Optimizing soundbanks by combining identical samples?

[whiplash]

The soundtrack for a little thing I'm working on is mostly ripped from an SNES game, and converted to IT using OpenSPC. They all use basically the same pool of samples. Wouldn't it be more efficient if mmutil could detect these identical samples (by hashes, maybe?) and simply combine them together?

I imported my soundbank.bin as a raw file into Audacity and heard the same samples over and over. I bet this kind of optimization would reduce my soundbank file from megabytes to a few kilobytes!

[mtheall]

It would be more efficient if OpenSPC would give you an IT that combines samples. You shouldn't be using music ripped from SNES games in the first place anyways. Can't you find some public domain music or create your own? Or do like I did and hire someone to write music for you. Who knows how many copyrights you are violating.

[whiplash]

It would be more efficient if OpenSPC would give you an IT that combines samples.

As far as I'm aware of, the IT format doesn't use soundbanks... unless you're talking about combining the entire soundtrack into a single IT file, which would require more than just some simple optimization on mmutil's part (AKA it would have to generate a file that would mark which patterns songs start and stop at, and I'd have to be able to import that into the source code somehow, a little more work than I want to do ATM)

You shouldn't be using music ripped from SNES games in the first place anyways. Can't you find some public domain music or create your own? Or do like I did and hire someone to write music for you. Who knows how many copyrights you are violating.

Consider it nothing more than a thought exercise, then

[mtheall]

Oh, I think I misinterpreted what you mean. You mean you have multiple songs, but they use the same pool of samples. I thought you meant a single song had multiple instances of the same sample.