fix: use LiveKit track.attach() for adaptive stream and switch screen share to VP9 single-layer

Camera was stuck at 180p because our custom <video> rendering bypassed LiveKit's adaptive stream observer. Replaced manual srcObject binding with track.attach()/detach() in VoiceUser, StreamTile, and PictureInPicture so the SFU receives viewport dimensions and forwards the correct H.264 simulcast layer. Screen share VP9 SVC with L3T3 spatial layers failed because hardware VP9 encoders (NVENC, QSV, VCE) don't support spatial scalability — Chrome silently degrades to L1T1. Reverted to VP9 single-layer (simulcast: false, no scalabilityMode). Also targets encodings[length-1] in applyOverdrive() for correct simulcast layer targeting.
2026-02-26 02:33:18 +01:00
parent 33e9198de6
commit 4bb69851d3
5 changed files with 55 additions and 24 deletions
@@ -122,17 +122,26 @@ export function PictureInPicture() {
    return 'Call';
  }, [currentVoiceChannelId, channels]);

-  // Video track attachment
+  // Derive the LiveKit Track from the selected stream's participant
+  const lkTrack = selectedStream
+    ? (selectedStream.type === 'screen'
+        ? selectedStream.participant.lkScreenTrack
+        : selectedStream.participant.lkVideoTrack)
+    : null;
+
+  // Video track attachment — use LiveKit's track.attach() to register the element
+  // with the adaptive stream observer (enables SFU layer switching by viewport size)
  // shouldShow in deps ensures re-run when PiP becomes visible (videoRef was null before)
  useEffect(() => {
    const videoEl = videoRef.current;
    if (!videoEl) return;
-    if (selectedStream?.track) {
-      videoEl.srcObject = new MediaStream([selectedStream.track]);
+    if (lkTrack) {
+      lkTrack.attach(videoEl);
+      return () => { lkTrack.detach(videoEl); };
    } else {
      videoEl.srcObject = null;
    }
-  }, [selectedStream?.track, shouldShow]);
+  }, [lkTrack, shouldShow]);

  // Initialize position to bottom-right
  useEffect(() => {
@@ -29,6 +29,7 @@ export function StreamTile({ tile, large }: StreamTileProps) {
  const isStreamMuted = streamMutes.get(userId) ?? false;

  const liveScreenTrack = tile.screenTrack?.readyState === 'live' ? tile.screenTrack : null;
+  const liveLkScreenTrack = liveScreenTrack ? tile.lkScreenTrack : null;

  // Quality badge state
  const [qualityBadge, setQualityBadge] = useState<string>('');
@@ -37,16 +38,18 @@ export function StreamTile({ tile, large }: StreamTileProps) {
  const [contextMenu, setContextMenu] = useState<{ x: number; y: number } | null>(null);
  const [qualityPopoverOpen, setQualityPopoverOpen] = useState(false);

-  // --- VIDEO ---
+  // --- VIDEO --- use LiveKit's track.attach() to register the element
+  // with the adaptive stream observer (enables SFU layer switching by viewport size)
  useEffect(() => {
    const videoEl = videoRef.current;
    if (!videoEl) return;
-    if (liveScreenTrack) {
-      videoEl.srcObject = new MediaStream([liveScreenTrack]);
+    if (liveLkScreenTrack) {
+      liveLkScreenTrack.attach(videoEl);
+      return () => { liveLkScreenTrack.detach(videoEl); };
    } else {
      videoEl.srcObject = null;
    }
-  }, [liveScreenTrack]);
+  }, [liveLkScreenTrack]);

  // Quality badge (poll every 3s)
  useEffect(() => {
@@ -34,16 +34,19 @@ export function VoiceUser({ tile, large }: VoiceUserProps) {
    return () => tile.videoTrack?.removeEventListener('ended', onEnded);
  }, [tile.videoTrack]);

-  // Attach Video
+  // Attach Video — use LiveKit's track.attach() to register the element
+  // with the adaptive stream observer (enables SFU layer switching by viewport size)
  useEffect(() => {
    const videoEl = videoRef.current;
    if (!videoEl) return;
-    if (activeVideoTrack) {
-      videoEl.srcObject = new MediaStream([activeVideoTrack]);
+    const lkTrack = tile.lkVideoTrack;
+    if (lkTrack) {
+      lkTrack.attach(videoEl);
+      return () => { lkTrack.detach(videoEl); };
    } else {
      videoEl.srcObject = null;
    }
-  }, [activeVideoTrack]);
+  }, [tile.lkVideoTrack]);

  // Context Menu
  const [volumeMenu, setVolumeMenu] = useState<{
@@ -46,6 +46,8 @@ export interface ParticipantInfo {
  videoTrack: MediaStreamTrack | null;
  screenTrack: MediaStreamTrack | null;
  screenAudioTrack: MediaStreamTrack | null;
+  lkVideoTrack: Track | null;   // LiveKit Track for attach/detach (adaptive stream)
+  lkScreenTrack: Track | null;  // LiveKit Track for attach/detach (adaptive stream)
 }

 export interface UserTile {
@@ -54,6 +56,7 @@ export interface UserTile {
  participant: ParticipantInfo;
  videoTrack: MediaStreamTrack | null; // camera only
  audioTrack: MediaStreamTrack | null; // mic
+  lkVideoTrack: Track | null;         // LiveKit Track for attach/detach
 }

 export interface StreamTile {
@@ -62,6 +65,7 @@ export interface StreamTile {
  participant: ParticipantInfo;
  screenTrack: MediaStreamTrack | null;
  screenAudioTrack: MediaStreamTrack | null;
+  lkScreenTrack: Track | null;        // LiveKit Track for attach/detach
 }

 export type GridTile = UserTile | StreamTile;
@@ -69,12 +73,14 @@ export type GridTile = UserTile | StreamTile;
 export function deriveGridTiles(participants: ParticipantInfo[]): GridTile[] {
  const tiles: GridTile[] = [];
  for (const p of participants) {
+    const hasLiveVideo = p.isCameraOn && p.videoTrack?.readyState === 'live';
    tiles.push({
      kind: 'user',
      key: p.identity,
      participant: p,
-      videoTrack: (p.isCameraOn && p.videoTrack?.readyState === 'live') ? p.videoTrack : null,
+      videoTrack: hasLiveVideo ? p.videoTrack : null,
      audioTrack: p.audioTrack,
+      lkVideoTrack: hasLiveVideo ? p.lkVideoTrack : null,
    });
    if (p.isScreenSharing) {
      tiles.push({
@@ -83,6 +89,7 @@ export function deriveGridTiles(participants: ParticipantInfo[]): GridTile[] {
        participant: p,
        screenTrack: p.screenTrack,
        screenAudioTrack: p.screenAudioTrack,
+        lkScreenTrack: p.lkScreenTrack,
      });
    }
  }
@@ -149,6 +156,8 @@ export function useLiveKit() {
      let videoTrack: MediaStreamTrack | null = null;
      let screenTrack: MediaStreamTrack | null = null;
      let screenAudioTrack: MediaStreamTrack | null = null;
+      let lkVideoTrack: Track | null = null;
+      let lkScreenTrack: Track | null = null;
      let hasScreenSharePublication = false;
      p.trackPublications.forEach((pub) => {
        // Detect screen share publication even if unsubscribed
@@ -164,8 +173,8 @@ export function useLiveKit() {
        if (!mt || mt.readyState !== 'live') return;

        if (pub.source === Track.Source.Microphone) audioTrack = mt;
-        else if (pub.source === Track.Source.Camera && p.isCameraEnabled) videoTrack = mt;
-        else if (pub.source === Track.Source.ScreenShare) screenTrack = mt;
+        else if (pub.source === Track.Source.Camera && p.isCameraEnabled) { videoTrack = mt; lkVideoTrack = track; }
+        else if (pub.source === Track.Source.ScreenShare) { screenTrack = mt; lkScreenTrack = track; }
        else if (pub.source === Track.Source.ScreenShareAudio) screenAudioTrack = mt;
      });

@@ -194,6 +203,8 @@ export function useLiveKit() {
        videoTrack,
        screenTrack,
        screenAudioTrack,
+        lkVideoTrack,
+        lkScreenTrack,
      });
    };
    processParticipant(r.localParticipant, true);
@@ -18,7 +18,7 @@ export interface OverdriveOptions {

 export interface ScreenShareBuildResult {
  capture: { width: number; height: number; frameRate: number };
-  publish: { videoCodec: 'h264'; videoEncoding: { maxBitrate: number; maxFramerate: number }; simulcast: false };
+  publish: { videoCodec: 'vp9'; videoEncoding: { maxBitrate: number; maxFramerate: number }; simulcast: false };
  overdrive: OverdriveOptions;
  contentHint: 'motion' | 'detail';
 }
@@ -63,7 +63,7 @@ export function buildScreenShareOptions(config: ScreenShareConfig): ScreenShareB
  return {
    capture: { width, height, frameRate: fps },
    publish: {
-      videoCodec: 'h264',
+      videoCodec: 'vp9',
      videoEncoding: { maxBitrate, maxFramerate: fps },
      simulcast: false,
    },
@@ -99,14 +99,19 @@ export async function applyOverdrive(
    if (!sender) return;

    const params = sender.getParameters();
-    if (!params.encodings?.[0]) return;
+    if (!params.encodings?.length) return;

-    params.encodings[0].maxBitrate = options.maxBitrate;
-    params.encodings[0].maxFramerate = options.maxFramerate;
-    params.encodings[0].networkPriority = 'high';
+    // Target the highest-quality layer. With simulcast, encodings[0] is the
+    // lowest layer; our overdrive must hit the top layer so the custom bitrate
+    // slider controls the full-resolution stream, not the quarter-res one.
+    // For non-simulcast tracks (single encoding), length - 1 === 0.
+    const idx = params.encodings.length - 1;
+    params.encodings[idx]!.maxBitrate = options.maxBitrate;
+    params.encodings[idx]!.maxFramerate = options.maxFramerate;
+    params.encodings[idx]!.networkPriority = 'high';
    (params as any).degradationPreference = options.degradationPreference;
    if (options.minBitrate > 0) {
-      (params.encodings[0] as any).minBitrate = options.minBitrate;
+      (params.encodings[idx] as any).minBitrate = options.minBitrate;
    }

    await sender.setParameters(params);
@@ -134,9 +139,9 @@ export async function startScreenShare(room: Room): Promise<boolean> {
      // @ts-ignore — LiveKit accepts frameRate at capture level
      frameRate: opts.capture.frameRate,
    }, {
-      videoCodec: 'h264',
+      videoCodec: opts.publish.videoCodec,
      videoEncoding: opts.publish.videoEncoding,
-      simulcast: false,
+      simulcast: opts.publish.simulcast,
    } as any);

    if (!track) {