MLV App - All in one MLV Raw Video Processing App [Windows, Mac and Linux]

Started by ilia3101, July 08, 2017, 10:19:19 PM

Previous topic - Next topic

0 Members and 10 Guests are viewing this topic.

Danne

Did some more tests. A few repeated questions questions
Tweaking exposusre signal, white level and highligh reconstruction and then adding:
cineon log
alexa wide gamut

Ok, this isn´t exactly mr color science working here but still it resembles the compressed log signal and alexa color gamut. Naturally now I´d like to add either some lut to this or start working straight into mlv app adding a curve, color etc but these tools are all greyed out? Does it has to be like this? So then I add the Allow Creative Adjustments. Complete new signal is added. Control is lost. See image below.


Allow Creative Adjustments added. Why is the whole signal changed into something rec709ish? Could starting point when enabling still be the exact same signal as before Allow Creative Adjustment was enabled? Or why are adjustments sliders disabled in the first place?


The alternate route seems to be to build a quick and dirty log signal straight in mlv app. Here´s an example of reinhard and alexa wide gamut and sliders tweaked for dynamic range:



Since we´re in full control and all sliders are still all open for tweaking we can still tweak and build upon the log signal.


I´m no color scientist so take all this with a grain of salt and use examples for further discussion perhaps.
I am really impressed how easy it is to dig out highlights and how the toolbox grown for getting creative solutions for almost any issue now.

Here´s the log to rec709 recipe if anyone want to test it. White level and transformation are left untouched:
https://bitbucket.org/Dannephoto/magic-lantern/downloads/log_rec709.marxml

ilia3101

it shouldn't get 'lost' if dark strength is kept at 0, see if it helps

Danne

Working! Need to compensate for the dark strenght set to 0(exposure to 0 and gamma pulled down) but control is now there. Great! Might be an idea to zeroe dark strength when a log is selected and maybe lower gamma accordingly? Not sure how it could be done without interfering with other workflows.

EDIT: Hm, we need dark strenght when adding back black so zeroing this maybe not the best idea. However hard to know that this is the culprit when enabling Allow Creative Adjustments.

cineon log with and without Allow Creative Adjustments added:


Addding back s-curve etc:


Luther

Danne, try a Log-C to Rec.709 LUT from Arri:
https://www.arri.com/en/learn-help/learn-help-camera-system/tools/lut-generator

Might be easier than doing a S-curve...
edit: you're clipping your shadows. This was already discussed on rawtherapee github.
But I agree with you:
QuoteMight be an idea to zeroe dark strength when a log is selected and maybe lower gamma accordingly?

Danne

Quote from: Luther on September 24, 2019, 06:27:48 AM
Might be easier than doing a S-curve...
edit: you're clipping...
Of course I clipped. It was quick and dirty demo ;).
I use ben turleys online lut calculator. So easy, and includes all I could ever think of :).
At least last time I tested...
Thanks for linking by the way.

DeafEyeJedi

5D3.113 | 5D3.123 | EOSM.203 | 7D.203 | 70D.112 | 100D.101 | EOSM2.* | 50D.109

Luther

Quote from: DeafEyeJedi on September 24, 2019, 07:48:31 PM
Thanks for acknowledging this. However, does it matter which file type we can export the 3D LUT format as?

No difference, it will only generate platform specific formats (3dmesh, clut, .vf, .cube, etc). MLVApp only supports .cube format, so you can choose either "Adobe" or "DaVinci Resolve" (both will have the same result)...

Luther

Nice find @masc and @Danne:
Quote
Piccante is a header-only C++11 open-source (MPL license v2.0) imaging library, specifically designed for High Dynamic Range (HDR) processing. Piccante is modular; this means new filters, algorithms, etc. can be added in a straightforward way to its core. Piccante is easy to use and to integrate in existing projects; to use Piccante you just need to include "piccante.hpp" in your project.
http://vcg.isti.cnr.it/piccante/index.php

Might be a better/faster solution than using ffmpeg blending...

Danne

I think hardly faster but might have a more suitable blending algorithm. The main issue with ghosting/aligning still persists though. A fast aligning added here would be nice.

Stousen

Quote from: masc on September 21, 2019, 02:01:48 PM
You can use the transformation tab. There is a horizontal stretch function.

Yes! But can I Crop an image in mlvapp?? For example, with 2x stretch I get 2:77:1 aspect ratio, want to crop it down to 2:39:1?

//Chris

masc

Nope. MLVApp can HQ stretch but not crop. But all NLEs are able to crop.
5D3.113 | EOSM.202

Stousen


Luther

@Ilia3101 I think the Rec.709 matrix is wrong. According to the ACES github, the XYZ>Rec.709 matrix is this:
Quote
R_out =  3.2409699419 * R_in + -1.5373831776 * G_in + -0.4986107603 * B_in;
G_out = -0.9692436363 * R_in +  1.8759675015 * G_in +  0.0415550574 * B_in;
B_out =  0.0556300797 * R_in + -0.2039769589 * G_in +  1.0569715142 * B_in;
But on MLVAPP processing.c it is:
Quote
/* GAMUT_Rec709 */
         3.2404542, -1.5371385, -0.4985314,
        -0.9692660,  1.8760108,  0.0415560,
         0.0556434, -0.2040259,  1.0572252
Not sure which one is correct.

Also, I have a request. Can you add AP1 to the Gamuts? This seems to be the matrix:
Quote
R_out =  1.6410233797 * R_in + -0.3248032942 * G_in + -0.2364246952 * B_in;
G_out = -0.6636628587 * R_in +  1.6153315917 * G_in +  0.0167563477 * B_in;
B_out =  0.0117218943 * R_in + -0.0082844420 * G_in +  0.9883948585 * B_in;

Danne


Luther

Quote from: Danne on September 26, 2019, 10:10:01 AM
Nice digging. Did you test these numbers and compiled, compared?

Tested now. Doesn't seem to have any difference at all ¯\_(ツ)_/¯
Where did you got the original matrix @Ilia3101 ?

anandkamal

Thanks for the wonderful software. Just curious to know if there will be an option for importing cdng files in the future versions :)

IDA_ML

Here is a short video that I filmed on the 5DMkIII using the amazing EF 85/F1.4 L IS USM lens:
https://vimeo.com/362342398
I wanted to test the stabilization and get an impression of the tone transitions and bokeh, so I filmed hand held at F1.4.  And since I did not have a VND filter at that time, I just waited until it got almost dark in the evening, so 1/50-th of a sec. shutter speed gave me the right exposure with the lens wide open.    I used the latest 5DMkIII experimental build found in Danne's repository and set the crop mode to 1360x2296 anamorphic mode at 10-bits lossless and 24 fps which allows continuous recording.   

MLV processing was done entirely in MLVApp 1.8, (unfortunately, I was unable to get version 1.9 working on any of the three Win7x64 desktops that I tried, the static version did not work either causing a crash upon opening a MLV file).  In version 1.8, I first applied the Alexa Log-C profile to all clips and activated the "Allow creative adjustments" option.  Then I made just a few very basic adjustments to get proper brightness and contrast (Exposure, Lighten, Curves) without clipping brightest and darkest areas, added some sharpness, set Chroma separation to 6 and that was it.  Then I exported the clips into ProRes 422LT, edited the video and added the music in DaVinci Resolve 15 Lite and exported the final video in the resolution of my screen (2560x1440) using the MPEG4 codec.  This keeps the file fairly small (only 262 MB).  I don't know if that is the optimum workflow in terms of image quality but comparing the video quality after upload to Vimeo with the original, I do not see any loss of quality.  So, I think, uploading in 4k, as usually people do to avoid quality degradation in YouTube, was not necessary here.   Please let me know what you guys think.  Also, your suggestions for a better workflow will be highly appreciated.

Personally, I think that MLVApp does a hell of a job with processing, grading and converting MLV files - the above workflow is simple and easy to use.  The developers of this amazing software have given us a very powerful RAW-video processing tool.  Thank you so much!

HHH456

Hi, thanks for all the hard work into ML and the MLV app.

I was wondering if technically there's any quality difference in adding (for example) the Alexa c-log LUT to the mlv app export vs. adding (for example) a flat LUT later in editor?
(Also, the same concept except upscaling in MLV vs. in editor?)

If I edit a multi-clip project with a fixed mlv app look, I've committed to that base look whereas the LUT could be removed or changed in editor.
Thanks for any info.

Luther

Quote from: HHH456 on September 27, 2019, 10:43:44 PM
flat LUT later in editor?
This will probably create banding issues. Always apply Log-C first and then do other processing.

Quote
(Also, the same concept except upscaling in MLV vs. in editor?)
It depends on your editor's interpolation algorithm. MLVApp has some pretty good ones, like Lanczos. I suggest you do that on MLVApp too, as Premiere Pro only has bilinear/bicubic interpolation (which is worse than Lanczos).

masc

Quote from: HHH456 on September 27, 2019, 10:43:44 PM
I was wondering if technically there's any quality difference in adding (for example) the Alexa c-log LUT to the mlv app export vs. adding (for example) a flat LUT later in editor?
(Also, the same concept except upscaling in MLV vs. in editor?)
Quote from: Luther on September 28, 2019, 08:36:08 AM
This will probably create banding issues. Always apply Log-C first and then do other processing.
It depends on your editor's interpolation algorithm. MLVApp has some pretty good ones, like Lanczos. I suggest you do that on MLVApp too, as Premiere Pro only has bilinear/bicubic interpolation (which is worse than Lanczos).

Correct. You'll easier get banding in your editor, because your editor has just 8..12bit footage (depends on codec) while MLVApp has the full 16bit working range (you import less, but after WB and other features used bitdepth will get bigger).
Lanczos, Sinc and AVIR should bring better results than bilinear/bicubic which is offered by most editors.
5D3.113 | EOSM.202

Dmytro_ua

Quote from: masc on September 28, 2019, 09:43:10 AM
Lanczos, Sinc and AVIR should bring better results than bilinear/bicubic which is offered by most editors.

What about Davinci?
R8 | Canon 16-35 4.0L | Canon 50 1.4 | Canon 100mm 2.8 macro
Ronin-S | Feelworld F6 PLUS

Luther


HHH456

Quote from: Luther on September 28, 2019, 08:36:08 AM
This will probably create banding issues. Always apply Log-C first and then do other processing.
It depends on your editor's interpolation algorithm. MLVApp has some pretty good ones, like Lanczos. I suggest you do that on MLVApp too, as Premiere Pro only has bilinear/bicubic interpolation (which is worse than Lanczos).

Thanks for your advice, will do.

masc

Quote from: Dmytro_ua on September 28, 2019, 10:45:21 AM
What about Davinci?
Quote from: Luther on September 28, 2019, 02:03:12 PM
DaVinci has a super-resolution algorithm. Better quality than Lanczos.
But I would only use it with DNG, otherwise you just super-scale your compression artifacts.  ;D
5D3.113 | EOSM.202

Luther

Quote from: masc on September 28, 2019, 06:23:44 PM
But I would only use it with DNG, otherwise you just super-scale your compression artifacts.  ;D
Only true for lossy codecs. ProRes is technically lossy, but there's no visual artifacts, so I think it's fine too.