Audio Critic

[jmercer]

Re: Re: Re: subjectivity

I dont think is that simple, how different? 99.5%? 88%? there most be some threshold. For example, when you rip a CD to MP3 the data is different, yet, a good encoding could be very difficult to detect by just hearing.

When you rip, the dataloss is algorithmically even, making it difficult to determine what's missing. And the better the algorithm, the less the loss is detectable - because good music compression algorithms target areas that can be "dropped" with a minimum of impact. (Or so the designers think. )

Randomized loss, OTOH, could have a dramatic effect. It could sound like a skip... or a particular instrument might drop out for a moment, or waver or even go off-note. I have no idea what the threshold is for a human to detect these variances, but consider that a good conductor can often detect a brief flat note from a particular musician in an orchestra.

 

[TjW]

Re: Re: Re: subjectivity

I dont think is that simple, how different? 99.5%? 88%? there most be some threshold. For example, when you rip a CD to MP3 the data is different, yet, a good encoding could be very difficult to detect by just hearing.

A CD image should be just that, an image of the data on the disc, complete with all the overhead of track times and so forth. If my understanding of rip programs is incorrect, then the first two files won't be identical, and the scheme is then not a good test. This could be because of the rip program, and not necessarily because of a change in the data.

I don't particularly want to argue about what the minimum perceptible difference could or should be. But an identical data stream is just that -- identical.

This is not like an analog signal, where it's possible to argue that the effect is too subtle for the most sophisticated equipment to measure, it can only be detected by human ears. In this case, the signal has _already_ been measured, and the data stream is the record of those measurements. If it can't be measured to 1 part in 65536, then it wasn't on the disc to begin with. You can't have less than a least-significant-bit change. That's as subtle as it can get. When the bit gets to the DAC, it's either a 0 or a 1.

 

[Wellfed]

...I was wondering, can the treatment you'd like be applied after the CD has had been "chiped" ?...

Yes, the disc can be treated with Vivid or some other treatment product after the GSIC application.

 

[TjW]

Re: Re: Re: subjectivity

I dont think is that simple, how different? 99.5%? 88%? there most be some threshold. For example, when you rip a CD to MP3 the data is different, yet, a good encoding could be very difficult to detect by just hearing.

A CD image should be just that, an image of the data on the disc, complete with all the overhead of track times and so forth. If my understanding of rip programs is incorrect, then the first two files won't be identical, and the scheme is then not a good test. This could be because of the rip program, and not necessarily because of a change in the data.

I don't particularly want to argue about what the minimum perceptible difference could or should be. But an identical data stream is just that -- identical.

This is not like an analog signal, where it's possible to argue that the effect is too subtle for the most sophisticated equipment to measure, it can only be detected by human ears. In this case, the signal has _already_ been measured, and the data stream is the record of those measurements. If it can't be measured to 1 part in 65536, then it wasn't on the disc to begin with. You can't have less than a least-significant-bit change. That's as subtle as it can get. When the bit gets to the DAC, it's either a 0 or a 1.

 

[DevilsAdvocate]

Re: Re: Re: Re: subjectivity

A CD image should be just that, an image of the data on the disc, complete with all the overhead of track times and so forth. If my understanding of rip programs is incorrect, then the first two files won't be identical, and the scheme is then not a good test. This could be because of the rip program, and not necessarily because of a change in the data.

I don't particularly want to argue about what the minimum perceptible difference could or should be. But an identical data stream is just that -- identical.

This is not like an analog signal, where it's possible to argue that the effect is too subtle for the most sophisticated equipment to measure, it can only be detected by human ears. In this case, the signal has _already_ been measured, and the data stream is the record of those measurements. If it can't be measured to 1 part in 65536, then it wasn't on the disc to begin with. You can't have less than a least-significant-bit change. That's as subtle as it can get. When the bit gets to the DAC, it's either a 0 or a 1.

Just curious here. What makes one CD player better than another? If they read the same 1s and 0s, would that mean the difference is in the way the data is output to the other audio components? Is it possible for one CD player to be "better" at reading 1s and 0s than another that would improve sound quality? (I already know that CD players can be "better" at reading 1s and 0s because the cheap CD drive at work will start to skip if there is even the slightest smudge on the disk. Never have that problem with the same disks at home. I can't say whether the is any sound quality difference because I use different headphones, PCs, etc.).

 

[TjW]

Re: Re: Re: Re: Re: subjectivity

Just curious here. What makes one CD player better than another? If they read the same 1s and 0s, would that mean the difference is in the way the data is output to the other audio components? Is it possible for one CD player to be "better" at reading 1s and 0s than another that would improve sound quality? (I already know that CD players can be "better" at reading 1s and 0s because the cheap CD drive at work will start to skip if there is even the slightest smudge on the disk. Never have that problem with the same disks at home. I can't say whether the is any sound quality difference because I use different headphones, PCs, etc.).

Well, the digital to analog converter can have a big effect, of course. There's usually a low-pass "reconstruction" filter after the D/A converter (Sampling at 44.1 KHz means you can only theoretically reproduce up to 22.05 Khz ), and maybe a buffer amp to bring the analog signal up to line levels.
All of these are opportunities to add noise and distortion. The correct digital data is as accurate as the representation of the signal on the CD is going to get. It's all downhill from there, the only question is, how fast or slowly?
Reading the bits better will help up to the point where all the audio bits are correct after applying the error correction. After you've got the same data you laid down, that's as good as you can do. The DAC doesn't know or care whether the data got corrected.
But certainly there are things on the CD player that affect how well the data on the CD is reproduced as an analog signal.

The thing I don't understand about devices to "improve CDs" is where are they getting the additional information to improve the reproduction? The sixteen bits per sample is all there is on the CD. Any change in that data would simply be a distortion of the previously-made measurements.

 

[DevilsAdvocate]

Re: Re: Re: Re: Re: Re: subjectivity

good stuff

Thanks. Hey, is the sound quality from a particular part of a CD always the same? Assuming you use the same system to play it. With vinyl, as records got old and used you would loose sound quality and get static or get a sort of "blurry" sound. Same thing happened with cassettes--they would get wornand scrathed and not sound as good. But with a CD I would think it would always be the same. CDs can get worn with scrathes and smudges which can cause clicks and jumps, but I have never noticed a difference in sound quality in the parts of the CD that play correctly. With CDs it seems to be an all or nothing deal: either you get exactly the same sound, or you don't--which would mean clicks or jumps. So from a CD you can't really get an overall "better" or "worse" sound like you could from vinyl. Is that thechnically correct?

 

[MRC_Hans]

Re: Re: Re: Re: Re: Re: subjectivity

*snip*
The thing I don't understand about devices to "improve CDs" is where are they getting the additional information to improve the reproduction? The sixteen bits per sample is all there is on the CD. Any change in that data would simply be a distortion of the previously-made measurements.

There in no way to improve the signal ON the CD. That is the reason this is a paranormal claim.

The only scope for improvement that exists is to improve the reading accuracy of the CD player, and this can be tested technically. Simply obtain the exact bit pattern on the CD, and compare the bitstream read by the CD player (requires some tampering). Equipment exists that can compare two bitstreams and display the difference. This would do away with all the listening nonsense.

A modern CD player should be very precise and have little scope for improvement (remember that we have CD-ROM readers that can read at more than 50 times the speed of a music CD, so the technology is VERY mature), but the only possible non-paranormal way of improvement is some kind of surface treatment.

Placing ANY gadget near to a CD, can ONLY work through paranormal means. Sorry to be so cynical here, but as I mentioned in another tread, this "chip" claim equals claiming to change the content (or readability) of a printed book by placing a device on top of the shelf.

Hans

 

[MRC_Hans]

Uhhh, and Wellfed: May I strongly advice you to do a blind test in private. It might save you some embarassment.


......And the JREF some time .

Just being cynical.

Try to think this cynical thought: The music business is very competitive. If a method existed to improve CDs, can you think of ANY reason manufacturers wouldn't be using it?

Hans

 

[DevilsAdvocate]

I still would like to hear TjWs opinon on whether or not a CD can have "reduced quality" or if it is an all or nothing case where you get the same sound and a bad CD can only produce things like skips and clicks. I simply don't see the sound quality of any CD could be worse or better than any other, with the exception of clicks or skips. But I'm no expert on these things.

 

[MRC_Hans]

I still would like to hear TjWs opinon on whether or not a CD can have "reduced quality" or if it is an all or nothing case where you get the same sound and a bad CD can only produce things like skips and clicks. I simply don't see the sound quality of any CD could be worse or better than any other, with the exception of clicks or skips. But I'm no expert on these things.

Well, you can have mine: On a CD (be it a music CD or a CD-ROM) is a pattern of bits, etched into a very thin metal layer. In effect the bits are represented by holes in the layer. A bit is either 1 or 0. A Cd player shines a laser-beam at the metal layer and depending on the way light is reflected, it determines whether each spot is a 1 or a 0.

Then, of course, there is a protocol for aligning and format, but the short story is that we obtain a tstream of numbers from the CD. Each number can be either right or wrong. If it was for some reason read incorectly, it may be wrong. Since there is a system of data reduncancy on the CD, the recovery software can correct for single wrong numbers (this is an old and well-known thechnique in electronic communication).

Numbers can be misread from the CD for a number of reasons:

- There can be faults in the etching. Especially with writable CDs, this is a problem.

- There can be scratches in the surface of the CD. Now, the bottom surface of the CD (non-label side) is the side from where it is read, but if you examine a CD, you wil lsee that it has a certain thickness, and that the metallic layer is really on the other side, right under the label. The laser optics focus through this transparant layer on the metal layer, and this means that scratches and dirt on the bottom of the CD are out of focus, making it surprisingly tolerant to scratches, dirt, fingerprints, etc.

The CD can deteriorate with time in various ways:

- The metal layer can be broken down by oxidation, delamination, and various other physical/chemical processes.

- The clear body of the CD can become opaque, due again to chemical and physical changes (espcially after exposure to heat and/or sunlight).

- The bottom surface can become scratched or contaminated.

In all cases of deterioration, the result will be an increased incidence of misread numbers. When reading faults occur, the correction software tries the following:

1) It tries to recalculate the number using the redundant information on the disk.

2) Failing #1, it tries to interpolate between good readings.

#2 can result in deteriorating sound, but not necessarily in audible skips.

So in conclusion: Poor reading accuracy on a CD might present as deteriorated sound quality.

Hans

 

[TjW]

Re: Re: Re: Re: Re: Re: Re: subjectivity

Thanks. Hey, is the sound quality from a particular part of a CD always the same? Assuming you use the same system to play it. With vinyl, as records got old and used you would loose sound quality and get static or get a sort of "blurry" sound. Same thing happened with cassettes--they would get wornand scrathed and not sound as good. But with a CD I would think it would always be the same. CDs can get worn with scrathes and smudges which can cause clicks and jumps, but I have never noticed a difference in sound quality in the parts of the CD that play correctly. With CDs it seems to be an all or nothing deal: either you get exactly the same sound, or you don't--which would mean clicks or jumps. So from a CD you can't really get an overall "better" or "worse" sound like you could from vinyl. Is that thechnically correct?

It should be, unless something physically changes it enough to overcome the error correction. Since it's a non-contact read, just reading a pressed CD shouldn't affect it at all.

Vinyl and cassettes had to rub the media against something, and that's bound to wear. Although I seem to recall reading about someone building a laser non-contact reader for vinyl.

You don't get a sound at all from CDs, you get data. The data is either correct or it's not. (And because of the error correction, you can tell whether or not it's correct) That data is from a previously-measured waveform, and theoretically, can be used to reproduce the waveform up to the accuracy with which it was measured.
You can never quite get there. The DAC isn't quite linear. The reconstruction filter isn't perfect. there's thermal noise in amplifiers, speakers have a finite slew rate, and so forth and so on. But as a transfer medium, it's good enough that it's all those little details in the reconstruction that are the limiting factor, not the data on the CD.

I wonder if the audiophiles have endless debates over whether something sounds better if it was downloaded over a T1 rather than coming into the computer on a 56K dialup?

 

[edthedoc]

Wellfed, we are very interested in whether you have conducted your own double blind trial. Please could you answer this question.

 

[Patricio Elicer]

Hello Wellfed, and welcome to the forums.

From the February 11, 2005 Commentary:

The Intelligent Chip automatically upgrades any CD/DVD/SACD disc when the Chip is placed momentarily on the top surface of the player while the disc is playing

(bold mine). Is that, in effect, all the treatment needed to change/improve the sound quality?, just to place the chip on top of the player?

 

[TheBoyPaj]

Wellfed, we are very interested in whether you have conducted your own double blind trial. Please could you answer this question.

He has (answered the question) and he hasn't (done the tests).


In reposnse to Gr8Wight:

This gentleman is an audiophile. He would definitely never think of messing with the innards of the cd player he uses to partake of his passionate enjoyment of his music.

He would never think of messing up his CD player? For a million dollars? What sort of sound system could he buy with that?

 

[IXP]

There has been some discussion about the accuracy of reading data off a CD. Consider this: We read programs off CDs all the time. Changing random bits in a program are likely to have dramatic effects if that piece of code is ever executed. Yet programs read off CDs seem to operate flawlessly (well, as flawlessly as program can.) Reading data from an audio CD should be at least as accurate given that it acutally has more redundancy than data CDs do.

TjW's comments about ripping a CD are right on. Ripping software can recover the exact bitstream from a CD and I would be very surprised if ripping two new CDs did not produce exactly the same image. MP3 encoding has nothing to do with it. The ripping software reads the exact bits, then an algorithm is applied to compress it to MP3 or WMA or whatever.


IXP

 

[IXP]

Re: the protocol negotiations ongoing here.
I find the suggestion that something could be put into the cd player to electronically detect the GSIC (or whatever you call it) silly. Isn't it our contention that there is nothing electronic or otherwise actually in the 'chip?'

You are not very imaginative then. It is not necessary to detect the GSIC, only to detect that something has been placed on top of the CD player. Here are some suggestions:

1. Placing the CD player on a sensitive electronic scale.
2. Placing an array of infrared LEDs on one side of the player and detectors on the other.
3. Using a camera trained on the top of the player that is monitored by a computer.

I do not think it is likely that this type of fraud is in the works, I am merely trying to point out that the protocol must prevent it. Given his proposed protocol, particularly that the chipping and test occur on his own system, I would gladly accept a challenge to distinguish the discs for a million dollars. This is why I called for doing the chipping on a system provided by the testers.

IXP

 

[drelda]

all CD transports are equal?

Hi,

Theres been lots of discussion here along the lines of "if the bitstream coming out of a CD transport is unchanged then the sound must be unchanged". I don't think this is quite true - the timing of the bits is very important too.

Its true I think that just about any CD transport is capable of accurately reading the bits on a disc (otherwise cheapo CDROM drives wouldn't work!) - but if a CD player delivers those bits with slightly the wrong timing it can make a very audible difference...

Don't misunderstand me - I'm extremely sceptical about this device - but it would be a mistake to fall into the "its digital data so nothing can make a difference" trap.

I must admit I used to wonder if the whole idea of one CD transport being better than another was an audiophile woo - but its easy to test so I did...

I took three CD transports (cheapo games console, cheapo DVD player and reasonably expensive audiophile CD transport) and did several blind tests - ie see if I could determine which was the expensive transport with the only variable being which player was providing the digital signal to a DAC/amp etc (I'm sure my protocol wasn't up to JREF standards but I think it was still sound).

The outcome was that there was a definite difference in the sound between the players and it was reasonably easy to pick out the expensive player - which (subjectively) did sound better.

However the interesting part was that the difference was a lot more subtle than I was expecting - and I think a lot more subtle than the CD transport makers would like to admit! But I was reassured that expensive CD transports are not just woo woo.

What was also interesting about this test was that when I first did it not blind (ie I did the switching of the cables / CDs myself) then I perceived a much bigger difference in the sound! I would have sworn I wasn't fooling myself about that - but its true that 90% of the perceived difference dissapeared when I didn't know which was which.

So to repeat - I'm not at all arguing that this device works (I am very sceptical) - but I am very sympathetic of anyone who becomes convinced an audio tweak makes a big difference without doing a blind test.

regards - Drelda

 

[MRC_Hans]

Sorry, but timing is not critical. A bit is sampled at mid time, ideally. If there is a timming jitter big enough, sampling occurs outside the physical location of the bit and a faulty sample occurs, but this is no different from sampling a wrong logical value(0 or 1) due to a scratch or any other reason.

Once read from the CD, the bitstram is retimed using the internal clock of the player (chrystal controlled to several decimal points).

The difference between cheap and expensive players can be real enough, but it is due to different quality of D/A convertion and filtering. Also fault-correction software might be better on some players. The best portable players have a large memory buffer allowing them to recover long dropouts in the bitstream due to bumps. They simply play from the buffer while the reading head rereads the missing parts.

ans

 

[drelda]

Sorry, but timing is not critical

Your right - I don't know for sure that its timing issues that make the audible difference between cheap and expensive CD transports - however :

1) I think there is compelling evidence that there is a difference (ie I really did the test).

2) It can't be the DAC in the player thats making the difference - I just used the digital outputs of each device and put them into the same processor (DAC/preamp)

3) It seems reasonable that differences in timing could make a difference - ie the 'if the bits are the same the sound is the same' argument isn't watertight

4) The people that make CD transports do seem to claim that timing/jitter type issues are important?

Given those I was assuming the key difference was timing - but I could be wrong. However I'd be surprised if the cause of the difference is that the cheap players are not delivering the correct bits (ie error correction etc). A cheap CDROM seems to read the correct bits reliably enough - so I've no reason to think the other devices can't?

Or are you saying that there isn't a real difference between the sound produced by different CD transports? I might be up for taking the pepsi challenge on that one myself .

- Drelda