xthread: (Bicycle)
xthread ([personal profile] xthread) wrote2007-04-02 01:48 pm

Music Industry develops Clue, news at 11...

Well, part of the music industry, at least.
EMI has announced that they will begin offering their collection without DRM, through iTunes, recorded at 256kbit AAC, beginning in May.
This is cool in a handful of ways - first and foremost, it means that a major music label is finally getting the clue that selling your customers what they actually want to buy (music they can play anywhere, anytime, can share with their friends, etc) is a better bet than trying to retain the ability to control what your customers do with their IP once they leave the room with it. Or, at any rate, if you plan to fight your customers to the death about it, prepare to die horribly.
Secondly, it's cool because the rips that EMI is going to make available are higher quality than the average customer's MP3 rip of their own CD would be, and in fact higher quality than the DRM'd versions (which will also continue to be available). Which makes this a fairly pure test of the proposition that 'DRM-free music is worth more than encumbered music.'
So, I think this is really cool (personally) because it now makes sense for me to buy good, archival-quality copies of tracks from iTunes when I have a spur of the moment desire to fill out my music collection with singles, because I know that I can buy a copy that's of sufficiently high quality that it's worth owning. This makes me happy.

The only remaining question is how EMI will be generating the 256kbit AAC tracks - will they be re-encoding them from digital and analog masters, or from already compressed cd masters? I hope it's the former, but that's probably rank optimism...

[identity profile] gomijacogeo.livejournal.com 2007-04-02 10:53 pm (UTC)(link)
or from already compressed cd masters

WTF? CDs are 44.1kHz, 16-bit stereo - i.e. 1.4Mbit. There's no compression going on.

[identity profile] xthread.livejournal.com 2007-04-02 11:17 pm (UTC)(link)
<audio geek>
When a cd is produced, the analog dynamic range of the original audio source is reduced to the dynamic range that can be represented by the CD format. This process is called 'compression,' and can involve throwing away a phenomenal amount of audio signal. Furthermore, digital masters are made at 48khz per channel, which is down-sampled to produce cds, at 44.1khz which shouldn't ever throw away any human detectable signal (since our ears run out well below 25khz, even for the most well-endowed listeners), but still makes me nervous about digital artifacting (for essentially no good reason, but still).

The CD format, while it is a reference standard of our day, actually leaves much to be desired.

<incredibly obscure reference>This message has been brought to you by the letter 'U' and the numeral '2'.</incredibly obscure reference>
</audio geek>

[identity profile] tronpublic.livejournal.com 2007-04-03 06:01 am (UTC)(link)
Agreed on down sampling, though I'm guessing that many modern recordings are digitally recorded and mixed in native 44.1KHz.

I would argue that the compression that is on CDs is not a limitation of the format. 96dB (the dynamic range of a CD) is a LOT of dynamic range. It's usually more dynamic range than is available between the noise floor of a typical room and the threshold of pain. It's definitely more dynamic range than typical music.

However, there is compression on most music that we find on CDs. This is usually done on purpose because most consumer audio systems don't sound "good" with that much dynamic range. For that matter, many recordings also have the bass EQ'd out so that it won't over stress systems that have no hope of reproducing it in the first place.

An annoying trend that I have seen in the last decade or so has been digital clipping. Grab any recent pop CD, rip it to a lossless format and check it out with a wave editor. You'll notice that the peaks are clipped. Big nasty digital clipping! With the dynamic range available on CD, this is totally unnecessary. Grumble.

And before you ask..

[identity profile] xthread.livejournal.com 2007-04-02 11:20 pm (UTC)(link)
No, I don't know if the psychoacoustic model that AAC uses can actually represent music sampled directly from a DAT or analog source better than it can represent music sampled from a CD source, but at least if they're going from DAT or an analog master, then it would be possible for the encoder to do something better, whereas if they're going from the CD master it wouldn't be possible, because some signal has already been discarded.

Re: And before you ask..

[identity profile] gomijacogeo.livejournal.com 2007-04-03 01:24 am (UTC)(link)
After thinking about it a bit more, there *might* be.

First, there are well known quantization problems at low signal levels with the 16-bit linear scale used by CDs (and DAT). I.e. a very, very quiet passage might still be discernable on an analog source, but lost in a linear 16-bit sample. Dithering helps, but a non-linear encoder would be better.

As most "born digital" studio work is at least 20-bit and more recent work is 24-bit, this helps tremendously if there is a non-linear encoding algortithm being used. Dunno if AAC does anything there.

Second, there are aliasing artifacts you can get going from 48kHz --> 44.1kHz --> <something> that you wouldn't get going from 48kHz --> <something> directly.

So, I take back my previous WTF.

[identity profile] toxgunn.livejournal.com 2007-04-03 02:36 am (UTC)(link)
I'll be curious to see what the mix of DRM@.99 vs. non-DRM@1.29 sales are.

If they in fact keep the whole album prices reasonable on the non-DRM flavor, I might start buying commercial music again. As is, I am highly resistant to paying for lock-in, and don't get to a record store often enough to expand the collection otherwise.

Good luck sir!

[identity profile] ewhac.livejournal.com 2007-04-03 05:13 am (UTC)(link)
I'll be curious to see what the mix of DRM@.99 vs. non-DRM@1.29 sales are.

How about this: I'm seriously considering signing up at the iTunes Store now.

cos: (Default)

[personal profile] cos 2007-04-03 03:10 am (UTC)(link)
However, one clue they didn't get: They're still charging too much. The Music industry is often complaining that online sales aren't a "success" because mostly people just by a few songs, not album-sized chunks. But they're charging $1/song, which makes album-sized chunks cost *more* to buy online than on CD in many cases (depending on how many tracks there are on the CD). When you buy the CD you get higher quality, no DRM, a pretty booklet with notes, and a durable backup, so of course people are not very motivated to spend as much or more to get an inferior product on line. Now they're increasing the quality and removing DRM, but increasing the price along with it. I think people would still rather pay less for a CD, with the booklet and the durable backup. So they'll continue to buy individual songs online in about the same numbers as they do now, while spending most of their music money on CDs.

This is because the music industry demands too much money per track for online sales (of which they give very little to the artists).

I think that 50 cents per track is about the right price to make online music sales a real success. 50 cents and no DRM, that is.

[identity profile] xthread.livejournal.com 2007-04-03 03:30 am (UTC)(link)
Well... only sortof... if you're buying cd singles, iTunes is cheaper, and if you're buying entire albums online, iTunes is also cheaper. It's only more expensive if you're comparing '12-18 arbitrary songs I mixed and matched on iTunes' vs. the price of 'an album' on CD that iTunes compares unfavorably.

And yes, there are are other properties of a physical album that the digital instance does not match, which is why I think we'll see albums around for the foreseeable future.