Skip to content

v2.5.2

Choose a tag to compare

View all tags
@twinsbox89 twinsbox89 released this 13 May 10:12
· 4 commits to main since this release
v2.5.2
b446114

Studio Sound Improvements

Sharper voice quality, lower latency, and fewer audio artifacts across the board.

Adaptive Jitter Buffer

Output buffering now self-tunes in real time — building headroom when the model is under load and shedding it during calm periods. The result is consistently lower latency without the dropouts.

Better Speech Time-Stretching

Rebuilt the OLA engine with WSOLA-style best-match search and RMS-normalised crossfades. Playback speedup is now nearly inaudible, even at aggressive rates.

Smarter Audio Enhancer

  • AGC reacts faster to loud spikes (10 ms attack) and recovers gently (50 ms release), eliminating gain surges after silence.
  • EQ pipeline fixed — multi-band processing no longer corrupts the signal path. Frequencies are automatically clamped to Nyquist; bad parameter values fall back gracefully.
  • Preset switching is now artifact-free thanks to transient suppression that fades out and recovers smoothly on every change.
Assets 2
Loading