w3speech format about wolvenkit HOT 8 CLOSED

@DellCliff

pub struct SpeechVoData {
    id: u32,
    duration: f32,      // duration in seconds
    unknown: u32,       // set to constant 4
    audio: Vec<u8>,
    cr2w: Vec<u8>,
}

duration can differ from audio length. it clips the audio if shorter.

from wolvenkit.

I can pretty much read/write everything else of that w3speech format, except for those 8 bytes.

from wolvenkit.

If you can help me out, I can send you a pull request for a complete reader/writer for those files.

from wolvenkit.

Hi
Thanks for helping with the project 👍 . Since I am coding most stuff myself I haven't really had time to code much w3speech related just an extractor. Rmemr has made a tool for it but its commandline only and is in rust (https://github.com/rmemr/w3.phoneme-extractor) you can probably look at that. If you can't figure it out then if you make a pull request with the reader/writer I will reverse engineer those 8 bytes after I have finnaly fixed the xbm converter.

from wolvenkit.

Thanks. I've looked at rmemr's stuff, sadly this particular repository only deals with audio files, not the w3speech format. He does have a file packer/unpacker on nexusmods which is not open source though. Bad luck. Guess I have to keep looking.

from wolvenkit.

Or just ask rmemr on nexus.

from wolvenkit.

@DellCliff I'll ask him to give some input into this
edit: we can just mention him I guess @rmemr

from wolvenkit.

@rmemr
Thanks, just what was missing. Gonna do some testing.

from wolvenkit.