GEM S2R

GEM have upgraded, expanded and enhanced their S2 keyboard synth and crammed it into a 3U module. Devoted S2 owner Mike Simmons discovers what he might be missing out on...

Even in this ever‑changing world, there are some events you can still depend upon; Christmas will bring a new jumper, the next door neighbour's home made wine will taste awful, and, sooner or later, the manufacturer of any half‑decent synth will start thinking about bringing it out as a rackmounting module. And so it is with the GEM S‑Series synthesizer which, since its launch in the latter part of 1992, has attracted considerable attention amongst those who know a good thing when they hear it.

General Music were not amongst the major league players in the synth market when the S2 was launched, and the fact that they should choose to move into an already competitive marketplace came as something of a surprise to many. True enough, they had a good reputation as manufacturers of home organs and keyboards, but the S2 was no tentative first step into the world of the budget synthesizer. The S2 and its 76‑key big brother, the S3, are full‑blown professional S+S workstations with specifications to rival the leading models from any of the more established manufacturers. With its 250,000‑event sequencer, two powerful signal processors and an operating system that's an absolute joy to use, the S2 was clearly a serious contender, offering, as it does, 16‑part multitimbrality and 16‑note polyphony. Furthermore, in common with a small number of other leading edge synths, the S2 is able to load samples into its internal RAM (via the optional Sample Translator software), where they can be used to augment the 208 existing waveforms.

Besides being very easy to work with, what has attracted many musicians to the S2 — including this writer — is the open‑endedness of its design and the startling clarity of the sounds. I originally found some of the machine's on‑board sounds a little 'safe' for my liking — more imitative than creative — but the ease with which it is possible to create new sounds (especially using Sample Translator) negates this criticism. The machine is also able to download new software enhancements, as they become available, via its integral disk drive.

Splendid though the S2 most certainly is, the S2R is rather more than just the same device repackaged in a new box. The 3U rack‑mounting module boasts a huge range of improvements and innovations which push it very firmly to the forefront of synthesizer technology. Given that the S2 was comprehensively reviewed by Martin Russ in the August '92 edition of SOS, I shall try to concentrate on the new developments included in the S2R and leave it to the interested reader to refer back to the original review for further information on the S2 itself.

The Package

The S2R's front panel is dominated by the internal disk drive, a particularly sizeable LCD, and nearly 50 buttons. The back panel offers two MIDI Ins, two Outs and one Thru, plus six audio outs; the first two audio outs are arranged as a stereo pair, while the other four can be configured either as further stereo pairs, as separate mono outs, or as a combination of the two. There are three more quarter‑inch jack sockets on the back panel, one for a dedicated volume pedal and two for assignable foot switches. One intriguing point is that on the back panel of the device there is a small knock‑out section of panel, measuring about two inches by half an inch, which looks as though it might, at some time in the future, accommodate a connector of some kind. What its purpose might be I can only guess — maybe a SCSI port upgrade?

Enhancements

Like the S2, the S2R is 16‑part multitimbral but now offers up to 32‑note polyphony. This is isn't the massive improvement it may seem, since the increased capacity has been achieved simply by separating the existing 16 oscillator pairs. What this means is that to make use of the extended polyphony, you have to use some single‑oscillator sounds, which are obviously less flexible than the dual‑oscillator sounds. You can also create dual‑oscillator sounds using two different waveforms, which wasn't possible in the S2 except for a few specific patches. Its introduction on the S2R is very welcome and greatly adds to the creative capability of the machine.

There are now over 500 sounds in ROM, as opposed to the S2's 350, and an upgraded version of the Sample Translator software is also installed in ROM rather than being available as an option. This allows the machine to read samples in a range of formats, including Sound Designer and Avalon, over MIDI as a sample dump, or direct from Akai S1000 disks. This obviously gives the user access to a huge range of sounds which can then be saved into memory and, ultimately, to disk as part of the machine's Sound Library. As it stands, there are 2Mb of sample RAM, though a future expansion is not out of the question.

...to get the most out of this machine, you have to be prepared to indulge in a spot of editing, as the presets only scratch the surface of what is possible.

It is perhaps worth looking at the way in which the machine stores sound data, since in this respect it is rather different from most other synths currently on the market. As mentioned, there are over 500 sounds in ROM and these are arranged to conform to MIDI 1.0, Version 4.2. What this means in practice is that the machine has the ability to access multiple banks of 128 sounds. Each of these memory locations is designated a Program number and a Bank number, and these can be called up over MIDI by a sequencer using Program Change and Bank Change commands. The sounds are loosely arranged into families — piano, organ, percussion and so forth — and the whole lot is configured in what General Music call a Sound Library.

If the S2 has one shortcoming, it's the inability to hold user data in memory once the machine has been switched off — there's no patch memory as we have come to know it. New sounds can be created, samples imported, footswitches assigned to specific functions, the whole setup configured to the user's specific requirements, but unless all that data is saved to disk, it will all be lost the moment the machine is switched off. In fact it's just like working with a computer with no hard drive — which I suppose is exactly what it is! General Music have now addressed this 'feature' by making it possible to install Static RAM chips on the S2R, which will hold data on power down. These were not available at the time of review, so for now we'll have to continue booting up at the start of every session.

What also makes the S2 a little frustrating to work with is its inability to distinguish between user sounds which had been saved to disk and ROM sounds which, though not physically on disk, appeared to be so. In other words, the user (all right, let's not mince words, this user) would create a few stunning sounds, save them into the machine's Sound Library and then save to disk. When attempting to work with those sounds at some later date, scanning the disk would show the user's sounds plus all the existing ROM sounds, which made locating them little short of a nightmare. Happily, this is not the case with the S2R which, more logically, only displays the sounds actually saved to disk. Furthermore, a Mask function now makes it possible to differentiate between the different types of sounds stored in the machine's memory. You can now choose between a display of all the sounds, the sounds that have been created using waveforms stored in ROM, or those created using waveforms imported via Sample Translator, or a combination of any of the above.

Besides saving the Sound Library to disk, it's also possible to save data relating to the Effects Library. The S2R boasts two signal processors, one dedicated to a range of reverbs, the other to a selection of delays, flangers, choruses and so on. All of these can be edited, and most of the parameters that a musician might need to work with are available, though the list is not entirely comprehensive. The editable parameters available on the delays, for example, simply feature Level, Feedback Gain, Delay and Filter. I wouldn't like to think of the on‑board effects as being the only ones available to the musician, but rather as supplements to a dedicated effects unit. Having said that, the effects that are available are all eminently usable, though there is a bypass button for the benefit of those hell bent on using external processors.

It is possible to configure the machine to work in Single, Layer, Split or Multi Modes. These are all fairly self explanatory and the sequencer user is most likely to be interested in the Multi mode. This offers 16 voices, or 'tracks', each of which can be programmed through six pages of parameters. These range from the familiar volume, transpose, delay, footswitch response, effect assignment, further splits and so on, through to a variety of Pan options, output routing and two pages of MIDI filters. There are also a further three pages to manage the overall performance of the machine, enabling the user to select which two effects will be available to the tracks; the volume of those effects; the sensitivity of the pitchbend; the footswitch assignment, and so on. There is also a new 'Rotary' speaker simulator function on the S2R which simply has two extremes of modulation — fast and slow — and the rotor speed ramps up and down between speed settings, just like the real thing.

The third of the performance edit pages presents a graphic display of seven consecutive tracks and allows the user to edit a range of parameters including volume, attack, release, filter 1, filter 2 and Pan. On the S2 this feature is complemented by a row of faders which allow a considerable degree of real‑time control over the device. The S2R is sadly devoid of these faders and the parameters in this case must be edited either from the numeric keypad or via a large rotary dial. It's worth mentioning the care with which the dial has been designed; it's accompanied by two buttons, one marked 'Enter' the other 'Exit', and these either confirm or discount any change in a parameter that has been made using the rotary dial. The three controls are nested together in a manner which makes their use entirely intuitive. In a very short time, the user becomes totally familiar with their feel and simply concentrates on what's on the LCD.

Sequencer

I am no great fan of hardware sequencers but, having spent a fortnight in a shack up a mountain with a GEM S2 and a dead Atari (you do your thing and I'll do mine, OK?), I came to realise how practical the S2's on board sequencer really is. Offering 16 tracks and the ability to display and edit just about everything on an event‑by‑event basis, it's possible to achieve some quite sophisticated results. Inevitably, the sequencer on the S2R is even better. It's the editing facilities that tend to make or break any sequencer, and those on the S2R are really quite comprehensive. It's now possible to erase, quantise and copy events not only between specified locator points as previously, but also within a particular note range, or both. While it was previously possible to erase a wide variety of events, the S2R user is now also able to delete duplicated notes — a very welcome addition. It's now also possible to copy tracks from one song to another.

I am no great fan of hardware sequencers but, having spent a fortnight in a shack up a mountain with GEM S2 and a dead Atari (you do your thing and I'll do mine, OK?), I came to realise how practical the S2's on board sequencer really is.

As you might expect, songs can be loaded and saved to disk, but the S2R also has the ability to load and save Standard MIDI Files. The S2 largely (but not entirely) conformed to General MIDI standards, but it is now possible to switch in a utility option on the S2R so that it conforms totally.

For the most part, disks written with an S2 can be read by the S2R. The only real limitation to this is attributable to the increased number of ROM sounds now within the S2R. Since users of the S2 are likely to have saved their own sounds immediately after the ROM sounds, it follows that booting up an S2R with an S2 disk is likely to overwrite some of the new on‑board offerings. This is not an insurmountable problem, of course, but simply means that the old user‑created sounds will have to be loaded one at a time into the S2R, then the whole new Sound Library saved to disk. S2 owners are likely to have more of a problem loading S2R disks, however, but General Music have produced software to facilitate the process. It should be noted though, that S2R Sound Libraries cannot be read by the S2.

Summary

Anyone looking for a module capable of producing high‑quality sounds would be doing themselves a grave disservice if they didn't check out the S2R. The sounds themselves are of a very high quality — clean, rich and very usable. As mentioned earlier, I felt that the original S2 tended to favour the more imitative sounds and, while many of these are extraordinarily convincing, I felt a little let down by the lack of more imaginative abstract patches. The introduction of some more textural offerings on the S2R helps redress this balance, and I also feel they help demonstrate more effectively just what the machine is capable of. Having said that, to get the most out of this machine, you have to be prepared to indulge in a spot of editing, as the presets only scratch the surface of what is possible. For those determined to avoid editing at all costs, more third‑party patches for the S2 are now becoming available, and the enhanced features on the S2R are only likely to stimulate demand further.

Working with this machine is a delight, and the comprehensive nature of its operating system means that it's just as much at home on the stage as it is in the studio. To miss out on the S2R just because of the connotations brought to mind by the General Music badge would be to do both it and yourself a grave disservice. Try one out and you could be even more surprised than the Teddy bears who had their picks nicked.