Virtual Microphones

My colleague at USC Chris Kyriakakis has studied the recordings that we have made in 10.2 and come up with a method of upmixing existing 2-channel recordings to an arbitrary number of channels with microphone positions that never existed on the original session. Use is made of the differences between the main stereo pair, and other microphones like hall mikes of our 10.2-channel recordings. By use of high-order mathematics, the difference between the two microphones in 3-D space may be found and applied to other different recordings that are dominated by main microphones. Thus hall mikes, and even spot mikes, not present in the original may be provided to the producer/engineer for new multichannel mixes.

105

4 Multichannel Mixing and Studio Operations

Tips from This Chapter

• Surround sound allows for enveloping (surrounding) the listener, not just spaciousness.

• Two basic approaches exist for front/surround sound: direct/ambient and direct-sound all around.

• Many microphone channels can be hard assigned to loudspeaker channels, while some need fixed panning in between two adjacent channels, and some may need the ability to pan dynamically.

• Panners come in various forms: three knobs, joystick, and DAW software. Panning laws, divergence, and focus controls are explained in the text. Work arounds for 2-channel oriented equipment are given.

• Panning is the greatest aesthetic frontier of multichannel sound. One major error is panning sources between left and right, ignoring center.

• Source size can be changed through signal processing, including surround decoding 2-channel sources, decorrelation, and reverberation.

• Equalizing for multichannel recordings varies from stereo: centered content will probably use less equalization, while direct sound sources placed in the surrounds may require more equalization than typical, in order to maintain the timbre of the source.

• Multichannel routing is usually by way of three AES-3 standard pairs, organized according to the track format in use, of which there are several. Jitter can affect reception of digital audio signals, and digital audio receivers must be well designed to reject the effects of jitter.

• The most common multichannel track assignments is: L, R, C, LFE, LS, RS. Several other track assignments exist.

107

• Multichannel audio for video is delivered by means of a special bit-rate reduction method called mezzanine coding so that it may be stored on the available digital audio tracks of video tape recorders, or by means of a digital audiotape separate from the video (double system) because most video recorders only have 4 channels.Time code must match the picture, and 48kHz is the standard sample rate.

• Mezzanine (lighter than transmission-coded audio) coding makes two of the digital videotape audio channels capable of carrying 5.1 audio channels, with multiple generation coding/decoding cycles possible and with editing features not found in the transmission codec. One such codec is Dolby E.

• SMPTE reference level is -20dBFS; EBU is -18dBFS.

• Multichannel monitoring electronics must include source-playback switching, solo/mute/dim, and monitor volume control, with calibration, individual level trims, bass management, and method of checking mixdown to stereo and mono.

• Outboard dynamics units (compressors, limiters, etc.) need multichannel control links.

• Several reverberators with stereo outputs can be used to substitute for one reverberator with five outputs.

• Decorrelators are potentially useful devices, including those based on pitch shifting, chorus effect, and complementary comb filters.