This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Auto-mixing

Other Parts Discussed in Thread: TLV320AIC3254, TLV320ADC3101

I am going to try to ask this question a bit differently.

Is there a TI solution that provides microphone automatic mixing?  That is, when the threshold is met by someone speaking into one microphone would lower the gain or mute the other microphones until some release period?

Thank you.

Phil

  • Hi Phil,

    We are looking into this for you.  Can you provide any more detail about the specifics of your application?  How many mics are you using?  What sort of performance are you lookig for (sample rate, bandwidth, etc.)?

  • Thanks, Tom.

    I'd like to have at least 8 microphone inputs. They can be part of one I2S multichannel stream but I then need to process each with respect to the others. That is, when one microphone is active (someone speaking) then the others either mute or gain is lowered. All mics are always mixed so this happens dynamically depending. In essence, it is AGC for individual mics with respect to the others. Sample rates are generally 48K but higher would be better, right?

    Thanks.

    Phil

  • Hi Phil,

    Thank you for the feedback.  Give us another day or so and we'll see what we can come up with.

  • The gold standard for automixing is the algorithm developed by Dan Dugan. Start there.

  • Hello,

    The TLV320AIC3254 is a stereo CODEC with an integrated Mini-DSP that could be used for developing this type of process flow.  Both microphones could be monitored and there are voice/audio detection blocks that I believe could be used to implement the mixing and muting. 

    At the time of this posting we have limited ability to support custom process flow development but it should be possible.  The TLV320ADC3101 is a 2-channel stereo ADC (possibly a better fit if you don't need the output DAC path), but has a little less processing horse-power and may or may not fully support the desired features but I wanted to throw it out there as well. 

  • I agree, Andy. Fascinating guy both in and outside of the lab. I have been studying his concepts for quite some time.

    Thanks for the heads up.

    Phil