Add audio encoder reconfiguration

[Qizot]

For some platforms (looking at you Apple) we can't reliably get the channel count from the media track, this information becomes available only when looking at the data inside of audio worklet callback.

The following changes allow for config creation only after receiving the first batch of samples.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 76a7da48-68d6-4153-8ed4-4d9e1cc43c31

📥 Commits

Reviewing files that changed from the base of the PR and between fb88a5c and fb72613.

📒 Files selected for processing (3)

• js/publish/src/audio/encoder.ts
• rs/moq-native/Cargo.toml
• rs/moq-relay/Cargo.toml

✅ Files skipped from review due to trivial changes (2)

• rs/moq-relay/Cargo.toml
• rs/moq-native/Cargo.toml

🚧 Files skipped from review as they are similar to previous changes (1)

• js/publish/src/audio/encoder.ts

---

Walkthrough

The audio encoder now creates its AudioConfig lazily from the first captured audio frame: the worklet message handler reads data.channels.length and calls a new #createConfig(). The prior #runConfig reactive path is removed. serve() defers encoder setup until #config exists and per-frame handling reads the full data.channels array, recomputing #config and resetting lastKeyframe when channel count changes, and skips encoding the first frame of a new configuration. Two Cargo.toml dependency feature lists were reformatted (no behavioral change).

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title 'Add audio encoder reconfiguration' directly matches the main change: audio encoder configuration is now created lazily from the first captured audio frame instead of upfront.
Description check	✅ Passed	The description explains the problem (channel count unavailable on Apple platforms until audio worklet callback) and the solution (deferring config creation until first audio samples arrive), which aligns with the actual changeset.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

✨ Simplify code

• Create PR with simplified code

Tip

💬 Introducing Slack Agent: The best way for teams to turn conversations into code.

Slack Agent is built on CodeRabbit's deep understanding of your code, so your team can collaborate across the entire SDLC without losing context.

• Generate code and open pull requests
• Plan features and break down work
• Investigate incidents and troubleshoot customer tickets together
• Automate recurring tasks and respond to alerts with triggers
• Summarize progress and report instantly

Built for teams:

• Shared memory across your entire org—no repeating context
• Per-thread sandboxes to safely plan and execute work
• Governance built-in—scoped access, auditability, and budget controls

One agent for your entire SDLC. Right inside Slack.

👉 Get started

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🧹 Nitpick comments (1)

js/publish/src/audio/encoder.ts (1)

135-145: Extract the Opus defaults into named constants.

32_000 and 20 are now part of the encoder policy, so keeping them inline makes later tuning harder and hides intent.

♻️ Suggested refactor

+const OPUS_BITRATE_PER_CHANNEL = 32_000;
+const OPUS_FRAME_DURATION_MS = 20 as Time.Milli;
+
 `#createConfig`(worklet: AudioWorkletNode, channelCount: number): Catalog.AudioConfig {
 	return {
 		codec: "opus",
 		sampleRate: Catalog.u53(worklet.context.sampleRate),
 		numberOfChannels: Catalog.u53(channelCount),
-		bitrate: Catalog.u53(channelCount * 32_000),
+		bitrate: Catalog.u53(channelCount * OPUS_BITRATE_PER_CHANNEL),
 		container: { kind: "legacy" } as const,
 		// TODO parse the actual frame duration instead of assuming 20ms.
 		// Opus supports 2.5–60ms but 20ms is the real-time default.
-		jitter: Catalog.u53(20),
+		jitter: Catalog.u53(OPUS_FRAME_DURATION_MS),
 	};
 }

As per coding guidelines, "Avoid using magic numbers; use named constants instead."

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@js/publish/src/audio/encoder.ts` around lines 135 - 145, Extract the magic
numbers in `#createConfig` into named constants (e.g.,
DEFAULT_OPUS_BITRATE_PER_CHANNEL = 32000 and DEFAULT_OPUS_FRAME_MS = 20) defined
near the top of the module and replace the inline literals (32_000 and 20) in
the bitrate and jitter calculations with those constants (use
Catalog.u53(DEFAULT_OPUS_BITRATE_PER_CHANNEL * channelCount) for bitrate and
Catalog.u53(DEFAULT_OPUS_FRAME_MS) for jitter) so the encoder policy values are
explicit and easier to tune later.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@js/publish/src/audio/encoder.ts`:
- Around line 122-125: The cleanup currently clears the encoder config via
this.#config.set(undefined) but leaves the derived catalog stale; update the
effect.cleanup block (the same place that nulls worklet.port.onmessage and calls
this.#config.set(undefined)) to also reset the derived catalog produced by
`#runCatalog` — e.g., call the catalog-reset method/property (the internal
`#catalog` setter/clear function) so that `#catalog` no longer exposes the previous
rendition after source teardown or reinit.

---

Nitpick comments:
In `@js/publish/src/audio/encoder.ts`:
- Around line 135-145: Extract the magic numbers in `#createConfig` into named
constants (e.g., DEFAULT_OPUS_BITRATE_PER_CHANNEL = 32000 and
DEFAULT_OPUS_FRAME_MS = 20) defined near the top of the module and replace the
inline literals (32_000 and 20) in the bitrate and jitter calculations with
those constants (use Catalog.u53(DEFAULT_OPUS_BITRATE_PER_CHANNEL *
channelCount) for bitrate and Catalog.u53(DEFAULT_OPUS_FRAME_MS) for jitter) so
the encoder policy values are explicit and easier to tune later.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 303c3d0b-f618-4cc8-8dc4-0ca6da4ac682

📥 Commits

Reviewing files that changed from the base of the PR and between 2083546 and fb88a5c.

📒 Files selected for processing (1)

• js/publish/src/audio/encoder.ts