For as long as computers have had soundcards they’ve also had some kind of synthesiser built in. Usually a General MIDI compliant set of uniformly awful sounds that are there to playback MIDI files in Windows – useful, but not exactly inspiring. Besides, these were FM based or wavetable sounds with no editing facilities of any kind which sort of defeated the object of “synthesis”.

In 1997 a small Scandinavian software company called Propellerheads released a little music program that revolutionised the computer music industry overnight. The program was called Rebirth RB-338 and was a beautiful simulation of a pair of Roland TB-303 bass synths and a TR-808 rhythm machine. It looked fantastic and sounded even better. Suddenly you could be the Chemical Brothers without having to invest thousands of pounds in vintage synth equipment. It’s amazing how two single oscillator mono synths and simple drum machine could have such an effect. You couldn’t even play them from a keyboard, the TB-303 used a step sequencer and the TR-808 used a row of 16 pads that you could turn on or off for each sound as it cycled through. This was precisely why Rebirth was such a success. At that time computer and soundcard technology was not up to the task of generating sound fast enough from software to make using a keyboard viable. Rebirth bypassed this problem completely by making it step-sequenced.

The success of Rebirth spawned two important technologies. Steinberg, creators of Cubase  had recently released their Virtual Studio Technology architecture which brought proper “live” mixing and real-time effects plug-ins to their recording software. Propellerheads wanted to find a way to use Rebirth alongside, or integrated into the VST engine. What they came up with was called “Rewire”. This was essentially a bridge between the outputs of Rebirth and the VST mixer in Cubase so you could route, mix and effect Rebirth alongside audio tracks. At the same time Steinberg was working on another technology, an extension of their VST engine which allowed them to plug-in software synths like they did with effects. These were called VST Instruments or VSTi’s. The success of VSTi’s hinged on the seemingly elusive ability to play them, in real time, from a MIDI keyboard. Steinberg’s driver technology ASIO (Audio System In/Out) was the key but it wasn’t quite fast enough yet. Working hard with a handful of soundcard manufacturers they released ASIO 2.0 and VST 2.0 in 1999 which, for the first time, managed to get latency (see box out) down to under the 10ms (milliseconds) mark which rendered VSTi’s totally playable. Steinberg also released their “Developers Kit” free of charge so that anyone with the desire and programming skills could create their own VSTi’s and VST effects. As a result there are hundreds and hundreds of freeware and shareware plug-ins as well as fully fledged, professionally produced, commercial ones.

Latency

This is really something quite simple and shouldn’t grip people with the kind of fear it appears to. Latency is the amount of time between a call and a response. In computer music terms this is the amount of time it takes for the computer to generate a sound when asked to. In practical terms it’s the time between clicking “Play” and playback starting, or it’s the time between hitting a key on a keyboard and hearing the sound from the computer, or it’s the time between plucking a string on a guitar and hearing the sound though the computer, with effects. In the bad old days you were talking about latencies of half a second or more which meant trying to play anything “live” was a joke. These days, with the right soundcard, latencies are so small that you don’t notice it and playing sounds on a computer has become as live as any other instrument. Latency is a natural function of computers, a built in delay to allow the CPU to process information and get it out in an ordered fashion. The delay is caused by a “buffer” which is a small piece of memory where information to be processed is stored and sorted. As data (or sound) comes in it fills the buffer while the CPU sorts it out, processes it and sends it in a constant stream to the outputs.The larger the buffer, the larger the delay/latency, the smoother the output. So for “real-time” processing we want that buffer as small as we can make it, however, this puts added pressure on the CPU as it has to work faster to maintain the same smooth output. If the data moves through the buffer too quickly then the CPU may not have time to sort it out and you end up with a glitch or crackle in the audio. You can see that there’s a bit of a balancing act going on. Generally, when playing instruments, a latency of 10ms or less is regarded as real-time. When mixing 30ms is completely fine so you’ll find that you can reduce the stress on the computer when mixing by increasing the soundcards buffer size resulting in stable playback and room for more effects – magic! To benefit from low latencies it’s essential that you have a soundcard or audio interface with drivers designed to do it (see previous article on soundcards).

Now you know what we’re talking about let’s have a look at a few.

Software Synthesisers

I don’t think there’s any form of synthesis left now that can’t be done on computer. Analogue, wavetable, subtractive, additive, granula, vector, acoustic modelling and more I haven’t thought of yet. The driving force behind many software synth innovations has been German company Native Instruments. One of their first releases was a model of the legendary Sequential Circuits Prophet 5. Looks superb and sounds amazing and is exactly like the real thing. The current version called the Pro 53 adds various cool enhancements. With the rest of the industry focused on producing analogue style synths NI then released a model of the Hammond B3 organ they called the B4, and again , it’s perfect. From there NI produced synths like the perculiar Absynth and the extraordinary Reaktor that have no basis in anything in the real world.

Reaktor lets you build completely new synths from the tiniest building blocks or just use the wealth of synths created by other users. They also managed to produce the synth that Yamaha wouldn’t called the FM7, based of course upon the famous DX7 and was even capable of loading old DX7 patches.

Many people had been producing simple analogue style farty bass synths in a Minimoog styli but it wasn’t until French company Arturia pulled their Minimoog V out of the bag that the legendary Dr Moog was happy to put his name to. On the release of their Moog Modular V Dr Moog said "Arturia's Moog Modular V is a high quality computer emulation of the analog modular synthesizer that Moog Music® originally intro duced... I am delighted that Arturia's Moog Modular V is adding a new dimension to an instrument tradition that has a special meaning to so many musicians." Arturia have since moved on to the Yamaha CS-80 and Arp2600. It’s interesting to note how important the visual look and feel has become. I’ve heard many great software synths that have failed because the interface didn’t look that great. 

Workhorse sound modules

Hardware manufacturers like Roland, Korg and Yamaha have resisted releasing software versions of their synths, so other companies have done it for them. Korg at least is getting with the program a bit with it’s superb Legacy Collection, software versions of the MS-20, Polysix and Wavestation, and the Digital Collection that adds the famous M1. The Trinity and Triton, though, remain for the moment hardware only. If you are after a Sound Canvas type module, like a Roland JV1080 or Korg Trinity then check out Sampletank and Sonik Synth from IK Multimedia and Hypersonic from Steinberg. These are huge synths concentrating on the kind of mass market sounds made popular by the hardware.

Software Samplers

If there’s one area where the hardware market has been completely decimated by software then it’s sampling. The sampler has been a vital part of any studio for twenty years, allowing the use of real instrument sounds, loops, hits and snatched performances in music production. An entire industry emerged creating CD’s of samples for your sampler, everything from saxophones to dripping taps, break beats to footsteps on shingle. In the space of a couple of years the two main hardware sampler players Emu and Akai would all but disappear. The main culprit was GigaStudio from Nemesys Music, now part of Tascam. Its killer technology was that it could use the computers hard disk to store and stream samples. The first version allowed for sample sizes up to 4GB’s (Gigabyte) in size. At that time hardware samplers commonly used 32MB’s of RAM for samples so the difference was staggering. Normally, if you wanted to create a sample of a real instrument you would record that instrument playing a couple of notes and put that into the sampler. The sampler would then speed up or slow down the sample to cover the notes in between. You may also make recordings at different strengths or velocity levels as the sound changes depending on how hard you play but with only a few MB’s of space there were limits. With GigaStudio came the GigaPiano which was an instrument made up of 1GB’s worth of samples. Each note on the piano was individually sampled and at multiple velocity levels giving an amazing amount of realism. As GigaStudio developed and other similar software samplers arrived hardware samplers simply couldn’t compete. I’ve installed GigaStudio systems into studios that have replaced entire racks of hardware samplers. To add insult to injury GigaStudio is completely compatible with hardware sampler formats so after a day of copying library across to the computer it is all instantly accessible off hard disk, no more searching around for that sample you had somewhere on a badly labelled floppy disk.

The main software sampler players other than GigaStudio are Kontakt from Native Instruments and Halion from Steinberg. Emu have recently rejoined the show as part of Creative Labs and have produced a range of professional music soundcards that include the EmulatorX software sampler that’s a direct descendant from their hardware samplers. Akai manage to hold on in the hardware market thanks to the live performance, hip hop and DJ markets. Their MPC range of samplers feature pads and built in sequencers for an altogether different beat making experience. All the soft ware samplers come with gigabytes of sample library, some of them containing the most amazingly sampled orchestral sounds. It is now possible to produce a completely convincing orchestral performance with sample libraries like the massive Vienna Symphonic Library and sample CD producer EastWest’s Symphonic Orchestra. Fantastic for the composer, being able to hear what his music will sound like without hiring an orchestra, but bad for orchestras and orchestral musicians? Maybe, maybe not. Talking to the ever level headed David Arnold, composer for many films including all the recent Bond movies, he finds that the quality of these libraries enables him to show the producers how good a real orchestra would sound before they invest in the process of doing it for real. This has resulted in more work for real musicians rather than less.

Romplers

That’s not a real term but it’s the best way to describe this wing of software instruments. These mainly come from the companies that used to produce sample CD’s. Rather than just selling the sounds for importing into your software sampler they’ve bolted them into their own software instrument interface so you can instantly play and use all the sounds and samples in your recording software. There’s no editing, sampling or synthesis going on, just the sounds with perhaps a filter or envelope effect, so they are essentially “Read Only” samplers or “Romplers”. This has resulted in a torrent of released from EastWest, Best Service, Zero-G and many more. Some are simply back catalogue re-releases whereas others are beefed up into new and useful collections of samples, loops and instruments.

MIDI Keyboards

With all this talk of instruments, latency and performance it might be worth mentioning the humble MIDI keyboard. Any keyboard with MIDI, whether synth, piano or electronic organ, can be plugged into a computer and used to play software instruments. All your computer needs is a MIDI interface – normally these come as part of a soundcard, but not always, so if you want to use an old MIDI keyboard make sure to check. Alternatively you can get a new MIDI controller keyboard that all come with a USB port for direct connection to the computer. MIDI controller keyboards can be simple one octave jobs all the way up to 88 weighted keys. Most now bristle with an array of knobs and/or sliders that give you MIDI control over the synths that you’re playing. It could not be simpler to connect up and get going and with a laptop and a MIDI keyboard you can take your entire synth rig with you.

There are tons of other software instruments out there covering whatever styles and sounds you want. To integrate them into your recording software you just have to be aware of the different formats:

Rewire – supported by Cubase, Sonar and Pro Tools

VSTi’s – supported by Cubase and Sonar

Dxi’s – supported by Sonar

Audio Units – Mac format for Logic on OSX

RTAS – supported by Pro Tools

So, read the back of the box and make sure it’s compatible with the studio software you’re running.

Propellerheads haven’t rested on their laurels either. Reason is a huge rack of synths, drum machines, samplers, effects, and mixers combined with a sequencer that is a completely self contained music production workstation. It retains the magic and ease of use of Rebirth but taken to many levels beyond. To top it all off Propellerheads discontinued Rebirth and have released it into the world as a free download from http://www.rebirthmuseum.com - how cool is that?