[NabiBukhsh-AI]

Summary

combine_audio_frames accumulates total_data_length to preallocate the destination bytearray before copying frame chunks in place:

total_data_length += len(frame.data)
...
data = bytearray(total_data_length)

However, frame.data is an int16 memoryview (format='h'). len(frame.data) returns the number of samples, not bytes. Because each 16-bit sample is 2 bytes, data = bytearray(total_data_length) preallocated exactly half the memory required for the combined frame.

During the copy loop:

frame_data = frame.data.cast("b")  # 2x bytes per sample
data[offset : offset + len(frame_data)] = frame_data

Python's bytearray slice assignment was forced to repeatedly grow and reallocate the underlying memory buffer on every frame, defeating the function's documented design:

"It efficiently merges the data by preallocating the necessary memory and copying the frame data without unnecessary reallocations."

Additionally, the docstring's example constructed AudioFrame(b"\x01\x02", 48000, 2, 1), which raises a ValueError because 2 channels with 1 sample per channel requires 4 bytes.

Changes

• Used total_data_length += frame.data.nbytes so the preallocated bytearray accurately matches the total byte size of all frame buffers upfront.
• Fixed the docstring example to use valid 16-bit PCM byte length and show combined_frame.data.tobytes().

Testing

Added unit test suite in tests/rtc/test_combine_audio_frames.py:

• test_combine_audio_frames_allocates_exact_byte_size: verifies that the allocated bytearray size equals combined.data.nbytes (fails on main where it allocated half the required size).

• test_combine_audio_frames_concatenates_samples_correctly: verifies sample concatenation.

• test_combine_audio_frames_stereo: verifies multi-channel frames.

• test_combine_audio_frames_single_frame: returns single frame as-is.

• test_combine_audio_frames_empty_raises: raises ValueError on empty list.

• test_combine_audio_frames_mismatched_sample_rate: raises ValueError on sample rate mismatch.

• test_combine_audio_frames_mismatched_channels: raises ValueError on channel count mismatch.

• test_combine_audio_frames_docstring_example: verifies the doctest example runs cleanly.

• uv run pytest tests/rtc/test_combine_audio_frames.py -v -W error (8/8 passed)

• uv run ruff check livekit-rtc/livekit/rtc/utils.py tests/rtc/test_combine_audio_frames.py (passed)

• uv run ruff format --check livekit-rtc/livekit/rtc/utils.py tests/rtc/test_combine_audio_frames.py (passed)

• uv run mypy livekit-rtc/livekit/rtc/utils.py tests/rtc/test_combine_audio_frames.py (0 issues)

[devin-ai-integration]

✅ Devin Review: No Issues Found

Devin Review analyzed this PR and found no bugs or issues to report.