From velhart@epas.utoronto.ca Sat Apr 9 22:45:17 1994 Date: 7 Apr 1994 04:07:39 GMT From: Nicholas Velharticky Newsgroups: rec.music.makers.synth Subject: Info for Synth FAQ (Roland Alpha Junos, D's, and U's) As I recall, someone was looking for information to include in a Synth FAQ. Here's my contribution (which is amenable to correction if anybody finds something in error). Sorry that the two reviews aren't in the same format. The Alpha Juno review was one I originally sent to a Synth FAQ compiled by Clarence Din; perhaps he still has a copy of that FAQ that he'd want to post to the group for revision/updating or handing it off to whoever recently wanted to compile a FAQ. I might still have a copy, which I'd be happy to post if Clarence doesn't want to or doesn't mind. This information is freely copiable, in case the person who maintains the Roland Product list wants to incorporate it into that document, etc. Here's what I know about the D- and U-series keyboards. No flames, please! I owned a D-20, and currently own a U-20, and I am amenable to correction about them. D-50/D-550: ----------- This was the first Linear Arithmetic synthesizer by Roland. It allowed you to combine synthesized waveforms with sampled attack transients and looped samples of many acoustic instruments, which was an economical way to simulate many "real" instruments without having a lot of sample memory onboard. The D-50 was bi-timbral, I believe, so you could layer or split two sounds, although there was a company that made a modification to increase the multi-timbrality of the D-50. The keyboard was velocity- sensitive and aftertouch-sensitive (channel aftertouch), and the built-in reverbs, chorus, and eq sounded nice. It was 32-note polyphonic, and the polyphony decreased, depending on how many partials were used in a sound. Using the maximum number of partials per sound, the polyphony was 8-note. Voice allocation was last-note piority. It had 64 presets in memory, and 128 waves in ROM. The samples were 16-bit, and it had 20-bit DAC's. The D-50 sported a 5-octave keyboard. The D-550 was the rack mount version. D-5/D-10/D-20/D-110 : --------------------- These were the first multi-timbral D-series synths that could do many instruments at once. The D-20 had an onboard sequencer (16,000 note capacity), a 3.5" double density disk drive, a 9-track sequencer (including drum track), 128 preset tones, 64 rhythm tones, and 64 user- programmable tones. The sequencer was linear (i.e., no cut and paste or repeat options) with quantizing capability, overdub capability, and punch-in/punch-out capability. The drum track was composed of a string of one-measure drum patterns, and track 8 could be used for overdubs on the rhythm track. I think there were 64 drum patterns stored in the D-20, half of which were preset, and half of which were user-programmable. User-programmed drum patterns, sounds, and/or sequences could be saved and loaded from disk. The D-10 was almost the same as the D-20, except that it didn't have a sequencer or a disk drive. It had a drum machine, though, as did the D-20, and you could play along with it. The D-10 was, presumably, for those people who already had a sequencer, and just needed a multi-timbral synth to be the sound source. The D-110 was the rack mount version of the D-10. It had stereo mix outputs, plus 6 extra outputs, I think, so more of its instruments could be processed independently. The D-20, D-10, and D-5 had only stereo outputs. Also, the D-110 did not have a drum machine built-in, as the D-10 did. All of the synths were 32-note polyphonic, theoretically, but practically, any decent sound (except a drum sound) would use at least two partials, probably three, and possibly four. This cuts one's effective polyphony to somewhere between 8 and 16 notes. The built-in reverb was adequate, but not so good as the D-50's, and none of them had built-in chorus. Although these synths did have a "performance" mode (layered and split sounds for performance, as distinct from 8 part multi-timbral setups for sequencing). they were not so much performance-oriented as the D-50, which had more real-time controls, such as a joystick for modifying volumes and balances between tones, portamento, and chase-play buttons. The D-10 and D-20 had velocity-sensitive keyboards, but were not aftertouch sensitive. They could respond to aftertouch via MIDI. The D-20, D-10, and D-5 were 5-octave synths with 196 tones onboard (64 of them programmable), 128 RAM combis, 128 ROM timbres, and 256 waves in ROM. The sounds on the D-10/D-20/D-110 were not as "clean, crisp, and sparkly" as those on the D-50: not so many bits as the D-50's sounds, and some sample aliasing in the higher registers. 12-bit samples and 12-bit DAC's, actually. But they sounded pretty good, and I can attest first-hand that limitations and drawbacks are not so apparent in a good mix as when the sounds are played on their own. My mixes sounded fine, the D-50 comparisons notwithstanding. The main worry I had was designing my sounds and playing sparingly, so that I could get the fullest mixes possible without running out of notes. It can be done, though. One just has to resist using many sounds with 4 partials, or playing monstrous two-handed piano parts legato, with lots of sustain pedal. The drum machine was the best feature of the D-20, I thought. Some great sounds, especially with some third party drum sounds that I got. But the internal piano samples leave a lot to be desired. The D-5 came out later, and seemed to be a stripped down version of the D-10: no drum machine, or effects; just multi-timbrality and a bunch of sounds. Velocity-sensitive keys. The sounds from the D-5 work with the D-20 (in fact, I greatly enjoyed getting the E-X-piano from the D-5's presets, as well as the Tapped EP. U-110: ------ This rack mount came out around the time of the D-110, or shortly thereafter, and it was pretty impressive at the time. 128 onboard sampled instruments, not just sampled attack transients combined with loops, but whole samples. Roland was able to fit that many samples onboard by using their proprietary RS-PCM technology (Resynthesized PCM). They would sample a sound via pulse- code modulation and then resynthesize it (using differential interpolation) and store that resynthesized sound (which was virtually identical to the original PCM sample, except that it took up a lot less memory) in the ROM on the U-110. RS-PCM allowed them to put many highly realistic sounds onboard while keeping the unit a reasonable price. There are 3 megabytes of samples onboard. The U-110 was 31-note polyphonic, had 4 PCM card slots, and had 4 or 6 outputs, I think. The big complaint about the unit was that it was noisy -- a bit too noisy to make it the sample-playback wonder of the year. U-20/U-220/Rhodes M-660/Rhodes M-760: --------------------------------------- The U-20/U-220 was the next generation of U-instruments, and it was QUIET. There was nothing wrong with the original sample data, since the same PCM cards (SN-U110-01 through SN-U110-15, now) played back on the U-20 with no unseemly background noise. The U-20/U-220 were the refinement that the U-110 needed to be a strong contender for studio work, where background noise is exposed in the septic, squeaky-clean studio surroundings. The U-220, a rack mount, only had two PCM card slots though (as did the U-20), and the polyphony was 30-note instead of 31-note. The Rhodes keyboards were clones of the U-20, with a slightly more understandable user interface. The Rhodes MK-660 was 61-key, like the U-20, and the MK-760 was 76-key. The keyboards were velocity- and channel-aftertouch-sensitive. There were 128 preset tones onboard, some of which used two tones in a velocity-mix combination, which cut the polyphony in half when using those sounds exclusively, such as the very nice acoustic piano tone in preset sound patch A-11 (which is the Acoustic Piano 2 tone). The sounds were good (the original samples were, variously, 8-bit, 12-bit, and 16-bit, with 16-bit DAC's -- and they can get much better dynamic range than the bit resolution would indicate, thanks to resynthesis and Differential Interpolation), but here are a few of my gripes: 1. It's too slow in dense MIDI passages. True, it does tend to chug in dense MIDI passages, especially when you are layering multiple timbres on the same MIDI channel, but that problem is somewhat alleviated in later ROM versions, at least in the U-20. Keyboard magazine recommends that you get at least the 3.1 ROM upgrade, which is a significant improvement in MIDI timing response over the earlier ROM versions. I have 3.03, which is the latest ROM version, I think. Someone at Roland Canada told me that the U-220's latest ROM is still 1.16 or something, but I could be misremembering. 2. The programming interface SUCKS. Yes, it true, the user interface is perhaps the worst that Roland has ever come up with (I look at the JV-80 and wonder why Roland couldn't have come up with those front panel ideas sooner). Wading through the many menus and submenus is a chore, as is trying to remember or figure out which menu the function you're looking for is in. Buy an editor/librarian (any editor/librarian!). 3. There are no filters (much less, resonant filters) on the sounds. Filters would really make the PCM sounds respond to velocity well, and allow for more flexible programming of the keyboards. 4. No third party PCM cards. Roland has a monopoly on the samples for the U-series. The third party sound cards that are available (from Sound Source, et al.), contain only sound patch parameters; if you want new PCM samples, you have to buy a card from Roland ($60- $70). There are 15 of them, but two of them are built-in to the U-20 (cards 8 and 9, I think), so you would only get them if you had a D-70, MV-30, or a U-110, I think. (BTW, Americal Music Supply was recently selling the SN-U110 PCM cards for the U-20 for only $25.00, but there was a limited supply of them, no April Fool's Joke! Seriously! -- Nick 4/1/94.) The way to check the ROM version on a U-20 is to put the keyboard into ROM Play mode, then hold down the JUMP and MARK buttons, and press ENTER. This enters the test mode, which should display the current ROM version. To get out of this mode, hold down JUMP and MARK, and press EXIT. You probably won't want to do anything else in test mode, unless you know what you're doing. The U-20 and its kind have chord memory and an arpeggiator for the curious; the chords can be transmitted over MIDI, but not the arpeggiator's notes. :'( D-70 (U-50)/Studio M (MV-30): ----------------------------- We hear that the D-70 was originally going to be called a U-50 by Roland, thanks to a bit of ingenuity, curiosity, and sticker-peeling by some D-70 owners out there (most notably, Nick Rothwell), but they decided to call it a D-something because it would probably sell more that way. The D-70 looks like a U-20 with an extra 15 keys, and some resonant filters slapped on. But before some D-70 owners protest: the D-70 does have a much more extensive synthesis engine than the U-20. More parameters of the sound are more finely editable than on the U-20. The D-70 reads U-20 sound cards, as well as its own special cards, and it does not read any D-50, D-20, or D-lower cards. So it's really a U-series keyboard in disguise, with a few more editable parameters. It is 30-note polyphonic, 6-part multi- timbral, including the drum part (as distinct from the U-20 and ilk, which is 7-part multi-timbral, including the drum part), and velocity- and channel-aftertouch-sensitive. The Studio M (MV-30) is a keyboardless work station with the same synthesis engine as the D-70. It reads U-20 cards and D-70 cards, and has a built-in sequencer with recordable fader movements for mixing down. It also does some form of tape sync, which was pretty impressive in the demo I saw a few years ago. 3.5" disk drive. -- Nick Velharticky | "Not only do I deny the allegation, velhart@epas.utoronto.ca | but I also deny the allegator." | -- The Rev. Jesse Jackson