The TiMax Software

As well as the technology involved in the delay matrix DSP processor itself, TiMax also includes a dedicated PC software application which provides showcontrol and an intuitive "domain-shift" that helps the sound designer engage creatively rather than empirically with the physics involved with independently time-aligning individual input sources to the speakers in the multichannel SOR system. At the root of this domain shift is what TiMax calls Image Definitions, which are programmed in their own screen in the TiMax software (See Fig1)

Fig 1: TiMax Image Definition Screen

Image Definitions, are created by the sound designer at show set up to provide a set of relative level and delay values given to speakers distant from the focus (or "anchor") of the Image Definition. The anchor is a "time zero" speaker, or an acoustic source such as a voice or a musical instrument etc. The delay values given to each speaker reinforcing the "anchor" are arranged such that the output from these speakers is delayed back "behind" that anchor source. Firstly, you have to time-align for the path length difference to the listener from the anchor speaker and the nearfield “reinforcing speaker”, and then secondly you have to overdelay to introduce the localisation effect. The overdelay is bounded by some well defined limits: the first 8-10ms of overdelay will introduce Comb Filtering and so are to be avoided, but after about 10ms there are another 15ms of "Haas Delay" available until the threshold of Echo Perception is reached at around 25-30ms.

This is programmed in the Image Definition screen using a combination of measurement and listening, first using a pulse generator plugged into a TiMax input and subsequently with program material such as speech. The delay is set at a listening position close to each “reinforcing speaker” until the direct and delayed signals fuse together and sound like they are coming from the distant "anchor" focus for that Image Definiton. Delays and levels are trimmed using a simple "crosshair" arrangement until the nearfield delayed speaker is as loud as possible without undermining the localisation effect. This process is repeated for all the speakers required in that Image Definition, and then repeated from scratch using other "anchor" sources for all the other required Image Definitions (up to a maximum of 32). It is possible to use speakers at right-angles to the anchor speaker's axis, but generally speakers directly behind the listener are less useful to include, partly due to the conflict between tonal differences caused by the shape of the ear but also due to the range of variation in time arrival differences for audience members placed between those speakers.

Fig2: TiMax Effects Screen

The Effects Screen has a large Work Area to import a bitmap plan of the auditorium or space, and then a "library" on the left hand side of available Image Definitions (yellow) and Input sources (green) to apply to them. Image Definitions are double-clicked into the work area and then dragged into position on the arena plan. Inputs are then likewise brought into the work area, adjusted for level (or faded in) using a big fader on the right hand side, and then applied to Image Definitions by being placed on top of them.

Dragging the input sources between Image Definitions then pans them round the space, and these moves can be recorded and played back automatically on a cue by cue basis. Up to sixteen sources can be statically localised or animated in each Effects Screen cue, and the TiMax software, which is a complex dynamic database, continually calculates using "circle of influence" algorithms what the multiple level and delay values need to be to create the localisations. It then instructs the TiMax matrix in real-time to set up or slide these values to create the static localisations or animated pan effects. The matrix firmware then uses some tricky interpolating DSP algorithms to continually adjust level and delay parameters independently between inputs and outputs to create these panning moves with minimal glitching.

The use of delay-based TiMax Image Definitions causes the pans to sound very smooth and makes them effective over the whole audience area. More conventional "discrete" panning around just individual speakers often creates gaps, or "holes" in the pans and draws the listener to the individual speakers themselves, as well as only working effectively over a relatively confined central sweet-spot.

To make all this manageable during a show, these complex cues are handled by a straightforward PlayList Screen showcontrol environment where cues can be set up to trigger from MIDI, SMPTE, RS232 or an internal TimeLine. This allows TiMax to be slaved to other external showcontrol environments or even to run subsections of cues to its own timeline, with an initial trigger coming in as host video playback timecode value for instance.. (See Fig3)

Fig3: TiMax PlayList Screen
(showing Input Trigger dialogue).

The Events Screen allows each cue to output a series of external showcontrol Events in MIDI, SMPTE, RS232 formats, as well as controlling up to two CD players in the TiMax ShowControl PC, triggering .wav file playback from the PC's harddisk, and can also trigger up to eight remote Event Relays. This can happen simultaneously to TiMax itself being triggered externally, so TiMax can work either as a showcontrol master host, or as a slave, or as a combination of both. (See Fig4)

Fig4: TiMax Events Screen

NEW to the TiMax software is the TiMax SoundTablet, a powerful and quick cue-driven sound effects editing package for anyone with a modern desktop, laptop or tablet PC. This rovides eight independent audio sample tracks per Cue, all with independently editable Play Zone, Loop Zone, Track Slip and Volume Profile. When used with TiMax dsp delay matrix, SoundTablet also provides a unique drag ‘n drop waveform-based pan programming function which allows complex pan moves to be set up and edited by placing Image Definitions onto the audio sample waveform.