The Yamaha TG77 Tone Generator arrived as a compact rack-mount solution for musicians seeking the distinctive sonic character of the 1990s without the footprint of a full-sized workstation. It served as the rack-mount counterpart to the Yamaha SY77 Music Synthesiser, offering the same core tone generation engines in a form factor designed for studio integration.
The instrument was built around two primary synthesis systems: AFM (Advanced Frequency Modulation) and AWM2 (second-generation Advanced Wave Memory). While these systems could operate independently, the TG77’s defining characteristic was their interaction, allowing for a synthesis method known as RCM (Realtime Convolution & Modulation). This capability allowed AWM2 samples to modulate AFM operators, creating complex timbres that blended the realism of sampled audio with the dynamic spectral control of frequency modulation.
The AWM2 engine handled 16-bit wave data sampled at either 32 or 48 kilohertz, utilising 24-bit internal signal processing and high-resolution 22-bit digital-to-analogue converters. This technical foundation provided a level of clarity and realism in acoustic instrument reproduction that rivaled compact disc players. The system included 4 megabytes of sampled waveform ROM, housing 112 individual built-in waveforms.
Beyond static playback, the AWM2 system featured a versatile digital filter system that allowed for real-time shaping of the sound. Users could layer and blend AWM2 waveforms with AFM outputs, and external wave cards provided a pathway for virtually unlimited expansion of the sample library.
The AFM engine represented a significant evolution from the original DX7 synthesisers, retaining the basic principles of FM synthesis while introducing substantial refinements. The standard 6-operator configuration was expanded to include 45 different algorithms, each supporting up to three user-programmable and independent feedback loops. Unlike the original DX7 operators which relied solely on sine waves, the TG77 allowed any of 16 different preset waveforms to be assigned to each operator.
The system also incorporated enhanced FM parameters, including six-segment envelope generators for each operator with adjustable delays and segment looping, as well as four-breakpoint rate scaling. These features, combined with the ability to modulate operators using AWM2 waveforms, created a synthesis platform with a vast range of spectral possibilities.
RCM synthesis functioned by integrating AWM2 samples directly into the AFM algorithm. The AWM2 sample could be filtered and enveloped before being used to modulate the AFM operators, adding further harmonic content to the complex waveform. The raw AFM output could then be further filtered before panning, effect processing, and final output.
This process allowed for a fusion of sample realism and FM expressive power. The system also permitted the mixing of the “straight” AWM2 output with the AWM2-modulated AFM output, providing additional textural depth. The TG77’s voice architecture supported this complexity through a system of “elements,” where each voice could be composed of one, two, or four elements.
Each element within a voice could be assigned either an AWM2 or AFM waveform, resulting in specific voice modes. Modes 1 through 5 produced pure AFM output, modes 6 through 8 produced AWM2 output, and modes 9 and 10 were dedicated to RCM synthesis. Mode 11 was a special drum voice mode, utilising 61 AWM samples, which was particularly useful in multi-timbre applications.
AWM2 elements featured a 5-segment amplitude envelope generator, enabling configurations such as patching an AWM2 piano attack to an AFM synth-brass sustain. Elements could be layered for hybrid sounds or assigned to different note ranges for split keyboard setups. Additional element-level controls included detuning, transposition, random pitch, portamento for AFM elements, microtuning, dynamic panning, effect selection, and controller assignments.
Dynamic timbre control was achieved through two advanced digital convolution filters available to each voice element. One filter operated as a low-pass type, while the other was switchable for either low-pass or high-pass response. Each filter possessed its own 6-segment envelope generator, allowing for a wide range of dynamic filtering patterns.
Users could combine low-pass and high-pass filters to create a bandpass response, or set both to low-pass operation with a 12-dB/octave rolloff slope to produce a steep 24-dB/octave low-pass curve. To emulate the character of analogue synthesiser filters, the digital filters included a resonance parameter that allowed the cutoff-frequency peak to be boosted into oscillation. In a four-element voice, this resulted in a total of eight filters operating simultaneously.
Stereo imaging was enhanced through dynamic panning, which utilised individual pan envelope generators for each element. This allowed the output from each element to be moved across the stereo sound field in various ways, from simple sweeps to complex motions. Panning could also be controlled via velocity, note number, or LFO control. The TG77 provided a bank of 64 preset pan memories and 32 programmable locations for user-defined pan patterns.
Playability was further refined through sophisticated dynamics control, where MIDI note velocity and after-touch pressure data could be assigned to control pitch, filtering, AWM2 modulation level, or a range of AFM parameters. Users could choose to control these assigned parameters in a positive or negative direction, and velocity switching allowed different elements to be triggered based on playing intensity.
The instrument included four essentially discrete programmable effect processors, comprising two for reverb effects and two for modulation. Several reverb effects allowed separate parameters to be programmed for the left and right channels. The effect processors could be interconnected in various ways to create parallel and series processing configurations.
The reverb effects list included 40 types, ranging from standard hall, room, and plate reverbs to more specific options like tunnel reverb, gate reverb, and feedback reverse. The modulation effects included chorus, flange, symphonic, and tremolo. These built-in effects provided essential ambience and spatial enhancement without the need for external hardware.
The user interface was designed to manage the instrument’s complexity through a large 240 × 64 dot backlit liquid crystal display panel. This display allowed several parameters to be visible simultaneously, with titles shown in large bold type and names spelled in full wherever possible to reduce ambiguity. Flow diagrams and bar graphs were displayed in graphic form for instant recognition.
A hierarchy of directory pages and a “jump” number system facilitated navigation between related functions, while a set of “smart” function keys aided movement through the programming environment. Data entry was supported by increment and decrement buttons, a numeric keypad, and a data entry slider for fast absolute value changes.
In multi-timbre mode, the TG77 could assign up to 16 different voices to 16 different MIDI channels, making it suitable for use with external MIDI sequence recorders or music computers. Sixteen memory locations were provided for complete “MULTI” setups, including voice-to-channel assignments, individual voice volume, note shift, tuning, panning, and effects. The instrument featured 8 individual outputs in addition to two pairs of stereo outputs, GROUP 1 and GROUP 2.
Elements could be assigned to individual outputs as well as either or both groups, allowing for separate routing of voices to a mixing console. For data storage, voices and waveforms could be saved to and loaded from compact Yamaha MDC64 type data cards, with a single card holding up to 64 voices. The TG77 included a complete MIDI implementation with independently programmable receive and transmit channels, a program change assignment table, and bulk dump functions for data transfer.
Features
- AWM2 (second-generation 16-bit Advanced Wave Memory) offers unmatched sample playback quality.
- AFM (Advanced Frequency Modulation) provides a dramatic advance in FM sound quality and programming versatility.
- RCM (Realtime Convolution & Modulation) achieves a new fusion of sample realism and the expressive power of FM.
- Versatile 1, 2, or 4-element voice architecture and complex envelope generators for extensive layering capability.
- Advanced digital filters for all-new dynamic timbre control.
- Multiple complex programmable envelope generators.
- Dynamic panning for enhanced sonic animation.
- Programmable aftertouch and assignable controllers.
- Four internal digital signal processors add essential ambience.
- Comprehensive display and data entry controls for intuitive programming.