[GH-ISSUE #666] Why was hop size of 10ms chosen? #394

New issue

Closed

opened 2026-10-20 09:58:56 +00:00 by bsakaguchi · 1 comment

bsakaguchi commented

2026-10-20 09:58:56 +00:00

Owner

Originally created by @jmasterx on GitHub (Jul 1, 2023).
Original GitHub issue: https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/issues/666

Hello!

I notice the models use hop size of 10ms, but a lot of papers for SVS suggest 5ms. Is there a reason this hop size was used?

Originally created by @jmasterx on GitHub (Jul 1, 2023). Original GitHub issue: https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/issues/666 Hello! I notice the models use hop size of 10ms, but a lot of papers for SVS suggest 5ms. Is there a reason this hop size was used?

bsakaguchi

2026-10-20 09:58:56 +00:00

closed this issue
added the
following up
label

bsakaguchi commented

2026-10-20 09:58:57 +00:00

Author

Owner

@RVC-Boss commented on GitHub (Jul 2, 2023):

Because the pretrained self supervised feature model RVC used only supports 16k-320hop. (And nearly all hubert series SSL feature only supports 16k-320hop)

@RVC-Boss commented on GitHub (Jul 2, 2023): Because the pretrained self supervised feature model RVC used only supports 16k-320hop. (And nearly all hubert series SSL feature only supports 16k-320hop)