← sources

Any tips how to make Minimax H3 voice consistent?

source: reddit (reddit:StableDiffusion)published 2026-09-23 04:49ingested 2026-09-24 02:39

retained text

So I’m making a 50-second clip using separate 10-second segments, but the problem is that the voice isn’t consistent.

Are there any tips to make the voice more consistent, or would it be better to use QwenTTS instead?

document fetched 2026-09-24 00:35 · canonical https://reddit.com/r/StableDiffusion/comments/1wnwd43/any_tips_how_to_make_minimax_h3_voice_consistent