Adjust segment duration calculation in stream (#51149)

* Calculate min segment duration internally

* Rename segments to sequences in StreamOutput

* Fix segment duration calculation in test_worker

homeassistant/components/stream/const.py

@@ -19,7 +19,11 @@ OUTPUT_IDLE_TIMEOUT = 300  # Idle timeout due to inactivity
 
 NUM_PLAYLIST_SEGMENTS = 3  # Number of segments to use in HLS playlist
 MAX_SEGMENTS = 4  # Max number of segments to keep around
-MIN_SEGMENT_DURATION = 1.5  # Each segment is at least this many seconds
+TARGET_SEGMENT_DURATION = 2.0  # Each segment is about this many seconds
+SEGMENT_DURATION_ADJUSTER = 0.1  # Used to avoid missing keyframe boundaries
+MIN_SEGMENT_DURATION = (
+    TARGET_SEGMENT_DURATION - SEGMENT_DURATION_ADJUSTER
+)  # Each segment is at least this many seconds
 
 PACKETS_TO_WAIT_FOR_AUDIO = 20  # Some streams have an audio stream with no audio
 MAX_TIMESTAMP_GAP = 10000  # seconds - anything from 10 to 50000 is probably reasonable

homeassistant/components/stream/core.py

@@ -105,7 +105,7 @@ class StreamOutput:
         return -1
 
     @property
-    def segments(self) -> list[int]:
+    def sequences(self) -> list[int]:
         """Return current sequence from segments."""
         return [s.sequence for s in self._segments]
 

homeassistant/components/stream/hls.py

@@ -36,7 +36,7 @@ class HlsMasterPlaylistView(StreamView):
         # Need to calculate max bandwidth as input_container.bit_rate doesn't seem to work
         # Calculate file size / duration and use a small multiplier to account for variation
         # hls spec already allows for 25% variation
-        segment = track.get_segment(track.segments[-1])
+        segment = track.get_segment(track.sequences[-1])
         bandwidth = round(
             (len(segment.init) + len(segment.moof_data)) * 8 / segment.duration * 1.2
         )
@@ -53,7 +53,7 @@ class HlsMasterPlaylistView(StreamView):
         track = stream.add_provider(HLS_PROVIDER)
         stream.start()
         # Wait for a segment to be ready
-        if not track.segments and not await track.recv():
+        if not track.sequences and not await track.recv():
             return web.HTTPNotFound()
         headers = {"Content-Type": FORMAT_CONTENT_TYPE[HLS_PROVIDER]}
         return web.Response(body=self.render(track).encode("utf-8"), headers=headers)
@@ -112,7 +112,7 @@ class HlsPlaylistView(StreamView):
         track = stream.add_provider(HLS_PROVIDER)
         stream.start()
         # Wait for a segment to be ready
-        if not track.segments and not await track.recv():
+        if not track.sequences and not await track.recv():
             return web.HTTPNotFound()
         headers = {"Content-Type": FORMAT_CONTENT_TYPE[HLS_PROVIDER]}
         return web.Response(body=self.render(track).encode("utf-8"), headers=headers)

tests/components/stream/test_worker.py

@@ -25,8 +25,8 @@ from homeassistant.components.stream import Stream
 from homeassistant.components.stream.const import (
     HLS_PROVIDER,
     MAX_MISSING_DTS,
-    MIN_SEGMENT_DURATION,
     PACKETS_TO_WAIT_FOR_AUDIO,
+    TARGET_SEGMENT_DURATION,
 )
 from homeassistant.components.stream.worker import SegmentBuffer, stream_worker
 
@@ -36,9 +36,10 @@ AUDIO_STREAM_FORMAT = "mp3"
 VIDEO_STREAM_FORMAT = "h264"
 VIDEO_FRAME_RATE = 12
 AUDIO_SAMPLE_RATE = 11025
+KEYFRAME_INTERVAL = 1  # in seconds
 PACKET_DURATION = fractions.Fraction(1, VIDEO_FRAME_RATE)  # in seconds
 SEGMENT_DURATION = (
-    math.ceil(MIN_SEGMENT_DURATION / PACKET_DURATION) * PACKET_DURATION
+    math.ceil(TARGET_SEGMENT_DURATION / KEYFRAME_INTERVAL) * KEYFRAME_INTERVAL
 )  # in seconds
 TEST_SEQUENCE_LENGTH = 5 * VIDEO_FRAME_RATE
 LONGER_TEST_SEQUENCE_LENGTH = 20 * VIDEO_FRAME_RATE
@@ -102,7 +103,8 @@ class PacketSequence:
             pts = self.packet * PACKET_DURATION / time_base
             duration = PACKET_DURATION / time_base
             stream = VIDEO_STREAM
-            is_keyframe = True
+            # Pretend we get 1 keyframe every second
+            is_keyframe = not (self.packet - 1) % (VIDEO_FRAME_RATE * KEYFRAME_INTERVAL)
             size = 3
 
         return FakePacket()
@@ -247,8 +249,13 @@ async def test_stream_worker_success(hass):
 async def test_skip_out_of_order_packet(hass):
     """Skip a single out of order packet."""
     packets = list(PacketSequence(TEST_SEQUENCE_LENGTH))
+    # for this test, make sure the out of order index doesn't happen on a keyframe
+    out_of_order_index = OUT_OF_ORDER_PACKET_INDEX
+    if packets[out_of_order_index].is_keyframe:
+        out_of_order_index += 1
     # This packet is out of order
-    packets[OUT_OF_ORDER_PACKET_INDEX].dts = -9090
+    assert not packets[out_of_order_index].is_keyframe
+    packets[out_of_order_index].dts = -9090
 
     decoded_stream = await async_decode_stream(hass, iter(packets))
     segments = decoded_stream.segments
@@ -257,11 +264,9 @@ async def test_skip_out_of_order_packet(hass):
     # If skipped packet would have been the first packet of a segment, the previous
     # segment will be longer by a packet duration
     # We also may possibly lose a segment due to the shifting pts boundary
-    if OUT_OF_ORDER_PACKET_INDEX % PACKETS_PER_SEGMENT == 0:
+    if out_of_order_index % PACKETS_PER_SEGMENT == 0:
         # Check duration of affected segment and remove it
-        longer_segment_index = int(
+        longer_segment_index = int((out_of_order_index - 1) * SEGMENTS_PER_PACKET)
-            (OUT_OF_ORDER_PACKET_INDEX - 1) * SEGMENTS_PER_PACKET
-        )
         assert (
             segments[longer_segment_index].duration
             == SEGMENT_DURATION + PACKET_DURATION
@@ -327,15 +332,21 @@ async def test_skip_initial_bad_packets(hass):
 
     decoded_stream = await async_decode_stream(hass, iter(packets))
     segments = decoded_stream.segments
-    # Check number of segments
-    assert len(segments) == int(
-        (num_packets - num_bad_packets - 1) * SEGMENTS_PER_PACKET
-    )
     # Check sequence numbers
     assert all(segments[i].sequence == i for i in range(len(segments)))
     # Check segment durations
     assert all(s.duration == SEGMENT_DURATION for s in segments)
-    assert len(decoded_stream.video_packets) == num_packets - num_bad_packets
+    assert (
+        len(decoded_stream.video_packets)
+        == num_packets
+        - math.ceil(num_bad_packets / (VIDEO_FRAME_RATE * KEYFRAME_INTERVAL))
+        * VIDEO_FRAME_RATE
+        * KEYFRAME_INTERVAL
+    )
+    # Check number of segments
+    assert len(segments) == int(
+        (len(decoded_stream.video_packets) - 1) * SEGMENTS_PER_PACKET
+    )
     assert len(decoded_stream.audio_packets) == 0
 
 
@@ -363,6 +374,9 @@ async def test_skip_missing_dts(hass):
     bad_packet_start = int(LONGER_TEST_SEQUENCE_LENGTH / 2)
     num_bad_packets = MAX_MISSING_DTS - 1
     for i in range(bad_packet_start, bad_packet_start + num_bad_packets):
+        if packets[i].is_keyframe:
+            num_bad_packets -= 1
+            continue
         packets[i].dts = None
 
     decoded_stream = await async_decode_stream(hass, iter(packets))
@@ -450,6 +464,7 @@ async def test_audio_is_first_packet(hass):
     packets[0].stream = AUDIO_STREAM
     packets[0].dts = packets[1].dts / VIDEO_FRAME_RATE * AUDIO_SAMPLE_RATE
     packets[0].pts = packets[1].pts / VIDEO_FRAME_RATE * AUDIO_SAMPLE_RATE
+    packets[1].is_keyframe = True  # Move the video keyframe from packet 0 to packet 1
     packets[2].stream = AUDIO_STREAM
     packets[2].dts = packets[3].dts / VIDEO_FRAME_RATE * AUDIO_SAMPLE_RATE
     packets[2].pts = packets[3].pts / VIDEO_FRAME_RATE * AUDIO_SAMPLE_RATE