Skip to content

Conversation

yuyun2000
Copy link
Contributor

Changes

  • Converted all logs in MeloTTS to English for better internationalization
  • Added debug switch for G2P (Grapheme-to-Phoneme) process logging
  • Implemented SOLA (Synchronous Overlap-Add) algorithm for improved frame transitions

Why

These changes improve the user experience for non-Chinese speakers, provide better debugging capabilities for the G2P process, and enhance audio quality by reducing artifacts between synthesized frames.

Testing

  • Verified English logs display correctly in various scenarios
  • Tested G2P debug toggle functionality by enabling/disabling while processing different texts
  • Conducted audio quality comparisons before/after SOLA implementation to confirm improved transitions
yuyun2000 added 2 commits May 6, 2025 16:48
Apply the Synchronized Overlap-Add (SOLA) algorithm to smooth the connection between audio segments output by the decoder, resulting in more natural-sounding transitions between segments.
- Convert all Chinese log messages in Lexicon.hpp to English for better international compatibility\n- Add a debug flag to control whether to display g2p process logs\n- Improve code readability and debugging experience
@Abandon-ht Abandon-ht merged commit e1d7e6d into m5stack:dev May 6, 2025
Copy link

@csukuangfj csukuangfj left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

太强了!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

3 participants