I am trying a similar thing, except I want to use it in an aircraft. I plan to use use gstreamer with the bandpass filter component (300hz to 3khz) with USB sound cards for in/out. (although the cheap USB dongles pass all the PSU noise right though)whatevs wrote:I'm considering using a Raspberry Pi as the basis of a noise cancellation unit.
The aim of the project is to combine a loud speaker, microphone and Raspberry Pi to create a small unit which can be situated next to a constant sound source to provide active noise reduction (working in a similar manner to 'noise cancelling' headphones).
The intended application is to reduce the noise created by an office server or a portable air-conditioning unit.
The aim of this post is to find out whether anyone has attempted anything similar, or simply whether anyone has any advice, hints or tips that could help me in the right direction.
Thanks in advance for any replies!
When I described my problem I wrote about noise reduction system being a separate unit that can broadcast FM on cochlear implant and hearing aids' frequencies. In my mind I was going to put the noise reduction system between the deaf kids and their FM receiver hearing aids, and the worst of the sound sources. I assumed this would be enough separation to largely offset any latency. The separation would be measured in meters.Jojoba86 wrote:If you place the microphone and speaker 1 cm apart* for example you get 0.3 ms to do the processing. Though that does sound a little optimistic[/i]
Hey cyrano,cyrano wrote:I have been tinkering with noise cancellation a very long time
redhawk wrote:There are several things you could do to improve your noise level issues by processing the incoming audio before it's sent to the Pi for recording.
Firstly make sure you have a good quality microphone and one sold as noise cancelling or uni-directional pick up they suppress noise better than traditional omnidirectional or cardioid.
Next you'll need a microphone pre-amp circuit with band-pass filtering to remove low frequency and high frequencies while faithfully maintaining the human voice with less background noise.
And finally a noise gate / expander circuit this adds more dynamics to loudness vs quietness of audio so background noise levels roll off faster or are totally eliminated.