Randomization and Crossfading of Musical Sections

[joseph]

@herrhass

I would need to assign a probability and a crossfade to each and every marker. I have 8 sections, so I would need to add 7 transition markers per section and define a behavior for each of them, which is really really time consuming and a not really scaleable approach.

If you’re planning to use transition markers and transition timelines, there is no way around this.

However, if you are able to use magnet regions instead of transition markers, you could drastically cut down the number of markers required, from 56 to 8. While a normal transition region tells the playback position to “go over there,” a magnet region tells it to “come over here,” so magnet regions are an effective way to cut down on the number of markers required for this kind of “from any section to any section” event.

The disadvantage of magnet regions is that they cannot be set to only affect the playback position at specific points on the timeline. The closest you’ll get is quantizing them so that transitions only happen on specific beats and bars, or setting them to be relative transitions so that the playback position always jumps to a position relative to its destination region that’s equivalent to the position it had relative to its source region.

Alternatively, if you don’t use transition timelines but rely on modulators or some other means of crossfading the instruments at the source and destination of a transition, you could create a single “logic point” containing eight transition markers at the same timeline position, and then just create a single transition marker back to that logic point at the end of each of your sections.

The disadvantage of using a logic point is that you would have to use asynchronous instruments to ensure the fade-ins and fade-outs of instruments play out properly.

@herrhass

Even when following through with this approach, there’s the possibility that there’s no transition at all because I would need to assign a probability for each transition marker which is below 100%.

You are mistaken. When multiple transition markers exist at the same timeline position, they are evaluated from top to bottom, and lower markers are evaluated only if the trigger conditions of all higher markers are not met. Thus, if you have seven transition markers, you can ensure they have equal chance of being triggered (and no chance of the playback position escaping all of them) by setting their probabilities as follows:

Topmost marker: 13%
Second marker: 14%
Third marker: 17%
Fourth marker: 20%
Fifth marker: 25%
Sixth marker: 33%
Seventh marker: 50%
Lowest marker: 100%.

This will ensure that each marker at the same timeline position has an approximately 1 in 8 chance of being triggered.