refactor documentation under test-content generation by codec

[jpiesing]

The documentation under test-content-generation is mostly what was produced organically during the process of creating the AVC test content. The audio isn't yet included. HEVC isn't yet included. The documentation should be refactored to separate what, if anything, is generic and what is codec specific so that @rcottingham can add the audio codec documentation.

[rbouqueau]

I've updated regularly the README. Audio and HEVC are included (also see the profiles folder).

Could you be more specific on what could be improved? We can discuss it here or people can make pull requests.

[rbouqueau]

After having a look I can already propose:

• Update the docx file in the root folder and move it to the Instructions folder.
• Organize the Instructions folder.
• Maybe rename Instructions as it is rather a documentation on what to test.

Opinion?

[jpiesing]

README.md is mostly about AVC video with audio having one subheading towards the end of the document. It doesn't say it's about AVC or indeed about video. If we expect all video to be the same then perhaps most of the current material could be moved to a new document dedicated to video and the audio moved to a new document dedicated to audio.

[jpiesing]

The pull request in #67 adds the audio descriptions into README.md . Work should be about to start on HEVC. Adding Audio and HEVC all into the same text file feels wrong to me. At least we should separate video and audio.

[rbouqueau]

Adding Audio and HEVC all into the same text file feels wrong to me

On the contrary I think it is desirable that the instructions are codec-independent. Then people should be able to add configuration files to control the script. Does that make sense?

By the way audio and HEVC are already present. I had just hoped that Dolby and AAC would also fit that mould.

[jpiesing]

Adding Audio and HEVC all into the same text file feels wrong to me

On the contrary I think it is desirable that the instructions are codec-independent. Then people should be able to add configuration files to control the script. Does that make sense?

By the way audio and HEVC are already present. I had just hoped that Dolby and AAC would also fit that mould.

Well the packaging might possibly be codec independent but I'm surprised you think the encoding could be codec independent. OK ffmpeg may be able to encode most codecs but not all of them and even for the ones it can encode, the command line may not be identical.

I agree that what can be common should be.

[rbouqueau]

IURC it was not my initial opinion either (I wanted to split the actual superscript into smaller entities). However the actual code works great (with AVC+AAC+HEVC+DTS-passthru) without adding to much pragmatical complexity.

I'm still a bit unclear about what/how to improve the documentation.

It would be great if someone with a need would try to execute the script.

[nlsdvl]

I've picked up the scripts and updated them for HEVC and HDR.

I ended up progressing on this issue :

• run-all.py script has been updated to be more versatile in content generation
• added a separate batch managenent script to archive and upload new content batches

#71