Cineon Log clips dynamic range!

Get answers to your questions about color grading, editing and finishing with DaVinci Resolve.
  • Author
  • Message
Offline
User avatar

Dmitry Shijan

  • Posts: 405
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Cineon Log clips dynamic range!

PostSun Oct 01, 2017 11:37 am

I use Cineon Log as my Timeline Log curve for BMMCC RAW footage for a while. I like the contrast response it produces without any additional adjustments. Some time ago i experiment with shadows recovery from underexposed images to get wide dynamic range images. As a Log to Rec compression i usually used custom LUTs generated in LUTCalc but with Resolve 14 Color Space Transform Tool now allow clipless compression of color space and gamma.
So i ended up that boosting Expose and than lowering Gain produces exact same result as need. It keeps highlights unchanged and recovers information from deepest shadows. So i have up to 6 additional stops of slightly nosy dynamic range.
During tests with difefrent Log curves i noticed that Cineon irretrievably clips highlights after such huge expose boosts and it can not be recovered by lowering Gain. So it seems it is just a natural limitation of Cineon which was actually was designed as a Film response curve but not a HDR editing curve.

Image
Image
Image
Image
Image


Just to compare this s how the image looks with Expose boosted to 5 without Gain adjustment
Image


Color Space Transform Settings used:

Image


And few samples from RAW sample shared earlier by other user in other forums thread

Image
Image
All my custom made accessories for BMMCC/BMMSC now available here https://lavky.com/radioproektor/
Offline

JosephSlomka

  • Posts: 75
  • Joined: Fri Sep 20, 2013 9:23 pm
  • Location: Burbank

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 12:28 am

Cineon was designed to hold the dynamic range of film which is HDR.
Since Cineon was created by Kodak a while ago there are many different implementations of it.
The Cineon version in VFX software, I am looking at Nuke right now, have no problem grading to and from large exposure differences in 32bit float color.
Offline
User avatar

waltervolpatto

  • Posts: 6188
  • Joined: Thu Feb 07, 2013 5:07 pm
  • Location: Burbank, CA

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 2:43 am

1) the original specification of cineon had been lost, Kodak never publish it. However there are reversed tables that can specify the mapping.
2) I'm not sure of what you're doing there: there are several implicit transformations going on. If you go from raw to any color space in 32 float, you will map out the range without truncate. Including cineon. (I don't see any reason why out should be different)
SuperServer 5039AD-I
C9X299-PGF - DDR4-2400 16x4 GB
i9-7920xCPU 12c 2.90GHz Water cooled
2x GTX 1080ti DeckLink Studio 4K
Window10 - Resolve Studio 15.0.1.003
nvidia dr: 398.82
Offline
User avatar

Dmitry Shijan

  • Posts: 405
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 2:46 am

Each Log curve is different but seems Log3G10 perfectly matches to Cineon/REDLog contrast response. Just slightly different offset. This is really smart and helpful for migration to new Log curve with minimal changes in workflow!

Image
Image
All my custom made accessories for BMMCC/BMMSC now available here https://lavky.com/radioproektor/
Offline
User avatar

waltervolpatto

  • Posts: 6188
  • Joined: Thu Feb 07, 2013 5:07 pm
  • Location: Burbank, CA

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 2:48 am

finally red is matching a 1990 standard!!!! In 4-5 years they will be in 2015.

Just expose the sensor more than usual if you're into hdr: taste the camera 200/400 instead of 800.
SuperServer 5039AD-I
C9X299-PGF - DDR4-2400 16x4 GB
i9-7920xCPU 12c 2.90GHz Water cooled
2x GTX 1080ti DeckLink Studio 4K
Window10 - Resolve Studio 15.0.1.003
nvidia dr: 398.82
Offline
User avatar

Dmitry Shijan

  • Posts: 405
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 2:57 am

waltervolpatto wrote:finally red is matching a 1990 standard!!!! In 4-5 years they will be in 2015.


It s funny but that Cineon 1990 standard is looking way natural than most modern LOG curves when you adjust contrast. It just makes image nice without additional headache. So Log3G10 is designed as Cineon successor.

Cineon encodes out to around 13.5 in float, which is great, but not amazing. Log3G10 goes out to 184.32 which makes it much more useful in the modern world.
All my custom made accessories for BMMCC/BMMSC now available here https://lavky.com/radioproektor/
Offline
User avatar

waltervolpatto

  • Posts: 6188
  • Joined: Thu Feb 07, 2013 5:07 pm
  • Location: Burbank, CA

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 2:57 am

I see, (v14), it look like resolve artificially clamp any mapped value above 1.

It might be because the accrual specification was not allow that is the first place, but it should not be allowed (to clamp)
SuperServer 5039AD-I
C9X299-PGF - DDR4-2400 16x4 GB
i9-7920xCPU 12c 2.90GHz Water cooled
2x GTX 1080ti DeckLink Studio 4K
Window10 - Resolve Studio 15.0.1.003
nvidia dr: 398.82
Offline
User avatar

Dmitry Shijan

  • Posts: 405
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 2:59 am

BTW 1D LUT REDLog3G10 to Amira709 generated in LUTCalc is still looks better than Resolve 14 gamma mapping node.
All my custom made accessories for BMMCC/BMMSC now available here https://lavky.com/radioproektor/
Offline
User avatar

waltervolpatto

  • Posts: 6188
  • Joined: Thu Feb 07, 2013 5:07 pm
  • Location: Burbank, CA

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 3:02 am

Dmitry Shijan wrote:BTW 1D LUT REDLog3G10 to Amira709 generated in LUTCalc is still looks better than Resolve 14 gamma mapping node.

Lutcalc will clamp.

Look "better" does not mean "right" ;)
SuperServer 5039AD-I
C9X299-PGF - DDR4-2400 16x4 GB
i9-7920xCPU 12c 2.90GHz Water cooled
2x GTX 1080ti DeckLink Studio 4K
Window10 - Resolve Studio 15.0.1.003
nvidia dr: 398.82
Offline
User avatar

Dmitry Shijan

  • Posts: 405
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 3:10 am

Yep, it is subjective. If compare Amira709 LUT just adds some kind of complicated film-like contrast and makes final image look more sonorously. But anyway i think that new color space and gamma transform method in Resolve 14 for sure is the best improvement of all.
All my custom made accessories for BMMCC/BMMSC now available here https://lavky.com/radioproektor/
Offline

JosephSlomka

  • Posts: 75
  • Joined: Fri Sep 20, 2013 9:23 pm
  • Location: Burbank

Re: Cineon Log clips dynamic range!

PostTue Oct 03, 2017 6:34 pm

waltervolpatto wrote:finally red is matching a 1990 standard!!!! In 4-5 years they will be in 2015.

Just expose the sensor more than usual if you're into hdr: taste the camera 200/400 instead of 800.

To be fair, Red has always used a cineon like standard for their RedLogFilm curve. It made using RED in VFX workflows significantly more simple than everything else.
Offline
User avatar

Robert Niessner

  • Posts: 1611
  • Joined: Thu Feb 21, 2013 9:51 am
  • Location: Graz, Austria

Re: Cineon Log clips dynamic range!

PostThu Nov 09, 2017 1:12 pm

waltervolpatto wrote:1) the original specification of cineon had been lost, Kodak never publish it. However there are reversed tables that can specify the mapping.


Walter, I am not sure but isn't this the original specification?
http://www.cineon.com/ff_draft.php
http://www.digital-intermediate.co.uk/f ... Cineon.pdf
http://www.dotcsw.com/doc/cineon1.pdf
http://www.fileformat.info/format/dpx/egff.htm
Robert Niessner
LAUFBILDkommission
Graz / Austria
---------------------------------------------------------------------
Blackmagic Camera Blog (German):
http://laufbildkommission.wordpress.com

Read the blog in English via Google Translate:
http://tinyurl.com/pjf6a3m
Offline
User avatar

waltervolpatto

  • Posts: 6188
  • Joined: Thu Feb 07, 2013 5:07 pm
  • Location: Burbank, CA

Re: Cineon Log clips dynamic range!

PostThu Nov 09, 2017 2:39 pm

Robert Niessner wrote:
waltervolpatto wrote:1) the original specification of cineon had been lost, Kodak never publish it. However there are reversed tables that can specify the mapping.


Walter, I am not sure but isn't this the original specification?
http://www.cineon.com/ff_draft.php
http://www.digital-intermediate.co.uk/f ... Cineon.pdf
http://www.dotcsw.com/doc/cineon1.pdf
http://www.fileformat.info/format/dpx/egff.htm


Yes it is, however that is not a "Kodak white paper" that is the work done to allow every one else to use the cineon files
SuperServer 5039AD-I
C9X299-PGF - DDR4-2400 16x4 GB
i9-7920xCPU 12c 2.90GHz Water cooled
2x GTX 1080ti DeckLink Studio 4K
Window10 - Resolve Studio 15.0.1.003
nvidia dr: 398.82
Offline
User avatar

Robert Niessner

  • Posts: 1611
  • Joined: Thu Feb 21, 2013 9:51 am
  • Location: Graz, Austria

Re: Cineon Log clips dynamic range!

PostSun Nov 12, 2017 11:32 am

I see, thanks Walter.
Robert Niessner
LAUFBILDkommission
Graz / Austria
---------------------------------------------------------------------
Blackmagic Camera Blog (German):
http://laufbildkommission.wordpress.com

Read the blog in English via Google Translate:
http://tinyurl.com/pjf6a3m
Offline

Greg Lee

  • Posts: 103
  • Joined: Fri Dec 20, 2013 6:09 pm

Re: Cineon Log clips dynamic range!

PostFri Dec 22, 2017 7:01 pm

Dmitry, those results are AMAZING! Sorry to ask, but is there any way you could post a step-by-step instruction for how you get there? I looked at the other links you listed, but i still can't totally track your steps.

Or maybe a YouTube video of how you do your magic? Obviously, no worries if that's too much, but I'd love to learn from the Master ;)
Offline
User avatar

Dmitry Shijan

  • Posts: 405
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: Cineon Log clips dynamic range!

PostSun Dec 24, 2017 4:37 am

Current basic workflow which is still under development:
Set project settings to V-gamut/RED Log3G10
- Adjust WB in RAW tab
- Cross Hatching Pattern Fix Node (X/Y 0.5px)
- Temporary Noise Reduction Node (if available)
- Chroma Noise Reduction Node (if available)
- Boost Expose with Lower Gain Node (if needed)
- Secondary WB Node
- ColorChecker correction Node
- Highlights Node (if needed)
- Contrast/Pivot Node
- Color Space Transform Node (Timeline to Rec709)
- Film emulation LUT Node
- Sharpen Node (if available and if needed to recover fine details smoothed by OLPF filter)

And some tips copy/pasted info from other threads:

In Micro and Pocket cameras RAW colors are very different from ProRes colors. There are two different implementations of RAW input colors in Resolve:
1. To get corrected colors from RAW you need to set project settings to YRGB and bypass input color space transform in project settings. It helps to get 100% exact same colors from RAW as from ProRes. But it also can make things worse in some situations.
2. To get original sensor data without additional factory color correction you need to set project settings to YRGB Color Managed. This will implement RAW data as is, without any additional correction. This needs manual ColorChecker correction to get reach saturated colors.

First i have odd results with ColorChecker but later i figure how to use it. It works better when timeline set to wide gamut color spaces and Log gamma curve. You should also setup preferences in Resolve Color Match tab according timeline preferences. As for me Panasonic V-gamut or Sony S-Log works best as wide gamut color spaces. And RED Log3G10 is great Log gamma curve to work.
Overall i suggest always work in wide gamut with log curve and add Color Space Transform Node with Gamma and Saturation mapping turned ON in the end of the chain to convert to Rec709 color space for export and monitoring.

Actually you can just create few reference ColorChecker shots in uniform daylight conditions and use them to correct and fix dimmed RAW colors in all footage. This gives you clear starting point for color correction and will match RAW colors very close to ProRes.

I came to the conclusion that BMDfilm log and color space is not a best setting for timeline. It works great for boosting Expose and lowering Gain but just very strange and unnatural when you adjust contrast. Arri LogC is always feels very low contrast. Cineon LOG (middle contrast Pivot 0.52) is limited in some situations. So i moved to RED Log3G10 (middle contrast Pivot 0.38) which have same contrast response as Cineon.
As a color Gamut i use Panasonic V-Gamut because it is the only usable color gamut to work with ColorChecker correction tool in Resolve.

RED Wide Gamut shifts colors to purple reds when use ColorChecker correction tool (probably bug)
ACES Gamut works very well ColorChecker correction tool but shifts colors to warm because D60 white point.
Rec2020 is more like delivery color gamut and if you use as Timeline color space it adds Rec2020 metadata to exported file even if final Color Space Transform Node set to Rec709. This produce oversaturation in some video players which can read that metadata.

To compress LOG to Rec you can use Color Space Transform node in Resolve with Gamma mapping turned ON.
Or you can generate custom 1DLUT in LUTCalc https://cameramanben.github.io/LUTCalc/ ... index.html
As Out Gamma i use Amira709 because for my subjective opinion it it gives best result. (You can also try Alexa-X-2 and LC709/A made by Sony as well)
For color gamut transformation i use Color Space Transform node in Resolve. In Resolve 14 beta they add New Saturation mapping to Color Space Transform node. This is great feature that finally helps to remove hard saturation clipping in extreme values.

Set input to BMDfilmBMDfilm, set timeline for example to V-Gamut with RED Log3G10.
Apply Cineon to Amira709 1DLUT, next apply color space transform LUT.
So i do all my grade in log gamma and in wide color space BEFORE transformation luts, but see final result in realtime in Rec709.
AFTER transformation LUT you can apply regular film emulation LUT and add luma sharpen. As a film emulation LUTs i use free GMIC Film Emulation collection converted from original HaldCLUT to .cube format to use in Resolve.

Boosting Expose for RAW is 100% same as boosting Gain for ProRes in Linear Gamma.
Lowering Gain must be done in LOG gamma. Each LOG gamma have different contrast response, BMDfilm and RED Log3G10 works best for this adjustment in my opinion.

To adjust Exposure in ProRes footage same as in RAW add two Color Space Transform nodes, and one more node in the middle.
Set first node input gamma Timeline Gamma in , Linear out
Set another node input gamma Linear in , Timeline Gamma out.
Adjust Gain in the middle node. This produce exact result as Exposure slider in RAW controls.

Boosting exposure and lowering gain is similar to boosting shadows but works in more natural and physically correct way.
Boosting shadows/highlights in Secondaries tab don't designed for so large and uniform adjustments.
Boosting shadows/highlights slider in RAW tab works same as Adobe Camera RAW or many other RAW processing apps. It adds destructive luma masks and ruins your footage source. So be aware of use it.

About 3 additional stops recovered from shadows is the real clean limit for BMMCC camera. Dig deeper and Fixed Pattern Noise became too visible over moving image.
Feel free to underexpose. But you are forced to use temporary noise reduction mixed with Chroma noise reduction, adding 0.5 XY shift to remove Cross Hatching and using OLPF+IR filter to eliminate wired moire/aliasing in RAW. Also use RAW+ColorChecker correction to get neutral starting point without shifted and desaturated colors. This will allow pixel perfect quality from BMMCC camera without downsampling.

And some screenshots directly from Davinci Resolve viewer:
Image
Image
Image
Image
All my custom made accessories for BMMCC/BMMSC now available here https://lavky.com/radioproektor/
Offline

Greg Lee

  • Posts: 103
  • Joined: Fri Dec 20, 2013 6:09 pm

Re: Cineon Log clips dynamic range!

PostSun Dec 24, 2017 5:02 pm

Wow, thank you!! That's an awesome Christmas present!

Thanks so much!!
Offline
User avatar

Dmitry Shijan

  • Posts: 405
  • Joined: Wed Sep 17, 2014 7:15 pm
  • Location: UA

Re: Cineon Log clips dynamic range!

PostSun Jun 03, 2018 3:53 pm

Here is also some extreme deep into the shadows tests with BMMCC. Timelne set to BMDfilm Log. Same workflow as described earlier. This s a 500W halogen lamp in almost dark room. ND64 filter on the camera. You can do the same Expose boost with ProRes by adding Gain node or even two Gain nodes in linear space. It is sad but FPN is the main limit in this camera and i still can't find a correct way to remove it even with manually created FPN sample patterns in the dark.
Image

Image
Image
Image
Image
Image
Image
Image
Image

And one more Expose boost attempt. This is BMPCC RAW sample from camera review article. Timeline set to BMDfilm Log. Expose 6 created by setting ISO to 1600 + Expose 5. Also applied my own Color Checker correction node from BMMCC.

Image
Image
Image
Image
All my custom made accessories for BMMCC/BMMSC now available here https://lavky.com/radioproektor/
Offline
User avatar

Marc Wielage

  • Posts: 4688
  • Joined: Fri Oct 18, 2013 2:46 am
  • Location: Hollywood, USA

Re: Cineon Log clips dynamic range!

PostSun Jun 03, 2018 10:54 pm

Dmitry Shijan wrote:Here is also some extreme deep into the shadows tests with BMMCC. Timeline set to BMDfilm Log. Same workflow as described earlier. This s a 500W halogen lamp in almost dark room. ND64 filter on the camera. You can do the same Expose boost with ProRes by adding Gain node or even two Gain nodes in linear space. It is sad but FPN is the main limit in this camera and i still can't find a correct way to remove it even with manually created FPN sample patterns in the dark.

I agree that if the BMD cameras are underexposed, you can run into sensor noise and FPN issues. (I think it's a miracle the cameras are this good for under $8000.) We solved that problem on a recent job by coming up with several different levels of Neat Video processing, but it took us more than 24 hours (at 2fps) to take care of a 90-minute film. I think it was worth it and the final results look surprisingly clean, compared to what we had to work with.

I think the cure for this is: don't underexpose the cameras. Rate them a little lower than what the printed specs say, overexpose at least 1/2 stop (if not a little more), protect the highlights, and use more fill light.
marc wielage, csi • VP/color & workflow • chroma | hollywood
Offline
User avatar

Cary Knoop

  • Posts: 699
  • Joined: Sun Mar 12, 2017 6:35 pm
  • Location: Newark, CA USA

Re: Cineon Log clips dynamic range!

PostMon Jun 04, 2018 12:55 am

Dmitry Shijan wrote:I came to the conclusion that BMDfilm log and color space is not a best setting for timeline.

Not to rock the boat (too much) but why ever use a timeline log format if large adjustments, especially in the blacks and shadows, need to be made? And worse, dealing with noise cluttering right above (and sometimes below!) the impossible zero point? Why not instead use a gamma closer to the target gamma?

With the exception of using ACES, but only for properly exposed material, I find using log in the timeline with the Resolve controls far from ideal.

Dmitry Shijan wrote:To adjust Exposure in ProRes footage same as in RAW add two Color Space Transform nodes, and one more node in the middle.
Set first node input gamma Timeline Gamma in , Linear out
Set another node input gamma Linear in , Timeline Gamma out.
Adjust Gain in the middle node. This produce exact result as Exposure slider in RAW controls.

In Resolve 15 this can now be done in a single node!
Offline
User avatar

Marc Wielage

  • Posts: 4688
  • Joined: Fri Oct 18, 2013 2:46 am
  • Location: Hollywood, USA

Re: Cineon Log clips dynamic range!

PostMon Jun 04, 2018 6:28 am

Cary Knoop wrote:Not to rock the boat (too much) but why ever use a timeline log format if large adjustments, especially in the blacks and shadows, need to be made? And worse, dealing with noise cluttering right above (and sometimes below!) the impossible zero point? Why not instead use a gamma closer to the target gamma?

I think these are all good questions. The wrong gamma settings can tend to increase noise, especially if you're crushing things down in one place and then amplifying them in another (depending on where and how it's done). For a lot of reasons, I tend to avoid timeline color space changes, but I do use the camera manufacturer's color science when I can to get it into reasonable film-like Log space. Something like Log-C is relatively easy to work with these days.
marc wielage, csi • VP/color & workflow • chroma | hollywood

Return to DaVinci Resolve

Who is online

Users browsing this forum: Adriano Oliveira, Dermot Shane, Google Feedfetcher, Johannes Hoffmann, Seth Goldin, Shannara and 31 guests