Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

The Catastophic Failure of MP3.com

By forrest in Op-Ed
Tue Oct 03, 2000 at 08:33:09 AM EST
Tags: Music (all tags)

The first attempts to harness any new technology try to make it look like the old technology.
I am sure we have all seen this phenomenon by now.

It would be hard to find a more compelling example than mp3.com, who have clearly modeled their entire approach to the market on the current mass-market music industry, failing to recognize crucial distinctions between their potential market and the music market which preceded them.

Because of their abysimal inability to gain even the remotest hint of a filesystem-checking clue, they are doomed to failure.
And I say, good riddance!!!!

Sponsor: rusty
This space intentionally left blank
...because it's waiting for your ad. So why are you still reading this? Come on, get going. Read the story, and then get an ad. Alright stop it. I'm not going to say anything else. Now you're just being silly. STOP LOOKING AT ME! I'm done!
comments (24)
active | buy ad
Sometime earlier this year, as Napster began to get more attention, I decided that it was important for me to find and support musical artists I like who are making their work freely available on the internet now.

I know very well that there are an incredible number of very talented musicians in the world, and that who happens to be well-known is a function of marketing and media hype more than anything else.

It was then that I first went to the one site I knew had the most freely-available mp3s in one place on the internet, mp3.com. I was shocked and disgusted to find the same sort of marketing and hype there, and no tools to help me actually locate music that I wanted.

The mass-market music industry depends on the econimics of mass production, in which costs are brought down by producing more copies. In this world, a CD that sells more copies is, from a business standpoint, a better CD. It is worth every penny of advertising that it takes to make certain artists (and, by extension, certain styles of music) popular, to take advantage of the economics of mass production.

The pressures on mp3.com are just the opposite. There is no advantage in trying to pressure potential consumers to like a particular sort of music, because the product is produced on demand. The optimal business strategy instead lies in trying to help each potential customer find music that s/he likes. Because they have failed miserably at this, I am not, and will probably never be, an mp3.com customer, despite the fact that there is probably a lot of music there I would like if I could only find it!!!!

Despite these market realities, every tool for finding music on mp3.com is based on a "popularity contest" model.

Yeah, you have a gazillion categories, and they like to brag about that, but in reality the categories are mostly meaningless. Artists, forced to try to "win" the popularty contest by hook or by crook, try and spread their offerings over as many categories as they can possibly get away with. You will find ambient electronica in the Spoken Word>Self-Help section. Abuse of the system? Nahh ... practically required by it.

Then they have payola, a system by which the poor artists (perhaps helped by deluded fans) can try to make their works noticed by paying money to mp3.com. I pity the poor artists who (literally!) buy into this, because they must not realize that they're paying the designers of a system to work around a fundamental flaw in their design. It's as nauseating as the illegal practice it's named after. (Payola historically referred to record companies paying radio stations to play certain songs.)

(I could go on ... that's just the tip of the iceberg, really ... but I kinda want to get done with this article.)

So what should mp3.com be doing?

They could have a database of objective facts about the mp3s they have, searchable in as many ways as possible. Some suggested categories of information:

  • Does the song have lyrics? If so, in what language?
  • What instruments are used on the recording?
  • Accoustic or Electric? Computer-generated?
  • How many musicians?
  • How long is it?
  • What's the Dynamic Range (ratio of loudest/quietest part)?
  • What's the average tempo? Standard deviation of the tempo?
As it is, there are no objective facts, only artists' self assessments, and no way to do a search that's even remotely sophisticated. For example, I would be very interested to find mp3s that with Chinese lyrics (because I'm studying that langauge) and with mp3.com's incredible collection of mp3s I find it hard to believe there are none. I do think I would never find any in a million years of searching the present system, though.

If they were really forward looking, they'd consider investing in research into AI systems of music catagorization.

Instead, they want to scream about how cool they are, and choose for me the latest music I don't want to hear, hoping that if they can shove it in my face hard enough, I've got to like it.

Mp3.com, reap your justly deserved obscurity.


Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure


Can you find music you like on mp3.com?
o Yes, easily 25%
o Yes, but it takes a hell of a lot of effort 19%
o No, despite the fact there are probably a lot of mp3s there I would like 37%
o No, I believe all the material on mp3.com is inferior 17%

Votes: 62
Results | Other Polls

Related Links
o mp3.com
o produced on demand
o ambient electronica in the Spoken Word>Self-Help section.
o payola
o AI systems of music catagorization
o Also by forrest

Display: Sort:
The Catastophic Failure of MP3.com | 27 comments (27 topical, editorial, 0 hidden)
Fuzzy copmlex problems and Collaborative filtering (4.55 / 9) (#1)
by wildmage on Tue Oct 03, 2000 at 12:22:25 AM EST

Finding music in the new digital age is a REALLY complex question. I've been mostly focusing my research into artist remuneration, but the nagging question of how someone is going to find good music keeps coming up.

Most of the pseudo-intellectuals I've spoken to in this area are confused themselves, but the common buzz word that keeps coming up is "Collaborative Filtering".

Basically collaborative filtering is the process of many many people contributing data to a music moderation system so others can use this information to find music that they might like. There are 2 pitfalls to this: the infamous "popularity contest" that most current implementations fall into, and the notion that technology can discover music better than people can.

My buddy, Nick Woolsey, has been pretty savvy on this concept and has written an essay on it with his suggestion of what he likes to call "Relationship Features". I haven't read it in a while, so I'll try to get Nick in here to explain.

We also had an interesting discussion at Fairtunes a while ago about collaborative filtering in which a lot of people popped out of the woodwork.

Some companies/groups that are involved in Collaborative Filtering:
moviecritic.com powered by Macromedia's software, Likeminds
Amazon.com's book finding features
Relatable who are supporting Freeamp
Grouplens at the University of Minnesota

This isn't an exhaustive list, but its a few things I pulled out of my hat.

And while we're poking fun at mp3.com, I'd thought I'd post some of their competitors in the indie artist scene.

Emusic's underling, IUMA

And of course don't forget Fairtunes who, in their ultimate wisdom, will set the precedent for artist independence on the internet... or not ;)

I'll be sure to come back here to discuss anything on anybody's mind.

Jacob Everist
Memoirs of a Mad Scientist
Near-Earth Asteroid Mining

Re: Fuzzy copmlex problems and Collaborative filte (3.00 / 2) (#6)
by El Volio on Tue Oct 03, 2000 at 09:33:59 AM EST

I think something like this must be where music selection has to go. Amazon.com is a good example. My perspective, of course, is not from a service provider standpoint, or even a researcher's, but from a user's standpoint. Personally, I have found Amazon.com's system to work well at recommending books and music I'll like. While my literary tastes are pretty plain-vanilla (science fiction, history, the occasional Clancy-like thriller), my musical tastes are a little more obscure (tropical Latin, Cuban folk, etc). Having found at least one that I like based on the recommendation of a friend, that gave me enough of a foothold in the system to get recommendations on other music that fits my taste, or is close enough that I like it (broaden out a little bit).

I've found some of this stuff on MP3.com, but it takes some work, especially because (as the author correctly points out) much of the music there is deliberately mis-classified. While it may be "encouraged" by the system, it's sure as hell frustrating to me!

Those links look interesting; I've only visited a few, like Riffage, in the past. I'll be interested to see if any of them have implemented even a rudimentary form of this type of system, and if they really have the variety of music I like.

[ Parent ]

Re: Fuzzy copmlex problems and Collaborative filte (3.00 / 1) (#7)
by codemonkey_uk on Tue Oct 03, 2000 at 10:31:43 AM EST

Good post, I was sitting here thinking that this problem is the same problem as the forum moderation problem. That is, instead of moderated messages your moderating music. Same problem, different subject.

I was thinking of reposting or linking to what I said in that discussion but you've said it so much better, and provided links to boot... :)

"The most savage controversies are those about matters as to which there is no good evidence either way." - Bertrand Russell
[ Parent ]

Some factual error(s) (4.20 / 10) (#2)
by fluffy grue on Tue Oct 03, 2000 at 12:40:18 AM EST

First of all, mp3.com does have the artists specify attributes of their music when they upload it. I should know. Though I don't know why they don't actually use this information in the search box (the reason they use for asking for all this is to make it 'easier for people to find the music').

Secondly, mp3.com isn't really intended to be the search engine anyway. Yeah, I know, they hype themselves up as that, but it's pretty obvious (to anyone with a clue who's actually put stuff on there) that mp3.com isn't in the business of helping artists be found, they're in the business of hosting and reproducing music as a service to the musicians. Things like their New Music Army, all of their artist guidelines, and the like all say, quite plainly, that it's up to the artist to get exposed - and mp3.com just makes it easier for an artist to get paid. Of course, to non-musicians they tell a totally different story, which is an injustice against both themselves and the artists.

That said, I do agree with quite a bit of your rant. Mp3.com needs to do a lot to actually make their stuff useful for the musicians. I don't think a fuzzy AI search would really help anything - but a search which actually uses the data which the artists provide would.

Now, from my point of view, I think mp3.com is doing a fine and dandy job as a content provider, and if I ever get around to finishing pointed little quill, I'll find out how good they are as a distribution provider as well. The artists who do well are the ones who hype themselves up, not the ones who expect mp3.com to suddenly be the magic bullet to lead fans to their door. Honestly, I've never even actively sought out music on mp3.com - all of the stuff I've downloaded from there is because someone (usually my cousin, who works there) has pointed me to something interesting, or because someone links to their site there and I'm curious enough to find out. IMO, the reason I haven't gotten any payback earnings is because, frankly, I haven't attempted to get them.
"Is not a quine" is not a quine.
I have a master's degree in science!

[ Hug Your Trikuare ]

Re: Some factual error(s) (3.40 / 5) (#3)
by forrest on Tue Oct 03, 2000 at 02:22:59 AM EST

First of all, mp3.com does have the artists specify attributes of their music when they upload it.

In my article I stated:

As it is, there are no objective facts, only artists' self assessments [...]

I don't think I made an error. The artists' statements are not objective facts. Although some try to present objective infromation, many do not.

The most common misrepresentation I find is that someone who uses the "saxophone" sound of their sound card claims his work contains saxophone (and something like 50% of all mp3.com's music must be "one guy playing with his sound card").

More common is simply the lack of opjective, factual information about the piece in question, in favor of some "artsy" type statement. I can understand why an artist might make such a statement (it's their perspective about their own creative process, that's valid) but it doesn't help me.

But as you point out, the artists' statements aren't searchable anyway. Even that would be a help.

[ Parent ]

Re: Some factual error(s) (3.50 / 2) (#17)
by fluffy grue on Tue Oct 03, 2000 at 03:06:04 PM EST

Ah, I see. Thing is, a synthesized/sampled saxophone is still a saxophone, it's just gone through an extra filter...

Is a sample of a guitar being played and mixed in with the music any different than were the guitar being recorded at the same time as the music? You do realize, of course, that in modern recording equipment, there's little to no coherence between the discretely-recorded tracks anyway, right? It's all split up in recording and mixed back together in mastering.

Granted, there's huge difference between FM/wavetable saxophone playing and a sample of actual play, but where do we draw the line between real and not?
"Is not a quine" is not a quine.
I have a master's degree in science!

[ Hug Your Trikuare ]
[ Parent ]

Re: Some factual error(s) (3.00 / 1) (#21)
by forrest on Tue Oct 03, 2000 at 09:44:03 PM EST

Granted, there's huge difference between FM/wavetable saxophone playing and a sample of actual play, but where do we draw the line between real and not?
When calling it a sax would make John Coltrane roll over in his grave?

[ Parent ]
Re: Some factual error(s) (none / 0) (#22)
by fluffy grue on Wed Oct 04, 2000 at 12:52:35 PM EST

See, you completely shrugged off my whole philosophically-minded question based on the sheer emotion of what you think a jazz musician of the past would think.

And how do you know he'd be rolling in his grave? Again, a sample of actual saxophone playing, integrated into other sounds which aren't necessarily intended to complement it perfectly -- well, that's what Jazz is, when you get down to it - mutual cooperation between differing musical opinions.

Again, look at my argument. In a modern studio, the saxophone is recorded onto a separate audio track and is mixed back in later. Can you then say that splitting it up based on riffs (samples) and rearranging that real playing isn't real saxophone, as an instrument? Then if you split it up based on note - is it then no longer a saxophone? And then you approximate it based on summations of basic waves (which describes both MP3 and FM synthesis) - is it no longer a saxophone? Where do you draw the line?

And it's not like saxophone is a privilege only held by the Jazz greats. Is saxophone in rock not saxophone? Is saxophone in breakbeat not saxophone? Is saxophone in classical not saxophone? Do you see my point yet?
"Is not a quine" is not a quine.
I have a master's degree in science!

[ Hug Your Trikuare ]
[ Parent ]

Re: Some factual error(s) (none / 0) (#24)
by forrest on Wed Oct 04, 2000 at 10:36:45 PM EST

My comment was somewhat facetious, and I didn't intend for you to take it at all as seriously as you did.

But, since you are are of a serious mind on this issue, I'll try to answer in the same vein.

To me, the sensible demarcation is the subjective one -- that it "sounds like a saxaphone". If, listening to it I think that it's really a sax, then you can probably call it that. (Note, I'm not suggesting that you should... but that's another issue.)

The saxaphone is a notoriously difficult instrument to imitate, which is kind of why I picked on it I've heard a lot of awful synthesizer sounds labeled "saxaphone".

Sampling individual notes doesn't get you there, because it's the breath control that really makes a sax "speak". I guess it would be possible to begin to get a usable approximation with individual note sampling combined with a MIDI breath controller -- but at that point, I have to ask why? What's the point in working so hard to imitate a saxaphone sound when you can just get a sax?

Anyone who's got enough skill to pull off a quality imitation has got enough skill to master the instrument, or at least communicate to someone who has mastered the instrument what's desired.

[ Parent ]

Re: Some factual error(s) (none / 0) (#25)
by fluffy grue on Thu Oct 05, 2000 at 02:34:11 AM EST

I'm aware of the incredibly fine control of the sound of a saxophone. That's why I took the trouble to go through the particular step of splitting up the playing into individual riffs - as a whole the fine control is still there (assuming that all of the emotions trying to be conveyed in the playing of the riff are present in various multitudes of sample). Also, the breath control can be somewhat emulated through resonant filtering and the like (which Impulse Tracker has, for example).

And, again, a riffwise sample of sax being played isn't the sax being played and recorded directly to its respective audio track, but it's not just a notewise sample of the play either - it's a middle ground. And it's not particularly hard to 'imitate' the saxophone being played, because there is no imitation - it's acutally being played, it's just being broken up and rearranged and the like.

And it seemed more like flamebait than facetious. :)
"Is not a quine" is not a quine.
I have a master's degree in science!

[ Hug Your Trikuare ]
[ Parent ]

Re: Some factual error(s) (none / 0) (#26)
by forrest on Thu Oct 05, 2000 at 02:26:14 PM EST

Well, I guess it's just you and me here now, fluffy.

I see you had reason to take my comment in a more personal vein. While my statement about misreprestentation came from my experience of a particular example that would indeed induce an increase in the angular momentum of a deceased jazz giant, you took it as an attack on an entire methodology of composition.

I have nothing against sampling as a method, although I have heard a lot of sampled music which I don't like. That's because such music uses a common idiom of abrupt, jarring cuts in rythmic repitition. It's like the audio equivalent of bad collage art. But hey, there's a lot of bad poetry in the world too, but that doens't make poetry intrinsically invalid.

I still don't understand, though, how you would put your heart and soul into getting your samples just so, and not want to tell everyone just what you did. I mean, if I made a picture by writing a program, I don't pretend that I drew it, and I don't think that saying that I did would do anything positive for its validity as a work of art.

Why would working with samples be any different?

[ Parent ]

Re: Some factual error(s) (none / 0) (#27)
by fluffy grue on Thu Oct 05, 2000 at 03:50:00 PM EST

Did I ever say I wouldn't say that I used a sampler as an intermediate step of the proverbial saxophone?

Also, although you might not consider the end result of your program to be art, don't you consider the program itself to be art? I certainly consider programming as an art form. Likewise, I consider the usage and careful arrangement of samples in a sampler to be an art form.

And I don't dispute that there's a lot of bad sampler "music" out there, but then again, there's a lot of bad saxophone "music" out there too. The choice of instrument is a means to the end, not the ends itself.

Also, I don't pour my heart and soul into getting the samples 'just so' - rather, I pour my heart and soul into using the samples more or less properly. Back to the 'made a picture by writing a program' link: composing using various synthesis means is like creating with GIMP and using a variety of tools. If you listen to my music, most of it is composed, and not samples just pieced together. The few places where I do use looped, riff-type samples rather than instrument samples (the drumloops in 'softspoken' and most of 'hypnagogic' are the only such things I can think of offhand), that's equivalent to, say, using a pattern fill in GIMP - the samples don't make the music, the usage of the samples makes the music. Likewise, the vast, overwhelming majority of my music is built up using more traditional synthesis/sequencing techniques - like using a brush to draw a line.

The only place where I really abuse Impulse Tracker's whole sample setup in any of my released music is the lyrics in "Ode To You" (don't listen to that if you're expecting to hear good singing, btw - I still need to work at the vocal tracks). I split up the singing into a few samples and play them as though it were just a single continuous WAV file mixed in with the audio stream - the only reason I split it up to begin with was because Impulse Tracker imposes a 2MB limit on individual samples. I'll likely be doing a similar technique but moreso with the last song I'm going to be finishing for the album (right now it's a tossup between two of them) - everything will be recorded and mixed together like samples, but the end result will be (hopefully) indistinguishable from if the whole thing were recorded live. Just as a hint, I learned to play guitar specifically for whichever song I decide to do. :) (Search for 'Celebrated Woman' on Everything2 if you want to read the lyrics for the song I'm most leaning towards.)

"Is not a quine" is not a quine.
I have a master's degree in science!

[ Hug Your Trikuare ]
[ Parent ]

Good Lord! (2.33 / 3) (#4)
by gblues on Tue Oct 03, 2000 at 03:15:28 AM EST

if I ever get around to finishing pointed little quill

Holy crap, qp, you're *still* working on PLQ? It's been what.. four, five years now? :)

gblues (yes, *that* gblues)

... although in retrospect, having sex to the news was probably doomed to fail from the get-go. --squinky
[ Parent ]
Re: Good Lord! (3.00 / 1) (#16)
by fluffy grue on Tue Oct 03, 2000 at 03:02:18 PM EST

It's not the same pointed little quill. The musicdisk pointed little quill was finished long ago. This is the album pointed little quill, which also includes most of the musicdisk.
"Is not a quine" is not a quine.
I have a master's degree in science!

[ Hug Your Trikuare ]
[ Parent ]

The problem and solution (3.60 / 5) (#5)
by deathwish on Tue Oct 03, 2000 at 08:59:19 AM EST

I agree about the problem, I always though how great mp3.com would be if I could just find the good music. But, I'm not sure about this solution. Honestly, I could care less about how long or how loud the song is, or how many people are in the band and what they play. That might be interesting information about bands I want to know more about, but it doesn't tell me anything about how good the music is and it doesn't help me to find bands I like.

Personally, I'd like to see ratings and user feedback. Let ever user who hits a bands site vote on what they thought of the music, and leave reviews too (kinda like imdb does). Then, I'd also like to see it track your votes so it can make recommendations. The site moviecritic.com does this, you basically rate how much you like various movies and then it makes suggestions on what you may like or dislike. It's not always right, but it certainly helps to weed down the field a bit.

Now this isn't a perfect solution, but it's the best thing I can think to do. The traditional method of finding music you like involved having it filtered through the staff of a radio station. I think we can all agree that this doesn't work. But when you eliminate radio, tv, advertising, etc, that leaves you to filter the music yourself. There's a lot of music out there, and I don't have time to hear it all, so that's the basic problem. But, if someone can come up with a good system to connect fans with music, I think they could stand to make a substantial bit of money :)

Re: The problem and solution (3.00 / 1) (#19)
by forrest on Tue Oct 03, 2000 at 09:28:46 PM EST

Honestly, I could care less about how long or how loud the song is, or how many people are in the band and what they play.
The idea is to collect as much information as possible, so there could be as many ways of searching as there are searchers. The categories I listed would all prove useful to me -- I tend to favor accoustic instrumental music that's not too fast and has a reasonably wide dynamic range, and has an interesting mix of instruments.

For example, bass clarinet and cello. If I knew a piece had that combination, I'd listen to it. I might think it sucked after I heard it, but I'd certainly want to seek it out.

Of course, other information would be useful to other searchers. I suspect that if such a database were available, as you played with the combinations of parameters, it would make suprisingly non-obvious (but correct) recommendations.

[ Parent ]

Exactly (3.00 / 5) (#8)
by DesiredUsername on Tue Oct 03, 2000 at 10:37:46 AM EST

I visited MP3.com...once. I noticed all the same lacks that you did and decided never to visit again. I still don't know where to go to find music I might like--Napster and friends' grep-style utility is just as bad.

A LOT of online sites have these problem. Take eBay: I can search by string but since the seller enters all the info as free-form text I can miss the buy of a lifetime because of a typo or a seller who doesn't know what they have (or doesn't describe it fully). Why not create a database of items and the seller can indicate which item they are selling (if it appears in the database)? Then I could search for video cards that run under Linux, support OpenGL and serve up ham sandwiches without relying on the seller to input all that information.

Obviously this idea needs some work to be practical, but why isn't anyone trying??

Play 囲碁
You may be right, but not in my mind. (3.20 / 5) (#9)
by rednecktek on Tue Oct 03, 2000 at 10:41:52 AM EST

I just mailed off a check ($14.00) this morning for 2 CDs (read: damn cheap) from Tapping the Vein. I found out about them from mp3.com. I probably would never have heard them outside the PA area if not for mp3.com suggesting them. Red Delicious - Emotional Blur is next week's buy (my wife only let's me buy 1 CD a week, otherwise I'd go broke!), and once again I was introduced to them by a CD from mp3.com.

mp3.com has introduced me to many artists I find new and exciting, not the same old sounds that producers think I want to hear. It's been quite a while now since I bought anything other than an unsigned band. I like the way I get my music now. I know the bands I'm listening to before I get the CD. I spend less on music no than I did before (no, I'm not buying less, just getting more). And I feel better about the bands I support. I can't (won't) say mp3.com did it all for me, but they were there to help.

forrest wrote:
I was shocked and disgusted to find the same sort of marketing and hype there, and no tools to help me actually locate music that I wanted.

I hope you get a clue about marketing in the Real World(TM). If you really feel that a site the size of mp3.com can be run without advertising, try it. Then we can talk about failure. As far as the tools go, I have never had a problem finding similar artists or types. I really don't see how you can not understand how to use them. Maybe you haven't tried.

So, you go ahead and rant, that's your right. But remember that when you rant you might find a rant coming back to you. A quieter rant, a more positive rant, but a rant notheless.

Just remember, if the world didn't suck, we'd all fall off.
Re: You may be right, but not in my mind. (4.50 / 2) (#18)
by forrest on Tue Oct 03, 2000 at 09:03:27 PM EST

Well, our tastes differ considerably. I tryed listening to Tapping The Vein, and I felt like I had just turned to the local "commercial alternative" station. The same goes for Red Delicious. I'm glad you can find music that you like easily, but surely you appreciate the fact that not the genre you favor is very popular these days, and people who like different sorts of music will have a much harder time.

Where is the next Kalevala, or Twelve Moons?

mp3.com has introduced me to many artists I find new and exciting, not the same old sounds that producers think I want to hear.
No, they're the sounds that mp3.com wants you to hear. What's the difference? They shovel songs into your my.mp3.com inbox, and the "popularity contest" design of the catagories is designed to create the illusion that the music at the top of the lists is actually better. It's the same sort of system.

Richard Nesdale, the one artist I thought was pretty good after a strenuous week's searching mp3.com, wasn't on any page 1 ... I chanced upon one of his songs on page 27 of some category. If mp3.com had their shit together, I would have been able to find his music in a day.

I hope you get a clue about marketing in the Real World(TM). If you really feel that a site the size of mp3.com can be run without advertising, try it.
Did you really think I was talking about banner ads? I think you're the only one in the galaxy who misunderstood me. I was talking about how much effort mp3.com puts into telling everybody how cool they are and trying to play up the "gee whiz" factor instead of working on really helping users find music they like.

I don't know about the planet you live on, but here on Earth, companies depend on selling their wares to a reasonably large customer base. A company that only bothers to reach "mainstream" customers such as yourself, and alienates all others, does not have a bright future.

[ Parent ]

Re: You may be right, but not in my mind. (none / 0) (#23)
by rednecktek on Wed Oct 04, 2000 at 02:46:47 PM EST

but surely you appreciate the fact that not the genre you favor is very popular these days, and people who like different sorts of music will have a much harder time.

You're absolutely right, I can't remember the last time I went looking for "A Maiden transforms into a Salmon."

they're the sounds that mp3.com wants you to hear

Pardon me?! I make up my own mind about what I want to hear, thank you very much. I suppose you also believe everyone is going to vote for Bore or Gush too (because they are the only two candidates, NOT!).

I was talking about how much effort mp3.com puts into telling everybody how cool they are and trying to play up the "gee whiz" factor

I didn't misunderstand you at all. Once again, if people don't come to the website to view the banner ads, you're out of business.

"mainstream" customers such as yourself

Well, that's the first time I've been called that before. Regardless of my musical taste or yours, companies that want to grow MUST appeal to the most common denominator in order to gain the greatest chance at a succesful business.

Although it is (and was) clear to me that you and I will not agree on this subject, instead of trying to rip down a company that clearly doesn't offer what you want, why don't you go find one that does?

Just remember, if the world didn't suck, we'd all fall off.
[ Parent ]
Listening to one such cool band right now... (3.50 / 4) (#10)
by Ricdude on Tue Oct 03, 2000 at 11:11:08 AM EST

A few months ago, one of my coworkers concocted a script to download all songs by the top 50 bands on AudioGalaxy.com. Most of it's total crap, of course, but there are about ten or fifteen bands that are easily as talented as anything I've heard on the radio. One of my favourites out of the whole mess is just a catchy guitar wank by some high school kid. The production on the track is total crap, but the foundation is in there. I've had the AudioGalaxyCharts in my playlist since then, and find it as good as any radio station in my area. As far as objective measurments of songs, does it really matter what the tempo is? And do you count 1/4 notes or 1/8 notes as one beat if the time signature is 7/8? Does it matter if it's computer generated? Or if I program the sequencer? Or if I digitally clean up a midi recorded performance? Do "Should I Stay Or Should I Go" by The Clash and "Caress Me Down" by Sublime count as english, spanish, or both? The local record store when I was in college would turn over managers every year, and each one had their own idea of which bands qualified for which genres. Is Rush hard rock? heavy metal? rock n roll? Are Boyz II Men Rap? Soul? R+B? I frequently wished they'd just alphabetize the whole store in one category (ok, soundtracks and compilations could be separated) and make it easier for me to find what the music I went there to buy. Personally, I think the best way to find new and interesting bands is through personal recommendations. Barring that, your only real option is to be exposed to a large quantity of potentially good music (broadcast and internet radio, mtv, mass download) and hope for a good track once in a while.

Re: Listening to one such cool band right now... (3.00 / 1) (#11)
by El Volio on Tue Oct 03, 2000 at 12:04:42 PM EST

But that's just what a lot of us want to avoid: having to wade through mounds of useless crap to find what we want. And for many of us, the music we're seeking may never make any top 50 list, or may be a genre obscure enough that it gets lumped in with a few other obscure genres, ruining the whole process. Your process works great for relatively mainstream types of music (rock, etc.) In fact, it seems to me that a lot of the subgenres are another way to try to drill down to what you're looking for. Unfortunately, artistic endeavours, especially the really interesting stuff, often defies such classification. When you see critics referring to "post-Seattle anti-modern industrial retro-grunge punk", you know they're reaching. And tastes often are not that specific; I like rock music, and I like Latin music, but I hate Latin pop (Ricky Martin, etc.), and I only like some of the overlap. And a lot of those two genres is crap, too.

Basically, this is all based on Sturgeon's Law, which states essentially "90% of everything is crap." I'm sure somebody will dig out the actual quote, but I can never get it right. Finding the good 10% is difficult, especially since what I consider to be that good 10% may differ completely from what you consider to be.

Personal recommendations are the way to go, and "collaborative filtering" (sort of like Amazon's "Other people who bought this also liked...") seems to me like a good way to solve it, or at least reduce the waste. At least that way, there's a good chance it's not crap, since somebody who agrees with me that Los Jubilados are among the greatest musicians of the last 50 years is likely to have a great deal of overlap in their tastes with me, but somebody who likes Marc Anthony or any of those "Spice Boys" groups probably will have very little in common with my musical tastes.

[ Parent ]

Alternatives? (2.00 / 3) (#12)
by Erf on Tue Oct 03, 2000 at 12:33:38 PM EST

So are there any alternatives to mp3.com that work better?

...doin' the things a particle can...

Holy totally opposite impressions Batman... (3.33 / 3) (#13)
by loner on Tue Oct 03, 2000 at 02:13:11 PM EST

Wow, I've got the totally opposite impression of mp3.com. Maybe it depends on the type of music one is looking for. I listen to Blues, Reggae, French, Celtic, Latin, etc. I just go to mp3.com, then to the category I want, pick a band from the list and click on their page, then if most of their songs matches the category, I listen to a couple of songs in lo-fi and if they're to my taste, I download the songs. I do wish there was a link to "download all songs" :)

Anyway, there is not tooo much new daily material in the categories I listen to so maybe that makes a difference. But to me mp3.com has got it exactly right as far the new "medium" of internet is concerned for music. What mp3.com brings me:

  • The ability to find and listen to new music of my taste no matter what part of the (connected) world I'm in.
  • A very convenient way for me to sample new songs and decide for myself which ones I like.
  • Web pages with direct links to other information on each artist so I can read more about them while I'm listening to their songs.
I do agree that one crucial thing is missing from mp3.com: a discussion thread attached to each artist where visitors can post comments. This is such an obvious oversight though I wonder if there is a compelling reason for its omittance.

What I'd love to see is a /.-style site to point out good mp3-bands and discuss them. Anybody with some time and hardware wants to start one?

[OT] Discussion site (none / 0) (#14)
by El Volio on Tue Oct 03, 2000 at 02:27:49 PM EST

This might be an idea worth exploring... I wonder if there are any such sites? If not, I might be interested in getting one started. Don't mind hosting it for a while, either.

Of course, you get into the whole classification debate as you start to organize the site, as well as the "M374LL1C4 R00LZ! UR B4ND SUX H4RDC0R3!"-type posting problems... Scoop would probably be a good engine for the discussion, and if there were a "Recommendations" area as well (implementing some of the collaborative filtering being discussed in another thread), then you might have something.

If you're really interested in doing this, e-mail me and we'll talk about it.

[ Parent ]

Chinese lyrics... (3.66 / 3) (#15)
by grifter17 on Tue Oct 03, 2000 at 02:36:21 PM EST

Have you tried browsing by region, if you go to golocal.mp3.com you can narrow your search down to Chinese alternative music for example. Strangely Hong Kong is not listed as a sub region of China, rather a sub region of Asia...

Re: Chinese lyrics... (3.00 / 1) (#20)
by forrest on Tue Oct 03, 2000 at 09:40:40 PM EST

Well, this is an interesting idea (thanks!) but a sampling of a half-dozen songs hasn't turned up any Chinese lyrics yet. So far, they're all instrumental ...

I'm afraid if I want to turn up anything, it will take a lot of time, as I have found typical of mp3.com ...

[ Parent ]

The Catastophic Failure of MP3.com | 27 comments (27 topical, 0 editorial, 0 hidden)
Display: Sort:


All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!