A new purpose rack-mount drum module from Korg combines onboard 12-bit percussion sounds, a pad-to-MIDI converter and a basic sequencer in a cost effective package.

IMAGINE YOU'RE A drummer or percussionist who wants to make use of electronics alongside, or instead of, acoustic instruments. You aren't satisfied with the inferior sound quality and limited sound range of most drum machines, but you haven't budgeted for a sampler and a pad-to-MIDI converter, and you'd like to do a little sequencing besides. Where do you turn? Quite possibly, to the Korg DRM1 Digital Rhythm Module.

It's a slick-looking single rack-space black box which contains drum pad and other inputs, a sound-generation system based on sampled (ROM) drum set and percussion sounds (23 built in) and a sequencer which records and plays back sequences of these sounds. It also speaks MIDI in both directions and has a stereo mix and eight monophonic audio outputs, four front panel card slots for adding sounds or program/sequence memory, a 2X16 LCD for parameter display, a hand-held remote control and no AC adapter - the power supply is built in. Quite a bit for the money, really.

Sounds, Voices, Programs

THE BUILT-IN SOUNDS of the DRM1 cover the standard range of beat box drum kit and percussion sounds, and all sound pretty good. In fact, this is one of the DRM1's major strengths. They're claimed to be recorded in 12-bit format, and I have no reason to doubt it. To run down the sounds, there are: four snare drums (a "fat" studio/drumbox snare; a tight high-pitched snare; a piercing rimshot; and a deep, solid gated-reverb snare); four bass drums (a tight, punchy acoustic bass; a deep acoustic bass; one which borders on electronic; and a gated-reverb bass); snare side-stick; dosed/open hi-hats; crash/ride cymbals; high/mid/low toms; cowbell; handclaps; high/low/muted congas; and high/low timbales.

I especially liked the assortment of snare and kick sounds, which helps to keep things interesting. The crash cymbal was recorded with a somewhat flat EQ for my taste and lacks brilliance when taken out of the stereo mix, but shows some real sparkle when brought out separately and EQ'd to emphasise the top end. Being ROM-based, none of the sounds are extremely long, but the critical sounds - cymbals, open hi-hat, mid and low toms - have been dealt with generously as far as recording time goes.

In addition to the internal sounds, all four of the front panel card slots can accept ROM cards available for Korg's DDD1 and DDD5 drum machines. Each of these cards may contain as many as eight sounds, adding up to a grand total of 55 sounds on line at once - although as you are about to see, you can't actually get at all of them at the same time.

The sounds are accessed via the 16 drum kit programs. Each program consists of 16 "voices", which can be thought of as cells which each hold a single sound, along with info which controls its response. For example, in program 5, voices 2-4 can all be assigned the hi tom sound, each with a different tuning, while in program 6 these voices may be doing something completely different. (These program voices are distinct from the actual physical voices, or voice channels, that play the sounds, of which there are twelve according to the manual.) To understand voices is to understand programs, so let's have a look at what a single voice can do.

First, each voice gets assigned one sound, either from the internal selection or from any of the cartridges. Arbitrary combinations are possible - thus, while no more than 16 sounds can be accessed from within the confines of a single program, you can create a program using the same sound in all 16 voices if you wish. Next, there are controls over the sound itself: tuning (127 steps covering a range of about an octave), amplitude decay time (15 steps), and overall sound level (15 steps including zero). You can also control where each voice goes (seven positions across the stereo output, or one of the eight individual outs). Each of these - tuning, decay, level and panning if the stereo output routing is used - can be set up separately for modulation by velocity, MIDI note number or an external pot pedal. The number of gradations of modulation sensitivity matches the number of increments in the basic setting (like the 15 for level response), and modulation can be applied in either a positive or negative direction.

I found in most cases that to achieve a full-ranged modulation effect spread evenly across the full range of the modulation source itself, I had to stay in the first 10-20% of the sensitivity range. For example, with note-number modulation of tuning set to 12, a sound plays in a normal chromatic scale tuning from a MIDI keyboard over the maximum one-octave range. The remainder of the sensitivity range up to 127 quickly compresses the modulation to a point where only the lowest and highest tunings are heard. Subtle modulation effects are particularly hard to achieve where modulation sensitivity is set in 15 or fewer steps.

"Programs: The programming scheme allows effortless mapping of voices to MIDI channels, or arranging of multiple sounds across a keyboard."

Rounding out the sound controls is the phase parameter, which imposes a slight time delay on a voice. This produces no audible effect on its own, but when two otherwise identical voices are layered together, the time delay results in a phase-shifted sound, which becomes a flanging effect if one of the voices is also detuned slightly. Sorry - no modulation of the phase setting is available.

Sadly, the tuning system of the DRM1 is not up to the quality of the sounds themselves. Evidently it uses the drop-sample method in which increasing numbers of data words in a sound are skipped over in a periodic way during playback in order to raise the pitch. Sounds are played back without sample-dropping and its attendant distortion only at the very bottom of the tuning range. This distortion is scarcely detectable in snare, kick and related sounds. On cleaner sounds, such as toms and congas, it appears as a ringing not unlike the clock noise heard on samplers at the bottom end of the transpose range. It's more or less subliminal except when one of these sounds is played in isolation. So, for these sounds it works passably well to have the "natural" pitch near the middle of the tuning range.

With cymbals however, any amount of transpose creates very noticeable distortion - beyond the first quarter-tone or so, the sound is only usable as a special effect, and is in sharp contrast to the clarity of the untransposed cymbals. Not surprisingly, the natural pitch of all cymbal sounds is at the bottom of the transpose range, and the factory programs are almost totally devoid of transposed cymbals. The unfortunate outcome is that your ear becomes jaded to these basically very nice but unvarying sounds. Also, the lack of any pitch-bending effect on the DRM1 is probably a result of the way in which it would highlight drop-sample distortion.

Dynamic filtering and reverse playback of sounds, features commonly found on samplers, are not present here, although the DRM1 provides control over voice channel allocation to a degree which puts many samplers to shame and is indeed another of its strong points. As mentioned above, each voice can be directed to any of the eight monophonic outs, in which case it disappears from the stereo mix out. If routed to the stereo mix, however, each voice can be separately set up in poly, mono or exclusive mode. These labels are confusingly reminiscent of MIDI terminology, but in fact accurately describe the way each mode works.

In mono mode, a voice appears only once at any time in the mix. If you play a mono voice twice in rapid succession, the sound of the first hit will be cut off and replaced by the sound of the second. In poly mode, on the other hand, both hits (up to 12, in fact) would be allowed to play at the same time. When applied to cymbals, toms and snare drums (or almost any percussion sound when played rapidly enough) poly mode eliminates one of the most irksome limitations of drum machines. Finally, exclusive mode functions like a mono mode that extends across all voices set up for that mode, such that only one of these voices can play at any one time. This is typically used to cause the closed hi-hat to cut off the open hi-hat, or the muted conga to cut off one of the other conga sounds.

Voice Triggering

HAVING SET UP all other aspects of the voices, the creation of a program is completed by setting up the details of voice triggering. The MIDI side of this is simple and powerful: for each of the 16 voices in a program pick a MIDI channel and a base note number, then set an upper limit note number which can be anywhere from 0-12 semitones above the base note. That's it. The upper limit determines how many different note numbers (or keys) will trigger that voice. For example, at the maximum setting of 12, any note between the base setting of C3 (MIDI Note 60) and the upper limit of C4 (MIDI Note 72) could trigger a snare sound, with different note numbers creating different modulations.

"Converter: The pad/voice system is particularly flexible in that MIDI channel and note number for each voice are generated from the pads."

This programming scheme allows effortless mapping of voices to different MIDI channels, or arranging of multiple sounds in any desired fashion across a single keyboard, including zone overlaps and very deep layers, and almost any sort of velocity or positional crossfade you can imagine. Remember that note number is a variable modulation source for tuning as well as other parameters, not a hardwired semitone stepper as is the case with most samplers. Across a dozen note numbers or keys, a sound can play at the same pitch or at very gradually increasing or decreasing pitches, as well as the more familiar chromatic steps. By layering four voices onto the same set of keys, it can even do all of these things at once.

And pads. For each voice you can select a pad input which will trigger that voice. Each assignment of a voice to a pad is either as a main voice, which the pad will normally trigger, or as a sub voice, which the pad will trigger instead of the main voice when footswitch or velocity-switch options are activated. Optionally you can have many main and sub voices per pad, subject to the limit of a total of 16 voices for each program. A modest two main voices on a pad allows for flanged sounds (as described above), layered sounds and velocity crossfading. Naturally, none of these effects can be used through a single mono output. There are only seven pad inputs, so you may find yourself using these multiple assignments more than you might expect. Voices can also be assigned to no pad - so that they are only triggerable over MIDI - or else just to get them out of the way.

Overall, the pad/voice system of the DRM1 is flexible and powerful, particularly in that the MIDI channel and note number settings for each voice are also used to generate MIDI output from pads - including layering and main/sub switching in response to footswitch or velocity (although other voice modulations are not echoed over MIDI). The presentation of pad assignment as a parameter of each voice, however, rather than the other way around (voices assigned as a parameter of each pad), although logical, can be confusing when program editing. There's no way to see at a glance which voices are assigned to a given pad, or which voices are not assigned to any pad. You see this information for but one voice at a time, and despite the size of the display, you must switch to a different screen to confirm which sound that voice is playing. Because there is no fixed number of voices that can be assigned to a pad, there is no way that the DRM1 can allow you to select a voice for editing by merely hitting its pad - you have to select that voice by hand. But if the same sound is assigned to multiple voices in the program, you will also have to step through all the pad assignments to make sure you're working on the voice you wanted. The possibilities are great, but it is neither quick nor easy to set up.

The Sequencer

THE DRM1'S SEQUENCER is aimed fairly and squarely at the drummer, not the programmer/composer. Although called patterns, sequences are clearly meant to be dealt with as complete pieces - there are only 16 with no provision for chaining them together. Recording takes place on a single track, as with drum machines, with input accepted from either pads or MIDI. Events are time-corrected as they are recorded, with resolution selectable in 10 steps from quarter-notes to 1/48th of a quarter-note. The position counter counts up in absolute quarter-notes, not bars and beats. Apart from copying an entire sequence, all editing is done while the sequencer is rolling. When you get to the desired point, you hit Yes to start erasing selected voices or deleting quarter-note pieces of the sequence, then hit No switch to stop. Mid-sequence tempo changes are inserted and erased in a similar way. To make this quicker, it is possible to hand-cue to the desired point. Unfortunately, there is no visible indication that the edit function is active.

While recording, an optional metronome counts off quarter-notes, with optional accenting every 2-8 beats. By stepping the position counter down into the negative range, a pre-recorded count-in of arbitrary length is set up - an irritation is that it must be re-specified each time recording is started. A separate tap-tempo footswitch input lets you control tempo while record/playback is in progress by tapping it out with your foot, and can also be used to tick off the record count-in manually, establishing the initial tempo in the process.

Interestingly, the sequencer records voice events, not sounds. If you select a new program after recording a sequence, it'll sound different when you play it back. While this does permit you to retune, replace and otherwise modify the sounds of a sequence through program edits, it also means that sequences cannot use more than 16 sounds at once, and that you cannot sequence one set of 16 sounds while playing live from a different set of 16. This being the case, it would be nice to be able to record program changes into the sequence but, alas, this is not possible. The sequencer refuses to record or play back MIDI events corresponding to keys on which nothing is mapped in the current program, but you can create a dummy voice for this purpose by mapping in a sound from an empty card slot, and thereby be able to sequence external MIDI gear without doubling the external part with the DRM1's own sounds.

"Sequencer: The sequencer is not suited to serious recording or composing, but to rhythm pattern experimentation or rhythm accompaniment."

Although it's not mentioned in the manual, the capacity of the sequencer seems to be about 2300 events - not a generous allocation, although patterns can be looped. Overall, the sequencer is not well-suited to serious recording or composing use, but it is just fine for rhythm pattern experimentation, rhythm accompaniment, or giving the drummer a chance to get out from behind the pads and do something else for a spell.

Other Controls

SENSITIVITY, TRIGGER THRESHOLD level and trigger inhibit time are adjustable per pad input - the DRM1 is compatible not only with any type of pad, but with audio signal triggering as well. Each pad can also be set to switch between its main and sub voices at one of 16 velocity thresholds. While useful, this would have been much more so had it been included as part of each program instead of being a fixed property of each pad, especially as you'll have to program around it in all the presets where you don't want it. As a bizarre twist, each pad can also be programmed for sequencer start/stop, continue/stop, or pattern up or down - this is strictly for those who are satisfied with actually playing six pads or less, although most drum pads are rather expensive to be dedicated to what is essentially a footswitch function.

The footswitch can do a number of useful things: sequencer start/stop, continue/stop, pattern up or down, program up or down, momentary or alternating main/sub switching on a single pad (although not all pads at once), triggering of a single voice (not pad) at a fixed velocity, and execution of sequencer real-time edits. It's really a shame there aren't two or more footswitch inputs, although as with pad inputs, there simply isn't any empty space left on the back panel for more.

The pot pedal can be used either as a tempo control for the sequencer or as a modulation control for one voice (don't be fooled by the ability to set up pedal control in each voice - only one voice gets it at a time). It affects only live voices, not sequenced ones, a helpful touch when playing along with a sequence. Finally, there is Omni on/off and base channel select for MIDI program changes, transmit and receive enable/disable for program changes, notes messages, system real time (start, stop, clock), system common (song position pointer and song select), as well as a global gate time adjustment of 0-512msec for transmitted note messages, obscurely referred to as Distance Time. Patterns and programs can be named, and can be saved and loaded using MIDI system exclusive transfer, or to and from an optional RAM card which can be plugged into card slot 1 - but which the manual claims is "not available". Puzzling.

Verdict

IN SPITE OF its flexible programs and voice modulation capabilities, the DRM1 has many imperfections. The major ones have been discussed - to this I will add, at risk of repetition, that the editing screens are not structured as well as they could be, and this brings a substantial fatigue factor into the programming process. Along these lines, I really don't consider a hand-held remote control with three-dozen tiny buttons to be the ideal front panel, especially since you really can't read the display from across a big room, and if the remote is lost or crunched (which could happen on the road), there's no way to control the DRM1. However, it is one of the keys to the size and low cost of the system though I'm sure that many users would have appreciated the inclusion of a large two-digit LED readout for drum set program numbers.

All in all, though, the DRM1 is a fairly full-featured unit which sounds good, is reasonably priced and is particularly well pitched for the drummer or percussionist taking those first steps into the realm of MIDI and electronics.

Price £749 including VAT

More from Korg UK, (Contact Details).