Derek Johnson plugs in two brand new GM MIDI synth modules, clamps on his headphones and waxes lyrical.

Yamaha TG100

Yamaha's TG100 module is a 28-note polyphonic, AWM (Advanced Wave Memory) Tone Generator, containing 192 ROM instrument voices and 10 drum kits. Some degree of programmability is offered and there are 64 RAM memory locations for user voices. The instrument is 16-part multi-timbral (16 different sounds can play at the same time) and conforms to GM System Level 1 Standard, allowing playback of MIDI songs recorded using other GM standard equipment. The TG100 has essentially been designed as a preset instrument, but simple editing can be performed on the 64 voices in the Internal voice bank. In keeping with recent trends in synthesiser design, the TG100 also has onboard digital reverb effects. This makes it very useful for budget, entry-level use and possibly for school use, since sounds may be enhanced without the expense of a dedicated effects unit.

The host computer connection on the TG100's back panel allows it to function as a MIDI interface for Apple Macintosh and PC compatible computers, obviating the need for a dedicated interface, which can cost from £50-£60 upwards. A stereo audio in facility is provided, which allows a further stereo instrument to be mixed in with the sound of the TG100 without recourse to an external mixer. This is a valuable feature for live performance but could also be beneficial in the home studio where mixer inputs are scarce.

The Package

The TG100 is packaged in the familiar half-rack sized format and is equipped with stereo audio outputs as well as the full complement of MIDI In, Out, and Thru sockets. Aside from the volume control and input level trim, the TG100 is entirely button driven, the readout being on a 16-character display. The unit can operate in three distinct modes, the first of which is General MIDI mode, when the unit will be configured to conform to the GM standard patch allocation. The Disk Orchestra mode matches Yamaha's own Disk Orchestra Collection of music software, while the CM mode follows Roland's CM64 patch map assignment. The sounds are divided into four Banks, and in multitimbral mode, sounds can be selected from any of them without restriction.

Voice Architecture

When the TG100 is switched on, it is automatically in Play mode. Selecting a voice bank is straightforward using the Cursor button and the + and - buttons, while the Part button scrolls through the 16 available voices or parts. You can then select which 16 voices from the 192 presets and 64 user voices are to be used.

Pressing the Edit button puts you into one of three edit modes; with the cursor placed in the MD position, Edit takes you into Multi Common Edit, which simply allows you to alter the reverb type (two halls, two rooms, two plates and two (preset) delays) and the overall reverb level. With the cursor in the CH position, the Edit button takes you into Multi Part Edit where each of the 16 parts can be assigned its own pan (stereo) position, EG attack rate, EG release rate, reverb send level and MIDI receive channel.

Finally, if the cursor is in the BK or PC# position, pressing Edit will take you to Voice Edit, where you can edit one of the 64 sounds in the internal user bank. Voices are made up from elements which can be thought of as sonic building blocks, but on the TG100, you can't actually recombine elements to create new voices; you're restricted to the combinations which Yamaha provide. Around half of the preset voices consist of one 'element', the rest being made up of two. Where a voice consists of two elements, individual pan and level can be set for each element, while detuning can be used to create a chorus type effect, for a fatter sound. Your new voice can be named, and a voice from any of the preset banks can be copied into the Internal user bank and then be edited. There is no way of creating a voice from scratch on the TG100 — you have to start from a preset and edit it, and the drum voices can't be edited at all.

The TG100 has no internal memory for edited voices, which means that any edited voices will be lost when you switch off. If you want to keep your edits, you'll need to resort to pen and paper to make a note of your settings, or save the contents of the memory to an external MIDI device such as a MIDI datafiler or computer librarian programme. However, the multitimbral setup parameters are stored in memory.

System Mode functions are accessed by pressing the Part and Edit buttons simultaneously. This allows access to the master tuning, velocity meter mode, MIDI System Exclusive functions (used for memory dumps), and a reset function which restores the TG100 to its factory settings.

TG In Use

The TG100's control layout proved to be both simple and logical, though the display itself is rather dull and difficult to read. Playing sequences recorded with another GM synth using the TG100's GM Voice Bank was moderately successful, in that the right type of instruments were played in the right places, but different GM instruments will inevitably impose their own character on a piece of music.

The on-board reverbs are perfectly usable, but having preset delay times is very limiting, especially if you want to sync the delay time to the tempo of your music. Having an individual reverb send level for each part is welcome and bypasses many of the limitations of having only a single pair of audio outputs when it comes to creating a properly balanced and treated mix.

The voice architecture of the TG100 is simple to grasp since there are relatively few parameters to worry about. Indeed, it is probably fair to consider the TG100 as, basically, a preset machine with limited editing facilities. The attack and release parameters in Part Edit Mode can go some way to creating a sound with a different character, and a fair amount can be done with the detuning of two element voices, but new sounds are never going to sound that far removed from the presets.

More detailed editing will be possible using external editing software; Yamaha will be providing editors for the TG100, free of charge. The Mac version, written with Opcode's Max, is imminent, with Atari and PC Windows versions following soon. These editors will allow the user to change elements within voices as well as making editing generally easier.

The Sounds

Luckily, the actual sound character is fine: these are good, quiet samples. Though there are not as many multi-samples as you would find on more sophisticated synths and sound modules, neither is there the same learning curve or cost. The pianos are very good, while the strings and brass sounds provide very useable examples. The WoodBass acoustic bass sound is very authentic and the GM spec also includes various other useful basses. The least good sounds are the distorted and overdriven guitars, while acoustic guitars are fine — in all there is a large number of friendly and genuinely playable sounds on board.

Given its modest cost, wide range of sounds and multitimbral capabilities, the TG100 represents excellent value, especially if you need to make use of the inbuilt Mac or PC MIDI interface.

Further Information
TG100 £349 including VAT.

Yamaha-Kemble Music UK, (Contact Details).

Boss DS330 Dr Synth

The Boss Dr Synth was one of the great surprises of the recent British Music Fair. It's the first non-drum machine sound source to come from Boss, Roland's MI-side alter-ego, and takes the form of a GM synth, the sounds being sourced from Roland's Sound Canvas series and crammed into a drum machine-style box. Indeed, the DS330 comes in much the same table top package as the DR660 Dr Rhythm.

Features

The DS330 features 156 GS sounds — GS is Roland's version of GM — and eight drum sets. The sounds are based on PCM samples, the vast majority of which seem to have a lot in common with Roland's own Sound Canvas. One thing to bear in mind is that besides being a 16-part multitimbral GM/GS module, the DS330 has an additional Single mode, where it behaves as an editable, single-voice synthesiser. The Dr Synth proves to be 28-voice (rather than 28-note) polyphonic — since several sounds are made up of two voices, actual polyphony will be a variable figure, with 14-note polyphony in the worst case.

While the number of editable parameters is not great, they are actually very useful in a creative way. Besides being able to alter the reverb and chorus settings, the filter cutoff frequency and resonance are variable, and there is even a basic envelope generator which adjusts the time variation of the cutoff frequency, featuring attack, decay, and release parameters. Additionally, there are vibrato rate, depth and delay parameters for each sound.

The on-board DSP offers reverb/delay and chorus, with three rooms, two halls, a plate, a delay and a pan delay. While there aren't the fine controls of dedicated units, you can alter (within a range of 0-127) the reverb time, delay feedback and level. Similarly, the chorus effect only has delay, rate, depth, feedback and level parameters, again within the range 0-127. Flange type effects are possible when the feedback is set very high, and even without the precise control over delay times, these are very useful effects.

In Single mode, not only can patches be edited, but Dual and Split mode allows two different patches to be layered or split. The split point is user selectable, and any sound can be chosen to go below the split, although interestingly, the DS330 initially chooses an appropriate sound for you. For example, choose split mode when your patch is Soprano Sax, and the lower sound will be Baritone Sax; if you choose piano, the lower sound will be Acoustic Bass, and if you choose one of the Synth Pads, the lower sound is a Synth Bass. It's all very thoughtful. Layered sounds can only have their volume balanced against the main sound, and both layers and splits share their effects with the main sound. Once a sound is altered, it stays that way: you don't have to consciously store it, and the DS330 retains its memory when it iis turned off. This goes for all functions on the instrument.

Control Layout

Though the unit resembles a drum machine, the buttons and pads are arranged quite differently. The dedicated liquid crystal display features a number of 'icons' that light up when you are in different modes, the top line showing the current patch name, and the bottom line the bank and patch number.

Sounds may be fattened by layering them with octave shifted or detuned versions of themselves, but the total polyphony is reduced. There's control over Reverb and Chorus in both Single and Multi modes, while it is possible to select Split or Dual modes only in Single mode. The Drums button selects the drum kits in Single or Multi mode. A useful but un-named button plays a Middle C on the currently selected part — a unique feature which allows the user to audition sounds quickly while editing, without recourse to a keyboard.

In Single mode, there is quite a large number of editable parameters, while in Multi, you are limited to level, chorus and reverb settings. In Multi mode, each part can be given a pan position, level, a different MIDI channel, a transposition, chorus and reverb 'send' level. The Utility button reveals a collection of global functions, such as a general transpose, master tuning, various Sys Ex storage parameters, a selector for Multi or Single Mode, and an LCD contrast control. The Value/Variation buttons are used to change parameter values in one of the edit modes or to select a patch variation, where there is one — this is a Roland GS feature.

The remaining 18 pads, which appear across the bottom of the Dr Synth, are used to select from the 16 banks of sounds and the individual patches within those banks. These are labelled (in pale blue) Piano, Chromatic Percussion, Organ, Guitar, Bass, Orchestra, Ensemble (string sections and choirs). Brass, Reed, Pipe (ie. flute), Synth Lead, Synth Pad, Synth FX, Ethnic, Percussive and SFX, and make patch selection very simple indeed. These buttons are also labelled in grey, for all the world as if they were drum pads — kick, snare, open hi hat, and so on. The reason for this is that when a drum kit is selected in Single mode, the pads will trigger up to 18 drum and special sounds. Additionally, the pads also send out program changes (and notes, when a drum kit is selected) over MIDI.

In Use

Soundwise, the Dr Synth has a lot in common with Roland synths of the recent past, in that the PCM waveforms at the heart of the system closely resemble those found on the recent Sound Canvas range while some appear to be versions of other D-series synths presets. Like the TG100, some patches are made up of more than one waveform, with inevitable polyphonic compromises.

In Multi Mode, the unit behaves like the GM section of a Roland Sound Canvas, which means 128 good quality sounds and 16-part multitimbrality. Variations mean a grand total of 156 sounds, which are also available in Single mode. Each part can be assigned any MIDI channel, with the default setting matching MIDI channels to correspondingly numbered parts. Any voice is assignable to a part, and although drums are factory set on channel/part 10, they can be reassigned to any other channel, and any voice can be reassigned to part 10.

Even without a manual, the DS330 is both simple and intuitive to use, which is just as well since the unit arrived before the manual. This invisibility of the operating system is one of the excellent things about the DS330, though you can't change the basic waveforms of the patches (remember the TG100?), or even name the patches you've edited. Edited sounds can't be used in Multi mode — but then it wouldn't be GM if you could. But if you look at the DS330 as two synths in one — a preset GM/GS module and a basic but programmable, non-multitimbral synth — then you won't be disappointed. I can put up with not being able to name patches (just about!), and since the filter controls mean that a new patch can be made to sound drastically different to the original upon which it is based, the lack of choice on the waveform front is not such a major problem. My test GM sequences were recorded using a Sound Canvas, so they obviously sound perfect on the DS330, barring a couple of slightly different percussion sounds.

Sonically, the Dr Synth is great: the pianos are perhaps a bit tinny, but the strings, basses and synth textures have a lot of depth and playability. Those of you familiar with Roland's atmospheric textures will not be disappointed: Favourites such as Bowed Glass, Fantasy, and Metal Pad put in an appearance, several of which have appeared on previous Roland LA synths while I feel the drum kits have an edge over those found on the TG100 — the TR808 samples in the DS330's drum kits are definitely from the horse's mouth! I should also mention that patches do not get cut off in the middle when a new patch is chosen. In all, the DR330 Dr Synth is a well thought out and extremely musical little instrument.

Further Information
DS330 Dr Synth £339 including VAT.

Roland UK Ltd, (Contact Details).