Menu

Show posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Show posts Menu

Topics - 50mm1200s

#1
TL;DR: Upscaling to achieve better youtube quality is a waste of time.

Following previous discussion, I decided to do my own tests since I couldn't find any good article about this.
The hypothesis: upscaling 1080p videos to UHD (3840x2160px) before uploading to youtube increases the video conversion quality for 1080p.
The argument for this is that youtube uses VP9 codec for resolutions higher than 1080p, instead of H.264.

Test Methodology

1- 20s video using 50D MLV with resolution of 1920x1080px
2- Processed with MLVApp (no sharpness or denoise filter) and exported to DNxHR HQ 8bit. This file was used as the reference file.
3- Interpolated the reference file to 3840x2160px on Premiere Pro v12.0.0 (with "Use Maximum Render Quality" box enabled) and exported to DNxHR HQ 8bit. I didn't wanted to use some better interpolation algorithm, like NNEDI3 or SuperResolution-based. Most people just use Premiere or FinalCut to do this.
4- Uploaded both and waited for youtube conversion. Note: youtube is now accepting direct DNxHR uploads.
5- Downloaded with youtube-dl using format code 137 (standard for 1080p H.264 files) and 248 (standard for 1080p VP9 files). Youtube-dl is a python script that gets the original files from googlevideo server, so there's not conversion in-between.
6- Converted to raw YUV420p using ffmpeg
7- Analyzed with VQMT tool (600 frames), to get MSSSIM, SSIM and VIFP values. Output to CSV.
8- Graphs using LibreOffice Calc

Non-Interpolated link: https://youtu.be/HtR-hJe-2Wo
Interpolated link: https://youtu.be/JMbeoiblla8


Note

Interpolation adds artifacts and downscaling can add what seems to be sharpness. As you can see in the example bellow, the "re-interpolated" crop (using Bicubic algorithm), seems to be sharper. This can affect the test, but I did not applied any sharpness on any of the files.



Results

Averaged from 600 frames. The closer to 1.0, the better:







Averages

MSSSIM:
Non-Interpolated H.264: 0.989555
Interpolated H.264: 0.989429
Interpolated VP9: 0.991146

SSIM
Non-Interpolated H.264: 0.956130
Interpolated H.264: 0.956211
Interpolated VP9: 0.962942

VIFP:
Non-Interpolated H.264: 0.635854
Interpolated H.264: 0.636947
Interpolated VP9: 0.677739

Percentage difference between the Non-Interpolated H.264 and the Interpolated VP9 averages:
- SSIM: 0.68%
- VIFP: 4.18%
- MSSSIM: 0.15%

Conclusion

The difference is placebo or induced by the added sharpness from the re-interpolation.
Besides wasting time, processing power and network bandwidth, you'll generate interpolation artifacts (I noticed some distortion in fine details on shadows). VP9 has a better compression ratio than H.264, this doesn't mean youtube keeps the same bitrate as H.264. In fact, that's the point of the lossy compression research: less data for the same perceptual quality, so you need less storage and load the video faster.
Haven't tested with HDR uploads (10-bit, Rec.2020).
#2
Does anyone know a software or research paper implementing something like this?

#3
At first I thought this was bullshit, because in 7 years working with photography I have never even heard of it, but then found this article from Zeiss:
https://lenspire.zeiss.com/photo/en/article/micro-contrast-and-the-zeiss-pop-by-lloyd-chambers/

Thought it would be nice to share, since we have been discussing lenses on this other thread lately.
#4
I've reached my saturation point today with Adobe software. I work with their products for 7 years now and they gave me nothing but headches. Missed a deadline because of bugs without fix. Is anyone willing to point me alternatives (except FinalCut, I don't work on Mac)?
I always find facinating how open source software work better than a multi-million dolar company like adobe. Not only their software are full of bugs, but also their support for the community is null. They leeched ideas from community and, besides DNG format, they gave nothing back.
Except for some crash's on Rawtherapee, open source software have almost perfect reputation from me. I heard Natron and Blender could make a good workflow, but I need to have the animation functions found on Premiere Pro, such as smooth motion curves and motion blur on objects (vector or png's). I'm also tired of Windows instability. Adobe and Microsoft is nothing but trouble for people that need stability. The only thing holding me on windows is Adobe produts (mainly Premiere and Photoshop).
#5
Does anyone know if Resolve and Premiere Pro has a tool like the Selective Color in Photoshop, where you can change specific color tones? I miss this feature when doing color grading in video...
#6
Thought it would be a good idea to share this:

#7
I'm using 50D professionaly for MLV video. So far it's working very well, thanks to all the hard work of you developers. I've noticed some bugs and missing features, maybe someone could answer me if these are really bugs or not:

1 - Sometimes while recording with MLV the camera gives an error. The message is something like "card/filesystem error". The card is formatted inside camera (exFAT, from what I remember). Everytime it happens I have to restart the camera (most of the time removing batteries).

2 - The Auto-ETTR module uses the "SET" button, the same as the MLV module, so there's incompatibility between the two.

3 - Dual ISO doesn't seem to work in movie mode, even though the MLV module is loaded. The message (in orange) says: "Dual ISO does not work in movie mode on your camera".

4 - No shutter fine-tunning. If find it useful to get exact 180 degree shutter on 24fps...

5 - A slight magenta 'corruption' in the midtones. I'm sure it's not a WB issue or chromatic aberration. I'm processing with MLVProducer, so it could be a bug on post-processing, haven't tested on other software.

6 - The maximum resolution is 1568px? When I try to reach a higher resolution it gives me the message "[resolution] is not possible in current video mode (max 1568)", so when I return to a lower resolution it goes to 1536px, instead of the supposed maximum 1568px. It doesn affect the functionality, it's just a bug it seems.



I can send the log files and a MLV sample if needed, just ask here and I'll do the upload.
I have some dumb questions too:
- The 10bit doesn't seem to be supported on 50D, right?

- Is it possible to have the optimal ISO values on the "ADTG experiment" without compiling it from scratch? I've tried to run the ISO research modules, didn't work.