My friends and I do karaoke with some stem splitting service. It works fine, but we have issues timing the songs. Does anyone know a service that can take the vocal stem and turn it into a follow along track? The lyrics don't have to be right, it just needs to cue us in at the right times.
I'm not aware of anything. You'd probably have more luck actually asking this in an OSU thread on /vg/ or something since proper karaoke never happens over the internet but rhythm games have to solve the same problem.
>>107746796bumperinos, I want to do exactly this but with audiobooks so I can read while listening
I'm only bringing this up since it's been 16 hours without a reply, if you run a song through whisper.cpp with the --max-len 1 argument and then mux the subtitle file into the karaoke version, with a blank video if needed, you'll get something that will at least be pretty close for the beginnings of words. Command might be something like ffmpeg -loop 1 -i blackbackground.jpg -i karaoke.mp3 -i whisper.srt -c copy out.mkvI also found https://github.com/nomadkaraoke/karaoke-gen but just from looking over the readme I'm not sure if it actually does that part you need.
>>107749220Thank you, I'll check there>>107752802I'll look into this, it seems so strange that there's not even a crappy "timed voice to text" program out there.
>>107746796I would recommend using a service like karafun, I think you can do lyrics and timing for custom songs with it. I haven't used it but it's a very good service.Otherwise you might have to use some custom open source thing out there.
>>107746796what does "Pope Bre" mean??
>>107754787That's an Umbreon wearing a shoe that kinda looks like the Pope hat. Pope Bre.
>>107754360It looks like this is possible, thank you so much
