search

technosaby / RedHenAudioTagger

MIT License
1 stars 4 forks source link

Common Audio processing for all Red hen's video #16

Open technosaby opened 2 years ago

technosaby commented 2 years ago

<< For Backlog, Out of scope for Gsoc project>>

@brucearctor Created this and put it to backlog as discussed.

brucearctor commented 2 years ago

Make sure to see what already exists... As I understand it, there are to be 2 audio processing pipelines your project, and one that already does something ... Ideally Red Hen can only extract audio once that is usable by both. That might [ or not ] be something accomplishable at this time. Would suggest looking at existing pipeline to exam potential commonalities -- if you discover the code, I am also happy to have a look.

technosaby commented 2 years ago

@brucearctor But is this in the scope of Gsoc? I was planning to take this up when all the requirements mentioned by Mark are met and so I kept it as Out of Scope.

brucearctor commented 2 years ago

Not a requirement... but nice to think ahead to minimize refactoring for you ,(or anyone)

technosaby commented 2 years ago

Sure. Thanks I will keep this in mind.

technosaby commented 2 years ago

@brucearctor This is the Audio pipeline of Redhen

brucearctor commented 2 years ago

Since there is not automated deployment configured in RedHen, the linked might be the Audio Pipeline that is running ... but I do not think that there is not a mechanism to 100% ensure that what is running on the compute is not different from what is in Git. That is a separate point, and worth more discussion with @turnermarkb and others -- how to get GH Actions, for example, to be able to deploy to CWRU/UCLA HPC environments. If the networking gets allowed, and then CD setup, you can be sure that what is seen in Git, is what is running in production/pipelines.

Great that you found code ...