One of the problems with Binaural recordings is that in order to fully appreciate the perceptual cues offered by the recording technique the audio must be listened to over headphones. While this may be an ideal scenario for individuals listening on the move or at their desk in an an office the perceptual cues that relate to direction etc can be easily lost or confused when being replayed over loudspeakers. This problem is caused by signals intended for one ear only, which is easily achieved through headphones, being received by the other ear. Inevitably this will cause confusion for the listener in respect of directionality and their ability to accurately place a sound source relative to a visual cue.
A system referred to as 'Cross Talk Cancellation' provides a solution to this....to a certain extent. It too suffers from a lack of individualisation for the listener as it is only an approximation of HRTF values. That said the approximation offered by the cross talk cancellation circuit may be equivalent to the differences in HRTF's between listeners listening through headphones. This is an area of research that is ongoing among many academics and it is hoped that through my own research some conclusions may be drawn that can inform or influence some future work.
Cross Talk Cancellation Circuitry
The process of cross talk cancelation or 'transaural' processing is achieved by 'feeding an anti-phase version of the left channel's signal into the right channel and vice versa, filtered and delayed according to the HRTF characteristic representing the cross talk path'. (Rumsey 2001: p 74). This suggests using phase cancellation and addition to prevent signals intended for one ear reaching the other. As mentioned previously the ability of the circuitry to achieve this task is limited by the fact that all potential listeners have varying head and torso shapes and as a result will affect the way in which audio is presented at the ears.
This image demonstrates how cross talk occurs:
This image demonstrates how cancellation circuitry attempts to eliminate the problem:
There is an argument that exists, however, that suggests that our ability to localise sound sources is made simpler when the sound source is accompanied with a visual cue. For example if we see something happening above us and to the right then we would naturally expect the sound source to emit from that direction. Our brain will, to some degree, convince our 'ears' that the sound is in fact coming from that direction. This may go some way to addressing minor inaccuracies brought about by this replay method.
Process
To produce a 'transaural' version of the sound track for the ice climbing sequence a post processing plug-in was inserted across the audio track in Logic. The parameters were set in accordance with the eventual positioning of the loudspeakers (60 degrees) and the audio was then exported.
The plug-in:
5.1
To produce a 5.1 version of the soundtrack the processed binaural (now transaural) audio was brought back into the session and placed on a surround compatible track. The 'Space Designer' plug-in was applied across the track which allowed various parameters to be set. What I wanted to achieve was direct sound emanating from the front speakers with a little reverb coming from the rear speakers to give the illusion of 'space'. The results were quite exaggerated but I feel that this will be appropriate for the research tests as the simulated 5.1 sounds associated with this will offer some further comparisons to be made. The audio stems were 'bounced out' prior to being muxed with the video using DVD studio Pro.
Space Designer Plug:
No comments:
Post a Comment