Skip to content

fix(rt): don't assume a default audio format #119

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
sam-s10s wants to merge 1 commit into
base: main
Choose a base branch
from
Draft

fix(rt): don't assume a default audio format #119

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 12 additions & 6 deletions sdk/rt/speechmatics/rt/_async_client.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,6 @@
from ._exceptions import TimeoutError
from ._exceptions import TranscriptionError
from ._logging import get_logger
from ._models import AudioEncoding
from ._models import AudioEventsConfig
from ._models import AudioFormat
from ._models import ClientMessageType
Expand Down Expand Up @@ -100,7 +99,9 @@ def __init__(
self.on(ServerMessageType.WARNING, self._on_warning)
self.on(ServerMessageType.AUDIO_ADDED, self._on_audio_added)

self._audio_format = AudioFormat(encoding=AudioEncoding.PCM_S16LE, sample_rate=44100, chunk_size=4096)
# Audio format is set when start_session is called with an explicit format.
# Deliberately None until then to avoid silently using incorrect defaults.
self._audio_format: Optional[AudioFormat] = None

self._logger.debug("AsyncClient initialized (request_id=%s)", self._session.request_id)

Expand Down Expand Up @@ -138,10 +139,10 @@ async def start_session(
... await client.start_session()
... await client.send_audio(frame)
"""
if audio_format is not None:
self._audio_format = audio_format

await self._start_recognition_session(
# _start_recognition_session resolves defaults (e.g. AudioFormat() if None),
# so we capture the resolved format to keep _audio_format in sync.
_, self._audio_format = await self._start_recognition_session(
transcription_config=transcription_config,
audio_format=audio_format,
translation_config=translation_config,
Expand Down Expand Up @@ -213,8 +214,13 @@ def audio_seconds_sent(self) -> float:
"""Number of audio seconds sent to the server.

Raises:
ValueError: If the audio format does not have an encoding set.
ValueError: If called before start_session has set the audio format,
or if the audio format does not have an encoding set.
"""
# _audio_format is only set once start_session receives an explicit AudioFormat.
# Failing here prevents silently computing with wrong defaults (e.g. 44100Hz).
if self._audio_format is None:
raise ValueError("audio_seconds_sent is not available before start_session is called with an audio format")
return self._audio_bytes_sent / (self._audio_format.sample_rate * self._audio_format.bytes_per_sample)

async def transcribe(
Expand Down
Loading