Questions to BF for moving sources and to Doppler effect and amplification

[coolco...@gmail.com]

Dear Acoular Team,

firstly I would like to express my appreciation for the fun and knowledge it brought me working with and learning from Acoular. 

I met this question of developing a Beamformer for a rotating line source und would like to get your professional opinions about the following questions:

1. Is the treatment of Doppler effect and amplification caused by a moving source implemented in Acoular?

2. the "moving gitter" BF-algorithm (BeamformerTimeSq in Example 2) can move the gitter in such a way, so that the gitter stays stationary relative to the source. In this way, the Doppler effect and amplification don't need to be considered, if I understand it correctly. But since my microphone array in the practice remains still, will I get reasonable results only when the source passes by in front of the array? I mean, because I cannot get my array to move with the source in the practice, BeamformerTimeSq is not the right algorithm for me, correct?

3. As far as I know, when dealing with moving sources in time domain, signal data can be split into such small time windows, that inside these intervals the data can be averaged for frequency domain Beamforming. But these intervals still have to be de-Dopplerized first, right? 

4. Is there already a Beamformer for moving sources in Acoular? 

Any comments to these questions will be sincerely appreciated. Have a good day!

Best regards,

Shelmy 

[Gert Herold]

While future releases will have considerably extended capabilities for dealing with rotating data, the current version can deal with moving objects in time domain.

To your questions:

1. It is implicitely implemented by the time domain beamformers that can follow a trajectory -- the focus is constantly updated, therefore the Doppler shift will not bee seen on the moving grid.

2. I assume you mean the BeamformerTimeTraj or BeamformerTimeSqTraj beamformers, which implement a moving focus grid. Those are designed to be used for an arbitrary trajectory, so you could apply them also on a source that passes by.

3. Yes, but this is not what is currently done in Acoular's beamformers.

4. Yes (see 2.)

Hope this helps, best regards,

Gert

[coolco...@gmail.com]

Good morning Gert,

thank you for your quick informative reply.

Yet I do not quite understand one point. The grid is moving with the source. The coordinates of the array is fixed at the centre of the gird. Doesn't it mean that the array also moves with the source in BeamformerTimeTraj or BeamformerTimeSqTraj algorithms? Why is this method still applicable analysing signals from a fixed array, like one I have in the measurement?

I look forward to your reply.

Best regards,

Shelmy

[Gert Herold]

The focus grid is moving along a trajectory (see Trajectory class).

The microphone array is fixed, therefore the distances to the focus grid change all the time.

You said that

The coordinates of the array is fixed at the centre of the gird.

This is not the case.

[coolco...@gmail.com]

Ok, now I understand. Thanks a lot!

[coolco...@gmail.com]

Hello,

can Acoular also implement 3D-Beamforming for moving sources?

I tried to combine the BF for moving sources with moving grid in Example 2 (BeamformerTimeTraj or BeamformerTimeSqTraj ) with the 3D-grid in Example 3.

But unfortunately I didn't get reasonable results. Therefore I would like to ask, what tricks I should take to make it possible.

Many thanks in advance and best regards,

Shelmy