Le DRM est mort, vive de musique numérique!

January 6, 2009 4:15 PM

Who would have known that that the death of DRM would come in the form of a press release? While MP3 stores are nothing new, with iTunes moving to a 100% DRM free catalog by the 31st of March this now cements a de facto standard of DRM free music in the marketplace. As a side effect it's now a near certainty that AAC will become the successor of MP3.
posted by Talez (135 comments total)

Successor of MP3? What makes you say that? I highly doubt MP3 will have a "successor" anytime soon, if ever.
posted by archagon at 4:19 PM on January 6, 2009 [3 favorites]

As a side effect it's now a near certainty that AAC will become the successor of MP3.

Perhaps, but 95% of people will continue to call it mp3.
posted by mannequito at 4:20 PM on January 6, 2009 [6 favorites]

As a side effect it's now a near certainty that AAC will become the successor of MP3.

No.
posted by selfnoise at 4:22 PM on January 6, 2009 [19 favorites]

I think the OP is right in that mp3 will become a minority format in the near future, but only because there are so many options out there. Is a stretch to anoint AAC as the successor, but then Apple pundits like making bold predictions (as well as Apple detractors).
posted by cjorgensen at 4:22 PM on January 6, 2009

Don't they make you pay to get the DRM-free version even if you already bought the DRM one? How lame is that?

And don't they still have DRM on the videos they will happily sell you from the store?

I like Apple. I have an iPod. I buy my songs at the Amazon MP3 store because they're better quality, cheaper, and plain old MP3s. iTunes merely drew even with this announcement. (Though Amazon is still cheaper for most songs.)
posted by smackfu at 4:23 PM on January 6, 2009 [1 favorite]

If MP3 is going to have a successor it will be FLAC.
posted by Jairus at 4:29 PM on January 6, 2009 [22 favorites]

Whatever format I end up getting my music in (almost always .ogg), I have to convert at least some of the songs to .mp3 anyway, as that seems to be the only format my mobile phone recognizes.
posted by Marisa Stole the Precious Thing at 4:30 PM on January 6, 2009

AAC itself is still owned by Apple, so device makers will have to pay licensing fees if they want to include AAC support. I suppose it could happen, but consumers would have to demand the ability to play iTunes music on non-iPods. Apple will do its best to stop anyone else from interfacing their players with iTunes as well as iPods do, so I don't see a big demand driver there.

I suspect, for lossily-compressed music, there will be probably never be a true MP3 replacement. Rather, someday, as bandwidth gets cheaper, the market might switch up to CD-quality "lossless" audio. FLAC might win that battle; it's the best-supported of the lossless formats.

AAC does sound better than MP3 at lower bit rates, but at the 256Kb level of both iTunes and Amazon, the differences are pretty minor. Even going from 256Kb to lossless isn't a dramatic difference; it's one you have to have decent audio gear to hear, and even then it can be hard to tell. The minor difference between AAC and MP3 will never drive a format change. I'm not sure 256Kbit to lossless will, either.

The broad music market may end up settling on 256Kbit MP3, forever. It sounds very good, doesn't take much space, and works on everything.
posted by Malor at 4:32 PM on January 6, 2009 [2 favorites]

smackfu, they do charge for the upgrade, but you pay the difference, not the full sticker price.
posted by Malor at 4:34 PM on January 6, 2009

AAC is not owned by Apple.
posted by dbiedny at 4:36 PM on January 6, 2009 [2 favorites]

Oops, sorry. But they still have to pay for it, so the rest of the argument stands. :)
posted by Malor at 4:41 PM on January 6, 2009

I like Apple. I have an iPod. I buy my songs at the Amazon MP3 store because they're better quality, cheaper, and plain old MP3s. iTunes merely drew even with this announcement.

This.

I just made a playlist of all my DRMed stuff I've bought over the years, did the math to figure out what it would cost to unshackle them, and said aloud, "Oh, fuck you, Apple."
posted by BitterOldPunk at 4:41 PM on January 6, 2009 [5 favorites]

Amazon has been doing this for a long time (--- not exactly earthshattering news here. I think Apple was one of the last places still selling DRM'd mp3s.

And yeah, the idea that AAC will replace MP3s is silly to me -- even though iPods have a ridiculously huge market share in dedicated music players, the number of other devices that play MP3s (receivers, televisions, game consoles, phones, car head units, toilets (probably somewhere), etc) makes it pretty unlikely.
posted by wildcrdj at 4:41 PM on January 6, 2009

you pay the difference, not the full sticker price.

Actually not quite. That WAS true when the iTunes Plus tracks first came out at $1.29; you paid the $0.30 difference to upgrade. But the Plus tracks dropped a while ago to $0.99, and the upgrade changed to a flat $0.30, or 30% on an album.

I did not know this either until I just looked up.
posted by smackfu at 4:42 PM on January 6, 2009

AAC is harder to say than MP3, though. Unless folks start pronouncing it as "ack".
posted by flapjax at midnite at 4:43 PM on January 6, 2009 [2 favorites]

Ah yes, I knew I'd find one somewhere: MP3 toilet seat.
posted by wildcrdj at 4:43 PM on January 6, 2009 [1 favorite]

> AAC does sound better than MP3 at lower bit rates

Actually, that's not true anymore (unless you mean lower than 128kbps, in which case maybe). Here's a recent 128kbps listening test.
posted by archagon at 4:45 PM on January 6, 2009 [1 favorite]

Also, most lossy encodings are surprisingly indistinguishable from lossless at even 128kbps, although it depends on the material.
posted by archagon at 4:47 PM on January 6, 2009 [1 favorite]

(By "indistinguishable", I mean when compared using a double-blind test.)
posted by archagon at 4:50 PM on January 6, 2009

I don't know of this "death of DRM" you speak of. DRM involves more than just music. What about locked-down devices like XBoxen and mobile phones? Software, video, e-books? Just sayin.
posted by naju at 4:51 PM on January 6, 2009 [3 favorites]

I really, really can't imagine AAC becoming the "standard" encoding format. As hard as it is for some to believe, the earth does not revolve around iPods / iTunes, for a lot of people.

My home stereo plays MP3s, it doesn't play AAC.
The USB-to-Radio gadget I've got in my car plays MP3s, it doesn't play AAC.
The sites I generally download music from, such as Archive.org, Jamendo, various MP3 blogs, almost exclusively offer MP3 or FLAC format. Not AAC.
Hell, my portable DVD player won't even play a lot of video files if the audio is encoded in AAC. Encoding as XVid+MP3 works like a charm, however.

AAC has been becoming more useful, as has OGG, in recent times, but saying ACC has replaced MP3 is like saying JP2 has replaced JPG.
posted by Jimbob at 4:52 PM on January 6, 2009 [4 favorites]

As a side effect it's now a near certainty that AAC will become the successor of MP3.

That's a pretty bizarre claim.

AAC itself is still owned by Apple, so device makers will have to pay licensing fees if they want to include AAC support.

You have to pay a licensing fee for MP3 playback as well, at least in theory.

Of course one of the big ironies here is that one of the main reasons that people chose MP3s for their DRM-free stories is so that they would play on iPods. But can most off-brand MP3 players play AAC files? I think to reach the widest market, people will have to continue putting out MP3s.
posted by delmoi at 4:53 PM on January 6, 2009

I'm pretty sure itunes/audible audiobooks are still drmed as well.
posted by aerotive at 4:54 PM on January 6, 2009

Of course, despite my annoyed post above, I decided to see if I had any upgrades available for the few songs that I did buy on iTunes. And now, somehow, Apple has $3.40 more of my money. So... Bravo Apple.
posted by smackfu at 4:59 PM on January 6, 2009 [1 favorite]

Looks to me like the apple.com/uk iTunes store has stolen a march & already sells it's mp3's DRM-free.

Why would that be so?
posted by dash_slot- at 5:04 PM on January 6, 2009

AAC itself is still owned by Apple, so device makers will have to pay licensing fees if they want to include AAC support. I suppose it could happen, but consumers would have to demand the ability to play iTunes music on non-iPods. Apple will do its best to stop anyone else from interfacing their players with iTunes as well as iPods do, so I don't see a big demand driver there.

As pointed out previously Apple don't own AAC. It's part of the MPEG-4 standard.

Even though you haven't heard much of AAC adoption it has been going on for quite some time.

* Almost every phone out there that supports MP3 supports AAC. Sony Ericsson have supported the format as far back as the K700 and W550 series while Nokia have had AAC support for yonks in Series 60 and Series 40. Mainstream Blackberry models (Pearl, 8800, Bold, Storm) all support AAC.
* All three consoles (PS3, Xbox 360, Wii) support AAC, the PSP and DSi both support it.
* Zens, Zunes and Sansas all support it natively.
* The Soundbridge and Squeezebox both support it
* Pioneer, Sony, Alpine, Kenwood, Clarion, Panasonic, and JVC head units all support m4as on CDs just like MP3s

The only mainstream players that doesn't support AAC natively? Windows Media Player (desktop and mobile) and even then Windows 7 will have native AAC support.

If any audio standard has a snowball's chance in hell of replacing MP3 it'll be AAC. I don't think there'll be any strong consumer demand for it but the natural quality increases and strong music store distribution of AAC files will result in people waking up one morning with a huge library of AAC files.
posted by Talez at 5:18 PM on January 6, 2009 [2 favorites]

Looks to me like the apple.com/uk iTunes store has stolen a march & already sells it's mp3's DRM-free.

Why would that be so?

80% of the music available on iTunes (the major labels) is DRM free and 256kbps from today. The other stuff is just signing contracts with the thousands of independent labels that distribute their iTunes music over the next couple of months.
posted by Talez at 5:19 PM on January 6, 2009

Also, most lossy encodings are surprisingly indistinguishable from lossless at even 128kbps, although it depends on the material.

At 128K I can almost always hear artifacts, most notably a flange-like/phase shift sound. At 192K and higher it's better, but there is still substantial loss of bass definition. It depends on what you're playing it on, the quality of the recording, the type of music, and who's the listener, obviously. But those are my biggest complaints about the mp3 format. Other formats like ogg are better, but lossless is still the real deal.
posted by krinklyfig at 5:19 PM on January 6, 2009 [3 favorites]

"The minor difference between AAC and MP3 will never drive a format change."

No, but by far the biggest vendor of music in the universe ever selling metric arseloads of AACs to everyone in the world ever might... ...is the point I guess Talez is making. Not that I buy that, necessarily.
posted by nthdegx at 5:21 PM on January 6, 2009 [1 favorite]

The biggest problem AAC had was its association with Apple's DRM. iTunes decoupling with that DRM is good news, but may be too late to help the format (which is actually pretty widespread now, but people have had time to associate "MP3" with "digital music.")
posted by ChurchHatesTucker at 5:21 PM on January 6, 2009

If any audio standard has a snowball's chance in hell of replacing MP3 it'll be AAC.
the only issue there is an iTunes bought AAC doesn't work in a S60 Nokia (for instance) straight away.

There are differences in file headers etc that make them slightly incompatible, it's inconsequential to fix, but still requires a fix.
posted by oliyoung at 5:24 PM on January 6, 2009

Seconding naju - this is merely one application of DRM, where we may see less of it in the future. However, I still can't download, nevermind buy, films or software in a non-DRM encumbered format, except through illegal channels.
posted by Brother Dysk at 5:25 PM on January 6, 2009

krinklyfig, it really depends on the codec. Many group ABX tests have been performed, and the general consensus is that if you're using LAME or OGG, 128kbps (maybe a bit higher) is probably going to be indistinguishable from lossless unless you have really good audio equipment or excellent hearing. Lossy codecs have improved substantially over the years, and "flange-like/phase shifted sound", to my knowledge, doesn't happen anymore. Are your MP3s recent and LAME-encoded?

Surprisingly, a lot of the differences that people hear are nothing more than the placebo effect.
posted by archagon at 5:27 PM on January 6, 2009

iTunes is dropping DRM?! HOLY MOTHER OF GOD
posted by exogenous at 5:28 PM on January 6, 2009 [1 favorite]

"The minor difference between AAC and MP3 will never drive a format change."

No, but by far the biggest vendor of music in the universe ever selling metric arseloads of AACs to everyone in the world ever might...

Call me when Pirate Bay users switch over. I would expect they're "shipping" way more MP3 files than Apple ships AAC files.
posted by GuyZero at 5:28 PM on January 6, 2009 [5 favorites]

krinklyfig, it really depends on the codec.

Yeah, that's a problem, too, because I have no control over what sort of codec a publisher may want to use. Better that they leave it true to the original master as possible and let me do the encoding, if I so choose. I realize that makes me a relic. I still buy physical CDs if there is no wav/flac option for download. So be it. I have to live with it, not you.

Surprisingly, a lot of the differences that people hear are nothing more than the placebo effect.

I hear a lot of people make this claim, but I have a lot of experience in running live sound and recording, and I trust my ears. For one thing, if you play digital tracks over a serious system in a club, like a modern DJ would do, you definitely hear any lossy issues. It particularly comes through in the low end, and especially with electronic music. And when you're dealing with massive soundsystems and huge subs and music which is written in the sub-bass range, it's not so much about placebo with the bass; it's what you feel in your chest and feet. This is one reason why DJs insist on lossless a lot of the time. And it's why Beatport offers wav (for an extra charge) as well as mp3, because the audience consists of more situations than just a teenybopper listening on their crappy headphones.
posted by krinklyfig at 5:34 PM on January 6, 2009 [1 favorite]

smackfu: "I like Apple. I have an iPod. I buy my songs at the Amazon MP3 store because they're better quality, cheaper, and plain old MP3s. iTunes merely drew even with this announcement. (Though Amazon is still cheaper for most songs.)"

Ditto. The iPod / iTunes combo is the best music management software + playback device around, mostly because they work well with each other. (I have some issues with the iPod design, most surrounding how hard it is to use with gloves on or without looking at the interface. But I understand it's designed more for the subway crowd than the cupholder one, and the proliferation of car head units with iPod integration has made this less of an issue than it used to be.)

But the iTMS has always struck me, pretty much from Day 1, as a way to separate fools and the very impatient from their money.

You can get a better selection, better price, better quality, a physical artifact that serves as a backup, and the ability to resell your music, by going to Half.com; more recently, Amazon has offered a superior product (DRM-free, actual MP3 files) at a lower price.

It's nice, I suppose, that they've breached the B&M wall and brought digital distribution (well, legal digital distribution — Napster certainly beat them to it overall) to the teeming hordes, and in doing so perhaps lowered the barrier to entry for indie bands and new talent, but iTunes has gotten steadily worse as it has become more focused on the iTMS and less on its core "Rip, Mix, Burn" (or Rip, Mix, Sync) functionality.

I love the iPod, but I'd be happy to see someone blow Apple away with a really superior product.
posted by Kadin2048 at 5:34 PM on January 6, 2009

This isn't an Apple thing. It's a greedy record company thing, where they tiered pricing in return for allowing Apple to distribute DRM free tracks.

I've always wondered why the record labels could allow DRM free on Amazon, but force Apple to include it. It's the same bits.
posted by Brandon Blatcher at 5:40 PM on January 6, 2009

Oops, sorry. But they still have to pay for it, so the rest of the argument stands. :)

No cheese. AAC is a standardised format with no licensing/patent fees (except for those who make AAC codecs). On the other hand, MP3 has a whole host of patent issues, with different companies claiming ownership and legal threats left right and centre.

Audio quality? AAC is better for the same bitrate.
posted by stepheno at 5:41 PM on January 6, 2009 [1 favorite]

I'm going to amend my previous statement: it seems that MP3 actually reaches transparency around 160kbps. Sorry. (128 should still be indistinguishable in most cases, though.)

I hear a lot of people make this claim
and I trust my ears

Is it still a "claim" if it's been verified by rigorous testing? I'm not doubting your experience, but "trusting your ears" can lead to ridiculous BS arguments for things like vaccuum tubes, resonating wooden volume knobs, vinyl over digital, etc.
posted by archagon at 5:42 PM on January 6, 2009

I've always wondered why the record labels could allow DRM free on Amazon, but force Apple to include it. It's the same bits.

The studios forced Apple to include DRM for the usual DRM reasons. Then Apple became the defacto place to go which gave Apple the upper hand in negotiations with the studios. So the studios wanted to go elsewhere, but now had to give the new parties a way to distinguish themselves from Apple (this is partly why it took Apple a year to shed DRM after they announced they would.)

It could easily have happened the other way 'round.
posted by ChurchHatesTucker at 5:44 PM on January 6, 2009 [1 favorite]

At 128K I can almost always hear artifacts, most notably a flange-like/phase shift sound.

I thought so too. Until I participated in a 128kbps listening test between different Lame encoder settings on Hydrogenaudio a couple years ago. On good equipment, with good headphones, I had to listen very, very hard for the differences between MP3 and original, and often couldn't decide which one was the MP3, especially on classical music samples with few transients. Now Lame is even better than when I did that test. There's no doubt that 1995 era 128kbps MP3s made by l3enc sound like shit, though. Check the results from the listening test archagon linked to above; modern MP3 encoders are incredible.
posted by zsazsa at 5:48 PM on January 6, 2009

Well, at least we know that *.m4p won't be the standard of the future.
posted by dunkadunc at 5:50 PM on January 6, 2009

Is it still a "claim" if it's been verified by rigorous testing? I'm not doubting your experience, but "trusting your ears" can lead to ridiculous BS arguments for things like vaccuum tubes, resonating wooden volume knobs, vinyl over digital, etc.

Well, I've already had this conversation, and I've learned it really isn't important convincing you or anyone else that I prefer lossless, or one thing over another. Who are you, and why is it important that you approve of my choices? But I will state that I do prefer certain things, although going around in circles is pointless and really unproductive for both of us. Why the fuck do you care?

I understand that some people can't hear the difference between tubes clipping and transistors, but I call bullshit on that. Hey, that's how you create *real* distortion on a guitar amp. Tubes do clip differently than transistors. I've actually tested this with an oscilloscope, and I've built amplifiers. I've also run the signal from vinyl and from a CD of the same recording through an oscilloscope. It's different. Volume knob material and ridiculously priced cables are pointless and for people who have more money than sense.
posted by krinklyfig at 5:52 PM on January 6, 2009

krinklyfig, I'm not trying to convince you of anything - just trying to clear up some misconceptions for others in the thread. In fact, I also prefer lossless. :)
posted by archagon at 5:55 PM on January 6, 2009

Talking about Mp3 quality vs. bitrate is a little misguided. An MP3 file is just a bunch of data that describes how to produce a soundwave. There are standard rules for converting the data into a soundwave, but there are no standard rules for converting soundwaves into MP3s.

The more data you use to describe the wave, the more closely it will match, but you can also be smarter about it. A really high quality encoder will produce better sounding Mp3s at lower bitrates then a crappier one. You can't just say "160kbps is transparent" With a shitty encoding, you might hear artifacts at 256kbps whereas an awesome one might be indistinguishable at 128.
posted by delmoi at 5:56 PM on January 6, 2009 [1 favorite]

Until I participated in a 128kbps listening test between different Lame encoder settings on Hydrogenaudio a couple years ago.

I've done some listening tests. Most of the time I can pick out the lossy format right away. For the music I listen to, it's very obvious. I only have a CD player in my car, and I burn everything from the CDs I buy so as not to ruin the original. But I used to burn a lot from mp3 for a couple years, and when I switched to doing it this way, I re-burned a lot of stuff I already had previously. I could hear things in the music I couldn't before, little details and dynamics, and the bass was much better. I did my own blind test with my car stereo. Sounds way better with lossless. So, anyway, I'm glad it doesn't bother you, but it does me.
posted by krinklyfig at 5:56 PM on January 6, 2009

Also, re: tubes, I meant for audio playback vs. digital conversion (I think? not quite sure how it works).
posted by archagon at 5:58 PM on January 6, 2009

AAC is better for the same bitrate.

Not really any more. A whole three years ago, Lame and iTunes AAC were statistically tied in listening tests at 128kbps. In the meantime, Lame has improved even more. More on listening tests. By the way, Amazon MP3s are encoded with Lame.
posted by zsazsa at 5:58 PM on January 6, 2009 [2 favorites]

delmoi, sorry. I meant transparent with the latest LAME codec.
posted by archagon at 5:59 PM on January 6, 2009

krinklyfig, to be honest, I find it highly unlikely that your music sounds "way better" in lossless on your car stereo - unless, of course, you have a really awesome car stereo. Might I suggest a proper double-blind test using foobar2000 and the ABX plugin?
posted by archagon at 6:02 PM on January 6, 2009

(Or unless you're using a really antiquated codec...)
posted by archagon at 6:03 PM on January 6, 2009

The more data you use to describe the wave, the more closely it will match, but you can also be smarter about it.

Yeah, but for me getting the original in non-lossy format is important. I can do 320K mp3 if I want, or flac or anything. But otherwise, I have to listen to whatever someone else decided to use. It's like letterbox vs. pan-and-scan. I can't understand why anyone would want to have some hack chop up a film so it fits the 1.33:1 tv screen, but some people don't understand why I can tolerate the black bars on the screen.
posted by krinklyfig at 6:04 PM on January 6, 2009 [1 favorite]

krinklyfig, to be honest, I find it highly unlikely that your music sounds "way better" in lossless on your car stereo

Well, you can come by sometime and check it out. But this is the sort of pointless argument I'm talking about.
posted by krinklyfig at 6:05 PM on January 6, 2009

I know a kid who transcodes all his music to 96kbps because he 'can't tell the difference'.
posted by dunkadunc at 6:07 PM on January 6, 2009

Oh, and re: 128/160kbps transparency, I meant variable bitrate (VBR). CBR is very inefficient.
krinklyfig, are you aware that 320kbps MP3 is still very much lossy?
posted by archagon at 6:07 PM on January 6, 2009

The only time I've found that low bitrate mp3s make a difference in quality is when you're using them with ableton or final scratch or some other app that manipulates the wave form, especially with time stretching. Lower bitrate mp3s seem to sound noticeably worse after any kind of digital manipulation.
posted by empath at 6:09 PM on January 6, 2009

For one thing, if you play digital tracks over a serious system in a club, like a modern DJ would do, you definitely hear any lossy issues.

I just wanted to second this. I don't really hear it at home or in my car, and only barely on good headphones. But something is noticably... less there... with mp3s vs cds or vinyl on a decent system (which includes the one in my garage ;))

I still use iTMS and mp3s and AAC and whatever, but I don't play out any more so it doesn't matter. If I did, and started playing digitally, I'd want at least 256k - or preferably FLAC - on my DJ tracks.

(does final scratch etc support FLAC? i don't even know.)
posted by flaterik at 6:09 PM on January 6, 2009

Don't they make you pay to get the DRM-free version even if you already bought the DRM one? How lame is that?

Not as lame as you having bought the DRM one in the first place.
posted by Flunkie at 6:10 PM on January 6, 2009

It's like letterbox vs. pan-and-scan.

Raising the issue of video brings about an interesting point.

Where are the people demanding lossless video encoding?

Everyone, at the moment, seems very happy with the quality produced by MPEG4 encoders. Hell, most people seem happy with the MPEG2 on DVDs. You don't see any puritans claiming they can't stand crappy lossy video, not understanding how anyone can stand to watch it, boasting that they like to have archives of the pure, original, lossless video source.

However, I can guarantee once we all have 50TB harddrives as standard, and 10-gigabit internet connections, these "lossless video" purists will appear out of the woodwork, and start telling us how unwatchable DVD or MPEG4 video is.
posted by Jimbob at 6:11 PM on January 6, 2009 [9 favorites]

> Well, you can come by sometime and check it out. But this is the sort of pointless argument I'm talking about.

The reason I'm asking is that no one, to my knowledge, has managed to consistently distinguish (via double-blind test) good bitrate LAME MP3s from lossless, especially on some dinky car stereo. If you really wanted solid evidence, you could perform a proper ABX test in foobar2000, as I suggested.

Not sure how that's "pointless".
posted by archagon at 6:12 PM on January 6, 2009

(Not to mention how happy people are to reencode video...but people who dare to reencode audio to a different bitrate, or between formats, apparently deserve to be infected with anthrax.)
posted by Jimbob at 6:14 PM on January 6, 2009

To be fair, Jimbob, we don't exactly have the ability to watch lossless video at the moment.
posted by archagon at 6:15 PM on January 6, 2009

Seconding exogenous' astonishment over the news. E-Books and X-Box notwithstanding, the iTunes Music Store is the place where by far the most people interact with DRM (and don't know it). This is a big deal.

On AAC v MP3: Our past experience with VHS v Betamax suggests it doesn't matter which format is better per se, it only matters what format the industry leader (here, clearly Apple) adopts.
posted by l33tpolicywonk at 6:15 PM on January 6, 2009 [1 favorite]

Not sure how that's "pointless".

What is it that you want from me in particular? I'm telling you my experience, and you're telling me I'm wrong. Well, not really much more to be said about it, right? Are you trying to prove something to me? Because all I'm doing is sharing my own experience.

If the planets all line up correctly and I use just the right codec, I'm sure it will sound pretty good. At 320K I have a really hard time telling the difference. This still doesn't mean my experiences were imagined.
posted by krinklyfig at 6:17 PM on January 6, 2009

I'd argue that the industry leader is piracy, by far.
posted by archagon at 6:17 PM on January 6, 2009 [3 favorites]

To be fair, Jimbob, we don't exactly have the ability to watch lossless video at the moment.

But, as soon as we have that ability, someone's going start claiming the old "lossy" video is unwatchable.

I remember my first ever MP3. It was a 96kbps encoding of Tommy the Cat by Primus. It came in a zip file, along with a very early version of Winamp, that I downloaded from a BBS. I thought it was the best sound I'd ever heard my computer make.
posted by Jimbob at 6:18 PM on January 6, 2009 [1 favorite]

Sorry, krinkly. I'm not trying to argue with you or prove you wrong, really.
posted by archagon at 6:20 PM on January 6, 2009

Who imagined up this format war? It ain't physical media like blueray & hd-dvd. All players must support mp3 & acc, period. A very good player should support ogg, wmv, etc. too.
posted by jeffburdges at 6:20 PM on January 6, 2009 [1 favorite]

I'd argue that the industry leader is piracy, by far

AAAARR! AAC! AAARR! Whar's me iPod?
posted by flapjax at midnite at 6:21 PM on January 6, 2009 [4 favorites]

To be fair, video artifacts are a lot more noticeable than audio artifacts.
Also, I don't think anyone's making the claim that lossy music is "unlistenable".
posted by archagon at 6:22 PM on January 6, 2009

On AAC v MP3: Our past experience with VHS v Betamax suggests it doesn't matter which format is better per se, it only matters what format the industry leader (here, clearly Apple) adopts.

What archagon said, though I wouldn't call it piracy. I'd call it "free exchange" or "p2p sharing" or something. (To me, piracy implies selling something that you don't own the rights to.)

I vote for FLAC as well, or another lossless, high-quality format, though it's hard to beat MP3 for universal hardware acceptance right now.

Also, DRM has been dead for a long time. Just don't accept anything with DRM and it's dead to you. It's not that hard. ... until they start cracking down hard on hardware, I suppose. But why not just keep our current devices?
posted by mrgrimm at 6:25 PM on January 6, 2009

Sorry, krinkly. I'm not trying to argue with you or prove you wrong, really.

Well, look, it's not a big deal, but what do you want me to say? I'm not going to start ripping my CDs onto mp3s which I then re-convert to wav to burn to a CD. That's like photocopying your digital prints of your pictures and then saving the scans. Right now I'm really not in the market for a portable ipod type thing. When I finally get one, it will play flac. I'm OK with this. Somehow seems to piss certain people off a lot when I tell them about my choices and why I make them. This is a bit perplexing, like the comic book-guy in reverse. Hey, listen to whatever you want to, and let me do the same.

The market doesn't know what to make of guys like me, but they keep making CDs, because there is a market. I think the physical medium is a bit archaic right now, but my hope with digital is to get better recordings, better media, not compressed and compromised. Up to the CD, the sound kept getting better. Post-CD, we seem to be going the other direction. Well, except now most people spend a lot of money on a big tv rather than a stereo system, per se, so there is blu-ray and HD. But people think I'm a fanatic because I prefer "high definition" audio and don't think it's strange that people spend several grand on a home theater.
posted by krinklyfig at 6:33 PM on January 6, 2009

Lol, the only way that double blind testing between MP3 (of any bitrate) and lossless audio would be considered "transparent" would be if the testers were completely untrained. Any audiophile worth his salt could spot the sound degredation easily. I have done so in the past, and don't have particularly well trained ears.

I will admit that the newer version of LAME do have simply amazing results, but don't kid yourself in the slightest thinking that it's indistinguishable from lossless! Even at maximum quality there is a significant difference (well to those who care about the quality of audio).
posted by parallax7d at 6:37 PM on January 6, 2009

krinklyfig, to be honest, I find it highly unlikely that your music sounds "way better" in lossless on your car stereo - unless, of course, you have a really awesome car stereo. Might I suggest a proper double-blind test using foobar2000 and the ABX plugin?

Remember that even then, you're probably not doing a good comparison. Computer sound is shit. To really, truly do it properly, and avoid the forced-48khz resample that so many cards do, you usually have to use a card that supports bitperfect S/PDIF. The cheapo Chaintech AV-710, about $25, has worked for that for a long time. You drive it in ASIO or kernel streaming mode with a good player, run a fiber or coax to a receiver, and use its DAC instead. Even inexpensive receivers will usually do a good job.

THAT will let you really test compression. Doing it with most soundcards doesn't mean very much; they have such poor native reproduction that the flaws of the underlying codecs are hidden. The X-Fi's pretty good if you put it into bit-matched mode. But it can also drive a bitperfect S/PDIF with some configuration effort, which is better.

To test your audio path, play a DTS-encoded 44.1Khz WAV file. If you have bitperfect audio, your receiver will jump to life and you'll hear multichannel music. If any component in your chain is interfering in any way with the signal, it will damage the compression, and you'll either get silence or static. Should anyone be interested, I have a DTS-encoded WAV file I use for testing; drop me a MeMail and I can send you a link if you need one.

This is easy on Macs, btw; set the system and iTunes volumes to max, turn off EQ, run your fiber, and you're golden. The onboard Mac DACs aren't very good, but external bitperfect is dirt-simple.
posted by Malor at 6:41 PM on January 6, 2009 [2 favorites]

krinklyfig: I'm not saying that lossy can replace lossless - not at all! In fact, I totally agree with your reasoning. As I said, I much prefer lossless to lossy and only convert to MP3 for my portable player.

With that said, the perceptual differences between LAME MP3 and lossless are very small, and I have good reason to suspect that you might be imagining them. If you have the time, I would highly suggest performing a double-blind ABX test in a controlled environment for your own benefit.
posted by archagon at 6:42 PM on January 6, 2009

parallax, NO ONE has been able to consistently ABX MP3s at maximum quality in a controlled setting, except for problem samples. Not a single person. Also, the Hydrogenaudio community (i.e., the guys that most often perform these tests) is generally very well trained. I'm afraid you're almost certainly wrong.
posted by archagon at 6:46 PM on January 6, 2009

Where are the people demanding lossless video encoding?

They're on YouTube, watching a hailstorm of pixel mash that I would have been outraged to see on my Quicktime 1.2-running LC II, or thumped my 1991 broken-ass no-aerial green-only TV for showing.
posted by bonaldi at 6:59 PM on January 6, 2009 [3 favorites]

If you listen to overcompressed pop music, no doubt you won't hear any difference between MP3 and lossless: there's no dynamic range in most of today's music and the recordings are probably craptacular anyway, because the kids are listening to it on their shitty car radios, ghetto blasters, and crappy ten-buck earpieces.

Go haul out something mixed by Alan Parsons during the early seventies, and tell me you can't tell the difference. When there's crystal-clear musical work with wide dynamic range (and, for that matter, a wide frequency range, what with the variety of instruments), and you're playing it through something with a bit of quality, the MP3 just doesn't do the job.
posted by five fresh fish at 7:06 PM on January 6, 2009 [1 favorite]

Sweetness. Not so sure about your AAC theory, though. MP3 may not be the most efficient way to encode music, but

1. came first
2. has an excellent encoder (LAME)
3. file sizes are not that big or big of a deal anymore
4. way more hardware and software support
posted by lemonjel at 7:09 PM on January 6, 2009

The samples in the ABX test I linked come from many sources, including classical music. "Hearing" the difference isn't enough - you really have to ABX it to get rid of bias and the placebo effect.
posted by archagon at 7:09 PM on January 6, 2009

The current lossy video formats are unwatchable.
Really.

Especially when you put i on a freaking iphone or something and end up watching a classic like Aguille on a 3" screen.

I'll make an exception for the ridiculously high definition formats, though they are a pain in the ass to move from one computer to another, and sometimes even to figure out how to get them to play. (I remember still the day we had to wrangle with the 8GB file size limit in FAT32...)
posted by kaibutsu at 7:12 PM on January 6, 2009

NYTimes blog says iTunes will go to two-tiered pricing -- 1.29 for current and popular, .69 for old, with record companies getting to choose the price point.

So can I get a refund for the difference on all my 99 cent DRMed music that will now be 69 cents? I want 30 cents per track back! (Or I want nice fat CD-quality 16-bit .aiff versions to be available. Hey, a guy can dream, right?)
posted by fourcheesemac at 7:21 PM on January 6, 2009

On another note, as most people have been saying Amazon is generally awesome for buying MP3s. It is always my first choice, even if I see what I want in iTunes, I'll usually buy from Amazon if it is avialable there, because it is cheaper and all of the MP3s that I've downloaded so far from them have been awesome, sound quality-wise. I think all encoded with LAME using the -V0 setting (info on LAME). All of my ripped CDs I encode at LAME -V2, but that's just to save some space, and it is still totally transparent (no one can tell the MP3 from the original).
posted by lemonjel at 7:24 PM on January 6, 2009

Hitler never complained about DRM.

I'm just saying.
posted by blue_beetle at 7:25 PM on January 6, 2009 [2 favorites]

NYTimes blog says iTunes will go to two-tiered pricing -- 1.29 for current and popular, .69 for old, with record companies getting to choose the price point.

They're wrong, it's three tiered and they'll still have the $0.99 price point.
posted by smackfu at 7:33 PM on January 6, 2009

As a side effect it's now a near certainty that AAC will become the successor of MP3.

And monkeys flying out of my butt will be the successor to the absence of monkeys-in-butt.
posted by Civil_Disobedient at 7:34 PM on January 6, 2009 [1 favorite]

you pay the difference, not the full sticker price

I found this mineral that I called a beat. Paid zero.
posted by Joe Beese at 8:08 PM on January 6, 2009

Thank you Malor! Honestly, for conversations like this we need a little flag that indicates if the person commented has ever actually heard a good audio system.

I'm not doubting your experience, but "trusting your ears" can lead to ridiculous BS arguments for things like vaccuum tubes, resonating wooden volume knobs, vinyl over digital, etc.

Case in point :)

Seriously archagon, what have you listened to and what do you know about the history of sound reproduction? I'm not buried in the audiophile thing myself, but I was pretty enthusiastic for a couple of years.

The basic story about CD vs. vinyl is that when CDs first came out, they sucked real bad. Sometime in the '90s improvements in mastering technology and CD player technology lead to a wholesale shift of opinion. It became, "CDs sound different, not better or worse, just different." At that time it was thought that only really really good CD playback equipment could get you there, but digital to analog conversion has been getting consistently cheaper and better since then.

Tube vs. Solid State amplification is a little more murky. They have different characteristics, and that creates vastly different sound. My friends citation II sure makes Willie sound fantastic compared to any of the mid-fi solid state gear I have though (different room of course, but the same speakers, and until recently the same CD player). On the other hand, you have Bob Carver's demonstrations, which suggest that solid state amps are capable of getting the sound exactly right if you just tune them right.

Volume knobs is just silliness, of course, but.. volume control isn't.

Anyway, back to the topic at hand for a second. delmoi is right on the money when he points out that you can improve the quality of the coding, and that bit rate alone is a pretty useless means of comparison. Beyond that, I don't really care. Hard drives are big, so there is no reason not to use a lossless format.
posted by Chuckles at 8:08 PM on January 6, 2009 [3 favorites]

Anyway, much more important than the file format, who did the mastering? I'm sure that any relatively good mp3 rip of Stardust on Columbia Master Sound gold CD would sound far better than a lossless copy of the mid-90s mass market version. However, that is as meaningless as comparing bit rates without considering the coding.
posted by Chuckles at 8:20 PM on January 6, 2009

Your favorite encoding rate sucks.
posted by Joe Beese at 8:28 PM on January 6, 2009

Where are the people demanding lossless video encoding?

1920 horizontal pixels*1080 vertical pixels*16 bits per pixel*30 frames per second*60 seconds per minute*60 minutes per hour = 417 GB per hour HD video in "lossless". And that doesn't include the audio track. So good luck with digital lossless video.

But the video purists are definitely out there. They are the one's crowing about the superiority of 35mm film to any video.
posted by Pastabagel at 9:06 PM on January 6, 2009

So good luck with digital lossless video.

And 10 years ago, when the average person had a dial-up modem, and maybe a 1.8gb hard drive (like I did in 1998), we would have said good luck to lossless digital audio. Why, you could only fit 2 CDs on that harddrive, and it would take you about 2 days to download a CD! It would never be a realistic option, so we made do with MP3s, and thought they were fantastic. Like people are making do with DivX, now?
posted by Jimbob at 9:20 PM on January 6, 2009

(lossless video).

Am I the only one who has trouble watching DVDs because the encoding?

In particular, I like to watch black and white movies but, annoyingly, there are always coloured fringes floating when I watch B&W movies on DVD. I got a highly-recommended BluRay Sony which decreases but doesn't remove the issue.

Even in colour, doesn't anyone else notice patterns crawl as the camera pans and that sort of thing?
posted by lupus_yonderboy at 9:24 PM on January 6, 2009

Ahh damn. I came in here to find out how people were responding to Apple finally living up to Steve's big claim that "ohhh we'd just LOVE to go DRM-free, but those gosh darned labels won't let us. Golly gee" and all I see is back-and-forth griping between self-proclaimed audiophiles and those unfairly labeled as needing their ears calibrated for optimal sound.

I guess I'll have to come back when that storm dies down and a new one stirs up.
posted by revmitcz at 9:25 PM on January 6, 2009

16 bits per pixel? Try 24, Pasta.
posted by Talez at 9:25 PM on January 6, 2009

Even in colour, doesn't anyone else notice patterns crawl as the camera pans and that sort of thing?

Admittedly, I've noticed what should be smooth gradients (say, under the ocean in Finding Nemo) turn into distinct colour areas, like I was looking at a 256-colour GIF or something. I always put it down to the shitty DVD players I've owned.
posted by Jimbob at 9:33 PM on January 6, 2009

Derail: I don't want to cause a panic, but did Phil Schiller look like he lost some weight?
posted by mazola at 9:49 PM on January 6, 2009 [2 favorites]

Lossy audio codecs, as I understand it, take advantage of psychoacoustical masking effects to achieve the major part of their compression efficiency. IOW, they filter out stuff that is thought as "won't be missed". However, before the widespread adoption of MP3s et al., most consumers were regularly exposed to music not subject to such filters. That's not the case with video. People weren't used to watching stuff right off a Digibeta deck via component on studio monitors. They used to watch interlaced SD via coax/composite on a CRT. For the masses, the ubiquitous format for video consumption has only improved. Perhaps explaining the lack of consumer demand for lossless video.
posted by Gyan at 9:53 PM on January 6, 2009 [1 favorite]

1920 horizontal pixels*1080 vertical pixels*16 bits per pixel*30 frames per second*60 seconds per minute*60 minutes per hour = 417 GB per hour HD video in "lossless". And that doesn't include the audio track. So good luck with digital lossless video.

You are confusing uncompressed with lossless compression. For example, FLAC is a lossless compression format that reduces files sizes ~45% according to wikipedia.
posted by euphorb at 10:49 PM on January 6, 2009 [2 favorites]

Can anyone comment on Jimbob's last comment:

I've noticed what should be smooth gradients (say, under the ocean in Finding Nemo) turn into distinct colour areas, like I was looking at a 256-colour GIF or something. I always put it down to the shitty DVD players I've owned.

How much of that is the player, and how much of that is the codec?
posted by mbatch at 10:53 PM on January 6, 2009

They used to watch interlaced SD via coax/composite on a CRT.

And most people used to listen to crappy FM (or even AM) radio, or cassette tapes, or poorly-cared-for vinyl played on very average record players. 128kbps MP3 of a CD is an improvement on all of them.
posted by Jimbob at 10:59 PM on January 6, 2009 [1 favorite]

That's a great point Gyan.

There is another way that the comparison breaks down. Nearly perfect audio reproduction has been possible for something like 40 years. Sure, most people don't know it -- as evidenced by statements like "back-and-forth griping [...] self-proclaimed [...] needing their ears calibrated" -- but nonetheless it is almost possible to fool a listener in the "is it real, or is it..." comparison. On the other hand, as much as recent improvements in video are astonishing, to actually capture a real range of human visual perception is just an overwhelmingly complex task that no one has come close to yet.

For me, the root of this discussion about quality is a sense of disappointment. In a certain sense the audio reproduction experienced by average consumers has been getting slowly worse and worse for the last 25 years. Don't get me wrong, I don't think the average has actually been getting worse, but in the 70s the new high tech thing was quality audio. Consumers, being what they are, gravitated to the new high tech thing. It turns out that the average consumer just never cared about high fidelity sound, and that is disappointing to people who'd like to see artistic attention, reasearch, and economies of scale directed at quality audio reproduction.
Of course I didn't really experience the hi-fi era, but I get that sense of disappointment from jaded older audio enthusiasts as they shake their heads over the latest piece of crap HTIB on sale for $150 at Best Buy. And, thinking about it, I guess cassette and 8-track demonstrate that even then consumers prefered convenience over quality, but...
posted by Chuckles at 11:11 PM on January 6, 2009

How much of that is the player, and how much of that is the codec?

Well, I'm not exactly sure what codec means in this context. I think the problems being noticed are mostly in the mastering process used on the specific DVDs, and to a lesser extent due to the player. Of course the better the reproduction gear, the easier it is to see the problems in all stages of content creation: from the lighting and camera work in the original, to the colour balance when the film was processed, to the digital transfer and compression algorithm applied to fit it on a DVD.

When the playback gear is good enough to reveal the limitations of source material, you can apply dither. Basically, little details don't really matter until you perceive something wrong about them. If the defects are random, the brain tends to fill them in automatically (and/or ignore them completely), but if there is something systematic about the defects the brain focuses attention on them. Dither overwhelms the systematic defects with random ones, making the systematic ones much harder to perceive. As long as the dither itself isn't too strong compared to the picture being viewed (or the sound listened to), it is effective.
posted by Chuckles at 11:53 PM on January 6, 2009 [1 favorite]

(does final scratch etc support FLAC? i don't even know.)

Final Scratch is dead at this point. Serato, pretty much took over the digital turntable market, but their codec support is limited. Traktor and Live have supported FLAC, AAC, and OGG for a while now. Although 320kbps CBR MP3 encoded with LAME, are generally considered the standard if you're going to play them on a decent club system.
posted by inpHilltr8r at 12:44 AM on January 7, 2009

Chuckles:

> The basic story about CD vs. vinyl is that when CDs first came out, they sucked real bad
> At that time it was thought that only really really good CD playback equipment could get you there

Yes, I know this. I was referring specifically to those who claim that vinyl had better sound reproduction and authenticity than modern digital simply because it's analog - a provably BS claim.

> Tube vs. Solid State amplification is a little more murky.

Again, I was referring to those who claim that tube amplification has more accurate sound reproduction.

> Anyway, much more important than the file format, who did the mastering?

Absolutely.

I think you misunderstood my intent, as we seem to agree on pretty much every point. I was trying to point out that simply saying "it sounds that way to me!" can easily lead to pointless "audiophile" claims, since the human ear is very easily tricked. I mean, some people even claim to hear the difference between different lossless codecs: FLAC is totally "warmer" than ALAC, don't you know! A dark and dangerous road indeed.

Perhaps I wasn't clear enough? Sorry!
posted by archagon at 12:56 AM on January 7, 2009

OK, I did a bit more research, and apparently Serato now supports AAC, and OGG, but not FLAC.
posted by inpHilltr8r at 1:07 AM on January 7, 2009

Even in colour, doesn't anyone else notice patterns crawl as the camera pans and that sort of thing?

Yes, but I also noticed the tracking lines on my VHS tapes, and the "squiggle" when the tapes got played too many times.
DVD was so much of an improvement over VHS that I'm willing to put up with some jaggies and weird banding.
Blueray, though it thoughtfully adds much more stringent copy-protection and even more incompatible player nonsense, retains many of the same problems. So, for me, it wasn't worth the investment.
posted by madajb at 4:22 AM on January 7, 2009

Regarding compression: I've got a huge hard disk (320 gigs) and not a huge number of albums - there's no reason for me to compress files.

I bought the 160kbps download of In Rainbows when the almighty 'head released it online. It sounds excellent on my ipod and midrange earphones, and on my 'laptop->70s stereo receiver'. Even now, over a year after it's release, even on my ipod, there are still little details to discover. A different set of earphones reveals other bits - there's stuff still to be found in that much-bitched-about-the-bitrate version.

I bought the CD last year and ripped it at 320, VBR, LAME. Listening to the CD, the 320 and the 160 on my home stereo (Mac -> 70s stereo receiver for the mp3s, CD separate -> 70s Stereo receiver for the CD) I honestly couldn't tell the difference.

Audiophilia seems like such a strange hobby to me. There seems to be a difference between being passionate about sound, and being passionate about music. Sometimes a person is both.

OTOH I recorded an album with a friend in February 2007 (not going to self link, but it's in the profile) and we made it available in ogg, 256 VBR, and FLAC - Maybe two of the three people who might have downloaded it appreciate that.

Good to hear Apple going DRM free, anyway.

Also Regarding video: I saw JCVD at the cinema last night. When large areas of the screen were white, you could actually see the pixels.
posted by Cantdosleepy at 4:57 AM on January 7, 2009

There seems to be a difference between being passionate about sound, and being passionate about music. Sometimes a person is both.

The thing about audiophiles, they believe in the holy grail, of perfect sound reproduction, that there is some way of them hearing the music as the musicians intended. Never mind that their stereo systems probably give better sound reproduction than the mixing desk and monitor speakers in the studio where the album they are experiencing was originally recorded.

Personally, I feel the only way to listen to the first Weezer album is to obtain a copy on cassette, high-speed-dub it to a second cassette, and leave that one on the dashboard of your car, in the sun, for three, maybe four months, before putting that baby in your crappy 2-speaker car stereo. That's how I first heard it. I was very, very disappointed to discover, upon hearing the CD itself, that they weren't lo-fi garage rock geniuses.
posted by Jimbob at 5:17 AM on January 7, 2009 [2 favorites]

On AAC v MP3: Our past experience with VHS v Betamax suggests it doesn't matter which format is better per se, it only matters what format the industry leader (here, clearly Apple) adopts

Just to echo some earlier posts, I don't think it matters much what Apple does, whichever format the rippers/seeders use is going to be the defacto standard. I'm not My friend isn't seeing a whole lot of unprotected AAC being put up @ Mininova or Demonoid nor does he expect to see that change anytime soon. I don't think a lot of people that buy from iTunes are that familiar with the formats anyway, they pay their money, get their tunes and rest is pretty much irrelevant. If there is going to be a shift I would expect FLAC to become more popular.
posted by MikeMc at 6:07 AM on January 7, 2009

The thing about audiophiles, they believe in the holy grail, of perfect sound reproduction, that there is some way of them hearing the music as the musicians intended.

Word. I've recorded in a studio. We had a guy mixing our songs, and we had a fair amount of input into how it was mixed and how it was sounding. It was, however, going through a certain pair of monitors from a certain desk, neither of which we chose.

Then it was sent away for mastering, and some other guy did something to the audio, which obviously we didn't have the opportunity for active input.

The idea that somebody with a valve-warmed stereo out there is getting to hear 'what we intended' with his $N000 system is nonsensical. The only way to really do that is to find out what monitors and mixing desk were used during the recording and to put an unmastered copy through that. In a room exactly the same size and shape as the one we mixed in. It's a nonsense.

And loads of artists are unhappy with the way their albums ended up. Ever hear Steve Harris talk about the first Iron Maiden record? Which is one of the fan favourites? If you wanted to listen to that as it was intended, you'd have to travel back in time and have him re-record it. Maybe that little sonic detail you love so much from A New Career in a New Town on Bowie's Low actually drove the man to distraction, and he wanted rid of it.

How do you think Spaceman 3 intended you to listen to 'Taking Drugs to Make Music to Take Drugs To'?

Scott Walker only listened to Tilt once, I believe, once he had finished it; standing at the bottom of a large staircase with huge speakers at the top.*

You can't hear the album in the way the artist intended it, and incremental increases in 'warmth' and 'transparency' and 'detail' go for orders of magnitude more money. It's a sucker's game.

*I don't exactly remember this anecdote, but it was certainly something like that.
posted by Cantdosleepy at 6:15 AM on January 7, 2009 [5 favorites]

You can't hear the album in the way the artist intended it, and incremental increases in 'warmth' and 'transparency' and 'detail' go for orders of magnitude more money. It's a sucker's game.

I thought I had seen it all until I saw an ad in Sound+Vision for iron filings or special sand (or some such shit) to fill your speaker stands with to dampen vibrations. Come on dude, do you really need to fill your speaker stands with lead shot to hear the album "as the artist intended it"? Is there anything out there the audiophile won't buy?
posted by MikeMc at 6:26 AM on January 7, 2009

I've noticed what should be smooth gradients (say, under the ocean in Finding Nemo) turn into distinct colour areas, like I was looking at a 256-colour GIF or something.

This can also be due to just not having enough colors in the rainbow. I think all the video formats are still 8-bit, so while you have 16.7 million colors in theory, you only have 256 shades of pure blue, or pure gray. If you want a gradient that goes from one dark blue to another dark blue, and covers a significant part of the screen, you have to work to avoid banding.

(BTW, this is why Photoshop has 16-bit color modes.)
posted by smackfu at 6:29 AM on January 7, 2009 [1 favorite]

I listen to each of my MP3s the way the artist intends it -- with a buck less in my wallet.
posted by mazola at 6:31 AM on January 7, 2009 [1 favorite]

I like my MP3s like I like my women -- in a plastic cup.
posted by archagon at 7:55 AM on January 7, 2009

I like my MP3s like I like my women -- 2-bit, non-proprietary, and widely distributed.

And compressed! I always forget compressed!
posted by mazola at 8:05 AM on January 7, 2009 [1 favorite]

Well I can tell the difference personally and the slightest whiff of compression at all has me bleeding from the ears in under a minute. It's all because I happen to be highly musical with a finely evolved ability to discern the slightest imperfection. I also happen to be a fully qualified acoustic engineer, a virtuoso cathedral organist, a record producer who masters elusively on vinyl and a jet-setting club DJ the weekends. Thankfully all this talent leaves me with a large sum of money and I like to spend most of it on hi-fi so esoterically sophisticated I will not bore you by explaining the least of its wonders.

But please, the rest of you, enjoy your compressed music; it's kind of cute.
posted by rongorongo at 8:21 AM on January 7, 2009 [1 favorite]

Yes, you're all obviously correct, and we should all be listening to music with two cans and a piece of string.

Many audiophiles get very stupid and fall for the placebo effect. But going from there to scorn of all people who happen to like good music reproduction is stupid. Not all sound is the same, and people really can hear compression artifacts.

You don't have to spend a ton of money to get good sound; a nice set of headphones, possibly a headphone amp, and a high-quality DAC can be had for between $300 and $500, and that'll give you outstanding sound reproduction. With full-size gear, $1500ish (maybe $2k if you want a big subwoofer) will give you a very good, very solid multichannel system that will last for many years. You certainly don't need to spend tens of thousands, and you absolutely don't need super cables or special knobs or any of that other twaddle, but there are highly noticeable gains up to around that price point. Past there, the returns become much lower, and the snake oil component becomes much higher, but basic competent gear makes a very large difference.

If you learned to listen to music on the computer, and don't have any other experience, your opinion is probably worthless. Computer sound in general is horrible. You won't really know it's horrible until you've spent a substantial amount of time with something competent. It doesn't take a ton of money to get competent sound. I get very good results on the computer with a $250 Onkyo receiver and a decade-old $500 surround-sound system. (to be fair, it'd probably be about $800 if I bought it new.) But if you don't even know you need to, and you've calibrated your ears to the garbage on most soundcards, you have no idea what you're talking about.
posted by Malor at 9:09 AM on January 7, 2009 [1 favorite]

Malor, you're absolutely right, but chances are high that most people still won't be able to hear compression artifacts (assuming LAME MP3 or OGG), except for problem samples. As has been demonstrated by numerous ABX tests, the difference between, say, 192kbps VBR for LAME MP3 and lossless is perceptually negligible.
posted by archagon at 9:17 AM on January 7, 2009

Also, scorning good music reproduction is NOT the same as pointing out the fact that the lossy/lossless divide isn't that great anymore.
posted by archagon at 9:18 AM on January 7, 2009

Maybe this is a good place to ask something technical of MP3:

Is MP3 necessarily lossy? i.e.

a)if one were to take an extremely simple sound, would the resulting MP3 necessarily lose some data?

b)let's say you encode a regular sort of music 'S' to a MP3 file 'M' @256k. When you play 'M' in any media player, the decoder, AIUI, generates the corresponding PCM samples and sends them to the soundcard driver. If you were to instead write this decoder output to a WAV file 'W', would the encoded MP3 @256k of 'W' be 'M'?
posted by Gyan at 9:45 AM on January 7, 2009

I myself don't find 192K VBR MP3 at all annoying. I can hear that it's MP3 on some tracks I have, mostly because the spatial cues are lost, but it certainly doesn't leap out at me as being awful -- I have to really listen, and even then it's subtle. I generally encode to 224k for my own use, on the theory that it's just a little more space and can't hurt. And this is true even via bitperfect; 192K MP3, as encoded by LAME, really is very good. Even 128K really isn't too bad anymore. I'm not sure if LAME has improved, or my ears have aged, but I tend to find most 128K tracks pretty acceptable, where I hated them ten years ago.

So, please understand that when I say that soundcards are shit, I mean they're really shit, not just crappy from some imaginary plateau high in the realm of Audiophile Assholery.

I'd be happier about the Hydrogen Audio tests if we knew that the testers were using bitperfect, though. I participated in one a number of years ago, and I didn't realize, at the time, just how bad soundcards were, so I'm sure I gave them bad data.
posted by Malor at 9:48 AM on January 7, 2009

Gyan:

a) Yes, a simple sine wave should be exactly the same pre- and post- MP3.

b) 'W' would not be 'M'. Due to rounding errors, successive re-encodings will lose quality. A few years ago, I did some tests. While I don't have the original for some reason, here's the same song re-encoded, with the same settings, 10, 50, and 100 times successively (warning: your ears will hate you if you listen to 50 or 100). For maximum horribleness, I think I used the worst encoder at the time, Bladeenc. I also used LAME for one test of 100 reencodings that just seemed to get really, really quiet.

<derail>Note that JPEG re-encodings are much, much better! I've re-encoded JPEGs at the same low quality of 40 100 and 1000 times with little visible degradation. However, once you start changing the quality levels back and forth between re-encodings, the results get much, much worse!</derail>
posted by zsazsa at 10:33 AM on January 7, 2009 [2 favorites]

The only way to really do that is to find out what monitors and mixing desk were used during the recording and to put an unmastered copy through that. In a room exactly the same size and shape as the one we mixed in.

Hell, most of the music I've been making lately (and I'm not claiming to be Kevin Shields or anyone, but still), was recorded with a shitty mic I picked up from a pawn shop, and constructed as I listened to it coming through a crappy soundcard (often in a laptop) through $30 headphones, then encoded to 256kbps MP3. And that's how I intend my 3 loyal fans to enjoy it, dammit.
posted by Jimbob at 11:08 AM on January 7, 2009

I like my MP3s like I like my women -- in a plastic cup.

That's really uncalled for, distasteful and not funny. There's no need for it.
posted by Brandon Blatcher at 11:14 AM on January 7, 2009

It's an Eddie Izzard reference, but even if it wasn't - distasteful? Huh?
posted by archagon at 11:39 AM on January 7, 2009

Yeah, I don't know what he's got his knickers in a twist about either, archagon. Props to Izzard.
posted by Jimbob at 11:44 AM on January 7, 2009

Malor, I'm not sure if you're replying to me anymore, but just in case:

> So, please understand that when I say that soundcards are shit, I mean they're really shit, not just crappy from some imaginary plateau high in the realm of Audiophile Assholery.

Yes, I know. I even had to get an external soundcard for my main PC because of background noise.

> I'd be happier about the Hydrogen Audio tests if we knew that the testers were using bitperfect, though.

Good point. I've seen the same results in individual tests by very experienced members, though.
posted by archagon at 11:58 AM on January 7, 2009

AAC won't be replacing MP3 because it's not much better than MP3 and every piece of hardware in the land will be playing MP3 for the foreseeable future. Any arguments regarding the patent situation are largely moot as those covering the decoding process expire relatively soon (IIRC in 2011), likely shorter than the length of time required for a format like AAC to catch on as the industry standard.

Ogg will likely continue to be the royalty-free format of choice for game developers. Sadly Apple likely succeeded in ensuring that Ogg Vorbis and Theora wouldn't be the format of the web anytime soon.
posted by HaloMan at 1:15 PM on January 7, 2009

archagon, I'm sorry, I never really directly talked about what you were talking about, largely because I agreed with you. For most folks, most of the time, 192k MP3 is fine, and the 256k that both Amazon and iTunes sell should be, if anything, overkill. I can't imagine that most people would reliably be able to differentiate between the two vendors.

I still like getting CDs and storing the images losslessly. Once it's stored as an image, I never have to rip again, even if the popular format changes. It's easy to regenerate any other lossy format from a lossless source, and I have plenty of space. Plus, my Squeezeboxes work nicely with CD images, so I can stream directly from the server with no effort. And I can burn a new CD that's almost identical to the original anytime I need one, so the images function as a backup as well.

Converting from one lossy format to another is a Bad Idea, because you end up with the accumulated errors of both. So I don't buy music in MP3 or AAC, as I'd be stuck with those formats permanently. Would it matter? Probably not, but I'm hedging my bets. Once those bits are gone, they're gone forever, and I've got lots of drive space. Once you've ground the steak to hamburger, there's no getting the steak back.
posted by Malor at 1:51 PM on January 7, 2009

Sorry, it was kind of hard to tell because you were partly talking about stuff I had included in my post. Or maybe I just suck at reading comprehension.
posted by archagon at 2:02 PM on January 7, 2009

No, I was sort of talking orthogonally to what you were talking about. My mistake, not yours. :)
posted by Malor at 2:47 PM on January 7, 2009

Brother Dysk writes "However, I still can't download, nevermind buy, films or software in a non-DRM encumbered format, except through illegal channels."

There are probably terabytes of free (in both senses) software available for legal download.

dunkadunc writes "I know a kid who transcodes all his music to 96kbps because he 'can't tell the difference'."

I crank the compression up on MP3 cds for my truck. It's 30 years old so what minimal sound deadening and windows sealing it came with is either hard or missing. It's equipped with a free flowing 440. The 6" speakers (only two) are mounted in the doors at foot level. I'm amazed that between wind noise, road noisy, the blissful sound of a half cocked 440, and the crappy acoustics I can even make out what song is being played.

euphorb writes "You are confusing uncompressed with lossless compression. For example, FLAC is a lossless compression format that reduces files sizes ~45% according to wikipedia."

And isn't there a lot more identical information, frame-to-frame, in most video? I'd bet even something unoptimized for video like .zip could probably half the size of most lossless video.
posted by Mitheral at 7:50 PM on January 12, 2009

Download VLC (an open-source, well-regarded video player) and poke around the myriad preference settings. Somewhere in there are some modes that show you how the unchanging/small-changing parts of the picture are modified from frame to frame. It's pretty cool.
posted by five fresh fish at 11:28 PM on January 12, 2009