Join 3,438 readers in helping fund MetaFilter (Hide)


rrrrrrrrrrr​aaaaaaaaaaa​iiiiiiiiiiiii​nnnnnnnnnnn
March 19, 2013 7:51 PM   Subscribe

The Beatles song Rain, 800% slower version.
posted by flapjax at midnite (51 comments total) 13 users marked this as a favorite

 
I get it! LSD!
posted by QueerAngel28 at 7:54 PM on March 19, 2013


So is that just 1/8th the speed? 800% makes it sound glacial.
posted by Obscure Reference at 7:58 PM on March 19, 2013


I was hoping for Type O Negative (Summer Breeze), but instead was treated to something like Earth (The Bees Made Honey in the Lion's Skull)
posted by 445supermag at 7:59 PM on March 19, 2013 [1 favorite]


I think that 100% slower would be stopped.
posted by found missing at 8:02 PM on March 19, 2013 [12 favorites]


So is that just 1/8th the speed?

Truth be told, I have absolutely no idea what the 800% means, and don't especially care, personally. That's just the title given the clip on YouTube.

800% makes it sound glacial.

It's probably appropriate then, because "glacial" is, I'd say, a very apt descriptor for the particular quality of this piece of music. As a side note, I'd have to say it's one of the most gorgeous pieces of drone music I've ever heard.
posted by flapjax at midnite at 8:08 PM on March 19, 2013


Just an FYI, paul stretch (the program that does these stretches) is more of a synth that uses the original track as a starting point -- similar to the way a vocoder works. It's not in any real sense stretching the song out (which would make it infrasonic, obviously). It's essentially generating random wave forms using a fast fourier transform of the song sampled at certain points for the harmonic content. I don't believe the process is even reversible-- I don't think you can recover the original song by speeding it up by the same amount.
posted by empath at 8:19 PM on March 19, 2013 [8 favorites]


At that speed it's not nearly as good as Bieber-800%. No joke - that's a beautiful track.
posted by justsomebodythatyouusedtoknow at 8:22 PM on March 19, 2013 [4 favorites]


After reading the askme about death and now listening to this, I don't think I'll ever sleep again.
posted by littlesq at 8:29 PM on March 19, 2013 [1 favorite]


Here's a technical explanation of how it works.
posted by empath at 8:33 PM on March 19, 2013 [2 favorites]


I don't think I'll ever sleep again.

Obviously what you need is an 800% slower version of this one.
posted by flapjax at midnite at 8:34 PM on March 19, 2013 [1 favorite]


I've been trying to make scary ambient music lately (as one does), and it's amazing how effective just babbling into a microphone and slowing it down is.
posted by lucidium at 8:34 PM on March 19, 2013


No joke - that's a beautiful track.

Sigur Ros playing Vangelis covers in a cathedral being washed away by the sea.

See also: the Jurassic Park theme.

Oh, and Judge Dredd.
posted by Sebmojo at 8:37 PM on March 19, 2013 [2 favorites]


The Bieber track is one of my favorites [1]. If I need to sleep through the long leg of my daily commute, it does the trick.

I've tried using PaulStretch on my machine and can't quite get it to work as well as it seems like it should. It's a really nifty tool, though.

-----
[1] Now there's a sentence I never, ever thought I'd type.
posted by MShades at 8:37 PM on March 19, 2013 [1 favorite]


PaulStretch. But... not for Mac. *sob*
posted by flapjax at midnite at 8:59 PM on March 19, 2013


Sounds like Sigur Ros, circa Vanilla Sky
posted by KokuRyu at 9:23 PM on March 19, 2013


Strangely beautiful. As is the Bieber version. Since I know "Rain" pretty well I feel compelled to listen the original Bieber track, just for comparison. Thanks?
posted by wallabear at 9:25 PM on March 19, 2013


Aaaannnnd that's what I get for feeling compulsive.
posted by wallabear at 9:32 PM on March 19, 2013 [2 favorites]


Are we certain this isn't Klaatu?
posted by davebush at 9:38 PM on March 19, 2013 [1 favorite]


How much more slow could it be?

None. None more slow.
posted by ODiV at 9:52 PM on March 19, 2013 [6 favorites]


flapjax at midnite: PaulStretch is open source, and someone has built it for OS X: http://music.cornwarning.com/2011/12/07/new-paulstretch-os-x-build/
posted by DanielK at 10:33 PM on March 19, 2013 [1 favorite]


None. None more slow.

But these go to -11.
posted by phaedon at 10:48 PM on March 19, 2013 [4 favorites]


It's essentially generating random wave forms using a fast fourier transform of the song sampled at certain points for the harmonic content.

Aside from the word "random", that sounds like it could be a description of any pitch-invariant timestretching algorithm. Is there something the paulstretch algorithm does which makes it a less accurate form of interpolation than others?
posted by Mars Saxman at 10:50 PM on March 19, 2013


Someone should make a noisy shoegaze version of this. Maybe these guys.
posted by palbo at 11:00 PM on March 19, 2013


I think 1/8th speed is the only I could play that Paul bassline (although I can't make it out on here).

I only heard the promo, but didn't Radiolab do an episode a short while ago where they played a classical piece ("Ode to Joy"?) in a similar fashion?

And if anyone feels like they need to let loose and hear the original version of "Rain"...
posted by TheSecretDecoderRing at 11:49 PM on March 19, 2013


Aside from the word "random", that sounds like it could be a description of any pitch-invariant timestretching algorithm. Is there something the paulstretch algorithm does which makes it a less accurate form of interpolation than others?

He randomizes the phases which is what gives it that 'choir of angels' effect.
posted by empath at 11:53 PM on March 19, 2013


My point was mostly that when you stretch it this much, the sound you are hearing is mostly artifacts from the algorithm and has very little to do with the source sound.
posted by empath at 12:11 AM on March 20, 2013


Mmmm. Artifacts.
posted by aubilenon at 12:30 AM on March 20, 2013


the sound you are hearing is mostly artifacts from the algorithm and has very little to do with the source sound.

My ears tell me different. The essence of Lennon's voice, the timbre of it, is readily evident to the ear. I could absolutely identify that voice as Lennon's if I didn't know who this was. It's that clear, and that reflective of the source sound.
posted by flapjax at midnite at 12:58 AM on March 20, 2013 [2 favorites]


It makes a Fourier transform and effectively blurs the sound out over however much time you want.
posted by dunkadunc at 1:40 AM on March 20, 2013


I am the Walrus
posted by quazichimp at 2:02 AM on March 20, 2013 [1 favorite]


PaulStretch. But... not for Mac. *sob*


flapjax: There's a version that works for mac, which I've used several times.
heeeeeeeereeyouuuuuuuuuuggggoooooooo
posted by dubold at 3:02 AM on March 20, 2013 [1 favorite]



There's a version that works for mac

Woot! Thaaaaaaaaaaaaaaaaaaaaannnnnnnnnnnnnnnkssssssssssssssssssss!
posted by flapjax at midnite at 3:21 AM on March 20, 2013 [1 favorite]


I have to go to work - can somebody tell me how it ends?

But really, this is a neat tool. I could see taking a clip of music, processing it down and using it as the basis to produce a normal-tempo piece. Kind of anti-sampling.
posted by SteveInMaine at 3:24 AM on March 20, 2013


I am the Walrus

Oh, wow, thanks for that! And that's for two reasons, the first being that, yeah, I like it! The second, though, is this: A few minutes ago, as I was listening, the very first Lennon voice entered. My daughter (12 years old and pretty familiar with the Beatles' music) was listening. I said, to her "this is the Beatles slowed down a whole lot. Can you guess what song it is?" After about 30 seconds, she said "We are the egg men?"

Oh my god, you cannot IMAGINE how proud of her I was at that moment.
posted by flapjax at midnite at 3:27 AM on March 20, 2013 [11 favorites]


Because.
posted by flapjax at midnite at 3:38 AM on March 20, 2013


Blue Jay Way
posted by flapjax at midnite at 3:50 AM on March 20, 2013 [1 favorite]


Slow Down

(heh heh)
posted by flapjax at midnite at 3:52 AM on March 20, 2013


I think that 100% slower would be stopped.
posted by found missing at 11:02 PM on March 19 [7 favorites −] [!]


And by extension, 200% slower would be backwards at regular speed. Of course, "Rain" has vocal and guitar parts that are already backward-tracked at regular speed, so that would mean these parts at 800% slower are, er... two plus six... carry the eleven... I am going to go sit down for a few minutes.
posted by ricochet biscuit at 6:25 AM on March 20, 2013 [3 favorites]


Can I get my favorite song slowed down to the point where I can just listen to one track for the rest of my life?
posted by orme at 7:18 AM on March 20, 2013


Would someone please run John Cage's As Slow As Possible through PaulStretch? kthx
posted by a person of few words at 7:20 AM on March 20, 2013


If you can't get PaulStretch to work, the tempo adjustment effect built into Audacity does fairly decent work.

Using that to speed this track up by 800% resulted in a quite recognizable facsimile of the original, contrary to empath's prediction.
posted by flabdablet at 8:34 AM on March 20, 2013 [1 favorite]


a quite recognizable facsimile of the original

Hey, I'd be really curious to hear that!
posted by flapjax at midnite at 8:40 AM on March 20, 2013


Are we certain this isn't Klaatu?

Ha! I'm old enough to remember the "Who Are Klaatu?" posters that appeared all over record stores when that rumor started. An interesting marker of just how different a world we lived in in pre-internet days. That's a kind of "information arbitrage" that simply wouldn't be sustainable in the current world. The rumor would bubble up in the morning, be all over the twitterverse by midday and be debunked by the evening.
posted by yoink at 9:10 AM on March 20, 2013 [1 favorite]


1. I am in love with SoundStretch. I will be incorporating it into a new drone project I'm working on. I don't know how I missed it before.

2. My cats now hate me 800x more than ever.
posted by item at 9:12 AM on March 20, 2013 [1 favorite]


Obligatory, after the above mention of John Cage: 4'33" of silence slowed down 800%.
posted by beagle at 9:16 AM on March 20, 2013


So paul stretch is fun, but it struggles with slowing things down by smaller amounts. Percussion goes all wrong or vocals go all wrong. Also everything sounds like a Sigor Ros or Godspeed song. Using DJ software to slow things down, say 90s euro trance songs from 130bpm to 90bpm - this is where the real fun lies.

Shameless self link time: DJ Sammy and Yanou featuring Do - Heaven (at 90bpm).

Also works pretty good with the Little Mermaid music.

The Zombie rave mixtapes that Mater Suspiria Vision did, really blew my mind. When Eduador dropped at 90 bpm BOOM.

I think witch house/drag wave/zombie rave might actually be dead; but I'm still a massive fan.
posted by 13twelve at 9:18 AM on March 20, 2013 [1 favorite]


the sound you are hearing is mostly artifacts from the algorithm and has very little to do with the source sound.

True, but only because there is so much more real estate to fill. It's like an attractive dithering pattern on a massively scaled-up GIF. If we step back far enough, the spirit of the original work remains, as evidenced by speeding this recording back up to its original tempo:

rain_restored.mp3
posted by tapesonthefloor at 9:41 AM on March 20, 2013 [2 favorites]


a person of few words: Would someone please run John Cage's As Slow As Possible through PaulStretch? kthx
Do you have any idea how long that copyright infringement case would run?
posted by IAmBroom at 1:11 PM on March 20, 2013 [1 favorite]


Ran John Cage's 4'33 through it.

Meh.
posted by Devonian at 2:32 PM on March 20, 2013


I don't believe the process is even reversible-- I don't think you can recover the original song by speeding it up by the same amount.

You actually can get the original song back, in a way -- it just sounds like really bad file compression. Everything is still audible though, including the vocals.
posted by speicus at 4:18 PM on March 20, 2013


It's essentially generating random wave forms using a fast fourier transform of the song sampled at certain points for the harmonic content.

This isn't... quite true. Since the song is recorded at a certain sample rate, to slow things down and keep it at the same pitch it has to interpolate between samples -- in other words it has to guess what's in between, because there's just no information there. This process isn't random (or it would sound like static), but it is arbitrary in a certain sense, so the maker of PaulStretch decided that it might as well sound nice by rounding out some of the harsher peaks and valleys. This is what "spectral smoothing" does, and yeah, it makes everything sound like Sigur Ros.
posted by speicus at 4:25 PM on March 20, 2013


« Older The Social In(Queery) Blog presents a criticism of...  |  "The Allman Brothers Band prod... Newer »


This thread has been archived and is closed to new comments