VA Acoustic Geometry Materials
Within the concept of VA there is the function to define material properties of rendered environment geometry. This is done via Steam Audio’s geometry tag component, which can assign properties of absorption / scattering / transmission to surfaces to adjust how incoming rays from sound sources interact with surfaces.
Within this session, there is the aim to recreate both dry and reflective environments along with distinguishing the impact of VAS volume size on the final render recording. The track used within this session, the word II by Shigeo Sekito (1975), has been picked due to its particularly resonant instrumentation and noise characteristics, as the recording contains vinyl artefacts; these will also be a good measure of impact of VAS characteristics alongside instrumentation.
With Unity Engine opened, a box VAS has been imported and assigned steam audio geometry with material corresponding to either a dry / mirror material for each render. Positioning from the receiver perspective of the source remains consistent on each iteration 1 render :
Iteration 1 :
Dry - VA Rec Dry It1.wav https://drive.google.com/file/d/1gGFy6exq5JmhcGS--KkpYBvu_djOIjRK/view?usp=share_link
Mirror - VA Rec Mirror It1.wav https://drive.google.com/file/d/13m36QOL5pge2T7hRZZLQAFBjFnlfIt_w/view?usp=share_link
The distinction here between dry and reflective materials is evident in both the increased amplitude of the signal through the mirror VAS, as well as in the drawn out tails and longer decay due to reverberating rays. As Amray models present a varying distribution of rays depending on source and listener position, the next iteration of these sources will be moved to a corner position and recorded from the same listener position as iteration 1.
Iteration 2 :
Dry - VA Rec Dry It2.wav https://drive.google.com/file/d/1inraoUzte1PcUJbSkTt4YeZ-9YOmUQCg/view?usp=share_link
Mirror - VA Rec Mirror It2.wav https://drive.google.com/file/d/1n7bfPRGd8Or1AgevY-LGuvOcm1c0Q3vD/view?usp=share_link
From this iteration it can be confirmed that variation of source position relative to listener position, imparts altered depth spectra characteristics. However whilst the room model is consistent, these changes in timbral characteristics are slight.
Between adjustments of VAS geometry including positioning / material assignments, the scene must be exported to rebuild the scene with new parameters. With iteration 3 of material testing, the source and receiver have been moved into a copy of the same room at 30% scale size compared to that used previously. Positioning is on the diagonal, with the source at a higher azimuth than previously. The smaller VAS should result in more reflections and higher amplitude recordings.
The increased reflections will be as a result of the lesser travel time of rays through the volume, with less simulated air absorption and denser distribution of rays. Upon initialising iteration 3, amplitude increased excessively to the point of distortion, therefore source volume has been reduced to 30% of that used in previous iterations.
Iteration 3 :
Dry - VA Rec Dry It3.wav https://drive.google.com/file/d/1y0ncsXN_dQT2Et4MbT8XNuNK8D3CkTW7/view?usp=share_link
Mirror - VA Rec Mirror It3.wav https://drive.google.com/file/d/1FtXTb-B45uMIQdCg_IspsbODW3lUXKjy/view?usp=share_link
On completion of iteration 3, the result maintained an increased amplitude with distortion occurring at highly resonant sections on the mirror material. With amplitude of both signals having increased, the impact of a receiver’s distance from the source is evidently impacting the pick-up similarly to conventional recording. The denser ray propagation does result in a more robust acoustic image over the wider spread present in larger VAS volumes, with heightened counts of received rays imparting a more complex and higher-definition VA capture.
For the final test iteration, the positioning of sources will be adapted with the audio source stereo signals being split to LR and positioned within the small VAS volume accordingly. The left stereo channel will be positioned to the left of the receiver, in-line regarding height, and into the room’s corner, vice versa for the right channel. This differs from previous iterations, where the source was positioned at 0° in front of the receiver perspective.
Iteration 4 :
Dry - VA Rec Dry It4.wav https://drive.google.com/file/d/1jfoR4cColsKZIZMwlSHs_5VKrI3H-j8p/view?usp=share_link
Mirror - VA Rec Mirror It4.wav https://drive.google.com/file/d/1ZZLFg31lIEbaICR6GGT7sf9KxSFEULI4/view?usp=share_link
In iteration 4, differences to amplitude are significant between the dry and mirror VAS renders, the audio contains more stereo spread than previously, however some elements of the composition remain in a central position. The overall timbre to the encoded acoustics is denser and resultantly higher in quality for it, the tighter packed cues from higher quantities of rays reflecting in the smaller volume VAS model, with the mirror material leading to a fuller sound with manageable resonant amplitudes.
