Spatial Sound in Virtual Reality

13 Recording with a Tetrahedron Microphone

A tetrahedron ambisonic microphone such as the RØDE NT-SF1 or the Sennheiser Ambeo needs a recorder with at least 4 inputs, one for each capsule. The Zoom H3-VR is a standalone device and will perform encoding and decoding in the device. It offers different recording modes: A-format, B-format (AmbiX or Fuma), binaural and stereo. When using a tetrahedron microphone, 4 files in A-format are obtained, these need to be converted to B-format with software. One importatnt aspect to remember is to link all 4 channels in the recorder so a single knob is used to control the gains of all inputs. It is extremely important to have the same levels in all channels to preserve the spatial image. These recordings will be used with minimal or no processing so the standard sampling rate of 48Khz and bit depth of 16 are recommended. When recording, try to set the levels to a range between -12dB to -6dB. When you are mixing audio in post-production and exporting, take care to make sure that those levels do not exceed -6dB. For other sounds recorded in mono (that will be added to the soundfield later), a higher sampling rate can be used. Higher sampling rates are useful when the intention is to process the sound using time-stretching, vocoding, etc. The final format of these sounds should be the same as the ambisonic recording.