Synthesis for the Masses

You'd have to be on a small uninhabitated Polynesian Island to not have noticed the synthesizer revolution over the last couple years. Keyboards and keyboardesque equipment such as sound modules and softsynths have been exploding to a point where there are actually more synthesizers in the hands of people that wouldn't consider themselves keyboard players than people who know what the black notes are for. (I'm just kidding about the black note thing, but you know who you are) The idea that you can press down a note and have it sound like just about anything you want, appeals to sound junkies of all types. Synthesizers can sound like an upright bass one minute and then a piano the next, and then some sound you've never heard untill then. Remember, I said "sound like". I wouldn't want to offend any real upright bass players, but, for me, I can play a much more realistic upright bass on a keyboard than I could on the real thing. That's the point, convenience. An extremely useful part as well is that many synthesizers can play all those sounds at the exact same time on seperate tracks. Non-keyboard players are getting into them, as a way to compose full songs to play over with their main instruments. You don't even have to be able to play a single note, you can draw notes on a staff in a computer, or use a computer music editor called a piano roll to draw in notes as blocks. This may seem like cheating a bit, but if what your going for is not trying to emulate real playing, you can create musical phrases that would be impossible to actually play, like intense techno or hip hop patterns. Making use of the music arranging power of a sequencer, you can get just about any song you've ever heard in a midi file, which is music in a form that a synthesizer can read as notes and performance data instead of the actual sound. You can then completely rearrange the song, change the instruments around, pull out the melody track, etc, so you can play guitar over top for example or make your own drum beat and bassline for a dance remix. This is only scratching the surface of the massive capabilities of modern synthesizers, but, of course, like any other instrument, they vary greatly in the types available, and not all synths are the kind that would be used solely to arrange a song, for instance. They come in a huge variety of forms many of which would be considered alien to new keyboardists and experienced players alike. In this article I will discuss some traditional and newer types of synthesizers as well as some technical terminology.

Multitimbral (timbre is really pronounced tamber, but everyone I've ever heard it uttered from has said timber, including pro types so I guess it's up to you) is one of the more common aspects of today's synthesizers. Multitimbral means "more than one sound" and though most keyboards can make more than one sound it refers to a keyboard's ability to play more than one sound at a time on seperate tracks. In brief, the way they work is that when used in conjunction with a sequencer, recorded performances play back on the keyboard as if you were still playing it. While that's going on, you can play and record more tracks. Sequencing is recording of the note's and foot pedal's action, etc. rather than audio recording which is recording of the sound itself. There are many advantages of sequencing compared to audio recording, one of which is size; midi (keyboard performance data) is tiny, one floppy disk can hold about 20 full length songs as sequence files at 16 tracks apiece. One floppy disk would hold about .5 seconds of a 16 track audio file. This makes midi files very internet friendly. Midi being primarily note data, is easy to rearrange and edit, and certain processes like quantizing (correcting note's timing) would be impossible without it. It also allows for flexibility in choice of instruments without being stuck to the original sound for instance. You can play and record on a flute sound for example, and change your mind at the end of the song and make it an oboe. This is not nessasarily instead of an audio recording which has other advantages. Sequencing is just before it, usually in the chain of events of recording a song. Music workstations are a common form of multimbral keyboard. They have a sequencer built in, but you can still use them with a computer sequencer or one in another keyboard if you like. Multitimbral just means it's able to be sequenced for more than one track at a time, so even if you don't have a built in sequencer you can still use an external one.

Technically 95% of synthesizers are multitimbral these days but if you go to a store and say "I'd like a multitimbral keyboard please" the store salesperson will probably assume you mean a keyboard with a wide array of sounds such as pianos, strings, drums, synthesizer sounds, etc. one that can do 16 multitimbral parts, and have any sound inside available with several button presses. The primary method these synthesizers use to make these wide variety of sounds is that they play samples from ROM memory. Samples are multiple recordings of notes and scales that can be derived from a wide variety of acoustic and electronic sources. The reason why a keyboard can sound like a piano for instance, is because you are playing an actual recording of piano notes. Even when you play an electronic sound on a ROM based synth you are playing a recording of something else that actually generated the electronic sound. The ROM (read only memory) is the storage for these sounds, it allows for the sounds to always reside in the memory of the keyboard to be pulled up at any time. However, ROM is expensive to produce because it must be custom made by the synthesizer manufacturer. The average ROM sample based synth has about 32 megabytes of ROM in which about 4-500 multisamples (instrument note recordings) are contained. 32 megs, however, is not enough memory to have even 20 full length note recordings of each instrument, so there are many tricks the manufacturers use to fit so many sounds into 32 megs. These methods do take away from the original sound's characteristics, however, are nessasary to be able to fit them all in at the same time. Samplers however, use RAM which is memory that can be written to, but generally empties itself every time you power off making sampler's use less convenient. With Ram, though, you can load sounds from sample libraries (generally, not included) on CD ROMS from C ROM drives (not usually included, either) You can, however, save custom banks of loaded sounds to hard drives for more convenient and organized storage. These, of course, are not usually included either. (geez) What, then, do you have to gain from a sampler? Because the sounds inside are non permanent, you can load much bigger memory sounds than would normally reside in typical ROM synth. In a synth with 32 megs of ROM, an average sound would be .05 megs of memory, whereas, in a sampler even with 32 megs of RAM, you could load specifically for the song you are working on, a few 10 meg sounds, and a few 4 meg sounds, for example. These sounds are hundreds of times larger in memory, so that much more of the original sound is retained. RAM is mass produced by computer manufacturers, so most samplers these days can have up to 128 megs inside. A sampler is the difference between the 10 or so sounds you have to load for this song being generally far more realistic than 400 sounds at instant access.

Not all synths are sample based however. There are also what I refer to as real synths. Sample based synths don't actually generate their sounds. They play back and process recordings of other things. They don't actually generate the sounds they play. A true synthesizer to me, and many that agree, is something that generates its own sounds. (usually electronic waveforms) These electronic waveforms can be manipulated with a far greater degree of control than a keyboard that is just playing back a sample of them. Many of the sounds that real synthesizers produce can not be fully reproduced with small memory samples (although samplers can do an ok job for some sounds) These include FM, wavetable, additive, and the widely popular analog type synthesizers that are making a huge explosion in the current market. Real analog synthesizers make sound from analog oscillators as opposed to digital microchips. When I said "analog type" synthesizers I was really refering to vitual analog which is a digital recreation of an analog synth. Real analog synths were the first electronic synthesizers and are, and were, very expensive to produce. Analog and virtual analog in addition to being used in more progressive rock are a key element to techno and electronic music styles.

Keyboards (and I use the term quite loosely) come in all shapes and sizes. Sound modules for instance have no keys (see what I meant about loosely?) and are played from other keyboards. Controller keyboards have no sounds, and are for playing other synths or modules only. A new breed of synthesizer can reside completely in the memory of a pc or mac. These are taking off due to their inexpensive cost and the immense power of today's computers. Some of them can do things no hardware counterpart could ever do, like play samples that would take over a gigabyte of memory for a single sound, and some of them make sounds no hardware based synth ever could before. Despite loving the keyboards of today's amazing technology, I'm still am a fan of the old real analog synths; To me they each have a unique almost acoustic character like pianos or guitars. There is so much to choose from right now that classifiying playing "keyboard" as one instrument is more impossible than it ever was, especially when it can often refer to playing somthing that has neither keys nor a board. Soon we'll just be thinking notes into a computer, I'm sure.