Waveform Palindrome Music Technique: Part 2
Writing a Full Song
In Part 1 we learned how we can turn any sound into a waveform palindrome. Using this technique, we can force an entire song into a waveform palindrome. If you were to pick a song at random, however, and force it into a palindrome, you'll soon realize that the second half will sound quite obviously reversed and not particularly musical, unless it was specifically designed to sound appealing even when reversed. Thus, when writing palindrome music, the only real restriction is making sure that whatever written will sound acceptable when reversed.
When you write a full palindromic song, you'll only need to write the first half, as the second half is simply the first half reversed. Keeping that in mind, it is important to keep the music varied and progress at an unusually fast pace as in the second half you'll be hearing it all again.
Throughout the rest of the tutorial, I will reference the "second half". The second half is referring to the second half a palindromic piece—that is, the reverse of the first half. Since when dealing with waveform palindrome music, the "first" and "second" halves are indistinguishable, I am really just referring to the half that is produced when you reverse the half that is originally composed by hand. The second half is perhaps the most difficult of a palindromic track, as it is the product half, and the half that is most likely to sound unacceptable. Since any music can be turned into a palindrome simply by appending the reversed copy (the second half) to the end of the original, a good waveform palindrome is one in which the second half sounds acceptable. As said before, composing waveform palindrome music is all about making the second half sound good.
A Palindrome in a Palindrome
Note the following English palindrome: "stack cats stack cats". It is simply one well-known palindrome copied twice. In fact, it can be shown that any palindrome can be appended to itself any number of times and still form a palindrome. What does this have to do with music? Well, music is quite repetitive by nature. Whether the notes are for a particular theme or a particular beat, music is given cohesion by a certain amount of repetition. What does this have to do with waveform palindromic music? To give a waveform palindromic track cohesion, as with any other sort of music, some repetition is required, and if you make the smaller, often repeated elements of the song also palindromic, they will sound the same on both haves of the line of symmetry. In other words, one very useful technique is to compose waveform palindromic pieces out of smaller palindromic units.
As a primarily electronic musician, good drum beats are vital to my music. Making the beat either palindromic or sound acceptable when played in reverse is important to making a waveform palindrome. To make reversible drum patterns, you have a couple options: either simply force it into a palindrome or design it in such a way that it will sound fine even when reversed. The first option is the easiest, and often turns out surprisingly well. Simply mix a reversed copy of the drumloop with the original, allowing an offset of 0. Yes, the drums will sound quirky and blippy with reversed drum hits mixed in, but the drumloop will sound exactly the same during the second half as it did in the first. The second option is to write two beats, one for the first half, and the other for the second half, and then mash them together, with one reversed. It will sound as just as blippy, but it will be different in the second half, and is good for adding variation.
Melodies too can be palindromes. You can either force the waveform into being a waveform palindrome, or instead use melodies that are note palindromes (that is, use notes that are reversible). As with drumloops, this adds cohesion to the piece. Just remember the pointers found in the section Note Durations, or what looks to you to be a palindromic melody might end up sounding like something very different on the second half.
The main danger with using too many sub-palindromes in your palindromic music is that it may sound too repetitive. The second half will sound end up sounding too much like the first. While consistency is important, variety is possibly even more important. Instead of using true palindromic melodies, using pseudo-palindromic melodies can often produce a desirable effect. By a "pseudo-palindromic melody", I mean one in which the first and the last few notes mirror, but the middle notes of the melody do not mirror. In this way, the melody in the second half will start the same as the melody in the first half, and thus produce some cohesion, but will progress in a different (but usually acceptable) manner, thus adding some variation.
Choosing Instruments
Since during the second half the instrument sounds themselves will be reversed, you will want to choose instruments that sound acceptable when reversed. If the particular patch or instrument that you choose has a great deal of gain-attack, it will sound somewhat parallel to release or decay during the second half. However, the reverse is not true. If the patch that you are using has any gain-release, the end of the release will become the beginning of the note, effectively breaking the time, and sound very out-of-rhythm in the second half. If the sustain level is quiet enough, a similar effect will occur. This issue is the very closely related to the issue described in the next section, Note Durations, as it occurs also when writing melodies.
With this in mind, any instruments with a lot of release or decay, such as any plucked or percussive sound, will be very difficult to get to work in a palindromic piece. Instead, the waveform palindromist must use mostly those leads and pads without any decay or sustain.
Similarly, it should be noted that any portmanteau will sound out of time, as the pitch-slide will not be when the new note is triggered, during the second half. Of course, a little portmanteau might be still acceptable, if the effect is not too noticeable.
Note: If you are not familiar with the concept of a ADSR-style envelope, then try reading through some of these documents: Wikipedia, Samplecraze.
Note Durations
Normally when composing music, the exact duration of a note is not as important as the exact time when that note is to be played. For example, if a note were to be held a half a second longer than intended, the result would be barely noticeable. However, if that same note were to be instead played a half second later than it should, it would probably break the musical time and sound very out-of-place. This fact becomes of vital importance when you are sequencing melodies. Even if you are using a good instrument (see above section), and even if you are using a note-palindrome melody, your melody may sound unbearable in the second half if you are not careful with programming your durations. This is because when a melody's waveform is reversed, note ends become note beginnings. If your note's ending is not in time, then its beginning in the second half will not be in time. In short, you should be sure to keep all of your notes ending in time.
In the image above shows a piano-roll or matrix view. The free/libre program LMMS was used as an example. The yellow bars represent notes, and the vertical grid represents musical time. Three melodies in 4/4 time are shown, each transposed an octave, and each spanning two measures. After taking the screenshot, I added two purple lines to separate the three melodies, and several blue and green circles to point distinguish note endings, using a free/libre image editor.
As you can see, in the first melody the notes are ending at odd times. If the melody were to be played in reverse, the timing would sound very broken, as the notes would be triggered between beats, and thus sound as if they were following several 1/128th note rests, or even completely off-time (depending on the snapping rules of your sequencer). The note endings circled in blue do not end on-time, and are not acceptable for a pleasant sounding waveform palindrome.
In the second melody, the issue is somewhat remedied, and possibly fully remedied if in this particular piece some syncopation is acceptable. As you can see, every note ends on a quarter-note interval. However, the two notes whose ends are circled in green will produce some syncopation in the second-half. This is fine if it fits with your music. If it doesn't, however, then the notes can be extended to being half-notes, and the issue in this example will be fixed.
Finally, in the third melody, the notes end precisely on time, in such a way that no syncopation will be produced in the second half. In most non-experimental genres, although you'll want to keep the notes lengths exact, you may still wish to introduce syncopation in the second half.
As mentioned before in the introduction to this tutorial, you'll almost certainly want to use a piano-roll or matrix view for entering notes, as traditional score notation is far too imprecise regarding note end-times.