Hello pd community!

I am actually doing a project involving a raspberry pi and pure-data, the aim is to create a sound level meter that calculate the equivalent level of the sound in dB (A-weighted). Basically, it is just a stockage of the level every seconds and ponderate on a 15mn basis. [if you don't know what is it : http://www.engineeringtoolbox.com/equivalent-sound-level-d_721.html]. The input~ will be provided via the wolfson audio card and the calculation on pure-data.

I started to create my patch but i struggle a lot, i would like to know if anyone would know existing patches that are used for that application : sound level meter, dbA filter, level stockage, any acoustics measurements,....

Anything could help

Thanks in advance!

Hi @buglover,

I am also using a RPI with the wolfson audio card. I have successfully compiled a RT-kernel and using PD with low latency settings. Really nice combo

What kinds of problems are you facing exactly? I am not sure if your question is about the maths or the patch?

Cheers~

Hi @EEight,

I studied the maths about signal theory, DSP, buffers, filters, ... I worked on a texas instrument card and codecomposer. Even if it was a few years back, I still have some notions of that. My troubles comes when i tried to applied that on pd, i am not even working on the pi already.

What i tried to do is : Input~ => dBA weighting filter => SPL + calibration => Leq 1 sec => storage => Leq 15mn
I was wondering is someone has ever done something similar....

I am currently at the first step with the dBa filter which is a IIR filter. I tried with fexpr~but without success (dBa_IIR2.pd. The next idea is to use FIR filter so i need to convert the coefficients via prony's method.

I thought using that distribution for the pi [http://forum.pdpatchrepo.info/topic/8476/raspi-wolfsonaudiocard-and-pd], is it better with a RT-kernel ? (sorry for the newbie question)

Regards

@HannaGen said:

a pre-cooked filter for the A-weighting curve

not sure if precise enough (although the error curves look really good!) - but here is a nice source for biquad coefficients for an A-weighting filter at 44.1 and 48 kHz:
https://jenshee.dk/signalprocessing/aweighting.pdf (from https://jenshee.dk/signalprocessing/signalprocessing.html)

In Pd, at 48kHz, this translates to:

[biquad~ 1.34731 -0.349058 0.965251 -1.3473 0.382051]
|
[biquad~ 1.89387 -0.89516 0.94697 -1.89394 0.94697]
|
[biquad~ 1.34731 -0.349058 0.646665 -0.383622 -0.263043]