Slashdot Mirror
Are iTMS's 128kbps Songs Worth Collecting?
pinchhazard writes "Randall Stross of the New York Times offers his opinion on iTunes Music Store's decision to offer downloads at only 128 kbps, and that decision's potential to affect collectibility of the songs. The article says that Apple makes the claim on its web site that "you'll get the full quality of uncompressed CD audio using about half the storage space."
Rhapsody, which offers encoding at 192 kbps, is compared."
I'd expect them to be roughly equal to 160 bit MP3s, but I'd expect 192 bit MP3s to be superior. That's just opinion, though.
(although I'm unsure what Rhapsody uses, maybe it uses MP3Pro which is pretty good).
I think that 128kbps is a little shortsighted from Apple, there will be losses in the audio at that rate. 192 kbps AAC would be preferable of course.
Then again, most people listen to music on cheap headphones, speakers, etc, or just want music in the background. In that respect 128 kbps AAC is way more than necessary, and beats a cheap FM radio totally (if only in that you don't have a retard DJ wittering on between tracks).
Music is just part of life these days.
I've ran a number of informal listening tests, and many of the people I tested cannot tell the difference between a 128kbps MP3 (LAME) and a 256kbps MP3 (LAME) consistently, even on good equipment. (Too much loud music as a teenager, perhaps)?
However, there most definatly are people who can tell the difference, and I am one of them. Personally, I like 200+ mpc (MusePack) files - MusePack seems to do a good job preserving the crispness, and "body" (don't know a better term for it) of the audio.
I allways encode my stuff in 320kbps variable bitrate, since HD space is cheap, and if I lose a CD, I can be sure that I have a nice quality copy of the song. Recently, I've tried using FLAC too. Even better (lossless), but takes 2.5 times more space or so...
I would only buy 128 kpbs songs from itunes if they had some kind of system where I could download FLAC versions later, when I have more HD space. You've paid for 'mechanical rights', just like with full quality CS's, so why not?
In a few years, when everyone is on broad band and storage costs half of what it does now, they will upgrade to "premium" 256 kbps songs. Lots of people will buy their collection again, just like they did with their record collection.
Allofmp3 suppots on-the-fly encoding to let you have 192 kbps Ogg's or whatever. Even FLAC or raw CD Audio is available, but only for some songs.
:-P
I have no idea how legal the site is where I live, but it's definitely legal in Russia.
Beware: In C++, your friends can see your privates!
In my experience in using my iPod, I'm more than happy with 128kbps AAC encoded rips of my CDs and am very happy with the audio quality of the stuff I've bought off iTunes.
Line eater? What lin
I say boycott any format that is any worse than the modern 192k (preferably better). If they can really do 128k without sounding any worse, that's fine. But based on the reviews I've seen, they haven't, so it's not.
Am I part of the core demographic for Swedish Fish?
Call me crazy, but I was listening to this one song over and over again. I kept switching between 128 mp3, 192 mp3, 320 mp3, and I'd listen to 128 AAC sometimes too.
I cannot tell the difference between CD and 320 mp3. There is a very subtle difference between CD and 192. 128 is a joke.
For AAC, I've found that importing the song at 192 is about the same as 320 mp3.
Give it a shot. Take some song that has some subtle sounds, like accoustic guitar, and listen to it. Import it from the original CD and listen to all the formats. It's surprising. I used the song "Battery" from Metallica because it has a mix of sounds. Specifically at the beginning where they're using 1 or 2 accoustic guitars.
Let's Look at the benefits of purchasing compressed online music:
... uh, that's it.
1. Immediate gratification.
Now, let's look at the disadavatages of purchasing compressed online music:
1. Lower sound quality. Everyone I have compared them for has asked "What's wrong with it?" after listening to the CD and then the AAC verison.
2. Codecs are changing very rapidly. You are investing a a fleeting software phenomenon that depends on the current and rapidily changing technology and the marketing whims of the computer and music industries. Soon there could be much better quality or with increased bandwidth CD quality. SOme sights now sell 24 bit flacs which you can burn using you regualr old DVDs and burner into DVD-A for BETTER THAN CD QUALITY.
3. Commercial CDs are inherently more stable than CD-Rs.
4. It is extremely difficult and time concuming to archive digital files for very long periods of time.
5. In most cases you get no liner notes or cover art.
6. You invite DRM.
7. For all the above, at a lot of stores, particularly iTMS, you PAY MORE for all these problems than a fine sounding CD, or a much better sounding DVD-A or SACD.
I'm not into instant gratification as the article implies. In this consumerist society, I'd prefer to be an intelligent purchaser rather than one of the many throw away consumers. I wait for the CDs to show up on half.com for $5, which ends up less than $8 after shipping. I receive my uncompressed CDs at less than 80% of your $10, FLAC it and I haven't sacrificed any quality of the art. I also have the original so that I can circumvent the next disruptive technology to come along. I know not everyone can wait, but I can most of the time. Oh, yeah, I don't own an iPod. Instead, I have a Rio Karma, so FLAC is portable at the get go.
Good point. $1 per song is no bargain. At 50 cents per song, it might be worth it if you can accept less than stellar quality. I can go to the record store and find jazz compilations with 24 songs or more on a real CD for under $14. And this includes liner notes, artwork, and a plastic case.
Maybe for older folks whose hearing has degraded somewhat. People usually cite an upper limit of around 20kHz. I can certainly hear a tone at 20kHz, from a good tone generator (not a cheap one with harmonic interference.) That alone puts the Nyquist rate at 40kHz.
What's more, although people may not consciously perceive higher frequencies, work has shown that people do subconsciously perceive them.
To quote (from the article I'm linking):
The author also notes such facts as that 40% of a set of cymbal's audio energy is above 20kHz. So a 96kHz audio recording (range=48kHz) is not unreasonable. But good luck finding equipment to really play it back on correctly
Article: There's Life Above 20kHz!
A preposition is a terrible thing to end a sentence with.
No, it isn't. I suspected it might be deficient due to the misleading quotes in the slashdot summary. The author of the article, Randall Stross, knows perdectly well the difference between lossy and lossless audio compression and demonstrates that knowledge in his article. Both AAC and MP3 use lossy compression based on perceptual characteristics of human hearing. They both try to gain space by "throwing out" what we wouldn't be able to hear.
The quote about using half the space refers to Apple Lossless Compression which is yet another entry in the lossless audio compression sweepstakes. So far there are at least three other distinct candidates: APE, FLAC, and SHN. Each does its own 2:1 (approximately) compression which allows for all the bits to be restored during decompression. But 128 Kbps is a 10:1 compression and it is lossy whether one uses MP3, AAC, or ATRAC.
The lossless standards take up about 5 times the space/bandwidth but don't suffer from any loss of quality with respect to the red book audio standard. Since disk space and network bandwidth are cheap and getting cheaper all the time why bother with lossy compression? You could also say why bother with a mere factor of 2 and just use straight CD files but that factor of 2 is free. $50 is still less than $100 but the audio quality is identical.
Apple made a brilliant business decision to move early and decisively with the tools and standards that were available. The challenge now is to evolve to lossless compression as Apple has begun to do and eventually raise the bar for the commercial standard for digitized audio in general. Both SACD and DVD-A are hopelessly compromised by DRM bullshit so who knows if that will ever happen.
Also the reviewer must be confused the iTunes can encode in many formats, including Apples lossless format that takes about half the drive space but the iTMS is 128khz AAC files only.
I have recorded the same tracks at varying rates and it is very hard to tell the 128khz ACC files from the uncompressed songs. Listening to them on most car stereos and on iPods in places that have even modest noise and you can't tell the difference.
If I really cared about the music I would buy the CD but having so many CD's in my collection I might not ever listen to again the iTMS is simple, fast and easy. What I like this month I might not like next month and who wants a large file on an iPod when you don't listen to it.
Unless some online store offers tracks over 192khz then they really don't compare with 128khz AAC tracks. Slashdot readers should check out the results of the online listening test.
http://www.rjamorim.com/test/index.html
Where do you shop? Overpriced boutiques where the thin waif of a sales person tries to sell you some idiotic membership card?
I downloaded a free song from iTMS last week.. "They" by Jem. I liked it so much, I went to Best Buy and bought the entire album for $9.99, the same price as iTMS and with art and a real stamped CD. And I got to rip it at my choice of quality.
Invariably, when talking about purchasing music online, someone likes to bring up FUD about Cd's that cost around $18-22. Reality check: Walmart, Best Buy, Amazon.com and countless other stores routinely sell music for well under $15 and closer to $10 when you get right down to it.
It's been posted over and over again, but people keeps forgetting. Apple's AAC is not the same AAC that you get when you buy a CD and rip it to AAC. Apple's AAC is encoded from master tapes, therefore, eliminating the master tapes -> PCM (CD) -> CCA process. The result is better than what you get at the same bitrate. However, it is debatable at which point the bitrate defeats the source.
Since the last update the Apple AAC encoder pulled way ahead of the others. just google for it.
I'd say 128 AAC WAS like 160 mp3, which is FINE for me, I can't tell on most music; HOWEVER, since the update, I'd say 128 AAC is more like 192 mp3. I can no longer hear the stuff I used to be able to hear at 128. Not that I have perfect hearing, but there was a very noticable change when the updated the encoder...
More importantly, would be to ask if Apple re-encodes their music store music when they get encoder upgrades...
All that said, I now encode at 160 for good measure; because my audio freak friends can't tell between 160 and uncompressed.
Democracy Now! - uncensored, anti-establishment news
People here are saying its like a 192kb/s mp3. AAC is about 1/3 better (see http://www.mp3-tech.org/aac.html) than mp3 (imho) which brings it closer to 160kb/s mp3. you've also got to remember that this isn't a lame vbr comparrison but a fraunhoffer cbr comparrison which has not had the massive open souce development efforts over many many years.
So yes, very adequete on iPod (especailly with those relatively average headphones that come with it), but for playing through any mid to high level hi-fi (depending on music type) it could be better.
Sure, if you want a whole CD.
.99 - 2.97 USD for the 1-3 songs I want, instead of 12 - 17 USD for the privelege of ripping those same 3 songs from a CD (assuming they're on the same CD).
.99 USD price tag.
I have bought about 30 songs since iTunes started. They were all singles.
I usually only like 1-3 songs per CD. So, in reality, I'm spending
Personally, I find the codec and bit rate fine, except for oldies. Some of those songs sound rather tinny. But more modern songs are good enough to warrant the
>Once freeze-dried, there is no way to reconstitute the music into CD quality for playing through a good stereo.
Most people forget the Beatles, Beach Boys, Motown , Phil Spector, etc. recorded music for *transister radios* In most cases, high fidelity is a red herring argument concocted by the music industry to sell more music.
The not-good-enough-for-home-stereo argument is nonsense. I play MP3s and AACs through my home stereo all the time--the sound quality is indistinguishable between CDs. And that is using only the sound card's line-out jack and a cheap connector.
Admittedly I listen to mostly rock. Classical music is more difficult to encode, given its more dynamic nature.
joab
First of all IANAE (Iamnotanengineer) but a former audiophile that used to do quite a bit of signal processing while in grad school.
I see what you are saying about introduction of artifacts going from a 48 kHz digital copy to a 44 khz digital copy and then compressing. What some posters don't seem to get is that processing digital is not the same as working with analog - you get essentially digitization artifacts of digitization artifacts if you are not careful.
However, I have a problem with your test about cutting off frequencies above 16kHz. Seems to me that this would be very hard to do without affecting the other frequencies in the waveform. Again this is because we are dealing with digital copies and thus can only adjust the signal by a single bit at single time point. So we have to worry about introducing more quantization noise. Furthermore the distortion would only be in a subset of higher (not important) and lower frequencies and not across the entire spectrum. This might be the the variance that you hear when you do this especially in quiet passages.
From what I remember, vinyl has a practical limitation of somewhere around 16kHz especially as you get near the end of the record where the grooves are more tightly spaced. However, vinyl still sounds better A/B (to my ears) at least compared to early CD's made from analog masters probably because of the lack of digitization artifacts. (Also, the early sound engineers probably didn't know what they were doing yet and there might not have been much signal above 16K in the masters anyway since the orignal engineers knew it was going to be for vinyl so there was not much benefit for digital to start with but that is another issue..)
Oh to get back on topic, I was listening to a friend's high end system and we both noticed the distortion in the digital MP3 streams that we got from his cable radio stations. Don't know what the bitrate was but I wouldn't want to buy it either...
That sterophile link is gorgeous. 'Using Science didn't agree with our subjective experience, so we stopped using it'
Of course everyone can hear differences between CDs and varying bitrates/encodings. Of course these differences are pronounced on more expensive gear. However, and this is important, this does not mean that the differences are there. Truth is, there usually isn't any difference. Your ears and audio processing circuitry are tricky devils.
Doing double-blind testing can substantially improve your music listening experience. When you find -- using science, not stereophiles's dowsing -- where the limits of your ears and equipment are, you can then enjoy those "lousy" recordings.
Also, why does someone always have to drag LPs into this? LPs suck. They don't sound better, they don't have "more" music on them, they have less. What they have is a bunch of analog artifacts that LP lovers confused with "better." But, those are vinyl-specific noise, NOT "more of the music."
There are some serious technical flaws to his article. Mainly that Rhapsody is using MP3 and iTunes uses AAC.... but this was my letter to the author:
Date: July 04, 2004 12:42:48 AM PDT
To: ddomain@nytimes.com
Dear Sir,
In regards to the article From a High-Tech System, Low-Fi Music, I have to wonder if you have any research or actual test data to back up your assertions. While I understand such references might be cut from a standard article in the interests of brevity, your conclusions seem based on a muddled understanding of the technology which makes doubt your sources.
First of all, your concept of loss-less versus lossy compression is errant. The digitalization of audio signals (analog wave forms) is inherently lossy. 8-bit digital filesystems cannot describe all possible audio signals. This was the primary criticism of Compact Discs release over vinyl that was the subject of debate in years gone by. (The RIAA was found guilty of anti-trust violations that drove the extinction of the more expensive to replicate vinyl discs, btw.) To offer the download of the raw digital audio signal (which is actually sampled at 1411 kbps, btw) from the Compact Disc is therefor technically not 'loss-less' (even if zip or stuffit compression is used on the file for delivery, which is what I infer you to be referencing by loss-less compression that expands to full size on the other end). This seemingly semantic point will become more important in a minute.
Secondly, compressed audio is not inherently lossy. Raw MPEG-1/Layer 3 & MPEG-2/Layer 3 (mp3) compressors at its highest bitrate (320 kbps) is NOT lossy. Audio sampling tests demonstrate that with a quality encoding, 320 kbps MP3 files have pretty much the same wave forms as their 1411 kbps WAVE source files. This is important to understand because the first line compression of MP3 is very similar to that of the zip & sit compression systems that are commonly called loss-less.
The second line compression is what is generally called lossy in the MP3 standard(s). It is important to understand that this compression system works by throwing out ranges of the wave form. It starts at one end of the wave form (that can be digitally encoded into the raw WAVE file) and cuts closer and closer based on the level of compression (bitrate).
For example, 256 kbps MP3 will cut out wave forms that are clearly completely outside of human audible range. I defy anyone (save for Superman himself) to listen to a 256 kbps MP3 file next to the source CD and blindly be able to tell them apart. Short of ESP, it can't be done. The average teenage male (research suggests they have the highest audio range of humans), on his best day (or obsessive audiophiles on high-end stereo equipment) would be hard pressed to distinguish anything above 190 kbps MP3. This point has been argued by some truly dedicated audiophiles, but I have yet to see anyone produce a blind test to refute it. The previous industry standard, 128 kbps MP3, represents what the audio range of the average human adult (which, I'll be the first to admit I find inadequate).
Finally, 128 kbps is not always 128 kbps. Your comparison of RealRhapsody & Apple's iTunes is invalid. While Real owns a RealAudio compressor that is based on the same AAC technology that Apple iTunes uses, Rhapsody doesn't use it. RealNetwork's audio database was inherited from Listen.com, which used MP3 files. AAC (part of the MPEG-4 standard) is largely a 2:1 compression over MP3 for the same wave forms. Your comparison of 128 kbps MP3 files from Real, are essentially compared to the theoretical equivalent of 384 kbps MP3 files... an audio range that exceeds the wave form capacity of CDs themselves.
In the future, it would be helpful to cite your research within your articles. There is far too much supposition and disinformation on these topics.
This IS the case. What the article and few people on /. have failed to mention is that the problem is more a matter of the output quality of the iPod than the bit rate of the compression.
When I got my iPod, I was hoping to replace my CD player by hooking it up to my receiver. This was trivial to do technically, but the sound quality was always poor. I experimented with many, many different sampling schemes (i.e., AAC and mp3 at various bit rates). I finally settled on AAC at 224 kbps, but the output from my iPod was still inferior to what I got from CDs. Then, one day I plugged in my Powerbook to my receiver via the exact same cables that I use to connect my iPod to my receiver. Low and behold, the sounds coming out were PERFECT. (FWIW, I have a Harmon/Kardan receiver with JBL speakers. Good shit.) That's when I realized that the iPod was not designed to be connected to high fidelity equipment. It's output was designed for earphones.
So, I complained and complained to Apple, and sure enough, one of the improvements in the last iPod update was "improved playback." And, I heard the difference as soon as I installed the update. It's still not quite hi-fi, but it makes my trips in my car much more pleasant. At home, I still use either CDs or my Powerbook, but I think complaining some more will get more results from Apple.
I have complained to Apple about the bitrate, also, but for $0.99, one does get a good bargain.
"...who search the reason of things
Are those who bring the most sorrow on themselves." --Euripides, The Medea
Try doing that "test" at 32 kbps and then you'll hear what 128 kbps sounds like to other people. We all have different ears. If you can't tell the difference then good for you, but others can, and very easily.
You can argue that there is no difference until you are blue in the face and the cows are home but that won't make it true. Psycho-acoustic music compression is more an art than a science. Many people still listen to analog cassette tapes (which probably sound better than 128kpbs MP3s). Most people just don't care about audio "quality". To Apple, even using 128 Kbps is probably considered overkill. It wouldn't surprise me if they could sell the idea of 64 or even 32 kbps downloads and most people wouldn't care.
I, for one, will never pay even one penny for music compressed with any form of lossy compression, even a good one like high vbr MPC. When the day comes that music companies stop producing uncompressed CDs, I will stop buying music forever.
OTOH, I would pay as much as the original CD for a FLAC, AL, or APE download assuming I had confidence in the transfer quality and the original source.
Quite an experience to live in fear, isn't it? That's what it is to be a slave.
Actually, iTMS downloads a "m4p" files, some sort of MPEG-4 standard for audio. Still lossy of course, but allegedly better than MP3.
Apple had an interesting demo at the World Wide Developers Conference this past week. They showed HD-quality video compressed with H.263 using the same bitrate as DVD. Very nice picture at a much-improved resolution. Not all 128Kbps are created equal. ;)
"I'll say it again for the logic-impaired." -- Larry Wall.
I'm "digital age". If I want music ("software"), I don't want to have to shove around physical objects, and I don't want to pay for all that shoving that goes on behind the scenes. If I buy the contents of a CD online it's got to cost (cost as in cost to me , but also as in cost to produce) less than what I pay for a CD in a shop. But I'd like to enjoy the same quality as those that bought the physical token. Why not?
One other possible explanation... iTunes has a sound enhancer which is on by default in the preferences - might want to check that out and see how much of a difference it makes. I don't believe the iPod has the sound enhancer feature, and I have no idea how much of a difference it makes, but it could be the last little explanation.
My English teacher once told me that two positives don't make a negative. Two words for her: Yeah, right.