chenjintao/lucky-cat
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fix(multimodal): stabilize downstream audio playback

Browse Source 
- add configurable prebuffer and adaptive rebuffer for packet jitter
- mute stalled output, auto-clear I2S DMA, and abort inactive playback after 3 seconds
- discard stale-generation packets and add optional downstream diagnostics
This commit is contained in:
chenjintao
2026-08-11 15:25:26 +08:00
parent b60a0373ba
commit 3ce9b38098
5 changed files with 703 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files
components/platform/src/multimodal/audio_pipeline.c
+46
View File
@@ -131,6 +131,26 @@
#define CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_IDLE_MUTE_MS 250
#endif
#ifndef CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_ACTIVE_STALL_MUTE_MS
#define CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_ACTIVE_STALL_MUTE_MS 3000
#endif
#ifndef CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_PREBUFFER_MS
#define CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_PREBUFFER_MS 160
#endif
#ifndef CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_REBUFFER_GAP_MS
#define CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_REBUFFER_GAP_MS 120
#endif
#ifndef CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_MAX_REBUFFER_MS
#define CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_MAX_REBUFFER_MS 640
#endif
#ifndef CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_DIAGNOSTICS_ENABLE
#define CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_DIAGNOSTICS_ENABLE 0
#endif
#ifndef CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_PLAYBACK_TASK_STACK
#define CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_PLAYBACK_TASK_STACK 12288
#endif
@@ -304,6 +324,8 @@ typedef struct {
bool downstream_output_primed;
bool downstream_response_active;
bool downstream_response_end_received;
TickType_t downstream_last_audio_tick;
TickType_t downstream_last_response_activity_tick;
esp_timer_handle_t downstream_idle_mute_timer;
QueueHandle_t downstream_playback_queue;
bool downstream_playback_queue_with_caps;
@@ -347,6 +369,29 @@ static uint32_t s_downstream_overflow_count;
static TickType_t s_downstream_overflow_last_log_tick;
static uint32_t s_downstream_queued_packet_total;
static uint32_t s_downstream_played_packet_total;
#if CONFIG_MULTIMODAL_AUDIO_DOWNSTREAM_DIAGNOSTICS_ENABLE
typedef struct {
TickType_t response_begin_tick;
TickType_t arrival_last_tick;
TickType_t first_arrival_tick;
TickType_t first_playback_tick;
TickType_t last_playback_tick;
TickType_t queue_empty_since_tick;
uint32_t arrival_gap_min_ms;
uint32_t arrival_gap_max_ms;
uint64_t arrival_gap_total_ms;
uint32_t arrival_gap_samples;
UBaseType_t queue_depth_min;
UBaseType_t queue_depth_max;
uint32_t queue_empty_events;
uint32_t queue_empty_max_ms;
uint32_t enqueue_failures;
uint32_t playback_failures;
bool summary_logged;
} downstream_diagnostics_t;
static downstream_diagnostics_t s_downstream_diagnostics;
#endif
static SemaphoreHandle_t s_downstream_playback_queue_mutex;
static bool s_downstream_playback_queue_mutex_initializing;
#if ( configSUPPORT_STATIC_ALLOCATION == 1 )
@@ -360,6 +405,7 @@ typedef struct {
typedef struct {
size_t len;
uint32_t playback_generation;
uint8_t *data;
} downstream_packet_t;