Issue link: https://digital.copcomm.com/i/1521235
M OT I O N P I CTU R E S O U N D E D I TO R S 77 duplicated into the spatialiser, in addition to sweetening the mix, it will highlight a certain directionality and awareness in the spatial domain. An additional high- frequency boost on those elements will further aid localisation. Mastering Although specialist tools for HOA signal-processing can be deployed to manipulate dynamic range, it should be addressed with a great deal of care and attention. For example, excessive compression commonly used in stereo or even surround productions to create punch and fatness, in 3D, can instead render the opposite effect. Dynamic range compression or over-limiting can also distort the spatial image of the sound field integrity, it is also extremely sensitive to any inconsistencies or unnatural anomalies in the spectral domain. In the case of most diegetic sounds, it is recommended that spatialised audio undergoes minimal dynamic processing to both maintain optimal localisation and of course, preserve the naturalness of the sound. Again, the parallel 2D playback stream allows the inclusion of sounds that fit better in a more processed sonic aesthetic. An Ambisonics file rendered out with insufficient headroom could easily clip the binaural decoder on a streaming platform and create an audible artefact or distort upon playback. As a precautionary measure during production, a multichannel limiter would typically be applied and set to -6dBFS, according to some plugin manufacturer's recommendations. However, it was found that in practice, setting between -2 to -3dBFS below the maximum peak level produces no audible artefacts. It's worth noting every program would have its own individual requirements. Furthermore, it is imperative to also utilise the 3D Audio Metering to monitor loudness-peak variation due to rotational HRTF attenuation, which might be by several decibels as it is frequency-dependent. Despite this, being able to use dynamic processing to master the head-locked stem separately allows it to be much louder in comparison to spatialised diegetic content, thus delivering an overall mix with the optimal methodologies across two mixing paradigms. Modern VR headsets come with improved in-built speakers and headphone output pre-amplification, however, it is not always guaranteed. It is important to optimise masters to meet certain thresholds to avoid situations where a user would need to max out the loudness setting and still not be able to reach the sufficient loudness level. Spatialised Narration Typically, a non-diegetic narration features off-camera and mixed in the head-locked playback stream. Occasionally, to make these moments more intimate and bring the viewers closer to action, the relevant portion of the narration can be externalised by placing the dialogue in 3D space just out of frame to immerse the viewer and give the impression of the storyteller standing beside them in space. As stated previously, a head-locked narration is generally preferable. However, this off- screen approach can be effective for short periods of non-diegetic narration and precipitates a sense of proxemic experience that can cultivate a deeper sense of presence for the audience. Spatialisation and Optimisation Traditional content production might combine large objects into a single sonically layered entity. For example, a large animal's paws, body movement and its growling and breathing or a large aircraft with engine and other mechanical sounds. In VR, this would reduce impact and immersion since an essential aspect of the immersive experience is the life-scale representation of objects. The user can get closer and further away from the object and get around its position. This means sound needs to be distributed across to maintain believable behaviour. Such sounds therefore need to be assigned to appropriate diegetic emitters with suitable attenuation related to the object's morphology and motion. Whether it's a creature or an aircraft, these objects represent different volumetric properties with various associated sounds. The object-staging principle appears to also apply in reverse, individual components could sound disjointed, and so the selection and treatment of individual sounds need to be heuristically tested in a 6DoF environment to subjectively identify and remedy any inconsistencies. Without dynamic 3D audio spatialisation, the sounds would appear flat with limited perspective and suffer from a lack of envelopment. This audio modality enhances a sense of presence and immersion. The approach of staging sounds with higher frequency content above the listener provides some exploitation of the pitch-height