I never used Arduino, but I want to use Arduino sensor because it is so compact, high quality, and modular. I'm planning to detect the position/velocity of a moving object with a dopler effect using multiple microphones and speakers. I don't care the device to be standalone, and so I was thinking a bunch of USB microphones connected to my laptop. I have been looking for cheap, modular microphones and I found the following: Electret Microphone Amplifier - MAX4466 with Adjustable Gain : ID 1063 : $6.95 : Adafruit Industries, Unique & fun DIY electronics and kits
I want to use it, but I have no background in EE and curcuits (I am a CS-only guy). It would be best if there is a USB-microphone connector for each of this microphone (and probably a driver & a python library), but it seems like I have to connect it to Arduino Board/bread board with a bunch of scary resistors here and there, and use a Arduino-specific language: http://www.arduinoos.com/2010/10/sound-capture/ http://www.dtic.upf.edu/~jlozano/interfaces/blow_sensor.html
Am I understanding correct?
Do you think using Arduino microphone with Arduino board is adequate? I don't need any program running in the Arduino board. What I just need to get multiple signals from these multiple microphones. I'm not sure this is the right choice.
If my plan is correct, do you think this Arduino kit would be enough for my project (+ microphones above)? http://www.adafruit.com/added
I'm planning to detect the position/velocity of a moving object with a dopler effect using multiple microphones and speakers.
What kind of the object, speed, size, distance? More details 'd help to give a suggestion. Why do you think you need multiple array, not just single emmiter-receiver?
I'm planning to detect the position/velocity of a moving object with a dopler effect using multiple microphones and speakers.
What kind of the object, speed, size, distance? More details 'd help to give a suggestion. Why do you think you need multiple array, not just single emmiter-receiver?
Hand.
Basically it's multi-microphone-version of SoundWave project from Microsoft. I want to get a finer resolution than the paper.
Yes, you missed that the amplifiers used in the links you posted are biased at 2.5V , (one has a pot, the other is fixed).
The adafruit microphone uses a single stage amplifier. The sound analyzer article you linked made a point to mention that the best approach is a two or three stage amplifier because of the bandwidth limitations of the op amps. In other words, 3 stages of 100 each is better than a single stage of 300 which would not be as clean of a signal.
More mic's doesn't provide finer resolution. It's depends on sound freq - space resolution , and fft size - speed resolution. Freq. they used in a paper 18-22 kHz is a maximum what sound card can process. You may moved it up to 40 k, but I think a cost of project would increase significantly. Easier to increase a fft size, from 2048 to 8192 or even 16384.
The only benefits you may get increasing a number of mic's ( up to 3-4) is to get 3D, instead of 1D shown in a paper. It'd become gesture sensing left-right, movement back-forth, and of course up-down - what they did. So to sample 4 mic's you better search USB sound card, that support 2 mics, and use two such cards.
Other options arduino DUE, which 'd sample a data fast enough from 4 linked mic's..
More mic's doesn't provide finer resolution. It's depends on sound freq - space resolution , and fft size - speed resolution. Freq. they used in a paper 18-22 kHz is a maximum what sound card can process. You may moved it up to 40 k, but I think a cost of project would increase significantly. Easier to increase a fft size, from 2048 to 8192 or even 16384.
The only benefits you may get increasing a number of mic's ( up to 3-4) is to get 3D, instead of 1D shown in a paper. It'd become gesture sensing left-right, movement back-forth, and of course up-down - what they did.
What I mean by "finer resolution" was finer resolution of motion detection, not the sound signal resolution increase. So as you said, I'm looking for the 3D. The idea is that if you have multiple microphones in different places, you will have different dopler effects from each position, like capturing (rough) depth from two cameras.
So to sample 4 mic's you better search USB sound card, that support 2 mics, and use two such cards. Other options arduino DUE, which 'd sample a data fast enough from 4 linked mic's..
I get the "USB sound card supporting 2 mics" idea. I guess I need to connect each of this microphone to an audio jack with wires? However, is there any better reason to use arduino DUE?
Yes, you missed that the amplifiers used in the links you posted are biased at 2.5V , (one has a pot, the other is fixed).
The adafruit microphone uses a single stage amplifier. The sound analyzer article you linked made a point to mention that the best approach is a two or three stage amplifier because of the bandwidth limitations of the op amps. In other words, 3 stages of 100 each is better than a single stage of 300 which would not be as clean of a signal.
I am sorry, I think you overestimated my knowledge. I really do not have much background to understand these. I re-read the article, and if I I understood correctly, you are saying that using a normal microphone with manually built amplifiers will provide a clearer signal than the mic-amplifier on the first link. And it seems like I am trying to build a sound card with multiple mic inputs. I was excited to the fact that there is a cheap microphone that can capture 20KHz frequency, but I do not think I have enough knowledge to use it for now.
Thank you very much raschemmel, Magician, and other people who read my long post.
It seems like I am not ready to ask a question on this forum, or use this microphone. I'll change the title of this post as "paused", and learn more about microphone, basic circuits, and Arduino, and then come back.
What kind of moving objects are you trying to detect? If it's a hand gesture, I designed a board that lets you detect gestures in the air like up,down,left,right swipes. You can check it out here: gethover.com
Actually, I was also looking into ways to create something that would let me detect moving objects for a side project i was working on. Tried out IR based implementations including some proximity sensors and wasn't really satisfied by the accuracy. That's what led a friend and I to create the breakout board that you see there. We just released the project and we're working on building a robust library that will enable positional tracking by x,y,z coordinates.