Yamaha TG33 Vector Expander
Yamaha's family of synths and expanders has a new member: a module counterpart to the SY22 keyboard. Nigel Humberstone celebrates the happy event.
Yamaha's continuing success and development of their AWM sample playback and FM synthesis has resulted in a vast number of their products adopting and adapting that form of technology. We have had the SY22 keyboard, SY55 keyboard and TG55 module, not forgetting the big daddy of them all, Yamaha's latest flagship, the SY77 workstation and its obligatory rack version. All in all, a rather tempting but at the same time confusing range of equipment from which to choose.
The new TG33 Tone Generator completes the current keyboard/module family structure of equipment by providing a rack-mount version of the SY22, although it has some significant improvements in its specifications over those of the SY22, and is unusual in being a 'desktop unit' that can also be rackmounted with the supplied brackets in a number of positions and angles. Although the TG33 itself is only 2U high, 3U of rack space are required for clear viewing of the backlit 32-character LCD and access to data entry, volume and on/off dials.
So, is Yamaha's latest offering merely an SY22 in a box? Well, really it's two SY22s in box. The TG33 contains 128 preset voices (twice that of the SY22), along with 64 user voice memory locations. MIDI specifications are doubled to allow the TG33 16-part multi-timbral operation as well as 32-note polyphony, an essential requirement nowadays for serious sequencing setups. 16 internal digital effects are provided: Rev Hall; Rev Room; Rev Plate; Rev Club; Rev Metal; Delays 1-2-3; Doubler; Ping-Pong; Pan Ref; Early Ref; Gate Rev; Delay & Rev 1-2; Distortion & Rev. These are not programmable, although you can adjust balance and send levels. As with most inbuilt effects on units of this type, they are a welcome inclusion if a little inflexible. Like a certain lager, they're good, but not that good.
Around the back of the unit are the standard MIDI In/Out and Thru connectors, an external 10v AC adaptor socket, and dual assignable stereo outputs. The layout of the TG33's 33 function buttons on the sloping front panel is neat if a little busy.
A welcome amendment is the inclusion of a data entry dial (omitted from the SY22), which can be used in conjunction with the [-1/NO] and [+1/YES] keys. The memory card slot has wisely been placed at the front of the unit. Cosmetically the unit's top shell is finished in a 'sandblasted' plastic surface which certainly feels rugged, and complements the overall design.
At the heart of the TG33 are the principles of Dynamic Vector Synthesis, using a joystick to control the dynamic mixing of voice elements. Vector Control is not a new or unique system — the design team at Sequential Circuits were the first to pioneer the concept with their Prophet VS synthesizer. Following Sequential's takeover by Yamaha in 1987, the idea has also turned up on the Korg Wavestation, and in a sense also on the Kawai K1 keyboard and K1M module.
The TG33 contains 128 preset voice patches, arranged as two preset banks. A TG33 voice can have either a 2-element (one AWM and one FM waveform) or 4-element configuration (two waveforms of each type). Through the use of vector control, these waveforms can be mixed and detuned, both manually with help from the joystick control, or automatically by internal editing.
Despite the TG33's appeal as a well-specified multi-timbral module, it will inevitably be judged by potential purchasers on the type and range of sounds that it produces. Most people's first impression is going to be formed on the basis of the unit's built-in demo, which in this case has a separate function button all to itself. This is an annoying trend that many manufacturers seem to follow, and one that I feel Yamaha may regret. The demo is appalling; it has an interesting start which leads into a predictably jaunty performance, ending with an out of time finale! But don't let this put you off — I persevered and was pleasantly surprised by what the TG33 had to offer.
Preset Bank 1 (containing 64 voice patches) features many ethereal and moody combinations, with names like Airy, Cosmo, Wispa and Neuro. In general, I would describe the style of sounds as a fusion of classic Yamaha SY and DX tones and D50 washes. You can almost sense the influence of people like Brian Eno on the programmers of these 'sonic soundscapes', but a wide variety of textures have been provided.
Sound effects are also featured; Mount presents you with your very own babbling brook and warbling birds, and 5.PM brings you rush hour in the city. Kit, as the manual says, "provides a fairly orthodox drum and percussion kit", which is very hard hitting and contains some tuned toms that are reminiscent of the themes from Brookside and Eastenders, as well as your own DJ 'scratch'. The TG809 patch attempts to emulate the classic TR808 and 909 drum machines — and doesn't do too bad a job either.
Preset Bank 2 (again 64 voice patches) contains a more classical collection of instrument patches. Here we have a very acceptable acoustic piano, along with a recreated DX7 electric piano. Patch 2.1, Gospel, is stunning in the authenticity of its recreation of a strident church organ — with aftertouch rotary speaker effect. A fair selection of brass combinations are provided, there's a very 'woody' clarinet, and a mediocre oboe. Strings were better than I had expected, especially Exel, described as "sophisticated classical strings".
There are of course some real let downs; Slap Bass and other acoustic bass sounds were not very impressive, and as with a number of voices there was an audible 'hiss' on the sound. The attempts at a classic guitar are also poor, whilst Omni is neither a "rich" nor "moody" string ensemble as the manual professes, but an inferior, buzzy string impersonation. Again on a critical note, I must mention the odd occasion where some of the TG33's voices suffered from a very noticeable 'graininess'. This was more pronounced in lower registers and with 4-element patches, where the combination of sounds became almost 'crunchy' and lacked definition. A good example was the Reed 4 waveform element in the Flute patch (No. 6.8), which had a distinct crackle. To some extent the build up of noise was eliminated with the application of reverb etc, but I hate to think that effects were being used to hide this defect.
"At the heart of the TG33 are the principles of Dynamic Vector Synthesis, using a joystick to control the dynamic mixing of voice elements."
In getting to grips with the manual vector control, it is helpful to imagine the X/Y axes as a compass, with North, South, East and West representing elements A (AWM), B (FM), C (AWM) and D (FM) respectively. Therefore with a 2-element voice you just have to wiggle the joystick between North and South to hear the separate waveforms being mixed and unmixed, or detuned, depending on which function you have chosen with the vector mode button. A 4-element voice doubles the number of waveforms that can be blended, and these lend themselves to quasi-quadraphonic type effects, which are further enhanced by the individual panning of each element. Using headphones, the spatial effects created by being able to globally mix four sounds or textures can be very mesmerising and hypnotic! After you have experimented with the vector control and practised some sonically stimulating sound sweeps (!), you can record your efforts into a preset voice.
In recording and editing each dynamic vector, 50 steps can be mapped to points along the vector path. These are dictated by your joystick movements, and the TG33 recreates these when playing back patches by 'sampling' the position of the vector control in graduated steps. The time between these steps is called the vector rate, which can be set at levels of between 10ms and 160ms. A short vector rate is preferable for quick control movements, in order to create smooth transitions, and vice-versa for slow movements. Alternatively, you can step-edit the vector levels by selecting the Voice Vector Edit mode. This allows you to numerically alter the parameters and decide whether the vector effect is to repeat or end at any of the steps between 1 and 50.
The vector joystick's X/Y co-ordinates are also transmitted via MIDI as controllers 16 and 17, so your vector movements can be recorded and played back by a sequencer.
Whilst still in the Voice Edit mode, several more parameters relating to the individual elements within a voice can be accessed by hitting the Element Tone button. You can determine the Wave Type used, Frequency Shift (pitch), Volume, 5-position Panning (unfortunately not available for the in-built Drum Set), Velocity and Aftertouch Sensitivity, Tone and LFO settings. There is a choice of five waveform settings for the LFO: saw up, saw down, triangle, square, and sample & hold, and you can set LFO modulation delay, rate and speed.
To aid editing, each individual element can be selected whilst others are muted by pressing the appropriate [ELEMENT ON/OFF] key. Similarly the Element Envelope function enables you to programme the amplitude envelopes for each element: Type; Delay; Initial Level; Attack; Decay 1 & 2; Release Rate; Level; Rate Scaling.
A useful feature within both the Element Tone and Envelope functions is the facility to copy envelope and element parameter data from another element to the current element, thus saving tedious programming time. Whilst on the subject of useful features, the TG33's Edit/Compare option allows you to review sounds prior to editing, and Recall Voice retrieves the last voice edited.
The TG33's underlying appeal must be its 16-channel multi-timbral capabilities, 32-voice polyphony and multi-play setups. Used in conjunction with a sequencer, up to 16 different sounds can be assigned to different MIDI channels in a Multi Play setup. The TG33 provides 16 internal memory locations (RAM) for Multi Play setups; you can also store them on MCD64/32 memory cards. Volume, detuning and pan position parameters are available for individual voices within an "orchestra" (as the manual refers to them), but effect type, balance, and send levels, along with the choice of output can only be defined for the overall group.
Dynamic Voice Allocation (DVA) is the TG33's way of ensuring that polyphony is evenly distributed, whether you are playing two notes on each of the 16 channels, or 32 notes on just one channel. But remember that 4-element voices will reduce the total number of notes available. (Of course, this process does not eliminate note stealing entirely, and there is no MIDI overflow facility.) The process of DVA can get a little baffling when you attempt to assign group setups to the unit's two pairs of outputs. Assign Mode invites you to select note ratios of 32/0, 24/8 or 16/16 for output groups 1 and 2, so that sufficient notes are available for the voices assigned to each group. So in order to route a voice to the second pair of outputs (which always bypass the effects section) for further processing, you need to first place it in group 1 or 2, and then assign either of those groups to output 2. This confusing procedure is not clearly explained in the manual (which by any standards is very readable and concise), and I can only wonder why a simple output mix function was not included for each channel.
In assessing the TG33's merits it is hard to be critical of a unit that offers such value for money in terms of polyphony — under £500, with 16-part multi-timbrality, 32-voice polyphony and a wide range of sounds. Yamaha have built upon the popularity of the SY22 keyboard and succeeded in producing a very powerful module. The Yamaha sound may not be to everyone's taste, whilst the feature of vector control may appear to some as a novelty addition to just another run of the mill synth module — but the TG33 has more to offer, and in the right hands is capable of producing some very stylish and unique textures and sounds. The sound quality is not comparable to, say, Proteus, but then we are looking at a different price range.
Yamaha have obviously intended the TG33 to be accessible to the first time synth/module buyer and 'inquisitive synthesist'. They have succeeded, but the module will be an equally attractive and valuable proposition within a more professional music setup, providing a powerful and creative tone generator.
£499 inc VAT.
Yamaha-Kemble Music UK, (Contact Details).
Review by Nigel Humberstone
Previous article in this issue:
Next article in this issue:
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!