CDs Dead In 2012?

[johnfull]

Even 32/192 is 'lossy' in the sense of distorted waveform.
But there is a definite improvement at 24/96.
Making copies of The Eagles 'Desperado' from vinyl could
hear an improvement in the crashing cymbals from 24/96
to the 'ultimate' 24/192. Otherwise, they are indistingishable.
The most revolting thing about CDs was the quasi-religious
incantation on every package that 'the high resolution will
reveal defects in the source material' - yeah, that's why
everything went to the dogs -- apology to dogs. Better mastering
and masking of deficient high frequencies with boosted midrange
and bass and dithering and etc eventually made a passable medium.
There's a growing awareness among the young, though, that we
threw something valuable away with vinyl and even tape.
Many albums are now being re-released on long-play 45 rpm double
disc sets. They sound stellar. The trick is to transfer them quickly to
a high-bit format. The studios won't do much better than that.
Many of the online direct downloads are disappointing, like they have
deliberately dulled them down somewhat so that they could retain the
originals. Mean, but understandable...

 

[MDS]

CDs are NOT compressed

CD certainly is a 'compressed' format.

Actually they are compressed (lossy) from the original analog or 24bit master.

CD (PCM) is not compressed nor is it lossy. Downsampling from 24 bit to 16 bit is not compression and sampling from an analog master is not compression. Downsampling and dithering from 24 bit to 16 bit is not compression; in fact dithering ADDS information (random noise at a very low level that perceptually masks quantization noise introduced by the analog to digital sampling process).

'Compression' can mean a few different things, depending on context:

1. Lossless compression is simply reducing the file size while perserving 100% of what you started with.

2. Lossy compression is reducing the file size by eliminating parts of the signal that a psycho-acoustic model deems inaudible. The technique is known as 'perceptual coding' and there are many models.

3. Dynamic compression is squishing the dynamic range thus increasing the average level across the track, making it louder. This is also not 'compression' in the sense of the first two.

Enough already...higher bit depth and higher sampling rates (if used originally, not resampled from something originally mastered with a lower bit depth or sample rate) don't sound better because CD is compressed - it is not. They can sound better because they capture more of the analog signal and the higher sampling frequency pushes any quantization noise into higher frequencies well beyond our hearing ability. Given that humans can only hear to 20 kHz and that ability rapidly degrades with age after about age 25, it really doesn't matter.

*quantization noise is the subtle errors that can be introduced by trying to assign a fixed integer value to represent a level when the 'real' level may actually be in between the fixed values (eg. a sample value of 32,000 assigned when it really should be 32,000.4). Dither is used to mask those artifacts.

 

[johnfull]

There is more to the sampling rate than the maximum frequency it can produce.
Nyquist is just part of the story. A sine wave produced at 16/44.1 will have artifacts
in playback that an otherwise identical sine wave produced at 24/96 will not.
Multiply that sine wave over two channels of varying frequency and you can easily end
up with a hash if the wave forms are jagged and not smooth. It's not just about frequency response...!...

 

[MDS]

There is more to the sampling rate than the maximum frequency it can produce.
Nyquist is just part of the story. A sine wave produced at 16/44.1 will have artifacts
in playback that an otherwise identical sine wave produced at 24/96 will not.
Multiply that sine wave over two channels of varying frequency and you can easily end
up with a hash if the wave forms are jagged and not smooth. It's not just about frequency response...!...

You have absolutely no idea what you are talking about. Feel free to believe whatever you want and do whatever you want with your personal music collection but don't throw around a bunch of technical terms that you clearly don't understand.

I'm sure you meant 'there is more to the sampling rate than the maximum frequency it can reproduce' but...

- Nyquist is the whole story.
- A 'smooth' wave is a single sine wave which of course is not audio or music.
- Sine waves are not 'multiplied' over two or more channels.

A musical waveform is a sum of a potentially infinite number of sine waves. Digital sampling is a process that takes a snapshot of the analog version of that waveform at regular intervals. The sample rate is simply how many times per second that waveform is examined - the higher the rate the higher the frequency that can accurately be captured. That's it.

 

[johnfull]

You are working from a theory and shaping your reality to suit it.
I'm coming at it from the other direction -- investigating the cause of phenomena that
I find interesting and/or repulsive. The lower the sampling rate of a sound, the more
it resembles a square wave. I have rudimentary memory of the sound of square waves
from my youth. The sound on a CD can break up if it is too complex, loud and concentrated in high frequencies. Crashing cymbals are the best test, but there are
plenty of others. Early CDs had the sound of shattering glass where complex music was
supposed to be. I began to investigate why this was so. Meantime, the priests of 16 bit have quietly modified their system over the years without ever admitting that it was
born with feet of clay and will never be more than burnished bronze. You really don't have to be so unpleasant if truth is on your side...

 

[johnfull]

Meantime, here's something GOOD about certain CDs -- if you have a decoder. I don't:

Quadraphonic CD Discography

 

[MDS]

You are working from a theory and shaping your reality to suit it.
I'm coming at it from the other direction -- investigating the cause of phenomena that
I find interesting and/or repulsive. The lower the sampling rate of a sound, the more
it resembles a square wave. I have rudimentary memory of the sound of square waves
from my youth. The sound on a CD can break up if it is too complex, loud and concentrated in high frequencies. Crashing cymbals are the best test, but there are
plenty of others. Early CDs had the sound of shattering glass where complex music was
supposed to be. I began to investigate why this was so. Meantime, the priests of 16 bit have quietly modified their system over the years without ever admitting that it was
born with feet of clay and will never be more than burnished bronze. You really don't have to be so unpleasant if truth is on your side...

Pure nonsense. You're right it is theory. Long established well-proven digital signal processing theory.

Sorry but if you're going to come here and spout off a bunch of audiophile nonsense you will be challenged on it. No offense, but I got a good laugh out of the statement' I have rudimentary memory of the sound of square waves'. Do you remember that a square wave sounds like buzzzzzzzzzzzzzzzzzzz?

 

[johnfull]

Yes, it's a most unpleasant noise.

No offense taken except for the challenge to my position.
Have you never encountered anyone who is unhappy with 16/44.1 sound before?
That seems a bit cloistered.
Wikipedia has an audio sample on the square wave page.
YouTube has many examples of the same.
The problem with square waves, besides being unreproducible, is that they create switching distortion and other artifacts like ringing.
It's easy to open a sound file in a graphical program, like GoldWave, and see the
difference in the shape of the wave as the resolution is changed, both sampling frequency in the horizontal axis and bit depth in the vertical.
You took issue with my shorthand of sine waves -- going from sine wave simplicity to
the complexity of actual sound (a sine wave is actual sound as well, though).
Some compression technologies alter the shape of the wave intentionally -- removing
detail at some fractal level (wavelet upon wave) which I don't understand at all.
But 16/44.1 sound is coarse compared to 24/96 in both appearance on a graph and to
the human ears that hear the difference between a square and sine wave.
Again, it is not disproof to claim to hear no difference and it's not convincing to trot
out A/B testing that is designed to disprove the difference. Compact disc sound can be
very clean and dynamic, no doubt. That fact masks the deficiencies inherent in running
so close to the edge, resolution-wise. You rejected everyone's attempts at analogy, so
I won't go there again except to say that human hearing is even more acute than sight
when it comes to detail. A coarse bitmap is easier on the eyes than a crescendo of music reproduced in an inadequately quantized medium. And that's not even going into the problem of sibilance.
Why fight progress?...

 

[Alex2507]

... human hearing is even more acute than sight...

That's amazing but consider that humans see the color red better than any other color.

 

[Adam]

That's amazing but consider that humans see the color red better than any other color.

No one man should possess so much power.

 

[Alex2507]

No one man should possess so much power.

It saves a group having to convene and determine that anybody that puts up 16 posts in his first 21 hours and manages to get MDS to step in to prevent him from making me any dumber needs to be easily identified by our eyes which are less acute than our ears. He's right though, I can hear those red chicklets and they are screaming.

Do you think red chicklets feel pain?

 

[Adam]

Do you think red chicklets feel pain?

If they do, it is only the pain of their enemies that they feel.

I wanted to try and counterbalance the red, but I've given out too much rep in the past 24 hours. Oh, well. BTW, it was only to experiment. No, not like Doug and Rick in college.

 

[Adam]

CDs might be dead in 2012, but this thread is on pace to beat that.

 

[johnfull]

Off to tryp-tophan the light fantastic.
Happy turkey day to those in the US.
Happy Thursday to the rest!

 

[Swerd]

I read this thread last night, thought briefly about wading in, and finally decided to avoid wrestling with a guy who could only hear himself. If music was BS, that guy would be an entire brass band.

Thanks to MDS for taking him on.

I found it funny that of all the recorded music formats discussed as being dead, that guy dismissed CD and liked SACD, DVD-A and hi-fi 45 RPM vinyl ! CD is the only one commercially viable during the last 30 years. SACD and DVD-A have been commercially abandoned for several years, and as far as I know, 45 RPM vinyl was never a commercial item other than cheaply made singles sold for 95¢.

 

[Stereodude]

The CD isn't going anywhere, but the record labels would like it to go away. They want to sell you a lossy digital file that you can't resell.

Actually scratch that... They don't even want to sell you a digital file. They want to rent you a digital file that you pay to play every time you listen to it. They're gradually moving that way with the whole "cloud" thing. The next thing you know you'll be paying per device to access your "cloud" collection.

 

[Stereodude]

Another year of development would have allowed 48 rather than 44 kbs, which would have eliminated the need for steep filtering at 20k.

You've heard of oversampling before right?

Pushing the sampling rate up on a PCM disc, though, takes a lot of stress off of the on-the-fly interpretation of the data stream, where sample and hold time is cut and aliasing is reduced.

What on earth are you talking about? A DAC doesn't do any interpretation of the data stream. It simply converts it to an analog signal and pushes it through a low pass filter to eliminated artifacts and return the signal to a smooth sinusoidal signal.

Actually they are compressed (lossy) from the original analog or 24bit master. As said before the CD's 16bit 44.1khz standard was a compromise in sound quality necessary to fit on a cheap to produce disc using the best affordable technology of the day.

Words mean things and it seems clear you're throwing around words you don't understand. CDs are not compressed. They contain 100% of the information in the original 16/44.1kHz audio data put on them. You can argue all you want that 16/44.1 isn't adequate to accurately capture the nuances of sound reproduction, but that's a totally different discussion from saying they're lossy or compressed. I'd love to see sources for your claims that Sony and Philips knew they were compromising sound quality in order to get something out the door though.

A sine wave produced at 16/44.1 will have artifacts
in playback that an otherwise identical sine wave produced at 24/96 will not.
Multiply that sine wave over two channels of varying frequency and you can easily end up with a hash if the wave forms are jagged and not smooth. It's not just about frequency response...!...

What sort of artifacts are these? Please tells us more. What is the manifestation in the output signal of these artifacts? Is it increased distortion, a decrease in signal to noise ratio, something else?

PS: DACs don't output jagged waveforms. They're low pass filtered so they're quite smooth.

You are working from a theory and shaping your reality to suit it. I'm coming at it from the other direction -- investigating the cause of phenomena that I find interesting and/or repulsive.

Talk about projecting.

The lower the sampling rate of a sound, the more it resembles a square wave.

Negative. It does not resemble a square wave. It resembles a low pass filtered version of the signal that when into the original analog -> digital converter. IE: one with high frequency information removed.

The problem with square waves, besides being unreproducible, is that they create switching distortion and other artifacts like ringing.

Which is precisely why DAC's don't & can't reproduce square wave. A square wave has infinite frequency content. Obviously in any digital system with a finite sample rate you can't reproduce something with infinite frequency content. There's a pretty important reason low pass filters are used in digital audio and sampling.

 

[sholling]

Words mean things and it seems clear you're throwing around words you don't understand. CDs are not compressed. They contain 100% of the information in the original 16/44.1kHz audio data put on them.

Do they carry 100% of the information contained in the original master? If not then while compression may not be the correct terminology it is effectively the same thing. Please educate me... if 10lbs do not fit in an 8lb bag what happens to the other 2lbs? I'm always ready to learn.

I'd love to see sources for your claims that Sony and Philips knew they were compromising sound quality in order to get something out the door though.

That shows a lack of understanding of technology and business. Nearly every new technology is a compromise to some degree in order to get something good enough to sell out the door at an affordable price point. The Internet wasn't in common use back then but there were discussions in the rags.

As for the failure of SACD and DVD-A they were doomed from the start by the record labels and their copy protection concerns and incompatibility with existing hardware.

 

[MidnightSensi2]

Do they carry 100% of the information contained in the original master? If not then while compression may not be the correct terminology it is effectively the same thing. Please educate me... if 10lbs do not fit in an 8lb bag what happens to the other 2lbs? I'm always ready to learn.


That shows a lack of understanding of technology and business. Nearly every new technology is a compromise to some degree in order to get something good enough to sell out the door at an affordable price point. The Internet wasn't in common use back then but there were discussions in the rags.

As for the failure of SACD and DVD-A they were doomed from the start by the record labels and their copy protection concerns and incompatibility with existing hardware.

Yes, your terminology is wrong, but you have the idea (which is more important than semantics).

Here is an example (free) of what is possible:
nin.com [download]- the slip

With the amount of bandwidth available, and storage being dirt cheap, higher sampling such as 24/96 is becoming more practical.

The real question is? Will anyone care? Perception of quality has really degraded in the past years, I'm not sure there is much demand for high quality recordings. This isn't just in music though, it's in magazines, print ads, fashion, house-ware, and so on. I think it will continue to be a niche, comparable to high fashion and so forth.

CD will die with a generation. I have this same conversation with car guys. The F1 transmissions have proven to outperform the manual gearbox, but drivers don't feel as 'connected' to the car. Will the next generation (which has little reason to learn to drive a manual) care (except for a small niche)?

 

[sholling]

The real question is? Will anyone care? Perception of quality has really degraded in the past years, I'm not sure there is much demand for high quality recordings. This isn't just in music though, it's in magazines, print ads, fashion, house-ware, and so on. I think it will continue to be a niche, comparable to high fashion and so forth.

What I can see becoming a good compromise from a marketing point of view is the audio equivalent of a "Blu-Ray plus digital copy" (yes both are digital), in other words you buy the FLAC or ALAC and it includes the right to download the MP3 or iTunes versions. It's a great marketing tool because it costs the seller almost nothing more and not everyone has a tool to convert the lossless format to fit their smart phone or portable media player. The popularity of DTS-HD MA and Dolby TrueHD shows that enough of us will demand lossless audio to support lossless formats. It may not matter to all or even most people but it will matter to enough. Will they do it? Who knows but if they don't we'll still have the boutique services like HDTracks.

CD will die with a generation.

In my opinion they would be nearly dead by now if DVD-A had not been hindered by copy protection concerns. The ubiquity of DVD players would have been a huge leg up if they would have played back through existing consumer grade DVD players and through existing S/PDIF connections. At least one car manufacturer did their part by offering a DVD-A as an option. But I agree the CD is going to be little more than a boutique item within a generation.