[seqfan] Re: Getting an integer sequence for a specific song

[Antti Karttunen]

On 6/16/19, Felix Fröhlich <felix.froe at gmail.com> wrote:
> Thanks for the replies.
>
> I agree that most popular songs are probably not of mathematical interst to
> add to the OEIS. I am also not sure complex songs (with a lot of
> instruments) can accurately be transformed into a sequence, though the
> basic rhythm probably could be. I had a song in mind from a favorite tv
> series of mine that has a simple basic rhythm that could most likely be
> turned into a sequence, but it is probably not of sufficient interest to
> add to the OEIS.

I think it's the opposite direction, turning sequences (and other
mathematical structures) to songs that's much more fertile field in
the long run.
For example, the discussion about the longest period of Life-patterns
on nxn toroidal grid is inspired by this sequence:

https://oeis.org/A179412

That pattern generates very nice dynamics / rhythm, when the rows of
the grid are read in binary and supplied as velocity-values for the
MIDI-notes, in this music machine I built from an old chess computer
and whatever parts and components:

https://www.flickr.com/photos/ermuggo/8118742971/sizes/k/in/pool-2089172@N20/

Note that the "melody" itself is generated from completely different
input, FPGA iterating over permutations of eight notes. Together, they
suffice to generate subtly changing algorithmic music for many days as
there are LCM(132, 8!) = 443520 rounds before it will repeat itself.

(Verilog-sources here: https://github.com/karttu/lifemidi )


Best,

Antti