Account for multi-frame Opus packets when computing packet duration (#439)

* Account for multi-frame Opus packets when computing packet duration

parseOpusTocByte only read the config field of the TOC byte and always
assumed a single frame per packet. Per RFC 6716 section 3.2, a packet may
carry 1, 2 or an arbitrary number of frames, encoded in the two low bits of
the TOC byte (plus the frame count byte for code 3), and its duration is the
frame duration times the frame count.

As a result, the Ogg muxer wrote granule positions that advanced slower than
the actual audio. Chromium's MediaRecorder packs three 20 ms frames per Opus
packet, so remuxing WebM/Opus to Ogg/Opus produced files declaring a third of
their real duration: a 5.7 s recording ended with a final granule position of
99840 (2.08 s). Decoders that trust the container then truncate the audio.

* Clean up

---------

Co-authored-by: Vanilagy <[email protected]>

src/codec-data.ts

@@ -2207,9 +2207,22 @@ const OPUS_FRAME_DURATION_TABLE = [
 
 export const parseOpusTocByte = (packet: Uint8Array) => {
 	const config = packet[0]! >> 3;
+	const code = packet[0]! & 0b11;
+
+	// A packet may pack more than one frame, in which case its duration is the frame duration times the number of
+	// frames it carries. See https://datatracker.ietf.org/doc/html/rfc6716, section 3.2.
+	let frameCount: number;
+	if (code === 0) {
+		frameCount = 1;
+	} else if (code === 1 || code === 2) {
+		frameCount = 2;
+	} else {
+		// Code 3: the frame count sits in the six low bits of the frame count byte
+		frameCount = packet[1]! & 0b111111;
+	}
 
 	return {
-		durationInSamples: OPUS_FRAME_DURATION_TABLE[config]!,
+		durationInSamples: OPUS_FRAME_DURATION_TABLE[config]! * frameCount,
 	};
 };
 

test/browser/ogg-muxer.test.ts

@@ -1,8 +1,14 @@
 import { expect, test } from 'vitest';
 import { Output } from '../../src/output.js';
 import { OggOutputFormat } from '../../src/output-format.js';
-import { NullTarget } from '../../src/target.js';
+import { BufferTarget, NullTarget } from '../../src/target.js';
-import { AudioBufferSource } from '../../src/media-source.js';
+import { AudioBufferSource, EncodedAudioPacketSource } from '../../src/media-source.js';
+import { EncodedPacket } from '../../src/packet.js';
+import { assert } from '../../src/misc.js';
+import { Input } from '../../src/input.js';
+import { BufferSource } from '../../src/source.js';
+import { ALL_FORMATS, OggInputFormat } from '../../src/input-format.js';
+import { EncodedPacketSink } from '../../src/media-sink.js';
 
 test('maximumPageDuration option', async () => {
 	const sampleRate = 48000;
@@ -55,3 +61,82 @@ test('maximumPageDuration option', async () => {
 	expect(pageCountWithoutOption).toBe(3);
 	expect(pageCountWithOption).toBe(23); // It created more pages
 });
+
+test('Multi-frame Opus packets', async () => {
+	const SAMPLE_RATE = 48000;
+	const SAMPLES_PER_FRAME = 960; // 20 ms at 48 kHz
+
+	const createOpusHead = () => {
+		const bytes = new Uint8Array(19);
+		const view = new DataView(bytes.buffer);
+
+		bytes.set([0x4f, 0x70, 0x75, 0x73, 0x48, 0x65, 0x61, 0x64], 0); // 'OpusHead'
+		bytes[8] = 1; // Version
+		bytes[9] = 1; // Channel count
+		view.setUint16(10, 312, true); // Pre-skip
+		view.setUint32(12, SAMPLE_RATE, true); // Input sample rate
+		view.setInt16(16, 0, true); // Output gain
+		bytes[18] = 0; // Channel mapping family
+
+		return bytes;
+	};
+
+	const createOpusPacket = (frameCount: number) => {
+		const data = new Uint8Array(2 + 3 * frameCount);
+
+		data[0] = (31 << 3) | 0b11; // TOC byte: config 31 (CELT fullband, 20 ms), code 3
+		data[1] = frameCount; // CBR, no padding, `frameCount` frames
+
+		return data;
+	};
+
+	const framesPerPacket = 3;
+	const packetCount = 10;
+	const packetDuration = (framesPerPacket * SAMPLES_PER_FRAME) / SAMPLE_RATE;
+
+	const output = new Output({
+		format: new OggOutputFormat(),
+		target: new BufferTarget(),
+	});
+
+	const audioSource = new EncodedAudioPacketSource('opus');
+	output.addAudioTrack(audioSource);
+
+	await output.start();
+
+	for (let i = 0; i < packetCount; i++) {
+		await audioSource.add(
+			new EncodedPacket(
+				createOpusPacket(framesPerPacket),
+				'key',
+				i * packetDuration,
+				packetDuration,
+			),
+			{
+				decoderConfig: {
+					codec: 'opus',
+					numberOfChannels: 1,
+					sampleRate: SAMPLE_RATE,
+					description: createOpusHead(),
+				},
+			},
+		);
+	}
+
+	audioSource.close();
+	await output.finalize();
+
+	assert(output.target.buffer);
+
+	const input = new Input({
+		source: new BufferSource(output.target.buffer),
+		formats: ALL_FORMATS,
+	});
+
+	expect(await input.getFormat()).toBeInstanceOf(OggInputFormat);
+
+	const sink = new EncodedPacketSink((await input.getPrimaryAudioTrack())!);
+	const packet = await sink.getFirstPacket();
+
+	expect(packet?.duration).toBe(packetDuration);
+});