arhiva7 - choose your destination
October 6, 2007 6:02 PM   Subscribe

posted by hama7 (81 comments total) 16 users marked this as a favorite
posted by ZachsMind at 6:05 PM on October 6, 2007

You've been waiting a long time for this, I imagine.
posted by puke & cry at 6:11 PM on October 6, 2007 [2 favorites]

that sure is a website!
posted by blacklite at 6:15 PM on October 6, 2007 [1 favorite]

Um, WTF? When you have six more words in your tags than in the text of your post, you maybe ought to think about, y'know, elaborating a bit more so people have an idea why they should click on the link.
posted by cerebus19 at 6:19 PM on October 6, 2007 [4 favorites]

Or you could look at the six tags and get a pretty good idea and save the snark. One of the two, ya know.
posted by lazaruslong at 6:22 PM on October 6, 2007 [1 favorite]

My Lord cerebus, I can't believe you're complaining! There are three extra words and a whole added hyphen in this hama7 post. He's clearly in a chatty mood today and you aren't even appreciating it.
posted by miss lynnster at 6:24 PM on October 6, 2007 [1 favorite]

Yeah, but in order to see the tags you have to click through to the post page. A good FPP should have something to grab your interest on the actual FP, don't you think?
posted by cerebus19 at 6:24 PM on October 6, 2007 [1 favorite]

As a musician, this proliferation of endless free music sites is a little disturbing. I like all the music, I don't love any of it to death and I fear that it's going to make it impossible to actually sell good music ever again.

Now, this'd be fine if we were going into a golden age where people could work a little at other jobs and then come back to their gentile homes and work at their music. But in fact it looks like the twilight of the middle class all over the world: "someone else can do your job for less".

Is this the future of music, unoffensive dabblings of a few people who can afford to do it (and of course legions of professionals putting out the old standards, classical, jazz, rock classics, for time to come...)?
posted by lupus_yonderboy at 6:25 PM on October 6, 2007 [2 favorites]

"hama7 at budweiser dot com"

posted by ZachsMind at 6:26 PM on October 6, 2007

"Is this the future of music...?"

It's the present. Where you been?
posted by ZachsMind at 6:27 PM on October 6, 2007

lupus_yonderboy: Are you implying that great musical works have been accomplished out of a drive to rake in cash?
posted by phrontist at 6:30 PM on October 6, 2007 [1 favorite]

I fear that it's going to make it impossible to actually sell good music ever again.

Thomas Edison didn't kill live music, and neither did Shawn Fanning. Sell the show.
posted by aaronetc at 6:36 PM on October 6, 2007

lupus: if people who have free time like to make decent music as personal entertainment, then it makes sense for there not be a huge market for people who want to do that as their job. There will of course still be people paying for shows, since those are enjoyable and demand time and logistics from the performer. If you're really good people will still pay. Immediately, it's the death of making big money from decent but not outstanding work. In the long term, it will be very hard to fund the time to become a truly skilled musical artist, just like today it is very difficult to find the money to become a great visual artist unless you're doing commercial graphics. Breaking into writing is quite the task now too, since so many people want to do it, are ok at it, and are willing to attempt it as a hobby.
posted by a robot made out of meat at 6:36 PM on October 6, 2007 [5 favorites]

posted by ZachsMind at 6:42 PM on October 6, 2007

Well, I like it. 7
posted by katillathehun at 6:58 PM on October 6, 2007

"Now, this'd be fine if we were going into a golden age where people could work a little at other jobs and then come back to their gentile homes "

What are you, an anti-Semite?
posted by mr_crash_davis at 7:04 PM on October 6, 2007

you maybe ought to think about, y'know, elaborating a bit

Ha! Hahahaaaaa! BWAAAAAHAHAAAA!!

Hama7... elaborate? Now that's funny!
posted by The Deej at 7:09 PM on October 6, 2007

a robot made out of meat: the trouble is that it's hard to really excel at music -- or any activity -- when you can only do it as a hobby.

People won't pay for outstanding work -- they'll pay for popular work. The guys at the top will be getting money, if only to advertise Pepsi.

I hate to say it, but I fear that music as a whole reached a point of maturity some time in the last two to three decades. I see music's childhood as lasting until very roughly 1950, its turbulent adolescence lasting about 25 to 35 years, and then a mature period that will proceed indefinitely. It's sad as I grew up right at the end of the wild energy of the adolescence.
posted by lupus_yonderboy at 7:11 PM on October 6, 2007

What does that make the B-minor mass? On your lifeline it's prenatal, precoital, heck, the carbon atoms are still inside the pre-hamburger cow!
posted by tss at 7:18 PM on October 6, 2007 [1 favorite]

Something suspicious about this website name... are we SURE this isn't a self-link?
I keed, hama, I keed
posted by wendell at 7:36 PM on October 6, 2007

lupus: How do you think that classical musicians get by? The end (or reduction) of instant-success from album sales for a tiny fraction of musicians isn't going to end music of any style.
posted by a robot made out of meat at 7:39 PM on October 6, 2007

I speak romanian and this website is still pretty opaque. Thanks for the tags hama7.
posted by jessamyn at 7:44 PM on October 6, 2007

lupus_yonderboy: I'm having great difficulty gaging how serious you are. Did you really just say, in effect, that the music of your generation was the peak, and everything before and after was just build up and then stagnation?

I would say that the most interesting areas of music are only now being explored seriously with electronics (where they have previously been somewhat gimmicky). The elements of music are pitch, time, and timbre the first two of which have been explored quite well but are being enlivened somewhat through the use of new technology (listen to some Squarepusher/Aphex Twin - tell me they weren't doing anything new). Timbre though has previously an area of limited control in music - and innovations in synthesis and interface design are allowing for expressive and creative control that was previously impossible with crude mechanical instruments.

Sorry, I'll get off your lawn now.
posted by phrontist at 7:49 PM on October 6, 2007 [1 favorite]

dele7ed 7hread
posted by Avenger at 7:53 PM on October 6, 2007 [3 favorites]

John 3:16

How'd I do?
posted by ZenMasterThis at 8:03 PM on October 6, 2007

I agree completely, phrontist. Timbre is already becoming a more important element in all music, and for me, it is rapidly becoming the most interesting thing to listen for. I listen to the Aphex Twin album Drukqs far more often now than when it came out because of this; each sound on the album is new and exciting in some way. Listen to Hy A Scullyas Lyf A Dhagrow, for a good example of timbre trumping pitch/rhythm. The melody is quite nice, but the sounds creating it are unique and captivating.
posted by synaesthetichaze at 8:05 PM on October 6, 2007

Would you believe John 5?
posted by ZenMasterThis at 8:10 PM on October 6, 2007

Also, The Books.
posted by synaesthetichaze at 8:11 PM on October 6, 2007

Some good stuff here. Thanks.
posted by xmutex at 8:14 PM on October 6, 2007

You can stream the tracks directly from here.

So I listened to a few tracks. There is a hell of a lot of mediocre electronic music floating around. This is some of it. Apologies in advance if anyone finds some really standout work here and points it oiut to me. I'd like that, really.

Also on you can find this album from ocp. It's depressingly good. A few years ago this would have been published on a real label, before the internet demonetized electronic music.
posted by fleetmouse at 8:15 PM on October 6, 2007 [1 favorite]

cerebus19: "A good FPP should have something to grab your interest on the actual FP, don't you think?"

Apparently it didn't need anything more to grab your interest.

Sheesh. This isn't marketing, after all. It's MetaFilter. If people don't like clicking, why do they use the internet?
posted by koeselitz at 8:35 PM on October 6, 2007 [2 favorites]

I'm not sure if this was posted here when it was first published, but I thought this message from Amon Tobin was pertinent to this discussion.

How big is Amon Tobin? Among modern electronic musicians, I suppose he's regarded up there among the biggest, with Squarepusher and Aphex Twin. On the one hand, it's genre stuff, but it's a very popular genre, and he's among the biggest names in that arena with numerous popular and well-reviewed albums to his credit.

A message from Amon Tobin
posted by PeterMcDermott at 8:39 PM on October 6, 2007 [1 favorite]

I am not even going to read any of the above comments to say that I HATE HATE HATE the hama7 style of posting here. I DO NOT WANT TO HAVE TO WORK HERE, I JUST WANT TO KNOW IF I SHOULD CLICK ON THE LINKY, MUTHAFUCKA.

That is all. Now, I will NOT click on his link, as per usual.
posted by yhbc at 8:44 PM on October 6, 2007 [1 favorite]

Oh changed my mind already. Found some good stuff. I'll shut up now.
posted by fleetmouse at 8:46 PM on October 6, 2007 [1 favorite]

Tell us how you really feel, yhbc.
posted by monju_bosatsu at 8:47 PM on October 6, 2007

As a musician, this proliferation of endless free music sites is a little disturbing.

Everything is free now,
That's what they say.
Everything I ever done,
Gotta give it away
posted by flapjax at midnite at 8:49 PM on October 6, 2007

phrontist: are you saying that electronic music has a greater range of timbre than "crude mechanical instruments"?

If so, I beg you to reconsider, because you're not only wrong, but dreadfully wrong. As interesting as the range of timbres that electronic music can create is, as other-worldly as its extremes can sound, it will never compare to what can be produced by purely acoustic means until some major achievements are made in computer technology. Think of it like the real numbers as compared to the integers: there are an infinite amount of both, but between any two real numbers there are an infinite number of real numbers. What's missing from that analogy also is the fact that the timbres of real instruments have much more depth to them, individually, than the timbres of electronic instruments. The imperfections of wood and metal contribute so much to a sound that just can't be replicated in the mathematical perfection of electronic instruments. Even basic overtones just take too much processing to be modeled very well.

I'm not putting down electronic music. Aphex Twin is awesome. There's a lot there. But if you're willing to give up on "crude mechanical instruments" for electronics, you just haven't pushed the limits of the real before you moved on for something shinier.

Penderecki's "Threnody for the Victims of Hiroshima" yells out from another room as I type this, pleading for the ingratitude to stop. Won't you listen?
posted by invitapriore at 8:50 PM on October 6, 2007 [1 favorite]

posted by moonbird at 8:51 PM on October 6, 2007

Tell us how you really feel, yhbc.

I think he's deliberately being a pretentious prick.
posted by yhbc at 8:52 PM on October 6, 2007

Thanks for the Amon Tobin link, PeterMcD. Everyone should read that.
posted by flapjax at midnite at 8:52 PM on October 6, 2007

posted by kittyshopping at 8:53 PM on October 6, 2007

I'm with Jessamyn - I spent most of the summer learning Romanian and feel I can speak it pretty well, but what the fuck is this site all about? Why publicize it when even those of us who can hack through most of the language still find it "opaque" enough not to get much out of it?
posted by Dee Xtrovert at 9:02 PM on October 6, 2007

It had good graphics and I could dance to it. I give it a 7.
posted by gummi at 9:09 PM on October 6, 2007 [1 favorite]

invitapriora: I am unconvinced.

It's likely that your recording of "Threnody for the Victims of Hiroshima" is a digital recording. Though some argue to the contrary, your auditory system lacks the resolution to distinguish between the discrete, sampled approximation of the analog signal and the original analog signal itself (were it possible to record and reproduce such a signal in the first place). Otherwise you would not bother with a CD, right?

The implication is that the "finite", discrete nature of digital audio is not actually a shortcoming. You mention that simulated audio cannot yet reproduce the complex acoustic effects that create the final tone of an acoustic instrument, and this is indeed true. Nevertheless, I feel it misses the point that people who are excited about the frontiers of sound in electronic music like to make. There's no reason electronic instruments should worry about the nature of vibrations in wooden boxes, brass tubes, taut strings, and concert halls. Any signal is fair game.

We may not have the technology now to make sounds that you find compelling. What's more, the more flexible the tool is, the harder it is to figure out how to use it effectively. Someday, though, someone will, and I'm sure it'll be a spectacular complement to all the ways we've been making music so far.
posted by tss at 9:10 PM on October 6, 2007 [1 favorite]

tss: You seem to have misunderstood. I completely agree that any signal is fair game. I'm only arguing against the marginalization of acoustic instruments, as if they were somehow pathetically limited in comparison, because it's not true.
posted by invitapriore at 9:16 PM on October 6, 2007

Let me add an additional note which may or may not be interesting. It's reasonable to suppose that in order to make compelling sounds electronically, increasingly sophisticated simulations of physical processes will be necessary to achieve adequate "richness" or "depth". After some exposure to perception research in my graduate program (mostly hearing talks from the lab next door), I am less certain this is the case.

Human visual and auditory processing certainly is optimized to process natural sounds (a loaded term if ever there were one), but not necessarily to reproduce them. That is, when you hear something, the representations built in your brain are not necessarily constructed so that the original audio data could be rebuilt from them---but also/instead for the convenience of various cognitive processes: language, emotion, perception of threats, memory, and so on. There is no theoretical reason why other, "unnatural sounds" could not be similarly provocative to the same auditory systems, in the same way that visual artwork that looks nothing like the savannah of our ancestors (and indeed is often much simpler!) can still dazzle us.
posted by tss at 9:27 PM on October 6, 2007

Who's marginalizing? Saying that electronic instruments have a "greater range of timbre" than analog ones is a fact, not a value judgment. Even though the violin can make a wide range of sounds, and even though many of them are fascinating to be sure, you simply can make more with a computer.

Nobody's saying throw out the bassoons; it's just that now we have some additional things we can play with. Telling us to hold back now seems a little like saying you can't cook Chinese food until you've graduated from a Cordon Bleu school
posted by tss at 9:34 PM on October 6, 2007

Sheesh. This isn't marketing, after all. It's MetaFilter. If people don't like clicking, why do they use the internet?

The question is, why should we click at all? I did click, and was faced with this text on a white background:

a log (graphic.sounds.likethis) netlabel (free electronics) almanahul (black/white)

Even after factoring out my own POV, I'm having trouble seeing this as best of the web.
posted by JHarris at 9:35 PM on October 6, 2007

A few years ago this would have been published on a real label, before the internet demonetized electronic music.

I believe I hate the word "demonetized" about 6x more than I hate the word "monetized", and that a lot.

Personally, I'd rather have 20,000 sounds out there for me to dive into and explore for free, than having to pay outrageous prices for some whitelabel vinyl.

And as a musician, I much prefer to just be able to get my music OUT there, rather than having to fight for a label of distribution.

Monetizing? Fuck that. Goes for lupus_yonderboy too - it's the same argument "professional photographers" level at Flickr. "How dare these amateurs put their photos out there under Creative Commons licences? They're going to kill real photographers like me - people will be able to get their hands on their crappy photos for free instead out paying me thousands for the exact same crappy photo!"

Music's just cool man; making a living off it has always, always been a rare and special thing. But until now, the ones making money off it were the only ones being heard. It can only be a good that the internet is letting us hear more of the musicians without commercial intentions, and making the musicians with commercial aims work even harder.
posted by Jimbob at 9:54 PM on October 6, 2007 [4 favorites]

posted by Effigy2000 at 10:25 PM on October 6, 2007

Basically, Warhol got his variables mixed up. The correct solution is: In the future, everyone will have 15 fans.
posted by Naberius at 11:23 PM on October 6, 2007 [9 favorites]

Wow, this debate about electronic vs. physical instruments is missing a fair number of facts. Here's the deal:

Computer synthesis of physical instruments just isn't very good right now. Oh, it's getting better, but our standard way of synthesizing even a piano is to digitally sample each note played on a real device, rather than simulating the entire system of hammer and string and reverberating chamber. We can't even interpolate very well between one note and another -- lots and lots of non-linearities that aren't correctly modeled simply by playing samples quicker or slower. The piano is probably the simplest instrument we'd try to simulate, and not only can we not do it very well, but the human ear can tell, very very easily.

Our brains are shockingly good at deriving nonlinear patterns. They're better than our code is. But remember, our brains can separate individual tracks in a mix, which is still an open and pressing problem in computer science (google Blind Source Separation or Computational Audio Scene Analysis).

That doesn't mean there aren't very interesting things happening in the electronic space. Just because we can't easily simulate natural nonlinearities, doesn't mean we can't invent some new ones. Electronic music is playing with entirely alien and unique timbral constructions, and one of the joys of listening to it is often our brains piecing apart just what's happening in this completely new signal being presented to us. Our brains are prediction engines, and to successfully predict something new and complex is a direct source of entertainment.

What I think needs to be respected is that a complex and fascinating signal can be made with a piano, a drum, a spoon, or a line of code. Alternatively, total crap can be made too. Depends entirely on the composer.
posted by effugas at 11:33 PM on October 6, 2007

Agreed, effugas. I hope I didn't give the impression that a computer could synthesize a convincing analog instrument yet---it can't. Digital sounds can adequately reproduce the sound of a piano, though, which gives the lie to any appeal to quantization as a reason for the inadequacy of electronic sounds. Eventually we will have a convincing bassoon, as convincing a bassoon as you can get on a CD anyway, but not now.

Interestingly, the human ear can tell when a present-day physical simulation fails, but that doesn't mean that better simulation is the only way to a convincing listening experience. There may be ways to "trick" the ear into thinking that you're playing a note on the piano even though the actual sound coming out of the speakers, measured by a microphone, is quite different from what a piano produces. For all the things our ears can detect, there are other things that it can't.[1] Exceedingly good physics is probably sufficient for great synthesis. For humans, though, it may not be necessary. Indeed, it may not be necessary in general. An example:

There are easier instruments to synthesize than the piano. A good candidate is something like the triangle, which is a comparatively simple system to simulate. In the "lab next door" that I referenced previously, one student is studying the synthesis of sound created by striking solid rods made of different kinds of metals. In a novel approach, however, she is not approximating the dynamics of the physical system at all; instead, she is creating a sophisticated statistical model derived from a large collection of samples of these kinds of sounds. She does not interpolate between samples in the traditional sense; rather, she finds a set of numerical bases that can be mathematically combined to produce different sounds. The "knobs" in this combination happen to correspond well to intuitive properties of the sounds, such as ringing and so forth.

Her sound reproductions are tested against current physical models (they are comparable) and unlike my "fooling humans" example above, the goal is to produce an accurate sound, not just one that sounds good. Nevertheless, the physics simply aren't simulated there.

The paper is here.

[1] Example: Take two sinusoidal tones whose amplitude slowly increases and decreases (i.e. there is no sharp attack or sudden release). You can vary the phase of the tones as much as you want and they will still sound exactly the same. C.F. also psychoacoustics; MP3.
posted by tss at 12:11 AM on October 7, 2007 [3 favorites]

p.s. for the experiment above to work, make sure the tones are different enough in frequency to avoid the creation of obvious beats.
posted by tss at 12:21 AM on October 7, 2007

Well, you can't convincingly mimic a piano on a synth, but there's no way to create the vast majority of sounds in electronic music with any sort of mechanical instrument.

Hearing Rank 1's "Airwave" the first time on a club soundsystem was kind of a revelation to me -- Granted, the lead synth is just a stack of detuned sawtooth oscillators playing chords, but the sheer ear-filling perfection of the sound tops anything you can do with strings. It wasn't someone mimicking an orchestra, it was MORE than an orchestra. (If you want to see the reaction it gets on the dance floor, 7 full years after it came out -- Rank 1 playing Airwave live in Toronto last year.)
posted by empath at 12:35 AM on October 7, 2007

favorited - just to annoy yhbc
posted by caddis at 7:06 AM on October 7, 2007

But until now, the ones making money off it were the only ones being heard.

Maybe that's a good thing. The labels functioned as a necessary filter. I trusted the taste of the people running Mille Plateaux (for example). They did the hard work of sorting through dozens of crap demos a day, finding the gems. I got good music out of it, and they got a modest living.
posted by fleetmouse at 7:26 AM on October 7, 2007 [1 favorite]

The one thing that has always limited me from appreciating electronic music more is my inabillity to hear passion (or excitement, emotion etc) expressed in the music. Whether its in the timbre of a voice, or the way that Joe Strummer stomped his foot at the beginning of "White riot" some deeper human subtext is (sometimes) present in the performance. Without that sub layer electronic music sounds cold to me. Can someone point me to some track or artist that might enlighten me?
posted by jlowen at 8:13 AM on October 7, 2007

tss: We shouldn't bother simulating traditional instruments, except perhaps to understand why they were emotionally resonant with people in the first place. Sound is just a signal, a mathematical function, and we shouldn't let our old methods of generating these signals hold us back.

jlowen: Mechanical and passionate are not mutually exclusive. Listen to Royksopp - What Else is There? (Trentemoller Remix). Email is in my profile.
posted by phrontist at 8:32 AM on October 7, 2007

(Also, miss you shows how highly minimal timbres can be extremely expressive.
posted by phrontist at 8:33 AM on October 7, 2007

Maybe try Amon Tobin's "Supermodified" album? I find a number of its tracks to be pretty evocative.
posted by tss at 8:37 AM on October 7, 2007

From Amon Tobin:

after the initial run it's unlikely that you will be able to order a copy even from online stores. this is because in-spite of more people having access to and apparently listening to my music than ever before, the predicted sales of the record were so low that it didn't justify the manufacture or distribution to any significant level.

I don't get that "even from online stores" part. Why couldn't Amon Tobin sell his album online?

Oh, he can. I mean, I see his general point, but that part seemed like a bit of an exaggeration.
posted by mediareport at 8:45 AM on October 7, 2007

phrontist: indeed, as I said. But I don't think I'm prepared to go as far as you. If it sounds good, it is good, right? Who cares whether it comes from a fine simulation, the latest psychophysics research, or somebody's serendipitous knob-twiddling?

I don't think we can say any method should be abandoned yet. It could be that simulating physical instruments (even fantastic, impossible ones) is the most fruitful way to explore a very large space---the space of all possible sounds. Maybe simulation more frequently leads to human-judged "more appealing" sounds than statistical prestidigitation or what have you, which could produce a few good things and lots of crap noises. Who knows? Computers give us so much flexibility, much more than we can handle now. If you ask me, the goal should be to improve versatility and ease of use for the artist.
posted by tss at 8:53 AM on October 7, 2007

jlowen: Paul van Dyk's album, "Out There and Back." Beautiful, emotional music.
posted by empath at 9:08 AM on October 7, 2007

CC'd music is not the death of labels, nor the death of the professional musician. This is the long tail of music. It's always been there, but now it has slightly longer arms, thanks to the internet. If you're really threatened by mediocre free music, the solution doesn't lie in being reactionary, the solution is to find new ways of distribution, in new ways of connecting to your fans, and in new ways of making money. Whether the threat is creative commons or internet downloading, the moral is the same, and it's one that the music industry needs to own up to eventually, as it has in the past when things have changed: There is no guarantee that any one way of making money will work forever. Radio changed things once already. Things are changing again.

That said, in the music I actively seek out and listen to, I don't see a lot of evidence that either of these things are killing traditional sales. I'm talking about bands which were only ever going to be releasing music on some of the most obscure indie labels anyway, and they continue to do so in the face of changing technology. Increasingly, this is accompanied by the bands themselves releasing their albums in whole or in part for free download from their websites. This far down the long tail, your problem is lack of exposure more than anything else.

The other day, I bought a bunch of music from a couple of more or less completely unheard of bands from both sides of the atlantic, off a label in based in britain and run by two friends who go exclusively by first names on the website, which puts out EPs in runs of 120 copies each. Free music is no threat to this sort of transaction. It barely exists in the same universe.
posted by Arturus at 9:35 AM on October 7, 2007

This is the long tail of music. It's always been there, but now it has slightly longer arms, thanks to the internet.


posted by fleetmouse at 10:19 AM on October 7, 2007 [3 favorites]

jlowen: try Melodies From Mars. It's a beautiful collection. Also "Rossz Csillag Alatt Született" by Venetian Snares, especially the track Öngyilkos Vasárnap.

I believe electronic music has the potential to convey emotion as any other genre of music. The nostalgia evoked by Boards of Canada is a perfect example. Or Kid 606's ambient work. The flexibility that artists like Amon Tobin have on sound is extraordinary. Sound itself has become the instrument these artists play. It is a different process then feeling and improvising on a saxophone, but the product is equally authentic.
posted by ageispolis at 10:44 AM on October 7, 2007

Statistical modeling is a numerical approach to simulating physical non-linearities. It can work quite well. But, lets be honest, physical instruments whose age is measured in centuries aren't yet getting correctly captured in code.

One major issue that isn't respected well enough is the UI problem: It's quite difficult to learn a musical instrument, but once you do, the capability of the player to drive the instrument in realtime is astonishing. We're just starting to get better code now for realtime PA, but some of the "soul" of music gets lost when it's not created in real time.

The scratchers get this; they actually have a low latency, highly interactive interface for music alteration. It could be better though. The other huge issue is that there is nothing more boring than a music performance with someone looking at a computer. Even music you'd normally never listen to can be made awesome live, just with a good stage presence. Fischerspooner gave up and just plays their CDs during their stage show. But there's something to be said for feeling the crowd, and subtely changing your music to match their rhythm.

We need better UI's for electronic music. We need them badly.
posted by effugas at 11:26 AM on October 7, 2007

When I DJ, I don't even like the crowd looking at me. I'd prefer to be off to the side somewhere fiddling with my laptop.

I do admit that mixing with CDs or vinyl is much more interesting to look at though. I think The Prodigy had this figured out 15 years ago, putting dancers and vocalists front and center and hiding Liam in the back.

Most DJs change my music to match the rhythm of the crowd. It's actually easier to do this DJing than it is with a live band, because you don't have to worry about syncing up with other people. If the crowd isn't with you, you just starting bringing the energy down till it matches, and then you go up from there.
posted by empath at 11:50 AM on October 7, 2007

(change THE music to match, etc..)
posted by empath at 11:50 AM on October 7, 2007

Statistical modeling is a numerical approach to simulating physical non-linearities.

Oh, I had a feeling you'd say that! No, it's not. When you learn a statistical model, you are not necessarily creating a system that tracks or models, in its "heart", more than a few superficial aspects of the physical system of interest. Here's why:

Statistical models describe the ways that data vary. As a simple example, consider a spread of arrows on a target. The data are the x,y locations of the arrows. We could choose to model this spread as a two-dimensional Gaussian distribution, one of several that say in effect that the arrow density falls off as you move away from a central point (the bullseye, let's hope).

Now, to generate more x,y data in the same pattern as the arrows, you can draw random samples from the Gaussian distribution. You do not need to simulate the 3-D trajectory of the arrow through space, the archer's arms, gravity, aerodynamic effects, etc. and so on. The physical system is not being simulated in any meaningful way. You might wonder why the system yields data that we can model with a Gaussian (or multivariate Laplacian, or whatever fancy distribution you want). Perhaps this derives from some interesting physical laws. The model, though, doesn't care.

In the case of the metal rods, the statistical model in the paper I referenced has discovered important kinds of variation in the input sounds. These variations, the "control knobs" I was talking about, may even correspond to physical properties of the rods. What they direct, however, are the combination of various basis functions---axes of variation, like the major and minor axes of our 2-D Gaussian ellipse in the arrow example.

I encourage you to examine the paper's figures and have a look at plots these basis functions. They don't say anything about the propagation of waves through different kinds of material, dampening due to friction, and so forth. Just as with the arrow example, the physical process is not being simulated.
posted by tss at 12:03 PM on October 7, 2007

That multivariate Gaussian link in my post above is not intuitive at all. Here are some plots of 2-D Gaussian functions. These are not probability distributions, since they don't integrate to 1, but the only difference is how tall the plots should be.
posted by tss at 12:17 PM on October 7, 2007


Do not mistake the verbiage for the mathematics. There's nothing wrong with numerically capturing statistical properties -- if it's stupid and it works, it isn't stupid. Our *brains* are ultimately doing precisely this -- coalescing nonlinear relationships from large amounts of sensory and partially processed input. That we can imagine an instrument playing a slightly higher pitch note, and the recognize the difference between that imagined sound and something that's just linearly interpreted is a pretty clear sign.

In that sense, you're totally right, and perhaps I wasn't clear enough. We're not creating a physical simulation of a violin in our brains, and comparing the output of that to the real thing. No, we're capturing across a large set of complex, not-yet entirely understood feature recognition and non-linear transformation functions.

My favorite example of this was the work done by James and Fatahalian at SIGGRAPH 2003: Precomputing Interactive Dynamic Deformable Scenes. Complex, visually clean simulations, all precomputed impulse response functions.
posted by effugas at 1:30 PM on October 7, 2007

OK. The term "physical non-linearities" was the sticking point for me then. I had assumed from this comment that you were referring to the nonlinear dynamics required for accurate models of physical systems.

It sounds like we are in concordance. E-bacon all around!
posted by tss at 1:49 PM on October 7, 2007

nostalgia evoked by Boards of Canada is a perfect example.

Seconded. The way BOC can stew up this strange, um, stew of creepiness and beauty that somehow perfectly evokes childhood is why they are one of my favorite musical groups.
posted by Bookhouse at 3:12 PM on October 7, 2007

People won't pay for outstanding work -- they'll pay for popular work. The guys at the top will be getting money, if only to advertise Pepsi.

This statement should not be in future tense. There's no "will" about it--this is happening right now, and giving music away is one method of circumventing the fucked up music industry. It is not, as you imply, the reason the industry is fucked up in the first place. If you want to avoid labels (which you must if you ever want to make money now) you have to market yourself, and that does involve giving at least some of your music away. Once people hear you and discover that you're outstanding, they will give you money. How do you expect to get there any other way?
posted by Saellys at 4:09 PM on October 7, 2007

This has to be the most commented hama7 post ever.

I have no idea what half of it was about. A perfect reflection of my day.
posted by From Bklyn at 12:42 AM on October 8, 2007

Um, this links to NOTHING.
posted by croutonsupafreak at 4:06 PM on October 8, 2007

Hama! Don't hurt'em!

croutonsupafreak said, "Um, this links to NOTHING."

I'm getting this:

"Forbidden - You don't have permission to access /index.htm on this server. Additionally, a 404 Not Found error was encountered while trying to use an ErrorDocument to handle the request. Furthermore you suck and your momma dresses you funny."

posted by ZachsMind at 10:17 PM on October 8, 2007

« Older Triumph of the Flow.   |   speculative landscapes and radical reconstruction Newer »

This thread has been archived and is closed to new comments