Is it possible to bypass the h5 file and get the sampling array "time data" directly

jinglo...@gmail.com

unread,

Oct 27, 2020, 9:23:01 PM10/27/20

to Acoular-users

Hi alls

I want to use acoular for real-time beamforming, so I want to avoid calling sample files in h5 format and obtain data array( dimension is numsamples *numchannels) directly from the outside. But in the operation, I found that it seemed impossible to avoid this point. All libraries of the entire acoular, such as Spectra.py, Traits.py, have functions to guarantee the usage of h5 file.
So I wonder if it's possible to create a subclass similar to TimeSamples to avoid getting sampled data from .h5 files.

Gert Herold

unread,

Oct 28, 2020, 1:58:45 AM10/28/20

to Acoular-users

It is not possible with the current version of Acoular to process a real-time input. We started some development in this direction for the upcoming version, but it is still in the early stages. The upcoming version will feature a class "SoundDeviceSamplesGenerator", with which it should be possible to use sound card input for processing. You would then use an object of this class instead of the TimeSamples object.

Vassilis Kazazakis

unread,

Oct 15, 2022, 9:43:30 AM10/15/22

to Acoular-users

Hey everyone, reviving this thread... how has this changed since 2020?

Is it possible to do real-time beamforming for an acoustic camera application? What would the CPU+RAM requirements be, say for a 6 mic setup?

Reply all

Reply to author

Forward