VA Mixing Session - Virtual Microphones
With considering directions this virtual rendering environment implementation could be taken for further post-production uses, one that comes to mind would be as a reamping setup where stems can be rerecorded with additional timbre and dynamics. To investigate the implementation of this practically, this session will investigate into the setup and configuration of applying virtual acoustic encoding in emulation of the re-amp process.
As previous encoding action research sessions the HRTF receiver has been placed on the movable camera perspective for ease of use and alignment during arrangement / placement. For this application however, it would be ideal to have a dedicated and fixed microphone detached from the camera perspective as this would be an appropriate configuration where the mic pickup placement is set in a tighter area unmoving in close proximity. Understanding that the HRTFs receiver operates omnidirectionally, if we are to focus this pickup directionally then an exterior geometry capsule would be required to occlude sound from directions that’re unwanted, to attain a focused pickup with a defined stereo image within the 360° soundfield.
Using sketchup to model the occluding virtual geometry capsule for the HRTF receiver, the forms have been constructed in such a way that portions of the receiver’s soundfield are blocked from incoming signal cues so that the image pickup is only in specified areas.
As the receiver is then placed into the capsules, it’s important to note the front facing direction of the listener perspective when aligning. In the default rotation facing vertically, the front focal of the receiver also requires rotating vertically to correctly line with where the microphone is facing. This spherical diaphragm / lens will now capture a set audio image of the VAS whilst obscuring incoming cues from directions outside of the frame. It is worth noting that front facing positioning is not a technical requirement, as the receiver is rotated so is the positioning of the occluding capsule / focal in its soundfield, meaning any zone of the 360° around the listener perspective can be rendered onto individually.
The green arrow illustrates the front facing focal of the listener perspective in the open zone of the sphere, meaning that in this configuration whatever incoming cues hit this zone of the soundfield will be present in the front centre of the rendered stereo field. Rotating the direction will move the open zone for incoming cues relative to the listener perspective. Occluding out the sides and back with the capsule will block incoming cues in those areas, resulting in an audio image recording that maintains the HRTF and VAS depth / positioning / reflection cues in the designated zone.
The rendered zone is defined exactly by the placement of occluding geometry, meaning the variable count of how exactly this could be applied lends to a large amount of creative freedom with rendering via custom virtual microphones. This application can even go further into timbral adjustments with considerations of occluding virtual geometry material and model design, as different virtual materials will impart different tonality to the final audio recording based off their absorption / transmission / scattering parameters. For these configurations, the capsule has used a high absorption and low transmission material assigned to geometry, which also must use a dynamic object component to move the occluding mesh on activation.
It has also been beneficial to add a camera gameobject for the perspective guidelines which it draws, to view the soundfield direction relative to mic orientation. This has doubled in use as it also provides a visual representation of what exactly is captured in the soundfield from the front facing listener perspective.
With the following demonstrations of these virtual microphones, a mono stem source has been used for the audio signal. The figures show the starting orientation for the microphones, alongside the captured stereo image representation in that position. Throughout the recording, microphone 1 fully rotates to occlude the soundsource once out of the capture zone. Microphone 2 rotates only the receiver sphere, with the capsule focal remaining stationary pointing to the source, as the captured image moves around the soundfield. Microphone 3 is oriented sideways, to block signals that come in directly from the sides.
Flood Lamp Mic - VA Mic 1 Virtual Mic Test.wav https://drive.google.com/file/d/17qej9XxQ1L_YUIszmBvIzGc8oN3tkhur/view?usp=share_link
Ballpoint Mic - VA Mic 2 Virtual Mic Test.wav https://drive.google.com/file/d/1UlkrYnOQSeLcEg66PVMYNdIfc8_ZuJvv/view?usp=share_link
Observatory Mic - VA Mic 3 Virtual Mic Test.wav https://drive.google.com/file/d/1us3fMAvK3Wn8Pj2YnRAVinyO69fuA8M3/view?usp=share_link
With these virtual microphones, specifications can be made on which specific zones of the full soundfield are rendered on with audio. As multi-tracking multiple renders together will increasingly add incoming cues in areas of the soundfield where no focal source is, it may be desirable to block these sections for some renders in such a way to reduce the overall noise floor buildup of reflections. This would lead to a final mixed soundfield which has focals on specific portions of the VAS from the listener perspective, still in any relative position with virtual acoustic cues maintained and received, whilst just keeping some sections clear for additional layers of instrumentation.
