The CinemaDNG Discussion (raw2cdng)

[marekk]

@marekk since when is the badpixel thing implemented? i have to look to newest raw-files.. thx.

It's not in raw file. There is a special badpixel_opcode in DNG tags. Look at chdk-dng.c file. With this tag dead/hot pixels can be removed automatically for example by Adobe Camera Raw.

[artyg]

nightvision04 is absolutely right.  Set it to "Clip" "As Shot" "BMD Film".  I'm grading an hour of 5D2 raw footage without LUTs learning resolve lite as I go and it looks very filmic.   rec. 709 will show pink fringing and even more moire.  The BMD film setting is a vast improvement.

Where this settings must be set?

[chmee]

@marekk it seems its just a flag for allowing repairing "bad" pixels - described with a number 0-4 - hmm. ok. will put it on toDo-list. but we have to check, if resolve/speedgrade is able to repair badpixels - so inbody masking is imho the better solution)

here is a version 1.1.8 with reelname, timecode and framerate in header - maybe it helps. the framerate is static on 24fps - because we're in testing mode - if it works i will change to variable framerates and timecode (if needed, starttime of filegenerate-datetime)

Download v1.1.8 here

regards chmee

ToDo-List 20.06.2013
(*) Roundtrip must work
(*) IliasG's idea of 10bit log (state 90%)
(*) strippedFiles
(*) Repair-Flag für badpixels? (simpler workaround: demasking in body)
(*) commandline

[peoplemerge]

Where this settings must be set?

Color pane, bottom left panel there's 4 icons, a circle, bar chart, 3 circles in a triangle, and a movie cam.  The last one, the movie cam, is for "Camera Raw".  below that is Master Settings pane.  Set the Decode Using to "Clip" etc as discussed above

[tin2tin]

@chmee Great to see you're still going strong! I haven't got resolve on this computer, so I can't test atm. Let's hope the mediainfo guys will add info subraction of cdng files.

[mediali]

@marekk it seems its just a flag for allowing repairing "bad" pixels - described with a number 0-4 - hmm. ok. will put it on toDo-list. but we have to check, if resolve/speedgrade is able to repair badpixels - so inbody masking is imho the better solution)

here is a version 1.1.8 with reelname, timecode and framerate in header - maybe it helps. the framerate is static on 24fps - because we're in testing mode - if it works i will change to variable framerates and timecode (if needed, starttime of filegenerate-datetime)

Download v1.1.8 here

regards chmee

ToDo-List 20.06.2013
(*) Roundtrip must work
(*) IliasG's idea of 10bit log (state 90%)
(*) strippedFiles
(*) Repair-Flag für badpixels? (simpler workaround: demasking in body)
(*) commandline

still not work  about timecode  problem

here is the xml file please have a look

http://www.istorm2010.com/temp/A-Resolve-out.xml

http://www.istorm2010.com/temp/B-premiere out(Importing A).xml

http://www.istorm2010.com/temp/C-Resolve out again (impotrting B).xml

[iunknown]

Yeah, tested, same issue, duplicate clips on the xml import into resolve.  Thanks for the effort chmee.

[nowhereman86]

I was wondering if anyone has had luck roundtriping from Resolve with the "AAF Roundtrip" function and using AVID to edit. I used RAWMagic to decode the RAW to CDNG, then followed the steps in this video: (

This worked perfectly for me! I edited a one minute short with multiple clips and it went back and forth between the programs without a hitch. Anyone else had luck trying this method?

[zach915m]

I'm on a PC so I can't test RawMagic.  But I did use the AAF roundtrip to test this and it didnt work...=(

[marekk]

OpcodeLists are allowed in CinemaDNG file format so maybe Davinci Resolve support this feature..

@marekk it seems its just a flag for allowing repairing "bad" pixels - described with a number 0-4 - hmm. ok. will put it on toDo-list. but we have to check, if resolve/speedgrade is able to repair badpixels - so inbody masking is imho the better solution)

[chmee]

"maybe" is the right description. As of version 1.1.0.0 of the CDNG-specification opcodes are optional by decoders, not mandatory. i put it on toDo-list, but there are some fixes before that..

by now i'm out for the next three/four days - because i cant solve the roundtrip-problem here on my netbook - when i'm back home i will analyse differences between (functional) bmcc-roundtrips and our stuff.

regards chmee

[peoplemerge]

@nowhereman86 this is a substantially different workflow from the final cut XML.  Interesting!  My resolve with fcp dailies crashes trying to export AAF with some of my existing projects but I'll try with smaller ones and the Avid MXF dailies.

[chmee]

right before implementing the 10Bit-log-Curve i'm in doubt, if some bodies have a blacklevel lower than 2017..?!

Someone any info for me?

[a1ex]

Yes, 1024, 1500...

It's computed on the fly; I have no idea what the absolute range is.

FYI: http://lclevy.free.fr/cr2/#wb

[chmee]

so it seems i have to compute the rec709-log on the fly also. ok. thanks.

regards chmee

[peoplemerge]

@nowhereman86 Same problem with importing/exporting Premiere Pro, which doesn't work with MXF files.  Just noticed Adobe has updates for all products, Premiere CC, which does support DNxHD MXF files!  However with both versions, the relinking still fails using both aaf and xml exports.  You might not notice if you only have 1-2 clips.  Can you confirm this works in Avid on a project with 7-10 clips?  Are you able to export xml from avid, and if so does the workflow still work?

[IliasG]

right before implementing the 10Bit-log-Curve i'm in doubt, if some bodies have a blacklevel lower than 2017..?!

Someone any info for me?

Yes, models older than  7D (5DII, 50D, 40D) have BL at around 1024. This in photo-raws, video-raws can be a bit different.

In any case for a correct calculation we need the full range limits

Except BL which is computed in camera by ML and accessible in metadata, there is the White Level which is set at a predefined value depending on model and based on photo-raw WL as used in CR2 to DNG conversions. This is not accurate enough. For example metadata for 5DIII say WL=15000 while histograms show 16382 ..

I think that samples for every model with burned areas are necessary to declare the WL based on real video-raw data ..   

Yes, 1024, 1500...

It's computed on the fly; I have no idea what the absolute range is.

FYI: http://lclevy.free.fr/cr2/#wb

I didn't knew a case of BL at 1500 !!.

Alex, as I saw on your 5DIII black frames the BL calculated by ML is truncated to the next lower integer so for values 2047.8-2047.9 it becomes 2047. Not a serious error for 14bit DNG, but after truncating to 12bit (as raw2cdng does for the moment) it can be a bit harmful ..

[IliasG]

so it seems i have to compute the rec709-log on the fly also. ok. thanks.

regards chmee

Or use predefined "caseA", caseB" ..

A small detail which might help if searching for "perfection" ..
While I started with BL-32, I used BL-31 to keep the BL=2048 at the center when mapping to 10bit (2047, 2048, 2049 all mapped to 10) so as to have the smallest error. With radically different range (WL-BL) this could change.   

[chmee]

@IliasG as the log-curve goes to full 14bit (16384) and the highlevel end is quite flat, i dont think a whitelevel-difference f.e. 15.000 or 13.500 is any significant problem. now thinking about two log-lists (one with ~2017 start and another one with ~1010) or really computing on the fly..

(like the caseA/caseB-thing much more )

regards chmee

[marekk]

I don't know how to test it now because Davinci Resolve doesn't open 14bit DNGs.. So I need a 12 or 16bit DNG with hot pixels and badpixel_opcode..

"maybe" is the right description. As of version 1.1.0.0 of the CDNG-specification opcodes are optional by decoders, not mandatory. i put it on toDo-list, but there are some fixes before that..

[revast]

hello, have tried your program in wine, under 1.5.31 with mono, it starts up perfectly, but the drag&drop seems to be disfuct. Really looking forward to the commandline version then.

[chmee]

@revast minimum requirement is .net4 compatibility.
http://appdb.winehq.org/objectManager.php?sClass=version&iId=17886

regards chmee

[nowhereman86]

@peoplemere Yeah, this workflow seemed to move along just fine with 10 clips in a thing I put together. I didn't use XMl though I used exclusively the AAF Roundtrip function and followed the steps in this video. I think the order in which you execute these steps is very critical since. If you have Avid and resolve, try it out and see if it works okay.

[zach915m]

Yeah that's the video I've been using too but the roundtrip doesn't work for me...

[fatpig]

are you aware that the progress bars lock after processing about 1/4th of the frames?
it still processes all of them, but the GUI greys out until its done if I click inside.