Sounds Sensational !
6 min read
Amiga Sounds Machine
Christopher HUMPHRIES explains why the AMIGA sound chips leave other computers dumbfounded
THANKS to the Paula sound chip, The Commodore Amiga has the most powerful sound synthesis hardware of any micro available today. No 8 bit machine of my acquaintance was as powerful as this. The Amiga’s nearest rival has to be the Acorn Archimedes, as Atari fell out of the running when it failed to put the much talked about Amy chip in the ST.
Both Amy and Archie could compete with Paula in terms of sound quality. Where the Amiga really comes into its own is the sheer practicality and ease of use that the sound hardware offers.
Back in the days of 8 bit home computing, sound consisted of little more than a varied assortment of beeps, hisses and gritly “hang” noises. As far as actually producing sound on a computer. I would use things called sound channels to produce musical notes of different pitch and volume.
In this respect most of the home machines of the time seemed superficially pretty equal. Each computer had between one and four or so sound generating channels, and most could also use some or all of these channels to produce something called “white noise” to make sounds which almost resembled explosions and gunshots.
And that is how sound quality on the affordable middle range of home computers stood in the early 1980s. Then things began to change. Slowly but surely sound demos appeared that were like no previous sound heard on these machines : music that played with REAL sounding instruments, humans singing, talking and shouting. Even programs that showed promise as genuine synthesisers of human speech.
I WAS very impressed, and surprised, for although these brief snatches of sound were muffled and more that a little on the tinny side, they were my first introduction to the world of digitising. That is, the recording of an analogue sound signal, such as you might find on an LP or magnetic casette tape, and its conversion to a digital format that a computer could store in memory and play hack at will.
The sound I was used to hearing on computers, those almost electrical sounding tones, were called square waves : perhaps the simplest form of sound that your Amiga can produce.
Sound travels to your eardrums as vibrations in the air, a series of pulses of consecutively lower and higher air pressure travelling at the speed of sound, which varies with respect to atmospheric pressure, hut at sea level is about 761 mph (about 1224 kmh). Sounds can be represented as a graph showing how the air pressure varies as time progresses.
The attributes of a sound are therefore related to the shape and overall nature of this waveform. For example, if it is regular and repetitive, the sound will resemble a musical note.
We know that different musical instruments are capable of producing the same notes, and yet at the same time sound entirely different in every other respect. This is because their waveforms, although repeating at the same frequency, are very different, this shows on a graph of their waveforms.
On a graph a square wave looks like a line following the battlements of a mediaeval castle. It goes up. along, down, along, up. along, and so on repeatedly. The distance between each repetitive unit gives the wavelength. By measuring the time elapsed on the graph the frequency of the note can be found.
The amplitude of the wave, that is the vertical distance between its peaks and troughs, represents the volume : the change in pressure of the air that the source of the sound has produced, The human brain translates this signal into what we perceive as sound.
But what about “real” noises?
People talking, dogs barking and the screech of car tyres on a wet road? What do these sounds look like if we were to draw them as waveforms on a graph? Although they have a frequency and pitch, they are very complex waves. Very little in them repeats regularly. If our simple wave produced by an 8 bit micro resembles castle battlements, the sound of a person’s voice resembles that of the most jagged mountain you can imagine.
What hope then for machines trying to produce waveforms of such complexity? Not much, I’m afraid. All they could do and this applies to the Atari ST also, and I intend no slur on that machine would be to use raw processing power to pour byte upon byte of data the “points” on the waveform graph into their sound chips as a series of different volumes of sound.
This technique works because it is essentiality volume or amplitude changes that define the shape of a sound wave. Unfortunately, the process ties up the computer’s main processor for an exhorbitant amount of time, restricting the user to either simple slow programs with reasonable sound, or good fast programs with only fair to average sound.
Since to update a sound with a frequency of, for example, 10,000 “up/downs” per second 10,000 cycles per second means 10,000 updates per second it is easy to see why most programmers steered clear of digitised sounds.
But what if a computer were to be introduced that could lake a digitised waveform of sound, a waveform about 16 times more smooth and realistic than any heard previously, and play that waveform without any burden on the computer?
It would need a large amount of memory since complex “real” sounds need upwards of about 8,000 bytes a second if they are to be convincing. It would need special hardware to handle the playing of these sounds. And to add an extra sparkle of reality, why not throw in a stereo option so that the computer could send sound to two separate sources at the same time?
Such a computer exists. It is the Commodore Amiga, designed by the brilliant and innovative Jay Miner as perhaps the current ultimate in computerised home entertainment. The Amiga can have up to four sound channels all playing separate and indépendant digitised waveforms. It can take these as two pairs and send them out the back of the computer as a stereo signal, each stereo line getting two of the computer’s sound channels.
This is the way in which recent chart hits have been made from sampling various musical pieces and sounds and mixing these together with a vocal track. The Amiga can do all this and more. Provided you buy the appropriate hardware which comes with its own software you can also sample sounds for yourself.
Good examples of these kind of sound sampling hardware packages are Future Sound from Applied Visions, and Perfect Sound from Sunrize Industries. Both will let you explore sound on your Amiga and help you exploit its full potential. They can be a lot of fun, since once you have sampled your sound you can then play it back at any speed or sample rate. By doing this you can produce different pitches from the same sample.
FOR those who prefer to avoid paying for such extravagant things as hardware add-ons, why not try a straight Amiga music package? I can guarantee that a music package on an Amiga will be like you’ve never heard it before: Real drums, real guitars and even real voices. These can be a lot of fun, and good examples are Aegis Sonix by Aegis, or Deluxe Music Construction Set by Electronic Arts. Both are very versatile and a lot of fun to use. as well as allowing serious musical composing.
Sound on the Amiga can be fun even at the hardware level, programming in 68000 assembly language. Machine code is perhaps a little too unwieldy for the easy expression of ideas and programs, but absolute power over the Amiga hardware can yield great rewards when it comes to sound processing.
Sources: https://archive.org/details/amiga-computing-magazine-001
Source of information – AMIGA Computing of June 1988 – Download of the magazine below:
