45 Minutes of 45 Time-Stretched Paul Stanleys All Stage-Bantering At Once
by the guy from The Loss Foundation, which normally releases about 3 minutes of actual music a year
5.12.13

So, right. For some reason, I took the delightful minor meme that is "45 Minutes of Paul Stanley Stage Banter" and split it up into 45 clips of different (carefully chosen) lengths. I then drastically time-stretched each of those (usually) very short clips to 45 minutes each, and assembled the final result into an experimental ambient "work."

The "work" starts with a single, obviously uber-stretched Paul Stanley rock-bantering, and adds one or two clips at a time until all 45 clips are heard playing together simultaneously for the better part of... 45 more minutes of Paul Stanley stage banter.

The result is nothing more than highly-structured silliness, but it's still semi-entertaining ambient listening which captures the dream-spirit of stadium rock quite nicely in my mind.

You can hear Paul scream-speaking his way through a ridiculously-extended helllllloooooooo Albuquerque / wellllll alllllriiiiiiiite just often enough, particularly as the piece begins and ends, to know where exactly it comes from. The subset of the KISS Army present at each gig cheers like crazy, possibly numbering in the collective millions when sonically piled atop one another; for forty-five straight minutes, they never, ever stop going nuts. The slow rhythm of stretched-out handclaps or floor-tom smacking generates a narcotic pulse for long stretches of the texture. Every so often, the band tires of Stanley's blather and hits a beautiful chord that seems to go on forever into Power Chord Heaven.

Plus, your chances of being able to actually fall asleep with this playing-- vs. the insufferable source original-- are much, much greater.

Headphones or halfway decent speakers very strongly recommended.

If you just want a quick sample, and don't intend to stick around for the full 45'51" of admittedly (but at least pleasantly) monotonous listening in total, check out the first five minutes on Soundcloud instead.

edit 10.23.18: You used to be able to hear the whole "piece" and also get the full download links to MP3 and FLAC files here, but for a few reasons involving my new primary page design, I had to change that all these years later. You should hopefully be able to stream, and also see the FLAC and MP3 download links for the full Stanlenchilada, at the top of this page. If not, try finding the 45 Paul Stanleys link from the main page of lossfoundation.com, click on that album entry, and you should see the direct download links then.

The process in ridiculous detail

The original "45 Minutes of Paul Stanley Stage Banter" is actually 45'51" in length - 2751 seconds total. I knew from the start that I wanted exactly 45 excerpts from those "45 minutes," so that they could all play time-stretched at once (45 Paul Stanleys, one for each of his "45 minutes"). I also knew upfront that each of the 45 excerpts in "stock" form ought to be of vastly differing lengths so as to have different degrees of time stretching going on when all the layers were time-stretched and put together.

I'd already decided to use some kind of Fibonacci-sequence-based pattern to decide the length of each excerpt and try to keep the results meaningfully organized. Splitting and spitting out the excerpts, however, was going to be WAY easier / faster in REAPER (my beloved editor / DAW of choice) if I could keep things rounded to whole seconds as much as possible.

So after some playing around with my spreadsheet and calculator to see how I could best use (or mostly use) golden-meanish principles in splitting up the 45'51" source, I ultimately decided on the following semi-arbitrary approach:
1) Split the full 45'51" into three equally sized clips of 15'17" (917 seconds) each.
2) Split each of the three new clips into further clips of lengths determined by the Fibonacci sequence, moving in Fibonacci order through each new clip, until the amount of time available in each clip is exhausted.

This resulted in fourteen sub-sub-clips per 15'17" sub-clip - each yielding two one-second sub-sub-clips, plus one sub-sub-clip each of 2", 3", 5", 8"... on down to a final sub-sub-clip 308" long (the full / "proper" 377 seconds from the sequence not being available).

Here are the exact timings of each resulting clip as they would appear in the original 45'51" file.

div		1	2	3
1		0:00-0:01	15:17-15:18	30:34-30:35
1		0:01-0:02	15:18-15:19	30:35-30:36
2		0:02-0:04	15:19-15:21	30:36-30:38
3		0:04-0:07	15:21-15:24	30:38-30:41
5		0:07-0:12	15:24-15:29	30:41-30:46
8		0:12-0:20	15:29-15:37	30:46-30:54
13		0:20-0:33	15:37-15:50	30:54-31:07
21		0:33-0:54	15:50-16:11	31:07-31:28
34		0:54-1:28	16:11-16:45	31:28-32:02
55		1:28-2:23	16:45-17:40	32:02-32:57
89		2:23-3:52	17:40-19:09	32:57-34:26
144		3:52-6:16	19:09-21:33	34:26-36:50
233		6:16-10:09	21:33-25:26	36:50-40:43
308		10:09-15:17	25:26-30:34	40:43-45:51

Long story short, given the length of the file, it was not possible to use Fibonacci integers like this and come up with 45 clips total - 14x3 only yields 42 snippets of audio. 42 may still be the meaning of life / whatever, but this all started as 45 Minutes of Paul Stanley Stage Banter.

So I took the whole of the original 45'51" file and took three more Fibonacci-determined chunks - 377 (at last! now I don't need to feel bad about the "308-second cheat"), 610 and 987 seconds respectively, from 0'00" in the banter file out. (This means I effectively had to use the first 1974 seconds of the full provided 2751 seconds of banter twice - heyyyyy, 1974 was a pretty darned significant year for KISS, if I recall.)

div
377		0:00-6:17
610		6:17-16:27
987		16:27-32:54

A bit later, thanks to REAPER's ridiculously fast / efficient marking and editing facilities, I had my 45 precisely-extracted clips of Paul Stanley (and the audience, and song starts, and other things). I started feeding these clips individually into Paul Nasca's Extreme Sound Stretch. (Perfect - Paul, meet Paul. welllllll allllllriiiiiiite)

Unfortunately, with seventeen different file lengths, it wasn't possible to do this as a bulk operation, and Paul 2's app doesn't really provide such amenities anyway. So this part took me a while.

I tried to stretch each clip to exactly 2751 seconds in length. Such precision is not quite possible with Extreme Sound Stretch. The app gets worse at guessing the final file length as the "stretch magnification factor" gets higher; actually, as it turns out, if you want a 46-minute file from a one-second clip, you need to tell Sound Stretch that you actually want just under an hour and a half of file output from your 0'01" source.

Otherwise, I didn't mess with any of the default settings in ESS; they were going to get me as close as possible to a superstretched Paul Stanley, and I felt like messing with things like the pitch level of his signature "banter tessitura" would have broken the spirit of the thing.

Hours later, from my 45 original excerpts 0'01" to 16'27" long, I had 45 newly stretched sound files that were all quite close to the original banter file's 45'51" in length... or close enough for rock and roll, anyway. I was usually able to get things within 30 seconds of 45'51" with trial and error, and most of my longer excerpts yielded stretched files that were right on the mark. (The total length of the necessarily-imprecise results: 34 hours, 8 minutes, and 44 seconds of banter-cum-baaaaaaaaaaaaaaaaaaannnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnntteeeeeeeeeeeeeeeeeeeeeerrrrrrrrrrrr.)

I now put everything back into REAPER and began shaping the "piece." I knew upfront that 45 of these things at a time would just sound like a white-noise mess, and I was right, so I at least adjusted the mess a little to avoid that as much as possible.

Beyond taking some liberties with the volume of each clip to make the mix as non-white-noisy as possible, I started with a semi-arbitrary hand-selecting procedure that introduced 1-2 stretched Stanleys at a time, one new Stanley about every thirty seconds. I also took a couple further / specific (but limited) artistic liberties right at the outset of the thing to try to give the finished product a somewhat interesting buildup of a five-minute opening (by which time fifteen of the stretched clips are already playing at once).

From there, I kept introducing one clip at a time every 30 seconds or so, until all 45 stretched Stanleys were bantering together at about 21 minutes in. You can barely hear the entrance of the individual "voices" because the texture is just too dense for that, even at the six-minute mark. (So I offer my REAPER screenshot of the finished project as proof of what's going on and/or that I do in fact have 45 simultaneous tracks of Paul Stanley in this thing:)

The piece has a strict "shutdown procedure" as well. At 40'00", after 20 straight minutes of 45 stretch-clips playing at once, I start removing "voices" from the texture one or two "groups" at a time, every thirty seconds, starting with the batch of three clips that began as 34-second excerpts.

Again, the disappearance of whole groups of stretched clips is surprisingly impossible to hear until perhaps the last minute or two. The three longest clips, treated as a group themselves, along with the six-member one-second-clip group are the last to disappear; the three-clip 308-second group-- which were also the first clips introduced at the beginning-- are the last clips standing as the "work" reaches an end at 45'51".

The individual stretched clips are kind of interesting listening too. I will probably release them sooner or later as Thirty-Four Hours of Time-Stretched Paul Stanley Stage Banter, and offer a bag of Detroit Rock Candy as a reward to anyone who can somehow prove to me that they listened to the whole thing.