Home > On-Demand Archives > Q&A Sessions >
Live Q&A with Chris Bore - An Acoustic Camera
Chris Bore - Watch Now - DSP Online Conference 2021 - Duration: 31:04
Thankyou for the presentation. Nice!
Another option for the microphones would be the ones that have an I2S interface. They are L/R so what you can wind up with is one input wire for every 2 microphones. A/D already done. An FPGA like an Artix 200T or a Zynq has plenty of DSP and memory blocks. I have the beginning of a system for my Digital class based on the PYNQ-Z2 and the AIY Voice Kit hat for RPi. Nothing great about it, just an example of interfacing in raw Verilog in a minimalist style. https://aiyprojects.readthedocs.io/en/latest/voice.html
https://www.adafruit.com/product/3421?gclid=CjwKCAjwkvWKBhB4EiwA-GHjFpPFKRZFfWpKdQjZuHVCuV6aYMxqAxy4JJ1f37vWFkgmrnLcNshT5xoCX0MQAvD_BwE
Chris, a question: You say at the very end that you don't care about the distance, it's the direction that you care about. Maybe the question you were answering wasn't fully stated, but I'd thought the point of coloration in the soundscape to reflect distance was one of the driving factors, and the reason for looking at phase shift. Could you elaborate?
Great start IMHO! Been thinking about the same thing as I sit in my back yard listening to birds in the trees. Did you consider setting the distance between elements of the phased array based on the frequency range. You may want to consider beginning with a linear array and then progress to a two dimensional array.
10:03:04 From John Phillips : As a radar engineer, I know God placed a head with sound-absorbing skin and hair between our 2 sensors. Have you trie resulkts after placing a sound-absorbing grid over your 4 x 3 array ? 10:04:53 From Timothy Young : Start with a 1-D array and get that working first. Also agree with what John is saying about having what we call an infinite baffle behind the mics. 10:05:23 From Leonard to Stephane Boucher(Direct Message) : did you get my msgs on LinkedIn? 10:05:48 From Stephane Boucher to Leonard(Direct Message) : I did. I will respond later 10:08:26 From Emanuele Ziglioli : What kind of microphones are you using? Most mics I find only record in the human audio range. It'd be cool recording near 400KHz in order to build a bat detector. See https://www.openacousticdevices.info/audiomoth 10:13:35 From Emanuele Ziglioli : Thank you, how do you go from real to complex? hilbert transform? 10:13:43 From John Phillips : Adafruit have similar $US7 mike+preamp. 10:14:28 From Michael Kirkhart : Here is a link for a potential substitute for the microphone used in this project: https://www.adafruit.com/product/1713 10:23:55 From rakhel : I heard your sensor is detecting the distance of the sound source from your detector is that right ? my understanding is you are just detecting the direction of the sound source and not the distance 10:32:18 From Leonard : don't fool yourself to me = don't drink you're own coolaid 10:33:16 From Emanuele Ziglioli : thank you!
Hello Chris,
I enjoyed your acoustic camera. presentation... In the not so distance past i took a leave of absence from school an did the phased array beam formers on the B2 Bomber... I was a kid in a candy store. Different frequency bands and different bandwidth... but same ideas... In fact, we do make acoustic camera.... they are called side looking sonars.... We located the Titanic as well as other ship wrecks using the acoustic cameras.... many fishermen use the acoustic cameras as depth finders when they are out doing what fisherman do... fish!
The other acoustic camera is the geo-phone (and hydrophane) arrays that certain communities use to probe the ground for likely oil deposits. and then there are the acoustic arrays used for earth quake detection and in a different time for nuclear events.
We have used a similar experiment to image a moving toy train measuring distance to engine at it moved in our field of view... I use to use a speaker and two microphones in class to show what a wavelength looked like in space... I would sample the output of the two microphones and move them apart while cross correlating the two time signals... I would display the phasor showing the transform of the single tone and watch it rotate as I moved the microphones further apart. When the phasor rotated a whole circle I would say... there is a wavelength! I would repeat using white noise and show the echoes from differnt surfaces of the class room.
Again,,, nice talk....
fred h