autotabber
A downloadable tool for Windows and Linux
A program for automatically transcribing diatonic harmonica playing into tabs.
This is a work-in-progress and may not work well for you.
Instructions
The default settings work well for me, but you may have to tweak them. Lower the MinVolume (or set it to 0) if the program is not picking up your playing. Consider leaving the other settings default first.
The buffer size is the number of samples of input collected before attempting to calculate the pitch. A buffer size of 512 corresponds to ~0.011 seconds of input. MinOccurs is the number of buffers a note must be detected in in a row to be outputted. If a note appears in less buffers, it is ignored.
known issues
- notes sometimes detected multiple times, when played only once
- notes usually detected only once if the same note is played multiple times in a row
- sometimes a higher note is detected than the one played
Contact / Feedback
- Discord server: dev-shed channel on harmonica hangout: https://discord.gg/Fp6j3Y9Fab
- Email: sebastianthuemmel@yahoo.ca
Status | In development |
Category | Tool |
Platforms | Windows, Linux |
Author | Seebass22 |
Made with | Rust |
Code license | MIT License |
Links | Source code |
Comments
Log in with itch.io to leave a comment.
This is SUCH an awesome idea. Thanks for inventing this! Is there a version for Mac?
Thanks! I'll upload a version for mac sometime in the future.
Thank you! Any luck with this? I'd love to actually try it out but I only have Mac.
Nice work! I have some problems with windows 11. I need to change the compatibility settings to windows 8. Do you use PYIN to extract the fundamental frequency of the notes? How can I contribute to this project?
Hi Filipe! Thank you!
I'll look into the windows issue.
I don't use PYIN to extract the fundamental frequency, only the barebones autocorrelation approach. I had planned to switch to YIN or SNAC, but I never got around to it.
I've never had anyone contribute to a project of mine, I'm new to this .
I'd gladly accept contributions! We can continue discussing this on discord.