VA Mixing Session - Sound Collage

For the purpose of generating additional texture to accompany the tape cassette EP which has been produced creatively alongside these sessions, a sound collage has been created to texture the noise between track transitions. This has been done to explore the abstract creative potential of VA use for soundscaping and ambient effects processing. In use are samples of tape sound effects from the archived USC Cinema Gold Tapes, arranged in a retro-futurist aesthetic composition which follows the listener's perspective through a number of scenes, traversed via an elevator.

The production of this textural track has also been done to stress test the capabilities of VA rendering, as the total track count far exceeds those in previous iterations. As errors such as audio tearing occur with high source counts, understanding the CPU load of associated encoding parameters underscores this implementation.

A 32-track audio source has been designed to accommodate all 32 tracks present in the composition. This has been formed as a rhombic triacontahedron with the listener perspective centred within it and sources surrounding, with some sections remaining static and others involving rotations of the listener.

These parameters resulted in renders which had a lower CPU load and less tearing artefacts, although this was only for approximately two minutes, on which major tearing occurs as processing fails to keep up with real-time. To work around this, the track’s first and second halves had to be individually rendered. Although this was a hindrance, it does draw the line for upper limitations of encoded rendering relative to the system in use. With faster processing capabilities these issues would likely not be encountered and track count could be increased further with hardware such as a GPU.

The final audio contains only the encoded signal renders and required 6 edits of cross fading to remove the 5 tears present along with the transition between first and second half, aside from this no additional editing or processing has been done aside from an EQ to pad high-frequencies above 10kHz to keep a saturated timbre and remove high-frequency pierces of noise. The stems had been balanced before processing, with fades and volume automation only used to emphasise or smooth small sections.

Mixdown of the final render: TapeTransitionTexture.wav https://drive.google.com/file/d/1kvAWudidzcYxmq0ZRg3s4SiBTIZqlCOr/view?usp=share_link