audio steering handsfree control
Has any body done a speech steering circuit using the ADAU1701?
I need back to back compressors/ gain blocks controlled by the audio as follows
When audio is received on the ADC 1 before ADC 2 the audio goes to DAC 1 and DAC 2 is reduced in level
If audio is received on ADC 2 before ADC 1 the audio goes to DAC 2 and DAC 1 is reduced in level
when audio is below a threshold on ADC 1 or 2 it is gated (normal compressor gate set up)
Regards
Audio Signal Processors
SigmaDSP Audio Processors
adau1701
Hello Anthony,
The problem I see here from the start is how to determine what is first? All we have in the DSP are samples coming in, some positive some negative. The DSP program only processes a sample at a time. So what would be required is to find some sort of impulse to trigger on to figure out which is first. This is difficult.
If you just use level then that can be done but that can have errors as well. It relies on the signal being louder on the closer mic. Not always true.
If the signals are broadband signals then you can filter out the audio to a low frequency signal that has a wavelength longer than the distance between the mics. Then a simple comparison of the absolute value of the signals will probably determine which mic picked up the signal first. If it is voice then that can be tough.
This is done in all these personal assistance devices but they have some rather high powered processing going on to find the direction of the audio, then tune a beamformer to amplify the audio in the direction of the sound. I have noticed that my personal assistant at home will sometimes be 180 degrees off when sensing the direction of my commands. It had picked up a reflection from a nearby object.
Anyhow, define your use case a bit more. Let me know what the mic spacing is.
Dave T
Please log in to reply.