search

ongyx / putao

Poor man's UTAU
MIT License
11 stars 0 forks source link

Unable to parse Kasane Teto's oto.ini #2

Closed ElonSatoshi closed 3 years ago

ElonSatoshi commented 3 years ago

I get this error when I try to render bad_apple.json:

UnicodeDecodeError: 'utf-8' codec can't decode byte 0x82 in position 2: invalid start byte

I get the same error even if I run putao with LANG=ja_JP.UTF-8. I'm using Fedora. How do I fix this?

ongyx commented 3 years ago

Most voicebanks oto.ini and sample filenames are encoded in Shift-JIS (because UTAU itself uses Shift-JIS), which makes it difficult to map UTF-8 hirigana lyrics in putao.

Since you mentioned you have Kasane Teto's voicebank, do you have the zip file? Try putao extract <zip file> which changes encoding of the oto.ini and filenames to UTF-8. Then the voicebank should be readable by putao.

ElonSatoshi commented 3 years ago

When extracting the zip with putao, I get this error: AttributeError: module 'putao.utau' has no attribute 'CONFIG_FILE'

DEBUG:putao:re-encoding install.txt to UTF8

ongyx commented 3 years ago

My bad, I'll go fix that and push a bugfix to Pypi.

EDIT: Pushed bugfix version 0.3.1.

ElonSatoshi commented 3 years ago

Can putao handle voicebank appends? The extractor says the zip file contains 3 voicebanks. I'm looking at the tougou voicebank pack.

ongyx commented 3 years ago

No, but you can extract the 3 voicebanks into seperate folders using the extract -t option. Appending voicebanks should be as simple as extracting all the voicebanks into the same folder and then merging the oto.ini files. I'll add an option to specify multiple voicebanks in the same zipfile to append into one.

EDIT: Okay, I pushed version 0.3.2 to add support for appended voicebanks.