Hi,
I pulled the latest versions of nrfconnect/sdk-nrf (1bd98d8) and zephyrrtos (fcaa60a)
I modified the nrf5340_audio/prj.conf file:
@@ -6,6 +6,8 @@ # nRF5340 Audio CONFIG_NRF5340_AUDIO=y +CONFIG_AUDIO_SOURCE_I2S=y +CONFIG_AUDIO_FRAME_DURATION_7_5_MS=y
I get I2S RX overruns on the gateway...
GW [00:00:51.358,612] <inf> audio_datapath: Drft comp state: INIT
--- 4 messages dropped ---
GW [00:00:51.384,124] <wrn> audio_datapath: I2S RX continuing stream
--- 3 messages dropped ---
GW [00:00:51.424,621] <wrn> audio_datapath: I2S RX continuing stream
--- 2 messages dropped ---
GW [00:00:51.440,368] <wrn> audio_datapath: I2S RX overrun. Single msg
--- 1 messages dropped ---
GW [00:00:51.464,355] <wrn> audio_datapath: I2S RX continuing stream
--- 4 messages dropped ---
GW [00:00:51.504,119] <wrn> audio_datapath: I2S RX continuing stream
--- 2 messages dropped ---
GW [00:00:51.510,864] <inf> audio_datapath: Drft comp state: CALIB
--- 1 messages dropped ---
GW [00:00:51.524,353] <wrn> audio_datapath: I2S RX continuing stream
... and audio decode errors on the headsets:
HL [00:00:48.180,999] <inf> audio_datapath: sdu_ref_us not from consecutive frames (diff: 12500 us)
HL [00:00:48.182,220] <wrn> audio_datapath: Decoded audio has wrong size
HL [00:00:48.192,382] <inf> audio_datapath: Drft comp state: INIT
HL [00:00:48.192,413] <wrn> audio_datapath: Decoded audio has wrong size
HL [00:00:48.193,115] <inf> audio_datapath: Drft comp state: CALIB
HL [00:00:48.200,988] <inf> audio_datapath: sdu_ref_us not from consecutive frames (diff: 12500 us)
--- 2 messages dropped ---
HL [00:00:48.221,008] <inf> audio_datapath: sdu_ref_us not from consecutive frames (diff: 12500 us)
HL [00:00:48.222,320] <wrn> audio_datapath: Decoded audio has wrong size
How should I enable 7.5ms frames for evaluation? From a cursory inspection of nrf5340_audio/src/audio/audio_datapath.c it looks like frames are assumed to be in multiples of 1ms.
Thanks,
Chris