This Year's Model
Korg DSM1 & E-mu SP1200
Are the latest samplers from Korg and E-mu Systems really new instruments, or are they simply updates on old machines? Lorenz Rychner counts the bits and checks the clocks.
Is often difficult to tell whether the latest update of an instrument actually represents a slimmed-down budget model an upgrade or a new presentation of the same instrument. Take the Korg DSM1 and the E-mu SP1200...
WE'VE ALL BEEN there: the ink on the cheque is hardly dry, the owner's manual is still unread yet your newest piece of gear is already out of date. A new model has been launched that makes yours look like something out of Noah's Ark. Did the manufacturer consider your situation by making the new model compatible with the old? Can you spend an extra tenner to upgrade yours? Exactly where do you stand?
Two new pieces of equipment taking their origins from some obvious sources, are worth serious consideration before you rush out to spend more of your hard-earned cash. Rather than the usual Appraisal, this comparison of old and new might give you a few surprises.
Digital Sampling Synth Module
SOME ASPECTS OF the DSM1 have a certain familiarity to them. It occupies three spaces in a 19" rack; it's 16 notes polyphonic, as opposed to the eight notes offered by its predecessor, the DSS1; its sampling rate can be set to 16, 24, 32, or 48kHz, and its resolution is 12 bits. Synthesised waveforms are permanently stored in ROM and additive synthesis, using up to 128 harmonics, is available - each harmonic adjustable in amplitude over 256 steps. And it's also a versatile analogue synthesiser to boot.
Does this make it a rack-mount version of the DSS1? No, it really is a new instrument. But it also reads DSS1 disks as if they were its own.
So what else is new? I'll give you the figures first: one Megaword (approximately 1.56 Megabytes) of sample memory (four times the capacity of the DSS1); faster disk interface (1 Megaword in less than one minute); a Play mode selector tab; an elaborate Combination mode for splits, layers, zones and overlaps; a facility for playing four Programs at once; four-way multitimbral operation with flexible voice assignment; 16 individual audio outputs plus one mono mix out; three-way gain switching for audio in (for sampling); and rotary pots for Mix (output) level, Trigger (input) level, and Record (sampling) level; triggers a user-defined note via audio input or from a footswitch. Uses the new 3.5" HD (high density) diskettes; two disk formats: Work and Performance; extended MIDI implementation, including new Sys-Ex messages and new samples can immediately be heard with pre-configured programming parameters of the user's choice. And to help you through all that lot the owner's manual is more helpful than many, and a separate reference card lists all the modes with their function pages. Time for a pause for breath, I think.
DSS1 users will search in vain for the two digital delays that they came to love on the DSS1, however, stereo sound treatment is now available through separate routings to individual audio outs. Also missing are the unison mode and the on-board equaliser, and waveforms can no longer be "drawn" on the LCD. Single oscillators are assigned one Multisound to each Program, where a Program on the DSS1 can draw on two Multisounds before the combined mix undergoes subtractive synthesis.
A CLOSE LOOK at the rear and front panels reveals a number of clues to the new operating system.
The rear panel contains only the 17 audio outputs (¼" jacks), MIDI In/Out/Thru, and the mains input.
The front panel is neat and functional, with the disk drive to the far left. Phones, sampling and triggering jacks and level knobs with gain switch are adjacent. The LCD is backlit and contrast is adjustable; the twelve mode tabs and the numeric keypad are centred; to the right are the six data entry tabs and the data entry wheel.
THE SMALLEST UNIT of sound data on the DSM1 is either a synthesised waveform, called Harmonics Data, which can be saved to disk, or an individual sample (recording) called a Sound, which is incorporated at all times into the next-higher unit, the Multi Sound. Harmonics Data is a single wavecyde, initially spread over the whole keyboard, with a total length of 1020 words. A Sound may occupy: 16.7 seconds at 16kHz sampling rate; 10.9 seconds at 24kHz; 8.3 seconds at 32kHz (default); or 5.4 seconds at 48kHz. Sounds can only be saved to disk as a Multi Sound file, unlike the DSS1 which recognises Sounds as a file type.
If a Sound is of maximum possible length, then it takes up the whole of a Multi Sound and therefore the whole of one Bank among the four Banks in RAM. While this may occasionally be desirable, it's much more common to assign several Sounds to a Multi Sound, with keyboard mapping and other parameters (several recorded pitches of a piano, or a number of drum sounds playing from adjacent keys and so on).
Once a Multi Sound has been assigned to one of the 32 Programs that reside in RAM as parts of a System, the linked Multi Sound and Program become a Timbre. Up to four Timbres are called up by one or more of the 32 Combinations that form a System. In Play mode, the user calls up Combinations (no longer Programs as on the DSS1).
The System is the highest file in the DSM1 hierarchy. A disk that was formatted as a Performance Disk contains only one System, whereas the DSS1 writes up to four Systems onto a disk. But the DSM1 System contains up to four times the memory of a DSS1 System and takes less than half the time to load. A separate Load Function loads DSS1 systems, but only the Multi Sounds assigned to Oscillator 1 are played. This requires rebuilding the DSS1 systems into several Timbres on the DSM1 if the same blends need to be reconstructed in Combination Mode. It is a difficult job but it's made easier by the many flexible operations on the DSM1.
THE DSM1 OFFERS a number of operation modes, but most are different from those of the DSS1. A Compare feature is available in many of them.
- Play Mode selects one of 32 Combinations. These are tunable in memory by +/- a semitone (approx). The screen displays the tuning and the assigned Programs for Banks A-D. Pressing Enter displays the MIDI channels for each Bank/Program. If the DSM1 is set to receive on a common MIDI channel, incoming patch changes select new combinations, wrapping the numbers around (incoming numbers 33, 65, and 97 all select Combi 01). When Banks are set to individual MIDI channels, then the incoming patch changes select new Programs for each Bank.
- System Mode loads Performance data from disks, saves current RAM to Performance disks, loads DSS1 Systems, sets the common MIDI channel, assigns playback of a selectable note by a footswitch for editing without a MIDI keyboard, and sets the Audio Trigger for a selected note.
- Combination Mode is the powerhouse of the DSM1. Its features are too varied to describe here in detail. To make life easier, Korg made these templates available: Single, Split 2, Layer 2, Split 4, Layer 4, Split/Layer, Multi. In addition, there is an open user configuration called Manual. Once the desired type has been selected, the user sets MIDI functions, transpositions, key windows (zones), keyboard mode (Poly 1 or Poly 2), number of voices (polyphony), detune, levels, voice allocation, individual MIDI channels, velocity polarity, velocity window, and output assignments for each component of the Combination.
- Edit System Mode loads Programs (with or without Multi Sounds) from Performance disks, loads Multi Sounds from Performance, Work or from DSS1 disks into Banks, copies Multi Sounds from Bank to Bank, erases Multi Sounds or whole Banks from RAM, and lists the Multi Sounds currently in RAM and the free space available per Bank.
- Edit Multi Sound Mode selects a Multi Sound from one of the Banks in RAM, selects the Program to be heard modifying the Multi Sound while editing, adjusts tuning, volume and filter cutoff for individual Sounds within the selected Multi Sound, assigns keys for playback of the Sound's original and of transposed pitches, sets the Sounds start, end and loop length, then recovers the unused memory, and renames/saves the edited Multi Sound to a Work disk.
- Program Mode assigns a Multi Sound to the oscillator, and deals with the "analogue" synthesiser features that make the samples musically useful. As users of the DSS1 know, this is a very sophisticated synthesiser, with separate six-stage envelope generators for VCF and VCA, selectable polarity for the VCF EG, Autobend, keyboard tracking (scaling) for VCF cutoff, VCA decay, and VCA release, plus lots of LFO, velocity and aftertouch features for all kinds of preset and real-time modulation.
- Sample Mode takes care of the recording process. Samples are considered to be and saved as parts of a Multi Sound. An existing Program can now be active during the recording, so that the samples can immediately be heard in the right context. A choice of automatic or manual memory and key allocation can speed up the process of recording samples that are meant to be, and remain, part of a Multi Sound. For this reason, the tuning, levels, filter cutoff, start/end/loop length, and memory recovery of the individual samples can be set right here, without having to be final.
- Harmonic Synthesis Mode lets the user choose from 17 wavetables, alternatively a new waveform can be built by assigning amplitude to any of 128 harmonics. Previously stored Harmonics Data can be loaded from a Work disk, and the data can be saved as a Harmonics Data file or as a Multi Sound. An existing Program can modify the waveform under construction, to assist the evaluation in the right context. A Bank must be cleared before this mode can be used.
- Edit Sound Mode takes over where Sample mode leaves off. A Sound is selected from within a Multi Sound, an existing Program can be assigned for monitoring, and the Sound can be treated for start/end/loop length, reversed, linked (joined) to or mixed with another Sound, the loop can be created with backwards/forwards or crossfade techniques, the pitch adjusted to match a loop (great for economical short loops), and the Sound can be viewed and edited in amplitude per address (sample point).
- Make/Remake Multi Sound Mode selects a Multi Sound from disk (DSS1, Performance, or Work disk) or from RAM, clears a Bank for the creation of a new Multi Sound, assigns a Program for monitoring the editing progress in context, erases or inserts or replaces Sounds within the Multi Sound, and renames/saves the Multi Sound to a Work disk.
"DSM1: A Performance Disk holds one System, whereas the DSS1 holds four Systems, but the System contains up to four times the memory and takes half the time to load."
- Disk Utility Mode formats disks for either Performance or Work, protects files in software, shows directories of the different file types, and deletes Multi Sounds and Harmonic Data files from disks. The disk drive can read DSS1 double-density/double-sided disks, but it can only write data to the new HD high-density disks mentioned before.
EVEN WITH THE introduction of the DSM1, Korg continue to support the DSS1, which is a different instrument with more of a live orientation. As their names imply, both instruments are at least as much synthesisers as they are samplers. The drawback to the DSS1 was a number of quirks in the software, and its limited memory.
DSS1 owners who're looking to upgrade their instrument will soon be able to add one, two or even four megabytes, with the addition of a SCSI port, and a hard disk of the user's choice. While this may not be cheap, it could be well worth saving up for. Building and rebuilding Multi Sounds and Systems - so far a cumbersome procedure - will be lightning fast, as will the loading of Systems during performance.
The DSM1 is also about to be enhanced with the SCSI and hard disk, but its memory will remain at one meg. The output configuration, the Combination mode features, the multitimbral capabilities, and the advanced MIDI implementation place it at least as much in the studio as on the road.
THE SP12 DRUM machine is no longer in production; in its place comes the SP1200. A look at the SP12 could help put the SP1200 in perspective.
While the original Linn drum machine precluded many users through its cost, E-mu's Drumulator offered sound quality and sophisticated programming facilities at a more modest price.
E-mu then announced a quantum leap forward in the form of the SP12. It arrived with a promise of being the professional's only choice. Although backed by the name of the company whose Emulator set a standard for roadworthy pro sampling keyboards, the SP12 fell short of that promise: users were frustrated by rigid memory allocation and unrealistic output configurations. Successive software updates slowly brought the instrument closer to, but so far have not realised that promise.
Despite these shortcomings, many professionals use SP12s every day because there simply hasn't been a comparable drum machine around. What else sequences ROM sounds, accepts external sounds, samples, and both reads and writes SMPTE?
So what's new on the SP1200? Well, it's certainly not its looks. The SP1200 lives in the same box as the SP12. The front panel shows the same cutouts for the operation buttons and play buttons, but the new lettering reflects some of the operational changes. The SP12 originally employed a cassette interface for data storage and was later adapted for an external disk drive. The user had to buy a Commodore 1541 disk drive, which probably holds the world record for lack of speed. Loading times of up to five minutes were not uncommon.
The SP1200 overcomes this by sporting a built-in disk drive, accepting double-sided double-density 3.5" microfloppies. As you might expect, access time is much shorter, disk formatting takes just over one minute, and the drive eliminates the need for a cassette interface. But users thinking of making the change from the SP12 to the SP1200 needn't worry about their cassette data becoming redundant. A special transfer mode can be activated by powering up the SP1200 while holding down the disk button. Once in RAM of the SP1200, the user sounds and sequences can be named and written to the SP1200's disk drive and used like any other user data. However, you can't transfer the ROM sounds from the SP12 in this manner. Only a sampling session will get them into the SP1200.
The machine's software resides on disk, which means that future updates can be made on disk, and a mode of the formatting command allows erasure and replacement of the program without touching user data already on the disk.
The rear panel hasn't changed much either. It shows ¼" inputs for Sample (audio in), Mix Out (mono sum), Channel Outputs 1-8, Metronome Out, SMPTE Out and In, Run (footswitch for remote run/stop), and Tap (for tempo footswitch in segment and song modes). MIDI jacks are the usual In, Out, Thru. A voltage selector takes one headache out of international tours; the world has become a small place for pro musicians lately.
Though it's not apparent, there is another change behind those individual audio outputs - it has to do with the output filters. Unfortunately, E-mu haven't seen fit to remedy the major complaint of most SP12 users; there are still only two outputs with dynamic filters (outputs 1&2), and four with static filters. None of them are adjustable. Since the dynamic filters are an absolute necessity to avoid the noise that becomes apparent late in many samples, most users assign two of the usual three tom-toms to outputs 1&2, wishing that there was a third output with a filter. The third tom has to be processed separately to match the others. What's new is the internal wiring; mono plugs give access to the unfiltered sounds and re-route the sound from the Mix Out jack. Ring/tip/barrel plugs access the filtered sounds (outputs 7&8 have no filters).
The SP12 had ROM sounds permanently on board; the SP1200 doesn't. You must first load sounds - which are still 12-bit - from disk every time you turn the power on. The overall memory has been doubled to 10 seconds which is subdivided into four blocks of 2.5 seconds. You have to load the longest sounds first, forcing you to squeeze the shorties into what's left in each memory block. You can't trust the memory readout as this shows you the available memory globally, not for each block. If the readout shows 1.4sec memory remaining and you need to load a sound that's 1.2sec long, you may still not be able to load your sound if that 1.4sec is spread across the blocks in small amounts.
ANOTHER LIMITATION IS the lack of communication through the LCD when it comes to copied sounds. I loaded the standard factory drum sounds from disk and deleted Crash 2 from location C8. When trying to load the same sound back into the same location, I got the message "Not Enough Memory" - it turns out Crash 2 is not taking up any memory, and therefore doesn't refund any memory when deleted, because it's a copy of Crash 1.
Copying sounds to other locations allows modified sounds to be stored without exceeding the available memory, but the list in the LCD doesn't indicate which sounds are taking up memory space, and which are memory-efficient copies. In my case, to get Crash 2 back into location C8, I would've had to reset the decay.
Dynamic output allocation is E-mu's way of overcoming the limitations of two sounds competing for the same audio output at the same time. It means that one sound will jump to the other output in a pair. It's not a very elegant solution to have sounds walk across the stereo field, seemingly at random - instruments of less than half the price solve this problem more successfully. On a more positive note, the new Global Tune feature is a (fairly self-explanatory) welcome upgrade.
Do these changes make the SP1200 a new instrument? I think not. Somebody at E-mu must have decided to "tough it out" with SP12 owners and those that I spoke to are resentful. (If you don't already own an SP12, you'll see this instrument as a new one and may not be bothered by these concerns.) Here is their consensus. Please draw your own conclusions.
Software updates are nothing new to SP12 owners; they would have welcomed the memory expansion and the software changes that accompany the new chips. Features like Global Tuning, changes in the MIDI implementation and so on could have been made as an upgrade to the SP12. In the hardware department, the only major change is the built-in disk drive - SP12 owners feel that a stand-alone drive could have offered the same increase in speed and flexibility. The reconfigured output wiring is something that many users had built in long ago.
Another interesting point is that the events surrounding the replacement of the SP12 by the SP1200 are in stark contrast to those surrounding the introduction of the Emax over the Emulator II. First of all, the Emax was a new instrument; secondly, Emulator II owners continue to be supported by E-mu. What happens once the Emulator III comes out remains to be seen, but even if support for the EII should stop the EIII is clearly a bigger step up than the SP1200 from the SP12.
THE BAD NEWS: SP12 owners will have to grin and bear it. Their machine is out of production, and they see the successor as an overdue update with a new name, rather than a new machine worth losing money for by ditching the old.
The good news: new buyers don't have to worry about the gripes of SP12 owners. They get a machine that works well in many areas, sounds great, comes with a detailed manual that includes many helpful hints above and beyond the call of duty, and includes a comprehensive MIDI implementation that advanced MIDI users will surely appreciate.
Prices Korg DSM1. £2395; E-Mu SP1200, £2199, both including VAT
More from Korg UK, (Contact Details); E-mu Systems UK, (Contact Details)
Gear in this article:
Review by Lorenz Rychner
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!