<HTML>
<TITLE>Sequitur Decay Pack v1.0</TITLE>
<CENTER><H1>  Sequitur Decay Pack v1.0 </H1><br> 
 --<em> MIDI processing filters for Sequitur 1.1.x --<br>
by <A HREF="mailto:civello@pacbell.net">Daniel Civello</A></em></CENTER><BR>
<HR width = 380>
<P>
<CENTER>
<strong><em>To get the most out of the Decay Pack filters, you should have the latest version of Sequitur (1.1.2) installed.  Download it from the <A HREF="http://www.bebits.com/app/1576">BeBits</A> page or directly from <A HREF="http://www.angryredplanet.com/beos/Sequitur">Angry Red Planet</A></em></strong>
<BR></CENTER>
<P>
<BR>
<strong>Installation:</strong> copy the contents of the 'filters' directory to your <em>$PATH/Sequitur-1.1.x-86/add-ons/Filters</em> directory.  The new filters will load automatically.  Run Sequitur (if it's not already running),  select Windows --> Filters, and start writing music! <BR>
<HR width = 570>
<BR>
<em>Decay Pack</em> is a set of simple filters written to aid in algorithmic MIDI composition.  The <em>Decay Pack</em> filters were designed specifically for music composition in Sequitur, and to maximize their usefulness, they should be used in conjunction with other Sequitur filters.   Please refer to the documentation distributed with Sequitur (or online at <A HREF = "http://www.angryredplanet.com/beos/Sequitur/Documentation">http://www.angryredplanet.com/beos/Sequitur/Documentation</A>) for a description of what filters are, and how you can use them to help create music.<BR>
<P>
Here's a short (174K) <A HREF="sample/pink.mp3">sample</A> mp3 file to give you an idea how these filters can be used.    The details of how this riff was created are at the bottom of this readme. 
<BR>
<P>
In this 1.0 release there are four filters. A good starting place for using these filters is the Edit pipeline -- that way you can visualize how each filter processes MIDI events.  The current release does not support dynamic play-through of the filters, but future releases will.<P>
<table border = "1"> 
<tr> 
        <td align="center"><img src="images/decay.jpg" ></td> 
        <td><strong>Particle Decay<BR></strong> 
Once a note is placed down, Particle Decay will add a series of descending notes until they reach the bottom of the keyboard.  This filter is basically a simulation of radioactive decay.  There is a degree of randomness in how this filter functions, which [may] affect the slope of the values, so it's a little more than just dividing the pitch by 2 for each new note. <P> <em>Suggestion: </em> Immediately after placing the first note, you can move the group of new notes around by clicking and dragging on any of the red notes.  This gives you more control of the slope of the riff.  Experiment with adding another filter after Particle Decay, for example, change your Quantize setting to place down 16th notes, and add an Example filter to the pipeline.&nbsp;
</td> 
<tr> 
        <td align="center"><img src="images/erode.jpg" ></td> 
        <td><strong>Erode<BR></strong> 
Erode is a very simple filter which places new notes at half the duration of the previous.  It has a very interesting effect with deep bass or drum patches.&nbsp;
</td> 

<tr> 
        <td align="center"><img src="images/dissolve.jpg" ></td> 
        <td><strong>Dissolve<BR></strong> 
Dissolve is just a slight modification from Erode, but can produce very different results.  Instead of just placing new notes at half the duration of the last, the next value is also one note lower, and 20% of the velocity -- so it creates a fading effect.  <P> <em>Suggestion: </em>Try placing a Duration filter after Dissolve so that new notes overlap the previous -- it creates a more natural and smooth fade.&nbsp;</td> 

<tr> 
        <td align="center"><img src="images/pink.jpg" ></td> 
        <td><strong>Pink Noise<BR></strong> 
Certainly the most interesting of the lot, this filter implements an algorithm originally written by Richard Voss.  For each note placed down, it creates a series "pink noise" MIDI notes.  A mathematician named George Birkhoff said, "...for a work of art to be pleasing and interesting it should neither be too regular and predictable nor pack too many surprises."  In a "white-noise" process, every value is completely independent of its past -- it's a complete surprise.  In "brown-noise", only the note increments are independent of the past, which in music, creates a boring, slow changing tune.  Apparently, what most listeners like best is music is where the succession of notes is neither too predictable nor too surprising.  Translated to mathematical functions, this corresponds to a frequency dependance plot (f^a), with the exponent a between 0 and -2.  Richard Voss discovered the exponents found in most music are right near the middle of this range, i.e. a = -1.    These functions were dubbed 'pink' because they fell right between "brown-noise" (f^-2) and "white-noise" (f^0).   In its current version, Pink Noise creates 20 notes for each note placed. <P> <em>Suggestion: </em>Since the notes coming out of this filter are of a random nature, using Sequitur's On Key filter right after Pink Noise is a great way to keep all the notes of your riff in the same key..&nbsp;
</td> 
</table>
<P>
<BR>
<H4>An Example Edit pipeline configuration for the Pink Noise filter</H4><BR>
The sample mp3 piece at the top of this document was created with the following configuration for the Edit pipeline.  It basically consists of three groups of notes generated by Pink Noise, and only minor editing.  The instrument used in this riff is an acoustic guitar patch from a Roland JV sound module, but try General MIDI patch #108, or #112.<P>
<img src="images/setup.jpg">
<P>
<HR width = 570>
<BR>
So, experiment and have fun!  Music composition using filters can be difficult at first, but when you get the hang of how filters feed into one another, it can be a lot of fun, and you can produce some really interesting pieces! <P>  Don't hesitate to <A HREF="mailto:civello@pacbell.net">email me</A> with any questions or comments.  <BR>
-daniel civello <BR>
02/28/01<BR>
<P>
<BR>
<HR>
References for Pink Noise filter:<P>
Madden, Charles. <em>Fractals in Music.  Introductory Mathematics for Musical Analysis.</em> No. 1.  Salt Lake City: High Art Press, 1999.<P>
Schroeder, Manfred  <em>Fractals, Chaos, Power Laws.</em> New York: W.H. Freeman and Co.,1991.
<HR>
<em>February 28, 2001</em><P>
</HTML>
