Hi, I am working on a AI-toy reference design which requires audio auto-recording for listening and recording user’s speech (.mp4 will be sent to MLLM)
In Arduino, Examples>AmebaMultimedia > Audio > AudioTriggerRecording
int currentVol = audio.micLevel();
if (currentVol > TRIGGER_VOL) { currentState = STATE_RECORDING; }
the example use audio.micLevel() > 7000 to detect the trigger,
is it possible to do auto recording starting from audio.micLevel()>7000 to audio.micLevel()< 1000
instead of setting recording duration of 3 ~ 5 seconds ?
what is the longest recording period ? maybe set LED flashing at the end of recording.
Your help will be very appreciated !
Hi @rkuo2000 ,
Thank you for your question and for sharing your use case.
The AudioTriggerRecording example is intended as a basic reference, so some customization may be needed for your application. In the meantime, we recommend taking a look at the Audio API and the AudioTriggerRecording example to become familiar with the available interfaces.
I will also look into your question and update you if I have any findings.
Thanks!
Hi @rkuo2000 ,
You may check out the provided audio_trigger_recording_miclevel_based example code to see if it fits your needs. The recording duration limit depends on the SD that you are using, for a FAT32 formatted SD card, a single recording file can’t exceed 4 GB, which is roughly equivalent to 550 hours.
Thank you.
audio_trigger_recording_miclevel_based.zip (2.1 KB)
OK, I will ask AI to study those related codes 