Timeline for Want a USB mini-microphone array. Is it possible?
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 5, 2012 at 14:28 | comment | added | Spacey | @ArikRaffaelFunke Thank you for the updates. The easiest thing for me now is to actually get an already made mic-array somehow. Actually to tell you, what I am trying to do is create an ICA demo, using real sound waves from the environment. Of course, the mixtures will not be instantaneous, so there are some variants of ICA, however that is the overall goal. | |
| Sep 4, 2012 at 20:38 | comment | added | ARF | @Mohammad Have a look if my edit answers any questions you may still have had. | |
| Sep 4, 2012 at 20:36 | history | edited | ARF | CC BY-SA 3.0 |
added 2465 characters in body
|
| Aug 26, 2012 at 20:07 | comment | added | ARF | @Mohammad From your comment I see that my comment was still not clear enough. I will try to update my answer with a more detailed explanation of the issue over the next week or so. Sorry for taking so long but this requires more than just a few minutes to explain from first principles. | |
| Aug 26, 2012 at 13:39 | comment | added | Spacey | @ArikRaffaelFunke I see what you mean about calibration, I misunderstood your statement. However, I do not see how one could remove the time delay from uncommon time base, without also inadvertently removing time delay from the channel, (which you want to keep). The only way is to make sure that distance to each mic from the clap is the same, but that is impossible for linear array > 2. Maybe if far enough (thus planar wave fronts ) it won't matter too much? | |
| Aug 26, 2012 at 13:31 | comment | added | ARF | @Mohammad What FakeName said... To make things maybe even clearer: The time base for any beamforming must be better than the half-period of the highest frequency component you want to take into account. I believe this is impossible to achieve this with independent sound cards by simply starting them recording at the same time. Thing are different if you have a single soundcard with multiple channels. | |
| Aug 26, 2012 at 12:27 | comment | added | Connor Wolf | @Mohammad - the issue isn't compensating for known time-delays, it's compensating for unknown time delays. Basically, there will be some processing-time for each separate USB soundcard, and it may vary from device to device (in which case it could be compensated for), or even not be constant/change depending on USB bus loading. | |
| Aug 25, 2012 at 20:22 | comment | added | Spacey | I actually have built into my laptop, a microphone array, I think is just two mics, but it might be a starting point, however I have no idea how to actually access their data with minimal fuss separately... | |
| Aug 25, 2012 at 20:21 | comment | added | Spacey | Thanks Arik. However, I am not sure why you want to compensate for time-delays here. For example, some adaptive beamforming as you know is delay-and-sum, so it takes case of time-delay issues automatically. The second thing (I think this is what you are saying), is that I do want a common time-sync. (ie, press a button, all microphones start to record at the same time), but I dont want them to have common time-delays. | |
| Aug 25, 2012 at 19:55 | history | answered | ARF | CC BY-SA 3.0 |