SOUNDMACHINE ST by Richard Karsmakers
Introduction; the principles of digitized music
Back in 1987, right in the beginning of the ST's era, a hackin'
group called 'The Desaster Area' made a couple of sound demos
("Oxygene", "Matt's Mood" and another one of which the title
slipped my mind) that excelled by their good quality and their
These demos consisted of massive datafiles containing digitized
sound data. For about 2 minutes of good quality music, 600 Kb of
memory had to be used - hence the reason that these demos only
worked on 1 Mb machines with a double-sided disk drive.
What exactly is 'digitized music'?
As you all know, sound consists of waves. These can vary from
pure sinuses to rather complex waves that e.g. belong to the
human voice or belong to explosions. There is a way to create
these waves on the ST as well, but therefore the analog sound
information has to be transformed into digital sound information.
This process is called 'digitizing' and the device you need to do
that is a 'digitizer' that usually contains an AD converter
(Analog-digital converter). This digital format is needed because
a computer only 'understands' numbers.
A piece of digitized music is called a 'sample'. Nowadays, the
technique of using samples in music is very common. Drum
computers use sampled drum sounds, and some groups ONLY use
samples (e.g. the Art of Noise and a lot of that current hip hop-
, Acid-and New Beat trash).
It is possible to digitize a piece of music and play it back
through the soundchip. The only register that is used here is the
volume register; playing back digitized sound is in fact just a
controlled version of the noise that would appear if you would
turn the volume of the soundchip to maximum and minimum in a very
But if a piece of music consists of several parts that sound the
same, it is possible to record these separately and play them
back in a specific order. This kind of mixing makes the music
appear to be longer, or allows the programmer to use less memory
This kind of 'mixed' music often appeared in games or demos.
Except for the fact that it takes a lot of processor time to play
digitized music (especially at a high frequency) and other tasks
of a program running the music will be harder to program, using
digitized sound is very easy. That's why I usually dislike
digital music when used in games; it just consumes a lot of space
and is very easy to program (even yours truly could do it).
But there is, as you could have guessed, an even more advanced
way of playing back sampled sounds. This method is sometimes used
by professional sound programmers and was first tried out by
Jochen of The Exceptions in the "B.I.G. Demo".
This way consists of single samples sounds (e.g. a guitar chord,
a bass drum, a flute note, etc.) and storing them in a much more
compact file. When playing them back, you can perform some
complex calculations that play the instrument sample on different
heights, thus obtaining different notes. When you want to make a
song with just a solo guitar, a bass guitar and a single drum,
for example, you would just have to sample ONE guitar note, ONE
bass guitar note and ONE drum sound.
This saves a LOT of memory space, since now you only need three
sounds instead of much larger segments of a song. And if you want
to use different sounds, you simply sample them, too. The program
will then play them back.
But what if you want a DRUM sound and a GUITAR sound to be
audible simultaneously? All the previous methods only played
'with one voice'. Since the ST has three voices, Jochen found out
that it is possible to play back three voices as well, by using
the individual channels.
Now we're getting somewhere. Not only can we now use digitized
music while consuming a great deal less memory, but we can also
use three channels together (in the "Amiga Demo" - March 1988 -
Jochen even uses FOUR channels by adding an extra one to the
third voice's data). Even advanced sound options (transpose and
portamento) are possible using this technique!
The only disadvantage of using three separate voices and
digitized samples is time: It uses up just about ALL processor
time there is (there's the answer why there are no other tricks
in the "Amiga Demo" digi-sound screen!).
Tommy Software's "Soundmachine ST"
Building upon the concept of THREE voices and single digitized
samples of musical instruments, Tommy Software's Ute Wickenhäuser
and Jürgen Piscol designed a program that would not only allow
efficient playback of this music, but that would also allow the
user to enter the notes themselves - just like on sheet music.
The "Soundmachine ST" was born.
The program is no doubt about the most revolutionary music
program to date on the ST. Ever heard of "Sonix" on the Amiga?
Well, "Soundmachine ST" allows you to do the same (only with
three instead of four voices since the Amiga simply has one
more). The main difference is the processor time used: The Amiga
has a 'digital soundchip' (whatever you want to call it) built
in, so it doesn't take much time to play digitized music. On the
ST, it does. It is quite impossible to do anything else at the
same time, or you'll have to compensate on the sound quality.
"ST Soundmachine" allows you to play samples, and therefore the
whole disk (plus a second one) of the package are filled up with
samples. They use a different format than the 'regular' sampled
sounds that are produced by sample programs - but that needn't be
a problem because conversion is possible.
There are two kinds of samples: First, there are the ONCE
samples; these sound once (e.g. drum sounds) and then fade away.
Next, you have the HOLD samples (e.g. violins) that just continue
to sound until they are no longer wanted.
Each sample can be assigned its own volume value, and each one
of course has its own length.
The user interface
Mainly, there are two screens inside "Soundmachine ST": The
screen on which you put down notes and special commands (more
about those later) and the screen on which you specify sample
volume, shape forms and such.
The whole program is mouse-controlled (very much in a way like
their old "Musix32" program), though many keys have been assigned
a function, too. You can load/save songs, samples and shapes
(shapes are means of getting vibrato effects), you can 'compile'
a song (so that you can play it back in your own programs!),
print playfiles (songs) and get e.g. info about free memory and
stuff like that.
It is easy to control, although it may look a bit messy and
incomprehensible at start (especially the shape/sample edit
The Special Commands
In between the musical notes (in the music notation), it is
possible to use special commands. These can function to slow down
playback frequency (thus allowing more time for other programs
running), jump to other measures, transpose, use portamento (one
note 'glides' into another one that's higher or lower) or legato,
or to change the playback speed.
These special commands make the "Soundmachine ST" into a
valuable sound programming tool!
The Beatmachine ST
There's also another program on the disk: "Beatmachine ST". This
program allows you to construct complex samples (e.g. rhythms)
for later use in the "Soundmachine ST". The program is also
mouse-controlled and easy to use.
"Soundmachine ST" comes with a 40-page manual (I got it in
German). It is clearly written, and it also explains (with source
material) how to use the music in your own programs - which after
all is quite important. It is illustrated as well. Technical
terms are evaded, and it starts with explaining the basics of
The "Soundmachine ST" is quite revolutionary (just like Tommy
Software's "Musix32" was at the time of its launch) a sound tool.
It is quite easy to work with, and has really powerful options.
The demo songs on the disk are already quite impressive
(especially PRAELUDE) and really made me sit back and be amazed.
A good product - like we're used from Tommy Software! The
samples that are contained on the program disks really allow you
to experiment to you heart's content - and if you find them
insufficient you can get the additional sound library disks!
Name: Soundmachine ST
Company: Tommy Software
Value for money: 8+
Overall rating: 8
Price: DM 148,--
Remark: Quite unique...
Hardware: Monochrome and color, all memory
Additional soundlibrary disks are available at DM 79.95 each
(there are two available; Soundlib 01 and Soundlib 02).
For info, you should contact Thomas Maier at:
Selchower Str. 32
D-1000 Berlin 44
The text of the articles is identical to the originals like they appeared in old ST NEWS issues. Please take into consideration that the author(s) was (were) a lot younger and less responsible back then. So bad jokes, bad English, youthful arrogance, insults, bravura, over-crediting and tastelessness should be taken with at least a grain of salt. Any contact and/or payment information, as well as deadlines/release dates of any kind should be regarded as outdated. Due to the fact that these pages are not actually contained in an Atari executable here, references to scroll texts, featured demo screens and hidden articles may also be irrelevant.