Roland's latest workstation incorporates features from their popular "MC" sequencers and "S"-series samplers - but is "workstation" the best description?

QUESTION: WHAT DO you get when you cross a keyboard with a sequencer and a sampler? Answer: a workstation. At least, this is Roland's latest angle on the workstation ethos. The company have already produced a synth-based workstation in the form of the D20, which ended up being undercut by the company's own D10 synth (essentially a sequencer-less version of the D20), which says a lot for the popularity of the workstation approach. Similarly, Korg's M1 Music Workstation has sold on the strength of its impressive synthesis abilities rather than any real claims to workstation fame.

Against this background, how will the W30 fare? Well, for one thing it has a much more powerful and flexible sequencer built into it, which is based on the company's existing MC300/MC500 hardware sequencers. For another thing, Roland claim that a sampler offers greater flexibility than a synth as the sonic basis of a workstation, as it's easier to sample synth sounds than it is to synthesise acoustic sounds. Broadly speaking this is true, though with synths like the M1 and Ensoniq's new VFX offering samples as the basis of synthesis (and the M1 even allowing you to access new samples via plug-in ROM cards), there's a gradual blurring of distinctions going on. However, it's worth pointing out that the W30 includes digital filtering and resonance, with the filter cutoff point controllable from a separate eight-stage filter envelope. If you start with a sampled sawtooth waveform, haven't you effectively got a traditionally-styled synth at your disposal? At the same time, as we all know, there are things you can do with a sampler that you can't do with a synth. Roland could well have a point after all.

Overview

AS WITH ROLAND'S S50, S550 and S330 samplers (to which the W30 is intimately related), the new sampler loads its operating system off a 3.5" DSDD System disk (v1.00 as of review) when you switch on the power; if the System disk isn't in the drive, you'll be prompted via the instrument's LCD screen to put it there. It takes 30 seconds for the W30 to become functional, after which time you shouldn't remove the System disk unless you want to load up existing sequences and samples. Why? Because the W30 needs to refer to the System disk whenever you select the sampling page or a sequence- or sample-editing routine (known as a Utility), as each routine overwrites the previously-selected one in memory. This is par for the course with Roland's samplers, but although each load only takes a few seconds it's still an irritation. On the other hand, this approach has the benefit of making the W30 open-ended, as new Utilities can easily be added in any future updates.

The W30 has been given a five-octave attack velocity and channel aftertouch keyboard, which has a shallow travel and a very comfortable combination of resistance and bounciness.

The new workstation has three modes: Performance, Sequencer and Sound. Each of these is accessed by pressing a dedicated front-panel button. Performance mode is the most straightforward (I'll come to the others later), and the one you should be in if you want to play any one of the Patches without regard to sequencing. However, this being an integrated system, you can play the current sequence from within this mode, and even fast forward or rewind to any position in the sequence (the sequence's status and position are indicated in the top right-hand corner of the LCD window).

You can scroll through a list of the Patches, and set an output level (0-127), octave transposition (+/-), audio output assignment (1-8 or Tone) and bend range (0-12 in semitone steps) for each Patch. Additionally, in this mode you can set the master tuning along with System configurations isuch as the routing of internal and MIDI data (which I'll discuss later under "sequencing"), MIDI sync transmission on/off and functions of the W30's footswitch and pedal inputs. These configurations can be saved to the System disk ..c and automatically loaded when you power up the W30.

The new workstation's front panel has a sleek, uncluttered appearance which is fast becoming a Roland style. The low-profile buttons divide into four functions: mode selection, sequencer operation, page selection and numeric/character keypad, while the single slider handles volume.

As with Roland's A50 and A80 MIDI controller keyboards, the W30 employs a generous 8X40-character backlit LCD screen with adjustable contrast. I found it both easy to read and easy on the eyes, and of course for operating purposes it's a massive improvement on the measly one- or two-line LCD windows we've had to put up with for so long. Finding your way around the display and editing the parameters are accomplished by two infinite rotary dials labelled "cursor" and "value" respectively, while you can also type in numerical parameter values directly using the keypad (a blessing when you're dealing with the large numbers that inevitably crop up with sampling).

Also, as on the A50/A80, Roland have adopted a system of five "soft" buttons, located immediately below the screen, which take on different functions depending on what page or window you're on. These functions are indicated along the bottom row of the screen, making it easy to find your way around the W30's features. Successive presses of the Exit button to the right of the soft buttons takes you back "up" through the pages until you reach the initial page of the current mode.

Easy though this might be, the W30 has 62 LCD "pages" (not to mention countless more windows within these pages), and moving around them all can get a bit laborious. Wouldn't it be nice if you could move directly to certain often-used pages? It would, and you can. When you press the User button, a scrollable list of the 62 pages appears in the LCD screen and each of the five soft buttons has a page number associated with it; pressing the relevant soft button takes you directly to its page. You can assign any one of the 62 pages to each soft button, while the currently-selected page can be jumped to by pressing the Enter button. Assigning five of your most commonly used pages to the soft buttons can be a great timesaver.

On the W30's rear panel you'll find MIDI In, Out and Thru sockets; mono sample input with a gain knob for adjusting input level; one footswitch and one footpedal input; eight individual audio outputs (with output one alternatively functioning as a mono mix output if required); a mono headphone socket which outputs the same signal as individual output one; a contrast adjust knob for the front-panel LCD window; and a (covered) SCSI port for connecting up a hard disk and/or Roland's CD5 CD ROM player (reviewed in MT, May 89). To make the SCSI ("scuzzy") connection you'll need the optional KW30 upgrade kit, which unfortunately will set you back a further £135. What are the advantages of getting connected? Well, a hard disk provides far greater storage capacity and greater convenience than floppy disk, plus faster loading times, while the CD5 CD ROM player offers you ready access to a vast (and still growing) library of samples. These aren't cheap options, but at least the upgrade path is there, should you want to take it at some stage.

Sampling

BEFORE YOU EVEN consider sampling your own sounds, the W30 can take advantage of the vast library of samples which already exists for Roland's S50, S550 and S330 samplers. For reviewing purposes Roland provided a good selection of S50 and S550/S330 disks to try out, and I can report that not only did they all load without a hitch, they demonstrated that if you buy a W30 you'll have potential access to a healthy variety of instrumental sounds, many of which are of very high quality. Incidentally, W30 sound data can be saved in S50 and S330/S550 disk formats for use in these samplers, though some features of the W30 will be lost in the translation.

But there's another reason why you don't have to start sampling straight away. In addition to its 512K word of RAM sample memory, the W30 has 512K of ROM sample memory containing some 70 samples. These offer a healthy number and range of drum and percussion samples, plus a variety of bass sounds (too many, perhaps), electric piano and a number of sounds which obviously originated in the synthesised domain (strings, brass, sawtooth, pulse). One of these Tone samples can be selected as part of a Tone (that is, with associated filter, LFO and amplitude parameters), and Tones in turn can be mapped across the keyboard to form a Patch (again with associated parameters such as key mode, velocity-switch threshold, bend range and aftertouch response).

In addition to the operating software, the System disk contains data for 128 Tones and 32 Patches specifically crafted for use with the internal ROM Tones (samples); these are listed on a chart which accompanies the manual. When you power up the W30, the first 64 Tones are loaded into Tones 33-96, and the first 16 Patches into the 16 internal Patch memories, along with Function data defining the eight Parts and MIDI data defining the MIDI receive options for each Part. The other Tones and Patches can subsequently be loaded individually into the W30. In this way you have a library of sampled sounds at your disposal straight from power-up. Incidentally, you also get three Sound and Song disks with the W30, which not only show off its abilities via pre-recorded sequences but also give you some more samples to start out with (in RAM, this time), among them acoustic piano and acoustic guitar.

The other advantage of the ROM approach is that you can use the ROM sounds to give yourself, say, a drum, bass, electric piano and synth strings grouping and still have the entire RAM sample memory free for your own samples. On the other hand, if you end up tiring of the ROM samples, or don't have any interest in them in the first place, you might wish that that 512Kword of ROM had been RAM instead.

The W30's sampler section (minus the ROM side of things) is essentially the company's S330 sampler in a different box, and of course given the benefit of a large, clear and informative LCD screen. The 5l2Kword of RAM memory is divided into two Banks, giving you 2X7.2 seconds of sample time at the maximum sample rate of 30kHz (you can't sample across the Bank boundaries). The only other sample rate is 15kHz (for 2X14.4 seconds), which might seem like a shortcoming when you're trying to offset sample quality against sample time, but the difference in quality between the two sample rates is less you might imagine, and quite often you can get away with the lower rate unless you're obsessed with cleanness. Of course, samples of both rates can be combined.

The RAM memory allows you to create up to 32 Tones. One limitation to watch out for is the fact that samples have to be in multiples of 0.4 seconds (30kHz) or 0.8 seconds (15kHz), which can be wasteful on space. Although sampling resolution is 12-bit, Roland's use of fixed-rate sample playback and differential interpolation ensures a quality of sound which is closer to 16-bit.

Sampling is straightforward: you select sample number, Bank, sample rate, sample time, original key and, if necessary, a pre-trigger time of up to 100 milliseconds (invaluable for catching -sharp attacks). Then you set a threshold level with reference to the displayed input signal level and select auto, manual or previous sampling as appropriate.

If required, you can then loop the sample by setting loop and end points, and call auto-looping and smoothing into play to help you out - though in practice these are no instant panacea for any looping problems you might be having.

"Resonance can be used in conjunction with the filter-cutoff off to create effects ranging from acid-type basslines to spacy synth sounds of the old school."

The resulting sample can then be called up as the source for a Tone. Here you can apply LFO control to the TVF and TVA (Time Variant Filter and Amplifier). In the TVF and TVA sections the Tone can be shaped by envelopes having up to eight stages, with level and rate settable for each stage. These envelopes control filter-cutoff and amplitude respectively, with further controls governing such features as envelope depth and key follow. Also significant in the TVF section is the resonance control, which can be used in conjunction with the filter-cutoff envelope to create a large variety of effects ranging from acid-type basslines to floating, spacy synth sounds of the old school.

Once you've defined your Tones, you can map any combination of them (from both ROM and RAM) across the keyboard. The way in which the W30 allows you to do this is particularly neat: a keyboard comes up on the display, you select a Tone and then play all the notes that you want that Tone to appear on. Then select another Tone and do the same thing, and so on. You can have up to two Tones on a key, and decide whether you want them in unison with detune, velocity-switched, velocity-crossfaded or velocity-mixed (in the latter case, a mix-ratio parameter allows you to set the velocity curve of the second Tone). Incidentally, you can select a global audio output assignment for a Patch, or select Tone output, in which case the output assignment of each Tone overrides the global output - in this way you can, for instance, route selected drum voices in a Patch through their own outputs for separate processing.

Selecting the Part Set page allows you to assign eight Patches to an eight-Part multitimbral configuration which can be utilised from the W30's sequencer or via MIDI input. Each Part has its own Patch, MIDI channel, audio output (global or Tone) and volume level assignments. You can select mix or multi audio output (mix transmits all Parts from output one, multi preserves the individual assignments), and choose dynamic or fixed voice assignment across the individual outs (in the latter instance, you select one of 22 voicing configurations).

Sequencing

JUST AS THE sampler section can load in existing Roland sample disks, so the sequencer section can load in sequences created on the company's MC300 and MC500 sequencers using MRC500, MRC300 and Super-MRC software, as well as song data from the S50, S550 and S330 created using SYS503, SYS553 and SYS33 software. You can also store W30 songs in the Super-MRC format for performance on MC300 and MC500.

However, the W30's sequencer doubles the Super-MRC software's eight "phrase" tracks to 16, and substitutes a tempo track for the former's rhythm track. Although each phrase track can store multiple MIDI channels, unlike the MC500's two independent MIDI Outs (which effectively give 32 MIDI channels) the W30 only has one MIDI Out.

The onboard sequencer has a storage capacity of 15,000 events, which can be divided between up to 20 songs, and a maximum resolution of 96 clock pulses per crotchet. To record into a track with a specific sound you must set the keyboard channel to that of the Part which contains the relevant Patch.

Working with the sequencer at real-time record level is not unlike recording with a multitrack tape recorder, particularly as the W30 has dedicated start/stop, record, fast forward, rewind, return-to-zero and jump-to-end buttons and a Locate facility which allows you to set record start and end points and ten locate points which can be jumped to when you call up the Locate window. If you want to start recording from any position within your intended Song, you first need to define the intended length of the Song before you start recording, by Inserting the relevant number of blank bars into one of the 16 tracks (a quick and easy process).

The W30 allows you to record and edit note, controller, patch-change, pitchbend, tune request, SysEx (up to 500 bytes at a time), and channel and poly aftertouch data: although the W30 itself doesn't generate or respond to poly aftertouch, it can be recorded from and played back to external MIDI instruments. You can quantise on record or after record, to most values from a minim to a 64th-note, and select a constant offset value (up to +/-100 in clockpulse steps), which can give a pushed or laid-back feel to a track, or a more deliberate delay. While you can mute any combination of tracks, you can't record mute settings during the course of each track; this is something I would like to see Roland add in any software update for the W30.

The sequencer provides you with various methods of recording: normal (you just press the Start/Stop button), key on (received MIDI note(s) instigate recording, and are themselves recorded), automatic punch in/out, auto punch in, auto punch out, manual punch in/out, and loop. The auto and loop positions are determined by the record start and stop points you set under Locate; loop recording is of the overdub variety, drum-machine style, though there doesn't appear to be any way of selectively erasing different passes through the loop.

The W30 has no dedicated metronome sound; instead you specify MIDI pitch, velocity and channel independently for upbeat and downbeat metronome strikes. The channel (1-16) can be directed internally or out via MIDI, so you've got plenty of choices for your metronome sound.

Micro-level editing is of the interpreted MIDI event-list kind, with one track at a time displayed; for clarity's sake you can be selective about what information you want displayed. It's in this mode that you can call up the Tempo track and input your tempo changes at any position in the track. Although you can't Play the sequencer in Micro mode, individual notes can be output as you scroll through them.

Higher-level track editing provides a great deal of flexibility. Erase, Copy, Quantise, Extract, Transpose, Change Velocity, Change MIDI Channel, Change Gate Time and Shift Clock all allow you to work on a selected MIDI channel and on any group of bars within a track: some of them even allow you to work on selected MIDI data and selected note ranges. Insert Measure and Delete Measure are sectional by their very nature, while Merge and Exchange operate globally on a selected pair of tracks.

Recording duets is an easy matter, as the sequencer can record the W30's keyboard input and MIDI input at the same time. As each track can record on multiple MIDI channels, you can play one Patch while your partner in crime plays another. You also have great flexibility in deciding where W30 and MIDI inputs are to be routed, with a soft MIDI Thru function allowing a MIDI wind controller, say, to be routed through to MIDI instruments hanging off the W30.

By setting TX Sync to on, you can use the W30 as a master for slaving another sequencer or a drum machine. The W30 also allows you to record into its onboard sequencer while it's synced to incoming MIDI clocks, so that, for instance, you can record parts while the sequencer is slaved off tape via a stand-alone SMPTE/MIDI unit. The W30's sequencer can respond to the usual array of MIDI Start, Stop, Continue, clock and Song Position Pointer sync data, plus MIDI Song Select for calling up one of its 20 onboard Songs. Incidentally, if you want to preserve a Song's tempo changes while slaved to tape, you'll need to be able to recreate those tempo changes onboard a SMPTE/MIDI unit.

Verdict

IT'S ALL TOO easy for seasoned reviewers (and seasoned readers, for that matter) to become blasé about instruments which seemingly recycle existing instruments in new packaging. However, the W30 has a lot to recommend it, not least in the sheer convenience of having sequencer and sampler integrated into a single system. In this sense it's greater than the sum of its parts; it also happens to be attractively priced in relation to those parts.

But what about the W30's claims to workstation status? Can you really record and play a complete song with the W30 alone? Well, here we come to the hoary old question of what "complete" means in this context. With its 16-voice polyphony (less if you're layering Tones), you're much more likely to ruin out of voices before you run out of sequencer tracks or memory, while the sampling section's fixed memory imposes its own limitations (it's worth bearing in mind that Akai's S950 and Ensoniq's EPS samplers both offer memory expansion up to several megabytes).

On the subject of sampler plus sequencer vs. synth plus sequencer, I would broadly agree with Roland's contention that the former offers more flexibility (particularly as the W30 has synthesis facilities too), while on a practical note it's a much cheaper proposition to add on a synth expander to a sampler than vice versa.

But the W30 really scores where other workstations fail down, namely in the sequencing department. By providing the W30 with such a well-specified sequencer, Roland have ensured that if and when you run out of voices on the W30 itself when sequencing, you can readily direct further sequencer tracks to external synth and sampler modules. Think of a workstation as a "control centre" rather than a stand-alone instrument which will provide you with every facility you need to make music, and you'll get a much more appropriate perspective on the W30. Of course it doesn't have all the sophistication and flexibility of a computer-based setup, where you can use much more powerful sequencers together with patch editors, composition programs, notation software, generic patch storage software and the like. If you've already established a computer-based MIDI system, or are attracted in that direction, then the W30 is unlikely to tempt you. But if you want to steer clear of all that stuff and keep things relatively straightforward but still have a system which is well-designed, flexible, powerful and easy-to-use, the W30 fits the bill very nicely, thank you.

Prices £1599; £135 extra for the SCSI option. Both prices include VAT.

(Contact Details)