Shop OBEX P1 Docs P2 Docs Learn Events
Problem finding audio direction — Parallax Forums

Problem finding audio direction

lardomlardom Posts: 1,659
edited 2012-01-26 20:45 in Propeller 1
My idea is to have a robot turn to the direction of an audio source using four electret mics instead of rotating a single microphone with a servo. After experimenting with a light to frequency converter I assumed I would see a similar difference in signal magnitude using electret mics. My first experiments were unsatisfactory. Apparently, generating a usable signal is more involved with audio than light. Audio is a new area for me. Please help.
audio location.JPG
674 x 484 - 34K

Comments

  • LeonLeon Posts: 7,620
    edited 2012-01-26 11:42
    Look at phased arrays and beamforming.
  • RaymanRayman Posts: 14,848
    edited 2012-01-26 11:54
    I think if you implemented it like the picture, it should work...

    Basically, using 4 megaphones (in reverse) to give directionality...
  • jvrproductionsjvrproductions Posts: 61
    edited 2012-01-26 15:50
    i think will be better if you "insulated or separate better the diferent audios" for example if you convert the white mics on your draw to walls and put the mics on the center that walls will help you to separate the audios. i hope you understuud with my english :-D . your actual design the audio tha must be collected on the mic a its colected on the micb too by the echo of the cone...
  • groggorygroggory Posts: 205
    edited 2012-01-26 16:03
    The light is very clear cut. It doesn't bounce. It's also cake to insulate it from the outside. Anything opaque will do.

    What is your audio source?

    How are you determining the amplitude of the signal from the electret mics?

    Could you post the results of your preliminary tests?
  • ElectricAyeElectricAye Posts: 4,561
    edited 2012-01-26 17:29
    Instead of depending on only the magnitude difference of the audio signal, could you also try the time of arrival differences? Humans use both magnitude difference and phase difference when trying to determine directionality.
  • lardomlardom Posts: 1,659
    edited 2012-01-26 20:45
    @ElectricAye,'time of arrival' sounds like a great idea. I would rearrange the mics in a "T" as opposed to an "X" to make use of both techniques. Two mics using 'time of arrival' would give me an angle and two other mics using magnitude would tell me if that angle, for example, was 45° or 135°. I would pass the result to a compass. The size of the audio signal should become less important meaning it could probably be done without using an adc.
    I will give it a try.
Sign In or Register to comment.