Slashdot Mirror
Vinyl To Signal the End for CDs?
PJ1216 writes to mention that vinyl seems poised to make a comeback in the music industry. Some are even predicting that this comeback coupled with the surge in digital music sales could possibly close the door on CDs. "Portability is no longer any reason to stick with CDs, and neither is audio quality. Although vinyl purists are ripe for parody, they're right about one thing: Records can sound better than CDs. Although CDs have a wider dynamic range, mastering houses are often encouraged to compress the audio on CDs to make it as loud as possible: It's the so-called loudness war. Since the audio on vinyl can't be compressed to such extremes, records generally offer a more nuanced sound. Another reason for vinyl's sonic superiority is that no matter how high a sampling rate is, it can never contain all of the data present in an analog groove, Nyquist's theorem to the contrary."
This statement is true, but completely irrelevant. The fact that a recording medium is analog does not mean that it is better at accurately recording and reproducing a sound than a digital medium. Magnetic tapes are also analog recordings. Putting a pencil on a string, hanging it next to a speaker, and having it draw a line on a moving sheet of paper is also an analog recording.
It's true that a digital recording can never contain the amount of data in a vinyl groove, but who is saying that all the data in a vinyl groove is more of an accurate representation of all the data extant in the original sound wave than a digitally sampled recording?
This is similarly irrelevant. Compression is a way of altering a sound wave, and has nothing to do with the final recording medium. Overcompression is a problem, but this is not an argument for vinyl over CD--it's just a comment on postprocessing techniques.
The Nyquist-Shannon sampling theorem states "Exact reconstruction of a continuous-time baseband signal from its samples is possible if the signal is bandlimited and the sampling frequency is greater than twice the signal bandwidth." More information can be found here. Wikipedia is your friend.
Only two things are infinite, the universe and human stupidity, and I'm not entirely sure about the universe - Einstein
A friend of mine and I had this battle about 10 years ago. He had a very high-end turntable from Linn and I had a CD player from Nakamichi. His argument was that vinyl retained a certain "warmth" and "depth" of sound that was lost in digital recordings. We played jazz, classical and soft rock tracks from various artists and the CD simply blew the turntable out of the water. The vinyl recording, even on his ultra high-end turntable and component stereo system, still audibly popped and crackled. The CD sounded absolutely clear and had an impressive depth of sound. The argument died for me that day. Technology is king.
>no matter how high a sampling rate is, it can never contain all of the data present in an analog groove, Nyquist's theorem to the contrary."
Sure, I could sample at 1 bazillion hertz, but if I'm only sampling at 1 bit I'm not going to be reproducing the original signal very well, since my sample size isn't high enough to differentiate the data I care about. And if I can't tell what data looks like, Nyquist can't tell me anything about how much sampling I need to do in order to capture it accurately.
Nyquist doesn't directly say anything about the sample size (8 bits, 16 bits, etc, just the sample rate (22 KHz, etc).
There's a show on Adult Swim called "Metalocalypse" - watch it. Then.. to quote their lyrics: Go Forth And Die
A record is a physical device as is the needle and player; it cannot reproduces all frequencies, only a finite range. You simply cannot create fine enough details in the vinyl to capture very high frequencies when the track is moving relatively slowly (70 RPM or whatever it is). If you know the frequency limits of the physical system, you could then sample at a high enough rate to exactly capture and reproduce the signal digitally.
For one thing, vinyl has always had a loudness standard: the bigger you make the grooves, the fewer can fit on the record. So LPs were most often mastered at levels appropriate for a 24 minute side. (Extended singles for club play, which have fewer songs on them, are often mastered louder.) Compact Disc Digital Audio, on the other hand, never had a concrete definition of the playback volume.
CD is more portable than vinyl and is often listened to in a moving environment. The loudness race started when portable audio players such as Sony Discman and car units first came out. Some used a cheap op-amp to drive cheap headphones; others were car units that played over the radio. Record producers realized that end users could barely hear Dire Straits' Brothers in Arms over environmental noise, and they pushed mastering engineers to push the levels hotter.
Also, vinyl equalizes the bass down before recording and equalizes it back up in the player's preamp, based on a standardized New Orthophonic preemphasis curve. The limiter algorithms to overamplify an audio signal while fitting it into [-1..1] in the flat-equalized time domain of CD are not optimal for a time domain equalized in New Orthophonic. It's the producer's job to approve a master, and hearing these suboptimal results on vinyl might encourage an ambitious producer to back off on the demands to the mastering engineer.
You'd do better to digitize all of your vinyl now. Just because CD reissues are generally incompetantly done doesn't mean that you can't make an effectively perfect digital recording of the signal your player produces when playing the vinyl before it gets damaged by wear and environment. It doesn't matter for the signal that goes to your speakers whether it is driven by record player or a DAC.
One thing about remastering is that the original recording may have been done with a vinyl-based idea of the threshold for perceptability. So they didn't bother with some aspects of the environment or accoustics of the studio which would make no difference in vinyl, but which come out clearly and distractingly on CD. Having it on vinyl in between effectively airbrushes out this junk. But that doesn't mean that you can't push the signal through vinyl first and then make a high-quality digital recording of the post-vinyl signal, and have your digital music player reproduce the sound of a vinyl record.
As an audiophile that read High Fidelity magazine during the time frame of the digital "revolution", I will try to add something useful to all this. Sampling highly complex audio waveforms, to convert to a digital medium will involve some loss. Compression, digitally, or analog, involves dynamic range loss. This may actually be required in some music, to prevent clipping, or a total loss of the quieter passages. So both sampling and compression become necessary evils in a "digital" age. The sampling level, and level/TYPE of compression are the determinants for overall quality. Unless you listen to synth/pop music, in which case differences are usually minimal, as the source material lacks the subtle nuances, and wider dynamic range inherent in non-synthesized music. Please bear in mind that the following, as well as the preceding are opinions and information subjective in nature, but largely true from an engineering and audiophile perspective. In the 80's I remember reading up on the Compact Disc format. While 44,100 samples per second, with 65,535 levels of volume may sound like a lot, even as a teen ager, I was unimpressed. Audiophiles were pushing for 96 Kilobit sampling rates, to ensure a more accurate representation of the analog wave form. Many wanted greater bit depth as well, since inaccuracies in digital reproduction would be buried under the noise floor. What we got, as consumers, was the aforementioned 44.1KHz, 16 bit sampling. The best that we could do was use strong "smoothing" in the analog output stages, and try to hide the "edginess" many complained of in the sound. Producers would do a digital recording, analog, mixdown, and drop it to the digital medium, at the end. This method, while seemingly over complicating things, would typically soften the sound somewhat on less expensive gear. High end gear would use extremely expensive analog output stages, or qould pass the signal to a dedicated D/A converter. Overall, I have never been happy with the CD medium for critical listening, especially recordings involving the female voice, or very complex high frequency content with wide dynamic range. Think of "unplugged" sessions, or orchestral recordings, when picturing examples. fiddy Cent is not affected by the CD format's limitations..... So where does that leave guys like me? In an age when kids are being taught that 128 kilobit data rates for MP3 is "CD quality, and simple, convenient formats, and lossy compression are fully acceptable, I am apparently a minority. MP3 is based on removing "data" that is expected to not be perceived. This removal is applied in varying degrees, depending upon desired data rate, or a variable rate, with an upper limit. I find the trend towards acceptable loss, in an already compromising medium, to be anything else but acceptable. It only takes one listening session of any preferred music on CD format, with a comparison to MP3 encoding at 320 kilobits data rate, to see a huge difference. At 128 kilobit data rates, it gets plain embarrassing. All of this assumes at least mid range equipment. If you are comparing the two digital formats on a $200 rack system from Wal- Mart, or on your PC speaker setup, it may not show the huge disparity in quality. All this being said, I hate the current CD format, and long for "albums to come out on DVD, with 96KHz or higher sampling, and 24/32 bit depth. While I have a custom D/A converter (PS Audio), and use B&W speakers, my total investment is WELL under $5000 for all of my gear. Perhaps when I cannot detect edginess or overall "grittiness" in the widely varying music I listen to, I will be happier with digital formats at the consumer level. I will never like the hack job that MP3 does to music. My opinion. YMMV. Some restrictions may apply. See your doctor if this post causes an erection lasting longer than four hours.
Analog on vinyl is not lossless. From Wikipedia:
. .
[snip]
Think of it as analog dynamic range compression.
The Christian Right is Neither (Christian nor right). See: Matthew 23, Matthew 25, Ezekiel 16:48-50
Aliasing noise is not related to sample rate or bit depth. A properly dithered 8bit 7KHz recording will have a high noise floor, and be severely band limited, but will not have aliasing artifacts. The noise floor is 'white' noise, and not related to the signal. The glitchy sound you associate with low bandwidth recordings is due to not dithering properly, data compression, or as using it as an effect etc. Have a play with some audio editing software some day. It's interesting how good the audio sounds at 12bit 35Khz or so. Modern records often have so little dynamic range that you could use 14bit and no one would notice. :)
Anyway, 192Khz converters are very cheap nowadays. I think about £2.50 or so.
It's the analog side and the clocking that makes the difference.
16bit 44.1Khz is fine for listening if the conversion to analog is adequate.
There's a lot of subtleties involved in going from Nyquist's theorem to actual practice. Some are related to problems of numerical analysis and others relate to how close you want the upper frequency cut-off to approach the Nyquist limit. The numerical analysis aspect is that the digital (discrete) representation is never exact, having said that, it is close enough most of the time (e.g. bass in mid-range). Getting usable frequency response to be close to the Nyquist limit requires use of 'brick-wall' filters which do bad things to time domain response - probably the worst case being an instrument like the triangle.Some of this is covered on the design and implementation of direct digital synthesizers.
Compression is the removal of dynamic range, and is actually REQUIRED for vinyl to get the low volume sounds out of the vinyl surface noise to make them audible.
BS. What's required is pre-emphasis (e.g. the 'RIAA curve' created ca 1950, back when the RIAA was doing something useful). To get a decent amount of recording time on vinyl, you don't want a consistently high recording level (requires larger spacing between grooves and may burn out the cutting head) - which argues against using compression.While a properly made CD will typically sound better than a vinyl recording, the article was correct in stating that CD's lend themselves more to overcompressing than vinyl and that has to do with the process of cutting the record (see points about groove spacing and burning out the cutting head).
It's true that a digital recording can never contain the amount of data in a vinyl groove, but who is saying that all the data in a vinyl groove is more of an accurate representation of all the data extant in the original sound wave than a digitally sampled recording?
The kicker for me showing a total lack of understanding of the technology is the popularity of USB turntables. They can't keep them in stock. Quick, someone show me any analog signal in a USB specification.. Analog is better.. Analog is king, Here use this USB turntable to enjoy your analog sound. What are they smoking? Nothing out the USB port of a turntable is analog in any shape or form. Who has a better low noise analog to digital converter, a consumer grade turntable or a CD mastering house?
Analog is king only because the mastering house slaughtered the conversion in the loudness war. If you check the links, the youtube link provides the best summary with an example of the problem which can be heard and seen.
http://createdigitalmusic.com/2007/05/16/loudness-war-music-over-compression-demonstrated-on-youtube/
http://my.opera.com/swerfot/blog/2007/08/26/loudness-war
http://en.wikipedia.org/wiki/Loudness_war
http://www.hydrogenaudio.org/forums/index.php?showtopic=55892
CDs are on the way out because the music on them is crap. Finding a decent recording in the pile of crap is why many simply avoid the contaminated format. USB turntables, even though you don't get analog, you also don't get the over compression, which is why the ability to play better source material is so popular. Analog has nothing to do with this argument. Destruction of the sound on compact discs in mastering is the problem.
The truth shall set you free!
The only difference with analogue is that the limit is gradual, with digital the limit is harsh. With a record, you are going to get a SNR of somewhere around perhaps 70dB if everything is right (quality equipment, new recording, no dust, etc). That means that noise will start 70dB below the loudest signal. However you'll find that the dynamic range is more than that. What happens is that you can hear things below the noise floor. Just because the noise is there, doesn't mean that you can't hear anything below it.
With a CD it's different, the SNR is 96dB and that's also the dynamic range. That is just the limit to how accurate it stores the samples. Also, low level samples get progressively more distortion since there's less bits for them. A 1-bit sine wave would, in fact, be a square wave.
So this is why vinyl is better right? Well wrong, because that problem with digital, really isn't.
You can solve it two ways. One is simply to increase the bit size. Use 24-bit and now you've got 144dB of dynamic range. Given that even the best converters are hard pressed to do over 120dB, as are human ears, you needn't worry. However you don't even need to do that. You do the analogue thing in digital. You just dither the signal. You take a high resolution signal, and dither it down to 16-bit. This lowers the SNR, but raises the dynamic range. So with 6dB of dither you'd have an SNR of 90dB, still very good, but you could expand the dynamic range to perhaps 114dB and eliminate the quantization noise entirely.
What it really comes down to is we can sample more accurately with digital with analogue, and we can easily store it past our ability to sample.
Tivo man!
TIVO
(I haven't seen it either)
0xB315AA8D852DCD3F3DCA578FD2E0BF88
"Has anyone actually done a real study on this...
...and actually determined that vinyl is better than CDs?"
Yep.. Many of them. TONS of them, in fact.
Ohhh! NOW you add conditions! In that case, the answer is no.
"This also reminds me of the age-old tubes versus solid state argument, and I don't think that one has ever been looked at objectively either."
Sure it has. Early transistor amps had lower THD but much higher IM distortion, which led to worse sounds. They were also prone to oscillation, which hurt the sound. Then there was the hard clipping at limit vs. very soft clipping of tubes, and you have lots of reasons that tube amps were better than transistor amps--in 1965.
Good transistor amps became completely transparent in the mid 1970s, by my estimation. That put them into the extreme stratosphere with the very very finest tube amps. Nowadays, a few hundred bucks and some good engineering will get you a transistor amp that is sonically neutral within its parameters. A few thousand will get you a tube amp that accomplishes the same thing. If you push either amp beyond its limits, or cut down either amp to be audibly flawed, you will get very different but very clearly understood and measured (and predictable) distortion models. Easy, straightforward, and proven for about a quarter century. Just don't tell the audiophools.
"People who do stupid things with hazardous materials often die." -- Jim Davidson on alt.folklore.urban
You believe wrong. They call them laser turntables.
http://www.elpj.com/main.html
We have enough youth, how about a fountain of SMART?
Well, the lossless format would be uncompressed WAV files... or AIFF if you own a Mac.
You could record your vinyl at 24 bit 96 KHz or better for storage and playback, then make down sampled copies at 16 bit 44.1 KHz to burn to CD, or save as high-bitrate mp3s for your portable devices.
24/96 really IS better than reel-to-reel tape ever was, and besides... reel-to-reel tape has been discontinued. The last manufacturer of high-end open reel tape (the "400" series) was Quantegy. (formerly known as Ampex) Read all about it. http://www.quantegy.com/
But if you insist, I have an Otari MX-5050 2-track and a stack of Ampex 467 & BASF "Studio Series" 7" & 10" reels I'll sell you. I haven't used them since I got my M-Audio Delta-66 card. (5 years?)
Wired seems to take all the standard audiophile BS hook, line, and sinker... "analog provides a warmer sound" (much more total harmonic distortion than a digital player), etc.
The argument about hot mastered CDs is particularly hilarious (reduced dynamic range). Basically, this is a result of crappy commercial pressure to sound louder, and is common but by no means universal. The fact that vinyl lacks this possibility is touted as an advantage. It's like claiming that a knife is better than a gun, because you can't shoot yourself in the foot with the knife.
For a devastating rebuttal of audiophile BS from a very experienced engineer, read Douglas Self's site: http://www.dself.dsl.pipex.com/ampins/pseudo/subjectv.htm
My bicyles
Sorry mwvdlee, you are wrong. Acording to Nyquist, EVERY signal that has a limited bandwidth (don't know if that's the correct english term), that means it contains no frequencies above a certain limit frequency, can be totally reconstructed out of a sampled signal with double that limit frequency as sampling rate. This is a perfect reconstruction if you leave noise floor out of the equation. With 24 bit or even 16 bit per sample the noise floor is practically unhearable and much better than on a vinyl record.
Please, read up a bit about digital signals and the Nyquist theorem, it is counter-intuitive, but it works. There are no "edges" in a reconstructed (played) digital signal!
You want to avoid polishing in the same direction as the temporal flow of the data. As you go around the disc in a circle you are moving ahead through time relative to what chunk of data correlates to how far along in the music you are. If you polish at right angles (from the hole in the center out to the outer edge and back) to the concentric rings of lands and pits (okay maybe it's just one long spiral like a record) any scratching you do (and that's what polishing is, replacing big scratches with much smaller ones) will not obscure sequential data bits, which means that the error correcting mechanism has a much better chance of working, whereas polishing along the same path which the laser beam will take risks obscuring several consecutive milliseconds worth of data.
For polishing CDs I recommend Wright's Silver Cream (originally intended for polishing silverware and probably available at your local grocery store).
I see even classic Slashdot is now pretty much unusable on dial up anymore.
I'll apologize in advance because I have a lot of nit-picking to do with your post.
... .
it just sounds fatter, warmer and... feels better... and it really IS the choice of
hardcore audiophiles.
I would like to know the science that makes something sound 'warm' or 'fat'. What defines a hardcore audiophile? Would they use the terms fatter or warmer?
. In contrary to CD's, the sound quality coming from a vinyl recording
depend on various external things
The needle used matter, (purist techno DJs and audiophiles spend insane amount of money on
their pickups, a good needle can really improve the sound)
To what level? How much money do you have to put into a needle before it reaches CD quality?
The quality of the actual
vinyl print matters a lot, for example the number of imprinted revolutions with respect
to the vinyl size, if we imprint 100 revolutions on a 12" disc, the soundquality
is generally improved compared to imprinting 500 revolutions. Further, the quality
of the overall manufacturing process and the vinyl material used matter. Further,
remember that technology is advancing within the field of vinyl record making and
playback, it has improved since the day CDs were born, today vinyl sounds better than ever.
I'll call this one a wash. You consider revolutions, the CD buyer can consider how the disk was mastered. 6 of one, half dozen of the other. Though it isn't an advantage of vinyl, it just means that you have to be careful what you buy. If anything that is a disadvantage.
You state that vinyl has improved since CDs were introduced. Did CD technology remain static during this period? The simple fact that the manufacturing process improved doesn't make the product superior to any competitor. The manufacturing process for wax candles has improved lightyears beyond what it originally was, but that doesn't mean you would use them to light your house today if you have electricity.
sum it up, depending on the circumstances - vinyl sound quality today is equal to or
better that CD quality, and vinyl sound will most likely improve as tech does.
This is a false statement. Vinyl is certainly not 'equal' to CD quality when you consider that to even come close to CD quality requires an investment of at least a thousand dollars. Compare a $20 CD player to a $20 vinyl record player. Not even in the same ball park. And to get 'better' than CD quality? You are going to be shelling out thousands of dollars for what is a marginal improvement at best. Your average CD in your average player will always sound better than your average vinyl record in your average player.
Vinyl sound will improve as the technology does... yes, I suppose, but the same is true for CDs...
And what do you prefer? a big 12" cover artwork of your fav band and a black
shiny thing that smells nice, is completely unique and cannot really be duplicated...
or a sloppy piece of cheap 12 cm plastic that only displays your geeky face when you
look at it, coming with with a CD sized artwork booklet?
I prefer not to think about smelling 12" black shiney things.
But kidding aside, what does album art have to do with the quality of the sound? And please forgive me, but something that is completely unique and not easily duplicated is not something I consider a strength. I like that I've taken my CDs copied them into a lossless format, stored that format on a server that I can access anywhere I go.
Out of modpoints but really liked a post? 1BDkF6TtmmeZ3yqXbz9yhdYVqRYnwFoXDj
Another reason for vinyl's sonic superiority is that no matter how high a sampling rate is, it can never contain all of the data present in an analog groove, Nyquist's theorem to the contrary.
Apart for being hogwash to begin with, it also reveals ignorance about how modern vinyl is produced. For the last few decades, the machine that cuts the master uses a digital buffer in order to be able to adjust groove widths to signal strengths (enough slack all the way through would mean very short play times).
Plus practically all mastering is done digitally today anyway.
sudo ergo sum
That alleged "data" in the analog groove is buried far below the noise floor of the best disc/reproduction system. The signal to noise ratio (in this context the same as dynamic range mentioned above) means any "data" that's allegedly on the disc is swamped by noise; the S/N ratio of the CD is figured as the ratio between the maximum sampled sine amplitude and the amplitude of the quantization noise. The quantization noise is the "step pattern" made by the discrete sampling, figured as subtracting the quantized signal ("sampled" to a particular amplitude representable by a discrete integer) from the original signal. You get the 96dB dynamic range often given for 16 bit sampling from the 2^-16 quantization noise (assuming full scale is 2^0), and 20*log10(1/2^-16)=96dB
The 44.1kHz/16 bit sampling of a CD is in no way an audio compromise, never mind when compared to vinyl. Higher sampling rates and widths are still useful to give more headroom when recording/mixing/mastering, but any reasonable recording fits well within a 16 bit/96dB dynamic range.
And, of course, there's a paper in the new (9/07) JAES doing double blind testing between new higher-resolution formats and good old CD-style sampling. No audible difference between the signal coming out of the player and one that undergoes a 44.1kHz/16 bit A/D/A conversion out of the higher res player.
Exactly right. A 22 kHz sampling rate means no frequency above 11 kHz can be reproduced (realistically more like 10 kHz with a good anti-aliasing filter). 44 kHz sampling rate cuts off at 22 kHz (realistically more like 20 kHz).
This is the Nyquist theorem, which says that the highest sound frequency that can be stored/reproduced in a digital signal is HALF the sampling rate.
If you have young and healthy ears, you should be able to hear many sounds above 10 kHz.
My bicyles