Loudness of sources

[Daniel]

Hello,

I am currently using the SoundScapeRenderer WFS module in a virtual
reality environment at university. During experiments I made the
following observation:
It seems that sources change perceived loudness when changing from a
reproduction region with many speakers to one with few speakers. In my
setup the front of the reference position has a more densely used
speaker array that the back (only every second speaker is used in the
back) so if a source switches behind the reference position it seems
that it is quieter.

Ist this phenomenon a result of the WFS-algorithm? Can the SSR adjust
it automatically? Or do I have to manually set weights in the
reproduction setup for the individual speakers?

If you need to I can provide a reproduction setup file.

Greetings
Daniel Löber

[Jens Ahrens]

Hi Daniel,

Yes, this is a property of WFS. The loudspeaker driving signals are purely local, which means that the calculation of a loudspeaker signal does not make use of any information on the other loudspeakers of the system. More loudspeakers simply means a louder signal. So, the short answer is: Yes, you need to set the speaker weights in the asd description of the reproductions setup yourself. There is no standard way of doing this (this is the reason why SSR does not do it automatically). I guess trail and error is viable approach.

Here's one more comment: The loudspeaker spacing determines the frequency above which spatial aliasing occurs (e.g. where unavoidable distortions of the wave front occur). This is also the frequency up to which the infamous pre-filter is supposed to be active. Now that your spacing changes with source position you might hear changes in the timbre, too. There is no way around this. Just to make you aware...

Best regards,
Jens

[Sascha Michael Spors]

Dear Daniel,

to add a bit more information, this is a result of the spatial discretization by individual loudspeakers as opposed to the continuous secondary source distribution assumed in the theory. In order to cope for the different distances between the loudspeakers, you would have to set discretization weights to the individual loudspeakers. The midpoint rule, as for instance implemented in the Sound Field Synthesis Toolbox for Python http://python.sfstoolbox.org/en/0.3.1/#sfs.array.weights_midpoint, is a good starting point. You can setup your array in the toolbox and calculate the required weights  with aforementioned function.

Another issue in WFS are corners. These may introduce level and spectral changes for sources moved around the corner. For a detailed investigation of this problem you might want to take a look on the paper/presentation linked here http://spatialaudio.net/non-smooth-secondary-source-distributions. However, the solution to this is not implemented in the SSR.

greetings,

Sascha

--
You received this message because you are subscribed to the Google Groups "SoundScape Renderer" group.
To unsubscribe from this group and stop receiving emails from it, send an email to SoundScapeRende...@googlegroups.com.
To post to this group, send email to SoundScap...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/SoundScapeRenderer/ae38147e-c7fb-423e-810c-9435bf82fb00%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.