For the audio, Anna has been looking into various kinds of microphones and ADCs and recently obtained a new ADC with a maximum sampling rate of 200kSPS. However, they have come to the realization that, without a dedicated microcontroller with a real-time OS, we cannot achieve this max sampling rate, since the Jetson runs Linux, a general-purpose OS. We will only be able to get a fraction of 200sps, but this could still be adequate for our purposes. We just need to sample at at least two times the maximum frequency we plan to capture, i.e. ~8kHz per microphone, so ~24kSPS overall for the three microphones. Anna will determine the most we can get out of our current materials ASAP, so that, in case it’s not enough, we can incorporate a dedicated real-time microcontroller to handle the audio.
For the integration aspect, the skeleton is already in place for the audio component to be inserted but the larger concern is how the ability of the mics determine how much more needs to happen in the CV component. With perfect pinpoint accuracy, the CV search algorithm can be quite simple but if we don’t have a good enough calculation then the CV is required to do more of the searching. That is a risk because doing computations will lead to a lower fps since more computations need to be done per frame.
Heather’s goal this week is to get the servo working, the base motor integrated into the physical product, and get the video transmission with two cameras finalized.