Slashdot Mirror
Does Portable Music Have to be Compressed?
FunkeyMonk writes "The Christian Science monitor has an article discussing the gap between music fans and audiophiles when it comes to portable music. Would you pay a few cents more to have lossless downloads from iTunes and other online music retailers? As a classical musician myself, I choose not to download most of my music, but rather rip it myself in lossless format."
...just with no quality loss. Perhaps the question is "Does portable music have to be lossy?"
I've challenged my local audiophile friend to a blind test several times and he refuses to give it a go [especially since he listens to the audio really loudly which will mask most tones anyways].
... use flac. But that's not what this article is about. It's about downloads for listening not remixing. And even then, uncompress the high bitrate mp3/mp4 to WAV, work with that [or store it as FLAC] and STFU.
192+ kbit mp3 with a decent codec (e.g. lame q=2) sounds just like the original for the music in my collection.
Yes [since I know someone will bring it up], if you plan to remix it
Tom
Someday, I'll have a real sig.
It depends on how you intend to listen to your music. If you're going to be listening to earbuds while you're outside or working out at the gym or whatever, then compressed files are fine. Enough ambient noise will be getting through that you'll barely notice any compression artifacts, if at all. However, if you intend to listen to music through a nice set of headphones or speakers in a quiet listening environment, then you'll want it to be as uncompressed as possible. The same generally applies for music with wide dynamic ranges, such as classical/orchestral music.
This guy's the limit!
What's the point? The bottle neck on MP3 players is not the audio files but the decoding/playback hardware and even more important the headphones. You simply can't hear the difference after a certain MP3 bitrate like you can on real audio systems with proper equipment.
Whenever I buy a new MP3 player I spend a few minutes to find the sweet spot where I simply can't hear any difference with a higher bit rate let alone lossless audio. This is almost always 128 kbps, even with quite good head phones.
Well, the poster of this article obviously doesn't consider CD quality to be "lossless." How far we've come from the OLD audiophiles who wouldn't touch anything that wasn't a meticulously cared for LP -- or better yet, reel-to-reel tape in your home rig.
How much longer before we consider 128-kpbs MP3's to be the "standard" for quality music, especially as we're moving to more and more of a "download on demand" compression crazed society?
Won't anyone think of the children!
This sig is a test. If this had been an actual sig, you would be reading something quite a bit wittier than this now.
So yes, some people out there would pay extra for a digital file that is uncompressed or losslessly compressed, but as most people use crap cans or speakers, most of those people would be wasting their money. If you want maximum fidelity, stick with the physical CD or vinyl.
I remember talking about MP3s with an audio engineer friend about a decade ago. As an engineer, he said that he would prefer MP3s to be mastered for the format, which means any limitations of the MP3 and other compressed file formats would be taken into account to minimize/delete any perceivable quality loss. For instance, the cassette version of a recording is mastered differently from the CD version, since tape has different audio qualities (the same also applies for vinyl versions). They don't just stick the CD master onto cassette tapes. On this point, I fully agree with him. However, it seems that all of the AAC/MP3/WMA files that you can buy are sourced from CDs, rather than being mixed especially for the format.
This guy's the limit!
From TFA
The sheer number of variations in compression technology. The array of audio file formats includes Apple's AAC and Dolby's AC3, as well as WMA, OGG, FLAC, AVI, and others.
AAC is not "Apple's". WMA is a container, not a compression codec. OGG is a container (usually used for Vorbis and FLAC), not a compression codec. FLAC is both a container and lossless compression codec. AVI is a container and not a compression codec. The man complains about audio quality, yet 4 out of 5 things that he discusses have "nothing" to do with audio quality.
For his own use, Mr. Goddard, like Willens, favors WAV, a "lossless" compression format that renders sound accurately but has some drawbacks - notably the tremendous amount of storage space it requires: some 50 to 60 megabytes per song, versus about two for an MP3.
Wav is not a lossless format. It is limited by in it's dynamic range (bits per sample) and sample rate. Compared to analog or a raw sound source, raw wav/pcm data loses a lot of the sound. FLAC and other lossless codecs produce identical byte-to-byte output when compared to wav/pcm.
I believe that this guys priorities are a little messed up. We should be focusing on lowering the noise floor, increasing the dynamic range, increasing the sampling rate, and getting the music industry to stop producing albums that are ultra compressed and "loud". You're not going to get decent fidelity out of an iPod when it is limited to 16 bit output and a 44.1/48khz sampling rate with a -90db noise floor. We need 24/96 players with a -110db noise floor, and a decent set of ear buds. Not that it would matter for consumers that listen to the typical tizz and boom being produced today.
BBH
Something else to keep in mind about lossiness and source files: If a recording is made and mastered in the studio at 96kHz/24-bit, the step to your 44.1kHz/16-bit CD is considered "lossy" since information is being discarded along the way. However, again, this is taken into account when mastering for the CD format. The DVD-A/SACD masters will be done differently. So in a sense, many CDs that people consider to be "perfect" source files have already been through a round or two of degradation. Is it something that they'll ever notice? Not likely, especially if they aren't aware of it.
This guy's the limit!
Sorry, I didn't format my previous reply:
There was a time when I couldn't hear a bit difference between a redbook CD track and the same song ripped as an MP3 at 192k. Then I went to school to get my BS in audio production. It is amazing how much more detail you can hear in music when you are trained to do so for four years. I would have never believed for a second that my advisor could hear things in music that I couldn't, until two years later when I 'saw the light'. Over time I began to pick out subtleties in music, even if I was hearing the piece for the first time.
All of the high end audio products generally have no benifit for the average consumer, but in a studio setting, when trained ears are listening, that expensive gear tends to be more valued. There is an inherent problem with this situation, though. Is it reason enough to justify buying equipment that is significantly more expensive because my collegues and I find it more pleasing to listen to, while the average consumer of the product can tell no difference? I don't have an answer to this, but I know that there is actually a growing market for DVD audio (with 5.1 mixes as well). On a DVD disk we can store music at such high qualities that it rivals the best master analog tape s out there.
The bottome line is that your ears are trainable. Listening to music is a learned process, much like wine tasting. At first pass, you may think is all tastes like sour grapes, but over time, with effort, you will discover flavors you never knew existed. For the record, I have been a part of quite a few 'blind' tests juxtaposing certain audio formats, and I can certainly tell the difference between an mp3 at 192/16 and a redbook track. Step than mp3 up to super high quality vbr, and I have some difficulties, unless the music is of the classical genre.
Doug
Does Portable Music Have to be Compressed?
Yes
Folks here are concentrating too much on listening rather than archiving. In my 33 year lifespan, I've seen vinyl, 8-tracks, cassettes, cd's, dvd's, DVD-A, SACD, and downloads come and go as formats. The reason I want lossless audio is not so that I can listen to it, but so that I can archive it. Sure, CD quality audio can't truly be considered lossless, but for most material, it is the best quality we've got that can be duplicated bit for bit. Sure, I've got a few hi-def audio disks in both DVD-A and SACD, but for the most part, my library is 16bit/44.1Khz. I don't expect that I'm likely to see sudden release of a higher quality digital format any time soon, so CD quality is the best I'm likely to see for most of the material in my library in my lifetime.
Given that I am 100% positive I will see at least 10 format changes in the years to come, I want to archive my cd quality digital files so that I can use them as a source for subsequent transcoding. Go through a couple of serial lossy encodings and it will sound way worse than analog to analog copies. Which is exactly why I don't want to download a 128kbit lossy compressed file. I would surely pay a little extra for losslessly compressed audio, but the music industry will never allow it because they are counting on selling the same music to me several more times during my life. I intend to do everything I can to stop them. That means archiving the 2,000+ cd's in my collection in a lossless format and only buying new material that can deliver 16bit/44.1khz or better.
And incidentally, I do listen with high quality headphones (Sennheiser HD650) when listening to headpones, through a high quality amp and DAC (headroom). I've got the FLAC files streaming around the house to various squeezeboxes, one of which is plugged into a very nice home stereo - probably not audiophile quality to a true audiophile, but very, very nice, and lovingly constructed piece by piece over many years. I can most definitely hear the difference between FLAC and the lossy files I transcode to (192 VBR, mostly), although I could surely transcode to a larger size and get files whiich I couldn't differentiate. However, my complete library is already 100GB when transcoded to 192kbit, so it already won't fit on an ipod, and that doesn't include video or audiobooks. The two copies of the library occupy 600+GB on my NAS. I do listen to the entire library, at least in the context that I frequently have things on random and Iike to have the largest possible library to get random selections from, so I tend to keep my 120GB laptop USB drive in my bag with the lossy library on it.
So, it isn't about listening to 'audiophile' quality recordings. It is about archiving the best possible quality in order to futureproof my music collection, which also allows me to use lower quality files on my portable.
"Using up to date encoders, for the vast majority of people, for the vast majority of tracks, 128 kbps is indistinguishable from source."
Particularly when listening on cheap speakers that are connected to a PC.
I mean, I wish I could listen to 64kb/s encoded music and say "sounds just like source" because it would be cheaper all around and I would be happy.
A perfect example (to me) is Sirius satellite. I like their programming. But their bit rates are so low that it sounds like shortwave radio. I have their service in the car, and if not for the talk stations I'd drop them. You certainly can't listen to music that poorly rendered and enjoy it. On the other hand, I hear people telling me "it's CD quality", so I suspect there are some people who really can't hear the difference. God bless them, they're much happier overall than I am.
You were mistaken. Which is odd, since memory shouldn't be a problem for you
Finally someone who makes some sense in this discussion. I agree with most points you made (I will disagree with one a bit further down). "One cannot hear a difference" is one of the most annoying /. memes to me, regardless of whether it is applied to lossy codecs or good audio equipment in general.
/. because it has been bugging me a long time: /. anymore? Sound recording and reproduction (that is, turning a complex air vibration into an electric current, storing it in some form, and later turning it back into an air vibration again that sounds as close to the original as possible despite this happening in a completely different room situation) is an extremely complex topic. And, like it or not, there is still a significant analogue part to this, and will be for the foreseeable future. This means that you have to live with the difficulty of interacting with the real world in a less deterministic way. Only recently has it become possible to simulate microphones, amps, and speakers digitally, and sound reproduction has benefited tremendously, especially by making good gear much cheaper. But until then the only way to become better was to design analogue gear and try it out, relying on basic measuring equipment and your ears to assess the sound quality. This IMHO is hardcore geekdom worthy of honorable mention on /. and not ridicule. There were and are serious practitioners out there like Nelson Pass or the naim guys who have dedicated d
About lossyness:
I agree with you that ears can be trained, and that you won't miss stuff if you don't know it should be there in the first place, or don't care whether it is. When I decided how I want to encode my music I did a quite extensive test and I found that to me even high-bitrate mp3 encodings made by lame can sound noticeably different from the CD. For example, I encoded the first track of Mike Watt's Contemplating the Engine Room CD. It starts with an e-bass solo, and using reasonable lame presets there were no artifacts and I certainly could hear the notes played. Somebody expecting nothing more will probably be happy with the compressed sound. However when you know how a bass can sound and listen to the CD, you realize that there is so much more in Watt's bass sound: it is full of harmonics that make the bass come alive and turn it into the recognizable Watt bass in the first place. And these harmonics are gone even in the highest lame preset. (And oggenc adds a nasty hiss which makes the song completely unlistenable.)
About equipment:
You said "All of the high end audio products generally have no benifit for the average consumer, but in a studio setting, when trained ears are listening, that expensive gear tends to be more valued", and that's where I disagree a bit because you make it sound as if only a professional sound person could appreciate good gear. I's agree that someone who is not particularly interested in music has no need for good gear. That's pretty obvious. If you're going to listen to music only as background noise while cooking, go with the cheap stuff by all means.
However I would argue that everyone who likes music and spends time actually listening to it will profit from good gear. To everyone who doubts that I can just recommend to grab a few favorite CDs and make an appointment at a good hifi shop for a listening session. "Good" means "a shop that has solid equipment from the lower to very high price ranges, but that will not rip you off by trying to sell you air conditioners."
Not directed at you, but I need to say this once on
To those discussion contributors who lose all ability to differentiate when they hear the word "audiophile": one cannot deny that wackos exist in this field. On the other hand, since when is being an analog geek not allowed on
"When I first heard Daydream Nation it quite frankly scared the living shit out of me." -- Matthew Stearns
What the consumer wants? More likely what the consumer is offered.
I don't recall being asked by anyone. ;)