The Sirius is a keyboard "groove-synth," featuring a subtractive hybrid-tone-generation synthesizer referred to as DTE synthesis introduced in 1997 by Quasimidi. The unit featured both real-time and step sequencers sequencer with pattern- and song-modes, capable of acting basic drum machine, groove-box, or sound-module. The unit is thus 7x multitimbral and has 28-voice polyphony across its 7 tracks, with track selecting a sound within 96 sounds of preset- or user-writeable sounds. Track/voice structure was organized as follows:
Kick drum- 1 voice sample through synthesis chain
Snare drum - 1 voice sample through synthesis chain
Hi-Hat - 2 voice samples through synthesis chain
Percussiondrum kit - 12 samples through common synthesis chain
Synth 1
Synth 2
Synth 3
...where the first 4 tracks can only load patches from 2 banks of 96 User patches of that instrument, but all 3 Synth parts can load from a common pool of 4 banks of 96 keyboard patches. The onboard pattern Sequencer allows for storage of 1600 motifs to freely assign to sound within Patterns, holding 142 ROM and 100 User Patterns. The Sirius can store holds 16 Songs, where each Song assigns
While the "genre" buttons remain fixed to a single purpose, the 7 Track buttons allow for per-track Part Select, Part Mute, and MIDI sequence routing. The 16 buttons perform multiple duties, including:
selecting 1 of 16 Songs
when running patterns w/in a Song, selecting 1 of 8 Patterns, and triggering 1 of 4 Breaks and/or 1 of 4 Special Loop tracks.
In addition to level and pan controls, each track could also be routed to 2 effects engines; 1 for reverb/delay, 1 for modulation delays. It also features an 11-band vocoder with flexible track routing to its carrier and/or modulator, and an advanced/programmable arpeggiator/gate-sequencer for automatic rhythms controlled by its 49-note keyboard. The DTE synthesis method combines a basic sample playback tone generator with virtual analog-like controls. While some of the waveforms are sampled instruments, most of them are merely single-cycle samples of periodic waveforms common in analog modeling synthesizers. The unit is technically a rompler, but its marketing and technical documentationrefer to it as virtual analog synthesizer. This is not entirely accurate, given the lack of any options to shape, sync, or ring-modulate VCO's common to real- and virtual-analog synthesis. While sequencing tools and the sounds of the Sirius were tailored toward 90's techno- and dance-music sounds and workflow, and the models in production held some unresolved limitations and bugs, the synthesis architecture and user interface of the Sirius feature unique innovations that may contribute to its niche appeal. Among these, include its particular handling of:
... the vocoder
* In addition to routing the onboard mic- and line-inputs for traditional "talking-synth" applications, any tracks could be routed to it as carrier, modulator, or both with the front-panel buttons.
* the vocoder acted as an 8th track, with level, pan, and effect-send controls.
...the "ARP" for rhythm creation, similar to the "motivator" of the Quasimidi Raven.
* in addition to storing the voice-synthesis settings, each program could store arpeggio settings for one of 3 modes:
** arpeggiator, with options for rhythmic value, constraining arp-length to pattern-length, and note-handling, using either simple monophonic sorting or one of 9 "UserArps;" user programmable 8-note poly-phonic note-sorting sub-sequences with per-note accent and swing.
** "gater," which rhythmically chops the voice amplitude in a rhythmical tremolo
** "chord," where the played keys will be re-triggered according to a fixed rhythmic rate, or according to the sequencing of a separate part
* any voice could be set to 1 of 16 preset arpeggiator settings. This allowed arpeggiation or rhythmic gating of drum parts, even in conjunction with the Vocoder.
* Despite each voice storing unique "arp" settings, the engine often exhibited bugs when attempting to operate multiple tracks with Arp settings simultaneously. Switching between tracks while an Arp was running may apply the previously used tracks arp behavior to the new track.
...randomization
* the unit featured a dedicated Random Sound button to randomize the settings for the presently-selected voice.
* any track could be reassigned to a new randomly selected Motif and sound Program from memory, facilitating improvisational creation and remixing.
...organization and inheritance
* the 96 voice programs available to each track were organized into 6-groups by character, both for chromatic voice and drum voice
* Sequencer was stored sequencer data as motifs-per-track within patterns, and songs as sets of patterns. Editing any motif or pattern data would change any pattern or song that used it.
* changing the character of any UserArp would change any voice that used that UserArp, and any Pattern- or Song or song that utilized that within it sequencing.
...knob Macros and multi-tasking
* a single Wave Macro knob selects the starting waveform. Synthesizer voices instantiate 2 detune-able instances of the same waveform, with no option to freely mix or match.
* the ADSR envelopes for both the filter and amplifier could have all relevant times/levels set from the detailed edit-menu, but the front panel featured an Envelope Macro that morphed the ADSR settings continuously among plug, sustain, and
* while synthesis knobs adjusted such properties of the presently-selected-track's sound, the 8 dedicated mixer knobs allowed for controlling level/pan/fx-send across 8 tracks simultaneously.
