Slashdot Mirror
Return of the Vinyl Album
bulled writes "NPR ran a story this morning about the comeback of vinyl. It seems that sales of new vinyl records are up about 10%; sales will approach a million this year (as against half a billion for CDs). NPR mentioned the popularity of a turntable with a USB interface — they didn't specify the brand; could be this one, or this — and speculated on other possible reasons for the resurgence. They mentioned sound quality and lack of DRM as possible causes. Sound quality can and will be debated, but DRM rates a resounding 'Duh.'"
I dispute that they "keep better". If you have an analog tape master, it will have a finite life no matter how much you pamper it. Thus the existence of techniques such as tape baking. The only way to preserve this tape over the long-haul is to transfer it to a fresh tape or other medium sometime before it is completely degraded. Every time you transfer the analog tape, you degrade it by a generation... doing the same with a digital master would give you an exact, or near-exact copy. You could do this as many times as you wish with no generational loss.
At the very least, you should immediately digitize the analog master so that you have the digital first-generation copy "forever".
W..w..W - Willy Waterloo washes Warren Wiggins who is washing Waldo Woo.
Amusingly enough they did it for Vinyl as well: http://www.elpj.com/main.html. Sure as hell aint cheap though.
"I have a porkchop, you have a porkchop. I have a veal, you have a veal".
One thing is true: vinyl will outlast CD in durability, and the error correction is much more robust on Analog.
I don't think so. the abuse a standard CD is subjected to would utterly destroy a record. how many people put a dozen naked Records on top of each other in a care that goe over bumps. 2 bumps and you have yourself a pile of useless plastic. I do the same to CD's and they last about a year with this abuse. Records last a long time now, because those who buy them treat them properly. CD's have finite lifespans because they are small, and versatile and thus often abused.
CD's and Records fail in different ways. A light scratch across the record will render every track with a regular periodic snap/pop or even render it unplayable. A light scratch on a CD may result in a bit of a skip or no data loss. A deep scratch on both results in an unplayable disc. Multiple light scratches on a CD will still often be playable and often without quality loss while the same for a Record renders it unplayable. Repeated play degrades a record, while it doesn't really degrade a CD. And Vinyl is not as mobile.
Also, You can back up a CD. You cannot back up a Record into the same format. Error correction on analog data is not more "robust" it's different. Critical failure on an anologue system is different then on digital. If I introduce random noise to a CD, I can digitally filter it out if I know how. The same type of error on a anologue signal might result in static. cleaning up such a systemic loss is hard on analogue. When the damage is mroe severe the digital may be unrecoverable while the analogue may still cary some of th edata. Digital has a recoverable area/damaged area rate that looks like a inverse log. 0-50% damage = 100% recover 50%+ = 0%. While Anaglogue has a linear decline.
"There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy."
105dB of stereo separation on a 16bit (~96dB SNR) medium?
Amazing!
Before you even concern yourself about the quality of any modern consumer electronics product with a well-known name, find out if that name is still worthy of its history. Many of the former greats in audio reproduction have sold out to Chinese manufacturers, sold their names, their brands, and the respect they earned in the marketplace. Now they're nothing more than marketing fronts, shadows of their former selves.
The higher the technology, the sharper that two-edged sword.
"If records really want to make a comeback, they'll come up with a nondestructive way to read the disc, like a laser beam. Oops, they did that. It's called a CD."
No, I think it's called a laser disk, since laser disks are a nondestructive analog medium.
like this? http://www.elpj.com/main.html
11 was a racehorse
12 was 12
1111 Race
12112
yeah, because everyone can afford 10k for a turntable
Thanks to file sharing, I purchase more CDs
Thanks to the RIAA, I buy them used...
I own well over 1,000 pieces of vinyl, and many of them sound better than the CD. This isn't because vinyl sounds better, but because either the master was damaged or poorly remastered for CD. It is amazing how poorly mastered some CD's are. Digital recording does not compensate for an idiot behind the sound board, in fact it makes it much worse.
"To those who are overly cautious, everything is impossible. "
There's actually 3 products on the market that does this. Final Scratch, Serato/RANE Scratch Live and MS Pinky's Interdimensional Wrecked System.
Depends on what CD you're talking about. The Loudness War has adversely affected CD quality for a decade now. The LP version of the latest Chili Peppers album doesn't have anywhere near the amount of clipping that the CD version has. It has a higher dynamic range.. this isn't "feels warmer" this is "measurably different wave forms".
I take it you haven't actually looked much then?
There's a bigger problem than mastering, especially where rock is concerned.
I started wondering about this a few months ago when I was transferring some of my ancient cassettes to digital with my ancient, but quality, tape deck. I was astonished to discover that two of my cassettes -- Sammy Hagar's Unboxed and AC/DC's Who Made Who -- sounded better than every CD in my collection. And in the latter case, we're talking about a 21-year old cassette that has had tons of play and sat in the car through ski trips and Summers in Texas. And here's the layman's version of what I figured out:
The more high frequencies you have interacting with each other, the more little squiggles you get in your waveform that are way beyond the range of human hearing. The squiggles come from the fact that the high-frequency sounds are out of phase with each other, and constantly shifting in phase relative to each other. Now although those little squiggles in the waveform are beyond what you or I can hear, our ears aren't looking for those squiggles -- they're looking for the harmonics that comprised those squiggles. But PCM encoding doesn't do that; it's trying to capture the actual waveform. Since those squiggles are well beyond the sample rate of a CD, they're fucking messed up. Sharp points get sheared off, the tiny squiggles disappear, and the sound wave that ends up being recorded gets aliased across a range of different frequencies.
When played back, instead of responding to a set of pure harmonic frequencies, our ears pick up a random assortment of frequencies near the originals. The crash cymbal goes "TSHHHH" instead of "TSSSS."
Now the standard objection to this argument so far goes something like this: Yeah, well man, vinyl and cassettes don't respond well much above, like, 18kHz!
There's two flaws with that objection.
First, an analog deck's frequency response may not go beyond 18kHz, but what happens is that those higher frequencies just fade beyond the noise threshold. They do not get mucked about with. And if you mix together multiple frequencies at 18kHz and below, record them, and play them back, you can deconstruct the result into the original waveforms pretty cleanly, because an analog recording device doesn't have quants that can go out of phase.
The other problem is that the pitches in question don't even have to be close to the Nyquist frequency (half the sample rate -- 22kHz on a CD, which is slightly above most people's range of hearing) if there's more than a few of them. And with rock 'n' roll is that you have tons of high-frequency shit going on. You've got the lead singer's shriek, the crash cymbals, the gate on the snare (and sometimes bass) drum, and the guitar on the fuzzy channel. All of those frequencies mix together, out of phase with each other.
The ultimate result of all of this is that high-frequency joy, those sharp high-frequency peaks in the guitar, the splash of the cymbals, the things that all combine together that made rock become the dominate pop music form for 30 years, disappeared once CDs became the main standard for audio. And rock, outside of the live show, hasn't sounded right ever since.
Yeah, vinyl and the mighty 1200's aren't going away but the digital stuff has got considerably better in recent years and hence more accepted. Pioneer CD decks have been standard equipment in UK dance clubs for some time. The future is a fusion of analogue and digital stuff, which in skilled hands can sound great. At the extreme pure digital is almost a different discipline though, you see artists specifically advertised as doing an Ableton Live set as opposed to "DJ set". The key is to use it properly i.e. as a live performance experimenting, mixing different media and reacting to the crowd, rather than laying down the track in the studio beforehand and turning up to press play.
(Not a DJ but a few of my friends are)
First off, CD quality in 30-40k. Really? Then why the hell haven't we seen it anywhere? You don't think Apple would kill to have a more efficient format for storing and distributing music?
Well I can't find shit on anything called ABBAcast. To me it sounds like you confused ABBA, an old Swedish pop band, with Abacast, a radio streaming service. Of course, Abacast doesn't have their own codec, they use WMA, AAC, and MP3. Well 30-40k WMA is NOT CD quality, not even close, never mind MP3. The good codecs, like AAC and Vorbis do a decent job at 128k, it sounds as good as CD for most music on lower end gear (like portable players) but still not even close. You are talking in the range of 256k with most codecs to get something that is reliably as good as CD on good gear.
Also you seem to have no understanding of the problems of encoding to a broadband format.
For one, you can't stack your channels right on top of each other. You'll find that you have all kinds of problems. Look at any real broadband system, and you'll find out they've built in space between channels for just that reason. For example TV channels are specified at 6MHz each. Within that 6MHz is a video, colour, and audio carrier. Channel 2 is from 54-60MHz. However, the video carrier is at 55.25MHz, the colour at 58.83MHz and the audio at 59.75MHz. Channel 3 is then from 60-66MHz. You might notice that means there's nothing in the lower edge of the range and you'd be right. The reason is that you don't want the channels bleeding in to each other. You need to leave space if the system is to work.
Next, you've got the problem of assuming that two stereo channels can be used separately. Errr, no. In any analogue system, two adjacent channels will have some amount of crosstalk. That is to say a signal on one channel will bleed over to the other to some degree. You'll notice that most stereo amplifiers specify this amount. Well, with records, it's pretty high due to the way that the stereo signal is recorded. It's horizontal needle deflection, not two discrete tracks. Not a big deal for stereo audio, it's highly correlated anyhow and we don't need a ton of separation to hear stereo, however it'll fuck with your encoding real bad.
Then of course we get to things like error correction, assumptions that the SNR is equal with regards to frequency (it's not) and so on. I'm not going to go in to all the problems in detail since it ought to be apparent at this point that you didn't think this through.
paul is using a cdj-1000 in the photo on that link
there must be some way outta here, said the joker to the thief
Interesting argument, but fundamentally flawed.
The simple fact is that neither analog or digital are perfectly accurate, but digital is more accurate even for loud rock music. Many many CDs today are mastered way too high, resulting in severe clipping of the waveform peaks - but that is the fault of ABUSE of the technology, not the technology itself.
Here's a small encapsulation:
Digital recording: you record from your source into an analog-to-digital converter, whereupon the bits are recorded as data. Individual tracks are mixed together, post-processing is done, but the sound quality remains "first generation" throughout. Copies do not result in degradation, so the final CD master contains the sound in as close a form to what the microphone heard as possible, barring any post-processing altering the waveform either by design (echo, reverb, chorusing) or accident (mastering at too high a level, resulting in clipping). If the multitrack original is not processed at all after the recording, then the mixed-down master consists of first-generation tracks mixed together.
Analog recording: you record from your source into a multitrack tape machine, making sure that the azimuth is correctly aligned and that the speed of the machine is constant. In order to get some sort of sound out of those microscopic 24 tracks squeezed onto a two-inch tape, some compression and equalization and noise-reduction has to be done. Oops, we've just compromised the sound signal, haven't we? Not extremely accurate.
Now those tracks have to be mixed down to a 2 track 1/4" master, because vinyl is not multitrack. But in order to do that, we have to make an analog copy from an analog master. Normally this would be a tremendous sonic problem, because when you copy from tape to tape you not only copy the signal, you also copy the noise onto a blank tape with noise of its own, thus effectively squaring the noise. Mix two tracks, get four times the noise, Mix four tracks, get sixteen times the noise. Solution? More noise reduction. That compromises the compromised signal even more. And any analog postproduction, such as adding reverb, requires still another generational loss coupled with artifacts from the analog processing.
Once we get the tape mixed down to two track, which is second generation, it has to be mastered. Cutting that analog tape onto a lacquer blank introduces even more compromise to the audio signal, because unlike a CD, you cannot record the sound wave "as-is" onto vinyl. The highs have to be boosted because otherwise their minute squiggles would be smaller than the width of the cutting stylus and they would be irretrievably lost, while the lows have to be attenuated because otherwise the cutting stylus would cut the groove straight into the adjoining grooves and the record would not play at all. So now we have a minimum of three analog (lossy) generations away from the studio master, and this final generation now introduces an equalization curve just to that the resulting vinyl can be played on its own reproduction equipment.
That lacquer disc now goes for a plating bath, where a negative matrix is pulled, being an exact copy of the lacquer but with its grooves raised instead of sunken. This matrix is then plated to produce a positive mother, which looks exactly like the original lacquer except that the grooves are microscopically larger because it's been plated twice. We are now five generations removed from the studio master - and we still are not finished. The mother is plated yet again to make the stampers, and the stampers are used to press the final record into vinyl - a thermoplastic not particularly known for its dimensional stability. Set the vinyl out in the sun for a few hours and see how accurate the sound is.
So now you have a vinyl record that is seven generations removed from the studio session master. Now it has to go on your turntable, and be subjected the tracking force of the tonearm assembly, as well as whatever sonic compromises come about because
One reason why an old tape may sound better than a new CD is that modern recordings get ridiculously heavily compressed at the mastering stage because the bands/producers/record company etc. want their album to be louder than everyone else's. Result: no dynamic range, no music, just loud loud loud.
Things were a little more civilised musically back in the days when tape cassette was a popular release format.