Sampling Percussion System
Not just a sampling drum machine, more a solid state tape recorder! As things start to heat up at the top end of the drum machine market, David Mellor gives E-mu's successor to their popular SP-12 a thorough pounding.
Latest in the Linn/E-mu tradition of heavyweight drum machines for pro users is the SP1200 - a sampling beatbox which gives you the facilities you need and holds back on the complexities you don't. David Mellor takes a look at an instrument that puts the 'mu' - E-mu - back in music.
Life, as they say, is becoming ever more complex. I remember an old cartoon series set in the world of the future, where 'button pusher's finger' had become a disease of epidemic proportions. Paradoxically, it's not the more buttons the greater the complexity - it's the reverse. The less knobs and buttons a piece of equipment has, the more difficult it seems to be to operate it.
In the world of musical technology, instruments are becoming ever more mind-bogglingly difficult for musicians to come to terms with. In the 'good old days', it was just a matter of making your mind up whether you wanted to be a pianist, violinist or church organist at the age of six or seven - then practising for ten years or so until you reached a good enough standard. You did your learning first, then all you had to do was play.
In this modern age, it seems that every day there is a new musical instrument to learn. OK - so you don't need a lot of physical dexterity to play a sequenced MIDI system, but you sure need those brain cells! The problem is that very often the musical parts of the brain and the mathematical parts seem to be in opposite hemispheres. Just when you have a great musical idea that you want to put down on tape - or into RAM - you have to switch over to the data processing brain cells to figure out how to do it. Inhibiting, to say the least.
Manufacturers of electronic instruments have come a long way in the last few years towards giving us all the facilities we need. (Although we still want more!) The next challenge is to make them simple to use - or at least make it possible to access an instrument on different levels of complexity. Some people may enjoy complexity for its own sake. Although I'm prepared to work hard to make my instruments work for me in the way I want them to, my aim is to make music - not to see how effective my brain can be as a space heater.
The purpose of the above is to try and put into perspective the coordinates occupied by the E-mu Systems' SP1200 in n-dimensional musical instrument space. Maybe I'm mistaken, but I am convinced that when E-mu were putting the SP1200 together, they were trying to reconcile the opposing concepts of functional capability and operational simplicity. In other words - it does a lot but it's easy for a muso to use. (Why didn't he just say that in the first place? - Ed.)
Whenever I get a new piece of equipment, I budget a certain amount of time for getting to know how it works. For instance, if I was getting new word processor software for my computer, I would allow at least half a day before I tried to use it at all seriously. (If I was getting a new video then I'd need a lot longer than that before I would dare to set it on the timer!)
Getting to know the SP1200 is like getting to know a piece of my mother's sponge cake - it just melts in the mouth! (E-mu can quote me on that in their advertising if they like.) Congratulations with first class honours must go to the manual, written by Craig Anderton. Would all you other manufacturers please look at this one and see how it should be done. It's written in a tutorial style that works - many have tried before and failed. You start at the beginning by making interesting noises and work all the way through until you have mastered the machine. Having said that, there should have been a reference section included as well, or a decent index. When you think you know a machine, you always have to go back and look up something that hadn't quite penetrated as well as it should. Oh well, no-one's perfect.
But on to the machine, what does it do that other machines don't? Well, there are no technological leaps to be seen, just tried and trusted technology - but it is a nice package. E-mu have put things together into a machine that will find a home in many top studios, and on many chartbound singles - hopefully, without the 'sample boredom' induced by the omnipresence of the Roland 707/727 and Linn machines.
Basically, what you have here is a sampler, a drum machine programmer, and a SMPTE synchroniser. OK, so the Linn 9000 can do all that - but you tell me of someone who has one in which the SMPTE option works? I can see this machine finding a lot of use in situations where someone wants to carry a single box into the studio, plug it in and use it - just as a drum machine. Although it could easily fit in with a full-blown MIDI set-up, somehow I don't feel that this is its market. I am prepared to stand corrected, however.
As a sampler, the SP1200 is capable of a total of 10 seconds sample time at a sampling rate of 26kHz with 12-bit resolution. The 12-bit format gives a good, but not perfect, signal-to-noise ratio as you would expect (although this can be improved, as I shall explain later). A sampling rate of 26kHz should give an audio bandwidth of between 10kHz and 12kHz, which I found adequate for normal drum purposes. The supplied cymbal sounds, for instance, were quite bright enough. If you are into the utmost in fidelity then perhaps you should investigate elsewhere, but for a straightforward drum track all was OK.
Unlike other drum machines which provide user-sampling as an added attraction, this machine has no internal sounds. The 32 sample locations must all be filled from external sources. Fortunately, E-mu provide a selection of five 3.5" floppy disks to get the new user started (see Table 1). Judging by past reputation, they will no doubt make many more available for those of us unwilling to get hold of real drums and sample them ourselves, or steal sounds from compact discs.
As a rhythm programmer, the SP1200 has most of the capabilities of any other machine on the market - and possibly a few more. As a SMPTE synchroniser, it promises to link up to a tape recorder in a way that will make separate sync boxes seem hopelessly complex. Look at my words closely: 'promises...'
The SP1200 is rather larger than your average drum machine but a good bit lighter in weight. The size helps because the front panel can look sufficiently uncluttered so that even the technophobes of the musical world ought not to be put off. It even has a few friendly faders.
The machine divides up neatly into several sections: Set-up, Disk, Sync, Sample, Programming, Performance and Master Control. Where shall I start? Ah yes, 'Performance'.
The first thing you want to do with any instrument is start making noises. As I said, the SP1200 doesn't have any internal samples so you have to load in a disk first. Disks in this case are double-sided double density (DSDD) 3.5" floppies. You can buy them from your newsagent or tobacconist.
Once you have loaded a disk, which is so straightforward that I won't bother talking about it, you have all 32 sounds at your fingertips - eight at a time, selected by the Bank A/B/C/D button.
Each of the eight trigger buttons has a fader above it which can control several different functions. One useful function might be to control the level of each of the sounds, that would be handy. If this is what you want, just find the button marked 'Tune/Decay-Mix-Multimode', it's right next to the faders. Each fader acts as a level control for the appropriate sound. Levels are indicated in a graphical display on the smallish backlit LCD and can be separately controlled for each of the 32 sounds - not just the eight you can call up at any one time.
In addition to this, the faders can be used to set either decay time or tuning for each sound. This is useful because you can have just one hi-hat sample in memory, for example, and use it for both open (long decay) and closed (short decay) sounds. The decay time you set is indicated on the LCD for each instrument. It's a pity you have to choose between decay and tuning, but you can't have everything - and at least you have the choice for each instrument, it's not a global decision.
I have to say that changing the tuning of an instrument was not always as successful as it might have been. If you own a keyboard sampler then you would expect instruments to sound properly at different pitches - it wouldn't work otherwise. In the case of the SP1200, sometimes detuning works, sometimes it doesn't. Don't expect to load a bass guitar sample in and be able to programme walking bass lines. On the other hand, if you just want to change the pitch of a snare drum then you will be perfectly OK.
As on the Yamaha RX5 and Kawai R100/R50, there is the possibility of selecting just one sound to appear on all the instrument keys. Here, there are two options - Multilevel and Multipitch. 'Multilevel', as it might suggest, puts the same sound on the eight trigger pads at different volume levels - useful for programming that good old snare crescendo cliche! 'Multipitch' does the same for note values, giving a diatonic scale on the eight pads. The value of these 'multi' modes becomes apparent when you start programming.
I couldn't resist breaking off writing to go back to the SP1200 and programme a quick sequence. It's so easy and you get a result very quickly with very little mental effort. Rather than work in bars, the SP1200 encourages you to think in segments, the default option being one segment = two bars of 4/4. I think it's good to get away from the idea of bars when you are programming a drum machine because it makes it too easy to programme a 'mechanical' pattern. If you consciously try to think in segments, and the machine keeps on demanding bar numbers from you, life soon becomes tedious. E mu give the punter what he wants. Most successful records which use drum machines seem to have two or four-bar drum loops, so the SP1200 invites you to work in this way. You could stick strictly to bars if you wanted to, of course.
A segment can be set to any length from just one beat up to 99 bars of 99 beats. You don't have to stick to a whole number of bars either, you can tack an extra number of beats on to the end of the segment to get any length you want. Indeed, if you do it in real time it need not be an extra number of whole beats. You can have, for instance, a segment of two bars, one beat and a little bit more - if that's what you really want. This process is known as 'dimensioning' the segment, and is dead easy to achieve. At a later stage you can shorten the segment, or lengthen it by copying, if you wish. So much for that. Let me tell you about my segment...
After an initial difficulty caused by my not realising that the SP1200 was expecting MIDI sync (why didn't I notice that it was flashing an LED at me?), I started up a segment at the default length of two bars. Just press 'Record' and 'Run' and you're away. Bass and snare were first, a basic Boom-Whack rhythm. For the hi-hat, I thought I would slip into something more comfortable... Multilevel mode. This gave me eight different levels of closed hi-hat to play with. I went for straight eighth-notes, changing levels at random. Good, so far.
I had a feeling that the snare I had programmed could be more interesting, so I erased it completely - three button pushes. I popped on the Tune/Decay mode. The snare was set to decay, so I set a long decay on the fader and went back into record. This, I put on the fourth beat of both bars. I went back and reset the decay to a fairly short value and recorded this on the second beat of each bar.
Although I was pleased so far, I had the feeling that a spot of cowbell would be good so I changed to the appropriate bank and found the right button. In Tune/Decay mode once more - Tune, in this case - I recorded a snappy rhythm on to the track while waggling the cowbell fader up and down. This gave me cowbells of random pitches! I liked this effect so much that I repeated the trick with a tom sound. I could go on to say that the track was so good that I sold it to Phil Collins for a fat sum, but that would be wishful thinking of the highest order. Nevertheless, I had achieved very quickly a rhythm pattern which would have taken an age to accomplish on my MIDI sequencer and sampler. It would have taken almost as long on the Yamaha RX5, which can do all these tricks, but you have to think about what you are doing much, much more.
I don't know whether the E-mu design team all drink Carling Black Label or whether they have had their parts refreshed by that other brew, but other handy little functions include velocity-sensitive instrument pads (if you want them to be) and auto-repeat, where an instrument will repeat at intervals determined by the quantisation setting for as long as you hold the key down.
A few segments later and I was ready to put a song together. I found that by thinking in segments, it was much easier to get a whole song into the machine. The ability to programme a segment of two bars plus a little bit is so useful where you have a phrase which needs that odd bit on the end. I always seem to end up with a lot of these in my music, so with my Yamaha RX11 I would end up making a one or one-and-a-half beat bar and tacking it on the end of another pattern to achieve the same purpose.
The SP1200 doesn't offer much in the way of finding your way around a song while you are putting it together or editing it, but there is a function where you can set the point where the song starts. It sounds a bit primitive but, after a while, you get used to it and there is no problem. Insertions and deletions are a doddle, and there are other goodies. Remember I was talking about setting levels for each of the instruments? Well, what I didn't say was that you can set up eight different mixes and store them. These can be recalled as part of a song. It's a little like automated mixdown. I definitely like this idea.
An even better idea is the system of tempo changing. Many drum machines seem to want you to stick to the same tempo, they make it so hard to change. Please note, all you designers out there, that I want to be able to store tempo data as part of the song - as on the SP1200 - and I want to change tempo in terms of beats per minute, not percentages. If you want to do it properly, see how E-mu do it. When you press the Tempo Change button, the display asks you for three pieces of information. Firstly, whether you want the track to go faster or slower (accelerando or ritardando for Italian speakers), then by how much and over how many beats you want the change to take place. No longer do we have to spend hours programming a gradual tempo change or put up with that unseemly lurching effect that sudden tempo changes bring.
Needless to say, following previous offerings from this company, there is adequate memory for segment and song data. If you go in for five hour gigs (I tried one once!) then you may be stretching it a little, but I think for most normal people there will be ample storage. 100 songs enough?
Enough about the programming, on to something else...
Think of the SP1200 as a drum machine where you can put your own samples and you will be OK. Think of it as a substitute for a keyboard sampler and you will be disappointed.
Let's face it, the major criticism of drum machines isn't so much that they sound mechanical - that's up to the programmer - but that you keep on hearing the same tired old samples again and again and again, ad infinitum. If I hear those same two agogo bells from that certain drum machine named after a Boeing jet, then I will certainly go stark raving bananas. Same goes for that cowbell on another machine from the same maker. They may be good sounds but they have been done to death, many times over.
It's one matter being subjected to the same samples on records and on the radio, but another manifestation of sample boredom is when you buy an expensive drum machine, then six months later you're saying to yourself 'I just can't get the sounds I want.' I've caught myself saying it.
It's not that the sounds aren't there, it's just that most people need to be excited by the sounds they hear to produce good music with them. If you get too used to the same set of samples, then it stands to reason that they are going to lose their stimulating effect. Drum machines with sampling capability are what the world has been waiting for and the SP1200, with its internal sample quota of zero, stands up for what it believes in.
In contrast to the programming sections of this machine, I thought that the sampling procedure could have been a little better thought out. Reason does prevail to a certain extent, however. There is no messing about with 'mapping' or 'key groups' on this instrument. There are four banks (A/B/C/D) of eight sounds, so you are invited to record a full complement of 32 samples starting with A1 and going all the way to D8. If a sample location is full, then you won't be invited to record over it until it is cleared. You are always prompted with the next vacant location if possible.
There is, unfortunately, no provision to record one overall 10 second sample if you wish, because the memory is divided up into 2.5 second chunks, one for each bank. Let's go through the procedure and see what happens...
There are eight 'jobs' in the sample mode, the first of which is VU - a simple meter. This is where you set the level - I'm not being too obvious am I? There is a knob for fine gain control, alternatively, job 3 can be called up which gives you the option of 0dB, +20dB or +40dB gain, which should cover most eventualities. Going back to job 2 - you see why I said it hadn't been properly thought out? - you can assign the sample to a location, which would be any of the eight positions in bank A, B, C or D. You will be prompted if you are likely to overwrite an existing sample. You can also set channel assignment here.
Job 4 is for setting the sample threshold, the point at which recording will begin automatically. There is a simple graphic display to help with this. Job 5 sets the sample length. You don't have to do this as you can truncate as necessary later, but you may find it more convenient now.
Job 7 - oops, had to miss one again - initiates threshold recording. Job 9 (would you believe that there isn't a job 8!) initiates recording instantly regardless of threshold. If you're not satisfied with the sample, then go back to job 6 which lets you try again.
Sampling isn't too difficult; in fact, it's quite straightforward apart from a couple of quibbles. Like why not have the job numbers in the order you want to use them? So much more sensible. The other thing is that when you take a sample the display tells you whether or not you overloaded the input. Why couldn't there be a marker on the VU display so that you would know when you were about to go over the top? As things are, you have to estimate whether the little bar has jumped far enough to the right for optimum level. I thought I might be able to put a mark on the display in the right place with a felt tip pen, but it seemed to change from sample to sample so I just had to guess. Oh, before I forget - wouldn't it be a nice idea if we could monitor the sound we are sampling? Come on lads!
Still, I mustn't get carried away. As I have been hinting, I like this machine and if I could trade in my sister for one then I might be tempted. But my story isn't finished yet and I don't have a sister anyway.
As you know, there are eight instrument buttons on the front. Perhaps you don't know that there are eight output jacks on the back, plus a combined mono output. It may seem that matters are little more complicated than eight bits of wire running from one end of the machine to the other, but this is not the case.
The best way to think of things is to imagine eight channels. Not eight identical channels, but different in subtle ways. Channels 7 and 8 are indeed straight copper wire, more or less. Channels 3 to 6, however, are filtered. Channels 1 and 2 have dynamic filtering where the cut-off frequency of the filter lowers as the sound progresses.
At first sight, this looks a bit awkward because you have to be careful of which sound you send where. For instance, it would be preferable to have hi-hats and cymbals going to the unfiltered outputs to preserve their brightness. Toms would benefit from outputs 1 and 2 because their long, bassy decay is a dead give-away of any noise the sampling system may be producing. I'm not quite sure where the other four outputs stand in terms of their 'wonderfulness' rating. Couldn't you just filter them on the mixer?
Anyway, there are two important points. Firstly, that it seems to work. On the supplied disks all the sounds seem to work musically and the custom samples I fed in seemed to work equally well, after a little thought. The second point is that you don't have to use the filtering if you don't want to. The output jacks are, in fact, stereo jacks. The unfiltered version of each channel is available on the tip of the jack, the filtered version on the ring. (I can imagine how many people will get by with just pushing in a mono jack halfway - I did.) You have a choice. The filtering seemed a strange idea but it worked and that's the important thing, and you don't have to have it if you don't want to - except on the mono output. The eight filtered outputs are combined to produce the mono signal, so if you intend to use this then you will have to think about channelisation a little.
There's no stereo output by the way. This may say something about E-mu's intended market - where 56-channel SSL desks are two a penny, if you know what I mean. A stereo output would be just a tiny bit superfluous.
Up until now, one obvious development that hasn't quite developed yet is the combined sequencer - or drum machine - and SMPTE unit. What could be simpler than having sequence and sync all in the same box? Yet still we have to suffer the awkwardness of having all your notes in one box and all your tempo and start data in another. I was told by one manufacturer that it would increase my versatility. Increase the number of products he could sell me, more like. So now at last we have it, the SP1200 combined drum machine and SMPTE controller...
We have it, but it's so slow. The SP1200 puts SMPTE code on tape all right and reads it off perfectly well - there's a display of hours, minutes, seconds and frames - but to catch up with what's on the tape takes forever, or up to seven or eight seconds which amounts to much the same thing. That's if you're going forwards. If you rewind how long will you have to wait? Until you catch up with where you were before you rewound, that's how long!
There is a way around this, which is to flip the Run/Stop button on and off. This way it will catch up more quickly, but still not nearly as fast as the £300 Nomad SMC-1 unit sitting on top of my piano, which doesn't need this kind of attention. Frankly, this isn't good enough, so I hope E-mu are working overtime to improve this aspect before paying customers complain. I should point out that finding the right place using MIDI song pointers is rather faster than with SMPTE.
It's a hard life reviewing products for Sound On Sound. Whenever I get something that I could make good use of, it has to go back straight away. If I can't make use of it in my system then it seems to hang around for ages waiting for someone to come and pick it up. That's life.
I would be happy for the E-mu SP1200 to hang around for as long as it likes. It's not very pretty but it's good at its job, a bit like a Channel 4 newsreader. It doesn't do everything, but it sounds good, it's dead simple to use and the possibilities must be enormous. I could go on but I've just heard those agogo bells on the radio again and I...
Price £2199 inc VAT.
Details from E-mu Systems UK, (Contact Details).
Review by David Mellor
Previous article in this issue:
Next article in this issue:
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!