Once upon a time, way back in 1986, there was a company called Greengate, and their 'bag' was the business of making a sampling and sequencing system called the DS3, and was based on the Tangerine, no sorry, the Apple II series of home computers. The Green men provided the punter with a PCB that even the most retarded musician could fit. They also provided the software to make it do it's thing. Software, as you will remember, was old 'computer bastard' terminology for the set of instructions that were on floppy discs used back in the 80's.

Although, they didn't say it themselves, the DS3 was very similar to the Fairlight CMI, in that it used 'pages' (computer bastard jargon for different sets of screens that come up on the monitor) for different functions. Five 'pages' were offered, that allowed the user to sample, edit, sequence and play (up to about twelve) sounds, among other things. Being clever chaps, Greengate designed the DS3 as a soft and hardware updateable system (hardware being yet more computer bastard words for the computer itself and any other bits of solid stuff) which meant that when the Green boffins thought of new ideas, they sent the user a floppy or a hardy update, and reasonably cheap they were. We shall take a retrospective look at two updates they offered back in '86, one being hard and stiff, one being soft and floppy.

Remember MIDI? Well Greengate thought they should incorporate it, and produced a plug-in MIDI card for the DS3. Perhaps I shouldn't say, but I used to own a DS3 and so remember the system quite well and recall feeling rather pissed off with the MIDI card, having paid £150 for it (remember money?), and then discovering several drawbacks. Let me tell you about them.

The biggest drawback was that the card did not send out MIDI information whilst playing back tunes or songs that had been put into the DS3's own sequencer, stopping me from using other synth sounds in conjunction with the DS3 samples. Stupid. Also, they designed the MIDI card using a DX7 (a primitive keyboard whose sounds were done to death) which meant that when I tried to play the DS3 from a Casio CZ1000, to my surprise, the samples were played out at about half volume. Apparently it was because not all MIDI synths give out the same velocity values over MIDI. Smashing.

I was in a studio one day, trying to run the DS3's sequencer by MIDI. It seemed to work, until I tried to record something onto tape. The DS3 just didn't want to know, in a nutshell. It didn't respond to start and stop information and although it did actually start running, the sequencer was most adamant that it wouldn't run in time with the rest of the track. I was impressed by the thought of a simple computer being able to analyse the song and decide that in fact the tempo was wrong and then demonstrate this by running at what it obviously knew was the right speed. Slightly taken aback, I thought of the other things I could buy with £150, but then realised that the MIDI card did let me play samples from the aged DX7, using the revolutionary (for 1986) velocity sensitivity, which did indeed add dynamics to the sound. Perhaps it was worth it after all, I couldn't be sure, but it was hard.

On a floppy note, I took delivery of the ADSR package. That consisted of some skimpy information, the disc itself and promises of wonderful things. According to the 'brochure', the package allows one to alter the Attack, Decay, Sustain and release portions of a sound's envelope, as well as merging different sounds together and performing digital surgery on samples. I was slightly shocked to find out how long-winded the whole process was. Let me give you an example.

I was in a studio one day, about to record a song using the DS3's sequencer, when we decided that the attack of the sound being used was too abrupt. No problem, I remember thinking, with the new ADSR software it will be simple to fix. I first had to exit from the sequencing 'page' and load up the ADSR software, taking another two minutes. Then I loaded up the sound in question, only taking about another minute. A graphic representation of the sound's envelope appeared on screen, and to make the attack slower/less abrupt, all I had to do was re-draw the envelope. This involves dragging the cursor along the attack portion of the sound to produce a more gentle slope. Now if that sounds rather hit and miss, I remember thinking the same thing. Anyway, the next step was to ask the computer (nicely) to calculate the new attack rate, which only took another good two minutes. When this was done, I tried to play the sound, but it was not to be. To 'play' the sound when using the ADSR package, you had to prod the number keys on the computer in order to hear anything. Brilliant design. Remembering that we were trying to record a song, I decided to get on with it. This involved saving the newly modified sound, taking only another couple of minutes, loading up the master disc for the DS3, going to the sequencing page, loading up the newly modified sound and original sequences. This whole rigmarole took a good ten minutes which really impressed me at the time. I then realised how advanced something like a Juno 60 was, for, instead of ten minutes, to alter the attack of a sound took about ten seconds!

Also on the menu was merging different samples together. Now this part of the package worked, albeit pretty slowly. I recall loading up a Vatican castrati choir sample anxious to combine this with my favourite power drill sound. The result was glorious. Digitally adding two sounds together worked a treat, although the rest of the package didn't. Lies, lies, lies. It did work, though not how I expected. Let me explain.

In theory, it was possible to create new 'interesting' sounds by adding up different harmonics, from the fundamental upwards. The problem was that there were no levels or numbers to gauge which harmonics were being used and to what level. In retrospect, I can remember being more impressed by the sound of my aged Casio VL tone, which sounded thick and lush compared to the thin reedy sounds that emanated from my hours of trying to create sounds using the ADSR package. Digital cutting up of sounds was also possible, known as 'inbetweening' in Greengate parlance. This (again) in theory allowed you to start with say a Trumpet sound, and then cut in a Bassoon sound to end. If Greengate had provided some samples of sounds made using the ADSR package, perhaps I would have been more inclined to spend (yet more) time trying to get musical results.

Speaking to the boffins at Greengate, back in '86, they assured me that they would be offering a software update to make the MIDI card produce MIDI output from the sequencer. They also promised a step-time sequencer back in '85 and that still hadn't turned up. I bought a DS3, I used my DS3 and produced some brilliant stuff (I thought) with it. However, I was frustrated by the slow workings of the system. Greengate said that they had new software to allow the user to go from one page to another without losing the sound/s being worked on, which would certainly speed things up, but did it arrive? My plea was to make it more musical and less 'computer bastard'. They were definitely getting there, but in '86 had a way to go. Perhaps in'87 they cracked it.