Data AcQuisition And Real-Time AnalysisScope - Spectrum - Spectrogram - Signal Generator
Software for Windows
Science with your Sound Card!
Contact us about
What is music? What is there about certain sounds or patterns of sounds that we find pleasing or interesting, while others may be unpleasant or boring?
These questions are not trivial to answer. In fact, it is unlikely that a discrete answer can even exist. People differ in their preferences, and any individual preference may change from moment to moment or develop over time.
But there are some overall principles, rooted in physics and physiology as well as in the more mysterious wiring of the brain.
In general, we find a sound pleasing if it is composed of frequencies that are integer multiples (called harmonics) of some fundamental frequency. If some of the included frequencies aren't integer multiples, we find the sound unpleasant. If many unrelated frequencies are present we call it noise. (So-called white noise sounds like a hiss.)
Two simultaneous sounds will be pleasing if their fundamental frequencies are related by a small integer ratio, like 2/3 or 3/4, which will cause their upper harmonics to align into a unified structure.
But even a pleasing sound soon becomes boring if it continues without change for a long period. This may be related to the well-known tendency of the brain to respond most strongly to a novel stimulus of any sort, and to ignore a constant stimulus.
This is true in general for all sensory systems, including vision, taste, smell, and touch, as well as hearing, and in organisms right down to the most primitive. We are all familiar with the common phenomenon of "getting used to" a persistent background noise, color, flavor, odor, or clothing feel.
On the other hand, it is easy to understand why a novel stimulus should command our attention, as indeed it commands the attention of every organism with a sensory system to detect it: A change in the environment may signal something that requires action, such as an opportunity to feed or mate, or the need to escape from danger.
So, simplistically, we might say that music consists of "mostly pleasant sounds that keep changing". But the ear/brain system also has, effectively, a (very) short-term memory for sounds, such that it detects the harmonic relationships between sequential sounds. As with simultaneous sounds, small integer ratios of their fundamentals are preferred.
This means that changes in the sound must also be constrained to a general harmonic pattern or they will sound discordant. We thus generally confine music to discrete notes, which have fixed relationships to each other.
Conventional Western music uses a system of 12 notes related by equal temperament: Given any starting frequency, the next note is higher by a constant factor equal to the twelfth root of two (2^(1/12)), or about 1.05946. If you multiply any number by this factor 12 times, the result is an exact doubling of the original... one octave. The 12-note cycle then repeats for each higher octave.
The main feature of this equal-tempered scale is that for any given starting note, there are several others that can be selected to provide reasonable approximations to small integer ratios. For example, there are 5 of these "semitone" steps between 'C' and 'F', giving a total ratio of 2^(5/12) = 1.3348, which is close to 4/3 = 1.3333. Likewise, there are 7 semitones between 'C' and 'G', giving 2^(7/12) = 1.4983 which is close to 3/2 = 1.5000.
So, we might adjust our simple definition of music to say that it consists of "pleasant sounds in a pleasing sequence", where pleasantness is a function of harmonic relationships.
Now things start to get interesting. Suppose we encounter some pleasing sequence of pleasant sounds, whatever that may be. When we first hear this, we immediately identify it as "musical". But with continued repetition the novelty wears off and the pleasantness fades, possibly becoming unpleasant to the point of torture. This happens sooner for short sequences like children's nursery songs than for long symphonies.
Our brains have highly-developed pattern detection abilities. We can perceive patterns on many different levels... patterns within patterns within patterns. And as with the detection of novel stimuli, we are also sensitive to changes in patterns.
To be pleasing, music must have variations over time, in order to excite our novelty detectors. But it can't have so much novelty that we can't find a pattern. (This is also true of games, books, movies, and most art.) Music thus lives on the frontier between order and disorder, stability and chaos.
Daqarta's DaqMusic allows you to explore this boundary. You would typically start with a stream of random equal-tempered notes from a Chromatic Scale, extracted by Pitch Track from an internal random source (or a Generator noise source), limited to a certain note range, and played by Pitch-to-MIDI at a certain Tempo.
With no other constraints, this endless meandering stream is somewhere between boring and unsettling. Limiting the Scale to (say) Pentatonic Major increases the general pleasantness by eliminating the less-harmonic combinations that can arise with Chromatic (all notes), but it still lacks purpose.
But it can really soar if a Changes script randomly raises that range for a while, every now and then, along with occasional driving sections provided by chords added to the second voice.
The Symphonic.DQM setup file is an example that uses all of the above. You may be surprised at how musical this can be, considering that it is just a few simple operations on a random note stream, without any predefined melody. It plays on forever, always changing, but after a while your brain's higher pattern detectors will notice that the overall "sound and feel" is the same. This may be fine for, say, background music in a video, but probably not for general listening.
Another path along the order/chaos frontier is used by the JukeBox.DQM setup. It generates two different patterns of 7-15 random notes each, and uses them in an AABABA sequence to create a random melody using a random instrument. It plays a couple of verses, holds the final note, rests, and then creates a whole new song with a different instrument. Unlike Symphonic there is a wide range of performances, but the occasional "good" performance is probably not enough to keep your interest through the intervening "losers".
To address that issue the "Glossy" family of setups each use 8 independent JukeBox-type voices, each playing a random melody of a random length, with a random instrument. The combination of melodies "works" because all use the same Scale, and the random melody lengths means the interactions are always changing.
The heart of the Glossy series is GlossyFish.DQM, which also sets random overall Levels for each solo voice performance, and randomly chooses to stretch all the notes of some voices to give them slower tempos. These features add to the combination of short-term patterns with long-term variations to keep listener interest.
Best of all, you can experiment with different Scales during a performance using hot-keys.
GlossyFishSticks and GlossyFishBones combine GlossyFish melodies with GlossySticks and GlossyBones percussion using the same multi-voice random pattern strategy for 6 non-tonal Percussion instruments.
The culmination of the Glossy series is GlossyTracks, with over 4 billion unique "tracks" of controllable duration, each with Scale, tonal instruments, and percussion instruments randomly chosen from editable text file list preferences, and tempo chosen with a selectable non-uniform random distribution.
Alternatively, PhrygidBlueJam.DQM moves a bit closer to the "chaos" edge of the frontier by dispensing with melodies completely. Each voice just plays 40-60 notes at random distances within +/-1 octave of the current random source note, then changes instruments. The range of notes accepted from the random source also changes every 40-60 beats, which shifts the central pitch. When the source pitch is above the limit it also causes some notes to be dropped, giving a staccato effect and even occasional dramatic pauses. The default Blues Phrygian scale keeps the whole ensemble sounding like a loose "jam session" of blues musicians.
You can use any of these setups as a starting point. Try changing tempo dynamically. Use Spectrum Curves to bias the note range, or toggle the Generator on and experiment with different noise sources other than White ... maybe even using a Play loop of a .WAV recording as the "noise" source. Create a random melody sequence as in the JukeBox and Glossy setups, but instead of continually replacing it, try incrementally modifying it. Or copy the original and allow it to reappear occasionally as a recurring theme after many modifications. Give the performance an interesting overall structure, perhaps like the movements of a conventional symphonic piece.
Welcome to the frontier!
Questions? Comments? Contact us!We respond to ALL inquiries, typically within 24 hrs.
Over 30 Years of Innovative Instrumentation
© Copyright 2007 - 2013 by Interstellar Research
All rights reserved