Management: how to develop/deploy audio -material? Large -files, separated -files or some other solution?

Question

I develop a large dev -file usually over 300MB easily. Then I categorize it and break it into parts (A) such as shouting, laughing, smiling and other descriptive things. This pieced material belongs to my Deployment -repository where fine-tuning is more easier, much easier than in fs-backuped Development -repo that contains the original massive files (B) and where you cannot easily edit or you need to kill the repo, to keep it in small size, or use no version-controlling. Now I feel a better way: some software to handle both A and B. How would you handle this step of processing the audio material into rich informative art-assets?

Helper questions

• Which alternative is better? And when would you use each?

AA) Deployed audio material broken into pieces that is easier to deploy for things such as games and movies

BB) Deployed audio material in large files from which you somehow get intervals, no idea how to do this, anyway good backups.

CC) some other methodology?

• Does there exist some software to handle this development-deployment -cycle for me?

• I currently use Audacity that is extremely slow for my purpose here, I would like to have tool by which I can denote different parts of the audio with labels. Generic name for such software?

Answer 1

You may want to try a different DAW to manage this project. Audacity stores its audio in many small files. DAWs like Reaper here store the audio in 1 .wav file per recorded track.

Audacity uses destructive editing, which means any time you edit the audio by cutting/pasting or adding effects, the audio is permanently changed. I don't know about other DAWs, but Reaper uses non-destructive editing for the audio files. This means that no matter how many changes you make to the audio, the original files never change, they are applied to the tracks in real-time. This allows you to make whatever changes you want without fear of losing the original recording and without needing to back up the entire project at every step.

When working with a large amount of tracks or effects you still have the option to render portions of the audio to stem tracks to save processing power. This will create a new file, but it does not need to be put in version control since you can regenerate it at any time from the original files and the project file.

I don't know of any DAW with version control integrated, but you might be able to write a plugin for and existing DAW.

Answer 2

It is usually important to move with the smallest footprint without sacrifying too much quality and work-hours. The below gathers tips how to keep production/development -material in small size. Even though you may be able to do the same thing as an expert but if your development files are significantly greater, it can result into a hit with development -- too large files can be extremely slow/booring to work with and fine-tune.

1. [Lossy and non-lossy (i.e. no data-loss) -audio-formats, thread here] Lossy formats such as MP3 should be used at the very latest point, probably deployment, because of the data-loss. Formats with non-lossy codecs such as FLAC, in contrast, result in no data-loss but may result in storage -penalty. Vectorized formats here and FLAC may be a good option during development while lossy-codeced formats such as MP3 better to keep file-sizes minimal.

2. [Vectorized audio -thread here] Create vectorized audio, a bit like using vector graphics instead of raster in Graphics design.