Roland D-110

This is an article culled from archive material, circa 1997.

Described as being "worth its weight in gold" when it first appeared around 1988, the 1U-rackmount D-110 was the first serious multitimbral sample-based machine on the market. Its heritage draws on the phenomenally successful (but bitimbral) Roland D-50 synthesiser, although it is more closely related to the MT-32, a table-top unit primarily aimed at the home market as an "expander" (as they used to be called).

The D-110 uses the same winning formula as the D-50: a layered combination of emulated analogue synthesiser with playback of ROM-based samples, usually short attack transients. Unfortunately, it lacks the D-50's true strengths (multiple onboard effects, parametric EQ and multiple-LFO modulation), resulting in a rather thin and static sound. That aside, the D-110 was a very capable box for its time: 32 voices, dynamic voice allocation, voice reserve, multiple polyphonic outputs. The MIDI response is also very tight.

The system exclusive implementation is clean and elegant, without the drawbacks of the D-50, although the packet-based transfer protocol causes problems for generic librarians. The D-110 was also the first of Roland's multitimbral machines to feature multiple edit buffers, a design which I personally like (exceptions, such as the Waldorf Micro-Wave, aside) but which induces violent fits in others. The patch heirarchy is rather convoluted, due mostly to the scheme used to select 256 tones using 128 MIDI program numbers via a table of indirection "timbres".

Despite the rather weak voice architecture, the D-110 is a capable little unit, ideal for "background" work in sequencing applications, although the front-panel programming interface is typical Roland issue, as is the manual.