Latest in the SY/TG line of descent, the TG33 module offers post-FM synthesis combined with the vector control pioneered on the Prophet VS synth. Ian Waugh expands on the '33.
Latest in Yamaha's SY/TG line of descent is the TG33 - a valuable addition to the range or another exercise in repackaging for Yamaha?
YOUR STARTER FOR ten - complete the third pair of instruments: SY77 and TG77, SY55 and TG55, SY22 and... If you answered "TG33" it means you've read the title of this review; I'm not impressed with your powers of deduction. Those of you who usually read magazine articles from the second line down, meanwhile, will still be wondering why the answer isn't TG22.
Although the TG33 has similar features to the SY22 it has more of them. It's 32-note polyphonic for a start (the SY22 is 16-note polyphonic) and 16-part multitimbral (the SY22 is eight-part multitimbral). Also, while the SY22's Multi mode is designed for performance, the TG33's Multi mode is edged towards use with sequencers and is more in line with Multi mode on the SY77.
But, like the SY22 (reviewed MT, July 1990), the essence of the TG33 lies in its use of Vector Synthesis and its claim to fame - in a world of multi-menued, multi-synthesis, multi-complex instruments - is its ease-of-use.
Rack-mounted Vector Synthesis raises an interesting practical problem: how to rack mount a joystick. Yamaha have dodged the problem by making the SY33 a module - although you can rack it using the rack lugs included with it. It doesn't rack vertically as you might expect, instead it sort of slides into the rack space. The lugs can be adjusted so it sits in or out of the rack and it can be tilted. I reckon you'll need at least 3U of rack space, 4U or 5U if you tilt it.
THE SOUNDS ARE arranged in eight Banks of eight Voices, which can be selected with the front-panel Bank and Voice buttons. There are two preset Banks, an Internal Bank (programmable) and room for a Card.
Voices are named using Yamaha's recent naming convention (evident on SY/TG77 sounds, too) which defines the basic nature of the sound with a two-letter prefix. SP is a Synth Pad, SE is a Sound Effect, ME is a Musical Effect, BR is Brass and so on. However, I couldn't find an explanation of the prefixes in the TG33 manual.
There are about 20 prefixes which should be able to contain most synth-generated sounds. This is an excellent idea. How useful it would be in librarian programs which let you assign attributes to voices. A clever program could assign them for you automatically. Regardless, I'd be surprised if other manufacturers follow suit. I await developments with interest.
Inside the TG33 are 128 AWM preset waveforms and 256 FM preset waveforms. In keeping with SY parlance, these are known as Elements. A sound can be constructed from either two Elements (one AWM plus one FM) or four Elements (two AWM plus two FM). These are assigned to the north, south, east and west positions of the joystick - labelled, A, B, C and D respectively. The AWM Elements always occupy positions A and C and the FM Elements occupy B and D.
One of the AWM sets contains drum sounds mapped across five octaves. There are some excellent sounds here including a record scratch, backwards cymbal, cuicas, an "oo" (or an "uh") and timbales. You may detect a soupcon of noise in a couple, but this would probably be lost in a mix.
The manual lists each Voice along with the waves and effects it uses and a comment about what the sound is, what it does or how to play it to best show it off. I like this documentation idea, I just wish all sounds on synths and sound cards/disks carried the same information.
YOU CAN ALTER either the Level or Tuning of the Elements by selecting the relevant option on the front panel and waggling the stick. This is fun and easy, too. But as well as altering the sound in real time, you can record the stick movements as part of the sound. Subsequently, each time you play a note, the sound goes through these same changes. This is accomplished by a sort of sampling process - the position of the joystick is "sampled'' at regular time intervals. You can set the "sample rate" (10-160ms) and 50 samples can be taken. You can edit the position of the joystick at each of these steps and, therefore, program vectors in step time. You can loop them, too.
As you waggle the stick, the unit transmits MIDI Controller 16 and 17 data which you can record in a sequencer to create long vectors. However, this can only be active on one channel at a time, so you can't program different vectors for different Voices and play them back simultaneously in Multi mode.
With all this crossfading going on you won't be surprised to learn that the sounds have a synthy edge to them, although there are very usable piano, brass, bass, organ and string sounds and so on. The TG33 has the "overlap" facility of the SY22, which means that a sound doesn't stop suddenly when you change patches but overlaps into the next. Very useful for live work.
YOU CAN EDIT the sounds using more traditional methods, too. As well as being able to select two or four waveforms for a Voice, there are Frequency Shift, Volume and Pan parameters. The Velocity and Aftertouch Sensitivity is adjustable and you can see the amount of Amplitude Modulation applied to an Element by Mod Wheel or Keyboard Aftertouch.
If you faint at the sight of an algorithm, fear not, because the only control you're given over the FM Elements is the TG's Tone, which adjusts the brightness of the sound.
As for envelopes - there are six preset types but the adventurous can construct their own using an Attack, two Decay and a Release rate settings. There are also Delay and Initial level settings plus Level and Rate Scalings. On a global scale you can construct overall Attack and Release envelopes for individual Voices.
"The idea of rack-mounting a Vector Synthesis instrument raises an interesting practical problem: how to rack mount a joystick."
One of my favourite functions is the Random Element/Level/Detune generator which randomly assigns Elements to a Voice or makes changes to the Level or Detune setting. Yet another easy-program option. Wouldn't it be useful if such random voice generation functions were included on more upmarket - and more complex - instruments? Steinberg's SY77 Editor includes a similar generation facility but taken a few stages further. How many K of ROM would it take to build this in? The TG's Voice Recall function recalls the last edited Voice, even if you've exited edit mode. A handy safety net.
There are 16 built-in digital effects including Reverb, Delay and Distortion along with Balance and Level controls which replace the SY22's depth setting. While they're not very flexible, they do a good job of enhancing the sounds.
THE TG33 HAS 16 Multi Play setups, and this is the main difference between it and the SY22. Rather than assign MIDI channels to the Parts, you assign Voices to the MIDI channels. It's ideal for work with sequencers, but it means you can't layer Voices when playing live (neither can you on the SY77) which would have been really neat.
Each Voice has Volume, Detune, Note Shift and Pan parameters and you can assign the Voices to one of two sets of audio outs. Although the unit has a maximum polyphony of 32 voices, four-Element sounds will reduce this. Voices are assigned dynamically and a modicum of care will ensure that you're never caught short. There's no reserve voice function, but you can dynamically assign a minimum number of Voices to the separate outputs.
UNDER THE UTILITY menus you'll find Master Tune, Transpose and Controller Reset facilities. There is Bulk Dump transmission and you can switch the transmission and reception of SysEx data on and off.
The effect of Program Change messages depends on the mode the TG33 is in. Basically, 0-63 select a Voice and in Multi mode, 64-79 select a Multi. However, the TG33 also supports the new Bank Change messages recently added to the MIDI spec, the first instrument I've come across to do so (apart from a MIDI mapping device from Quasimidi). This uses Controllers 0 (MSB) and 32 (LSB) to transmit a Bank Number. This must be followed by a Program Change number for the change to come into effect. That's a total of three messages but it means you can select any Multi, Voice or memory area.
The Bank Numbers as implemented on the TG33 don't seem to follow any particular pattern: 0, 1, 2, 4 and 5 select Voice Play and one of the banks; 16, 17, 20, 32, 33, 34 and 37 select the Multis and various voice/memory areas. This may be due to internal memory organisation or to provide compatibility with other (unreleased?) Yamaha units. It'll be interesting to see how the Bank Change messages fare. What is required is for a synth to transmit a message (messages?) when you press a Bank button. But if the numbering isn't consistent we're going to have the same problem we had when voices were first put in banks and selected by Program Change numbers 0-15, 32-47 and so on as on Casio CZ synths. The TG33 sidesteps this problem by not transmitting any Program Change messages at all.
The card slot can take MCD32 and MCD64 memory cards for additional Voice and Multi storage. The TG33 can also read SY22 Voice cards and, as there are already several on the market, there's a ready-made library of extra sounds. As the SY22 doesn't have effect balance or effect level controls, these are set to their default values.
Is it just familiarity or are Yamaha manuals getting better? This one has a couple of tutorial sections to get you started, followed by a reference section containing the nitty gritty. That said, the instrument is fairly easy to use and doesn't have a mountain of parameters.
YOU CAN LEVEL a few criticisms at the TG33 although, given what it sets out to do, it's debatable whether or not they would be entirely justified. For example, it has no filter - but this would further complicate the programming and increase the price. We can dismiss comments about lack of control over the FM Elements and lack of individual output sockets for the same reasons.
The biggest niggle I have is the two-line LCD and a touch of the multi button-push syndrome which results in some operations being a little involved. Still, the TG33 doesn't have nests of menus and if this was the most complex synth I ever saw I'd be a happy reviewer. When all's said and done, operation is relatively straightforward.
The change in Multi mode operation (from the SY22) marks the TG33 as a sequencer's expander. The extra polyphony testifies to this although you can use it live, too - the SY77 uses the same Multi arrangement. But sure, it would be useful if you could layer sounds. Take into account the lower price, and the loss of a keyboard seems like a very good tradeoff indeed.
It's difficult to play with this unit and not thoroughly enjoy yourself in the process. To paraphrase Atari's slogan - power without the complexity!
Price £499 including VAT
More from Yamaha-Kemble Music (UK) Ltd, (Contact Details).
Review by Ian Waugh
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!