jketreno/ai-voicebot
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Fix s16 assertion (media was being provided as flt)

Browse Source 
Signed-off-by: James Ketrenos <james_git@ketrenos.com>
This commit is contained in:
James Ketr 2025-09-01 15:05:05 -07:00
parent ba6aa697a4
commit 75b0417ad3
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
voicebot/main.py
View File

@ -391,12 +391,12 @@ class SilentAudioTrack(MediaStreamTrack):
"""Generate silent audio frames"""
pts, time_base = await self.next_timestamp()
# Create silent audio data
samples = np.zeros((self.samples_per_frame,), dtype=np.float32)
# Create silent audio data in s16 format (required by Opus encoder)
samples = np.zeros((self.samples_per_frame,), dtype=np.int16)
# Convert to AudioFrame
frame = AudioFrame.from_ndarray(
samples.reshape(1, -1), format="flt", layout="mono"
samples.reshape(1, -1), format="s16", layout="mono"
)
frame.sample_rate = self.sample_rate
frame.pts = pts