
There is a working build for the 7D in the 7D alpha2 topic.
The produced files (ISO 100/1600 and ISO 1600/100):
http://pel.hu/down/100_1600_7D.CR2
http://pel.hu/down/100_1600_7D.DNG
http://pel.hu/down/1600_100_7D.CR2
http://pel.hu/down/1600_100_7D.DNG

In these files, the white level is close to 13000; in my 7D test sample it's 15382. Normally you get these values with intermediate ISOs (160, 320).

If you change the white level in cr2hdr.c to 13000 instead of 15000, these files will be converted without pink highlights. Now trying to figure out how to get this value from exif or what settings to use to get raw values higher than 15000.

a1ex

Updated the converters (download from first post):

cr2hdr:
- fix memory leak, tested with 100 GB of pictures
- white level autodetection (should fix pink highlights)
- exiftool fix from jpaana, not tested but looks OK

raw2dng:
- fix memory leak, not tested (should be OK with large clips)

I did not address the black level issues yet.

If you are having issues with image quality, please upload the raw files (*.CR2 or *.RAW).

Danne

Some more examples of what dual iso can achieve. The shadows pop out like never before. Mostly use interval between 100-1600 and then use ETTR to find exposure settings. Works perfect most of the time. From a visit in Copenhagen.
Due to bandwith reasons and uploaded images being too big i decided to post a youtubelink containing the stills instead. Enjoy.




Magic 7D

Quote from: Danne on July 30, 2013, 08:52:56 PM
Some more examples of what dual iso can achieve. The shadows pop out like never before. Mostly use interval between 100-1600 and then use ETTR to find exposure settings. Works perfect most of the time.

It's incredible!
Sorry if this is a noob question, but what is ETTR...? I was actually going to ask you how you know what exposure to do before.

@a1ex thanks a thousand for your hard work!!! How can I send you a cookie?  :)
A MAGIC 7D and 5D III + Lenses in the range of 15 - 400mm. Man it's shining.. it's a Lantern oh it's Magic, Aha it's Magic Lantern! ;)

Danne

ETTR is a great function that require some testing out. It,s in the menu. You can search on the forum on how to work with it.
I just hit ETTR on and then proceed to dual iso where I set the interval lets say 100-1600 or 100-3200. WHen I hit the set button the mirror flips and the ETTR is making its magic (calculating exposure to the right). Don,t be surprised if the iso and shutter changes. I believe the intervals stays the same though. Not sure. Than take the picture.
You could always do it manually by selecting the first iso for the highlights and then the other iso for shadows. If there,s a lot of contrast usually between 100-1600 or bigger works nice. Even 100-6400 could work sometimes.
Yup, it,s kind of magical. Feels like another camera atm :)

Walter Schulz


IliasG

Quote from: a1ex on July 30, 2013, 04:37:20 PM
In these files, the white level is close to 13000; in my 7D test sample it's 15382. Normally you get these values with intermediate ISOs (160, 320).

If you change the white level in cr2hdr.c to 13000 instead of 15000, these files will be converted without pink highlights. Now trying to figure out how to get this value from exif or what settings to use to get raw values higher than 15000.

Some Canon models (I think all APS-C with 18Mp) have lower white levels for ISO 100 than the rest ISOs. For 7D it's around 13500 for ISO 100 and 15300 for the rest "integer ISOs".

It would be very useful if you find a way to declare this from *.CR2 and it's exif :)

The sad thing is that some channels (B and G2 in my samples) have raw values higher that this white clipping point making the detection difficult even if we have burned areas ..

a1ex

dcraw uses 13584 (hardcoded). And yes, since the white points changes with ISOs, it makes things difficult; for now, I just use the smallest value from burned areas, if it's above 10000. Probably a bit overkill, maybe 13000 should be a better starting point?

sletts02

It would be fantastic to see some video examples of this. Perhaps indoors with windows to see how it can compensate for blown windows/doors.

Also, could someone try recording in auto exposure movie mode and report back it it works with dual ISO - I will be forever grateful!

a1ex

Right now, auto exposure won't work in raw video, because the digital ISO gain is not applied to raw data.

But if you include the post deflicker algorithm in raw2dng, you may even get something usable. Worth trying IMO.

IliasG

Quote from: a1ex on July 30, 2013, 09:36:33 PM
dcraw uses 13584 (hardcoded). And yes, since the white points changes with ISOs, it makes things difficult; for now, I just use the smallest value from burned areas, if it's above 10000. Probably a bit overkill, maybe 13000 should be a better starting point?

Dcraw is ineffective regarding WP, it uses a single value for every ISO and runs into the known problem, either pink highlights (ISO 100-160-320-...) or 1/3 stop burned highlights.

The best solution is to have a lookup table with the correct clipping point for every case. Keep in mind that we have data scaling (and higher WP) even when shooting at wide apertures (wider than f/2.8) with Canon glass ..

Adobe DNG converter uses 13000 as WP for 100-160-320 .. and 15000 for 200-400-800 .. For 250-500... I think it goes up to 16000.
But takes no care for wide apertures so in this case up to 1/3 stop highlights get lost ..

I do think that if you use a single WP then 13000 is a safe bet. Although you loose 1/3 stops of highlights for ISO 200-400-.. and maybe more for 250-500.. and when using wide apertures.

13584 is on the limits for ISOs 100-160-320-.. because when there is raw denoise (like "black frame subtraction") then we take not a single clipping point but a distribution extending -/+ 100 raw levels from the center value .. it's the same case as in Guillermo's  study .. http://www.guillermoluijk.com/tutorial/satlevel/index.htm

a1ex

Well, I'd use 13000 or higher. If there are no burned areas, the only side effect should be picture being a little brighter than normal, and maybe a few clipped pixels (and highlight recovery should catch them). If there are burned areas, it will use the true white level.

arrinkiiii

Quote from: a1ex on July 30, 2013, 05:35:34 PM
Updated the converters (download from first post):

cr2hdr:
- fix memory leak, tested with 100 GB of pictures
- white level autodetection (should fix pink highlights)
- exiftool fix from jpaana, not tested but looks OK

raw2dng:
- fix memory leak, not tested (should be OK with large clips)

I did not address the black level issues yet.

If you are having issues with image quality, please upload the raw files (*.CR2 or *.RAW).

The new cr2hdr works very good, no more pink highlights in the photos taking with the 7D. Tomorrow will take decent photos.

Thank you a1ex  :D

feureau

Btw, is there a way to make a drag-and-drop a bunch of cr2 to convert to DNR using cr2hdr? (batch convert)

a1ex

On Linux I run cr2hdr *.CR2

No idea if this works on Windows, but try it.

Danne

Tried a little quick and dirty test filming in 3x crop mode with a 14mm samyang lens. I used 6400-100 setting in a very dark kitchen. Surprisingly I can,t find any moiré or aliasing. The picture is totally clean! Check out the movielink. Genious work.



Also a link to my still pictures made with dual iso on youtube.


Audionut

Quote from: feureau on July 31, 2013, 08:43:08 AM
Btw, is there a way to make a drag-and-drop a bunch of cr2 to convert to DNR using cr2hdr? (batch convert)

Drag and drop a bunch of them on the executable.

feureau


lourenco

These are not extreme examples, but show practical use for me at the track. 100-800 ISO. 










It was a big help for this shot.


5D Mark III, CF Lexar 1000X 32GB, 24-105 F4L

ShootingStars

Nice shots! Is it me or does it seem these "merged" shots seem kinda unrealistic? Or is it because we have rarely seen this. Looks like HDR. What's the difference?