Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › Which format is closest to the master: vinyl, CD, DVD-A, or SACD?
New Posts  All Forums:Forum Nav:

Which format is closest to the master: vinyl, CD, DVD-A, or SACD? - Page 2

post #16 of 64
Digital is just analog at a lower resolution. We feel and hear in digital - neurons firing ; it's binary. There was an interested thread a while back where someone mentioned that the method of transporting audio within the brain; or from the nerves in the ear to another processing point in the brain; was essentially a < 20 kbps digital link when it came down to it. The point was that if we could model our primitive digital compression schemes closer to what our brains use, we could get some incredible compression by todays standards. How true that is, I don't know, but it was interesting.

The downside to digital is the finite resolution; with 16bit / 44.1kHz, that is a tangible downside. With 24 bit audio / sampling rates provided by DVD-Audio, it's a massive increase in the raw quantity of the data. (re: 16 bit, we have 65,536 possible amplitudes; with 24 bit we have 16,777,216 possible amplitudes to represent a sound. That's 256 levels of volume in a 24 bit recording for every 1 bit in a comparable 16 bit recording).

With 96 kHz sampling rates easily blowing past twice the frequency capacity of human hearing (theoretical 48 kHz vs 15-20kHz for human) - and don't mention 192kHz, I'm curious what analog proponents consider a downside to digital mediums? A/D & D/A? Getting better & cheaper each day...


Rob
post #17 of 64
Quote:
Originally Posted by robzr
The downside to digital is the finite resolution; with 16bit / 44.1kHz, that is a tangible downside. With 24 bit audio / sampling rates provided by DVD-Audio, it's a massive increase in the raw quantity of the data. (re: 16 bit, we have 65,536 possible amplitudes; with 24 bit we have 16,777,216 possible amplitudes to represent a sound. That's 256 levels of volume in a 24 bit recording for every 1 bit in a comparable 16 bit recording).
AFAIK the resolution of any system is limited by its own noise. An infinite resolution system simply does not exist (contrary to what is often claimed). 65536 possible amplitudes translates to 96 db dynamic range. Any analog system has more noise, i.e. lower resolution.


Regards,


L.
post #18 of 64
Quote:
Originally Posted by robzr
I'm curious what analog proponents consider a downside to digital mediums? A/D & D/A? Getting better & cheaper each day...


Rob
Apparently, it's more than just specs for analog lovers. But that's for another thread (I already went OT once in this thread. Bad habit of mine )

Oh, and Leporello, what I meant by 'real alternative' is that specs that good would make a big enough difference that there wouldn't be these little A/B "which version of this album sounds the best" comparisons that I see often done with sacd, dvd-a, CD, and analog It would make the choices for high-end audio down to either analog, or 128-bit 1000khz digital ...Oh crap, I went OT again :\ Technical discussions like this just suck me in
post #19 of 64
Thread Starter 
Quote:
Originally Posted by Leporello
An infinite resolution system simply does not exist (contrary to what is often claimed). 65536 possible amplitudes translates to 96 db dynamic range. Any analog system has more noise, i.e. lower resolution.
I disagree. Actually analog audio, especially vinyl, has quasi infinite resolution, as it isn't based on samples and quantisation, but a continuous «data» flow. Resolution seen as the amount of data per second available for defining a sonic signal, not S/N ratio.

If you would divide the mentioned dynamic range of 96 dB by 16,777,216 amplitude values instead of 65,536, that would indeed mean higher resolution, although the noise floor would still be the same. Now I know that in the actual PCM world resolution is always coupled with S/N ratio, I guess because the amplitude steps between neighbored amplitude values are standardized. Given this, the audible resolution of a 24-bit system is only increased towards very low amplitudes with the effect of making dithering obsolete; signals within a more normal amplitude range in fact aren't more accurate than with 16 bit. Under this premise the main merit of DVD-A lies in the higher sample rate, not in the higher bit depth.

post #20 of 64
At some point, the number of data points, or samples becomes unimportant. Movies are only 24 samples/sec. TV is only 29.97 samples/sec (60 interlaced fields, but only 30 different samples). Both tiny ranges considering the amount of analog info our eyes collect.

44,100 was picked because the theoretical range of continuous samples yielded a 22K high end.
post #21 of 64
Thread Starter 
Quote:
Originally Posted by Dave1
At some point, the number of data points, or samples becomes unimportant. ... 44,100 was picked because the theoretical range of continuous samples yielded a 22K high end.
Anyway, 44.1 kHz is not enough. Or more precisely: a steep low-pass filter at 20 or 22 kHz creates ugly artifacts within the audible range. So in the case of the sampling rate the decisive factor isn't data density, but passband characteristics.

post #22 of 64
There was an article in the September 2004 issue of AbsoluteSound.

"
74 - TAS Roundtable
The Sound of Analog Tape, LP, CD, SACD, and DVD-Audio
Our ongoing look at important topics finds Mark Levinson,
Harry Pearson, Doug Sax, and Jonathan Valin discussing and debating the sonic merits of various audio-storage formats—from analog tape to today’s high-resolution digital media. Robert Harley moderates.
"

I would not pay $7 to download the issue in PDF but basically all of them agreed that:

-2" analog tape is superb
-vinyl is beautiful
-DSD (SACD) mostly sounds very good
-PCM at 24 bits (DVD Audio) sounds very good if it is WELL CLOCKED
-CD - I don't think they even talked about it
post #23 of 64
Quote:
Originally Posted by robzr
Digital is just analog at a lower resolution. We feel and hear in digital - neurons firing ; it's binary.
No it's not & no we don't. And no amount of argumentative upsampling will change the inherent incorrectness of those statements.
post #24 of 64
I guess that would depend on how good your hearing is. 20k HZ is way above what I can hear. I know that there is the opinion that harmonics above that can still influence the audible range, but I have listened to tests to see if I could hear any difference, and I could not.

How many analog systems truly reach over 22K? I don't think many.

For mr the proff is in live music. I do not hear any additional information from live music than from a well recorded CD. I am not missing upper harmonics.
post #25 of 64
i believe cd would.
post #26 of 64
Thread Starter 
Quote:
Originally Posted by Dave1
I guess that would depend on how good your hearing is. 20kHZ is way above what I can hear. I know that there is the opinion that harmonics above that can still influence the audible range, but I have listened to tests to see if I could hear any difference, and I could not.

How many analog systems truly reach over 22K? I don't think many.

For me the proof is in live music. I do not hear any additional information from live music than from a well recorded CD. I am not missing upper harmonics.
Is this post addressed to my passband argument? If yes: the mentioned artifacts aren't missing harmonics, but crippled transients in the audible range. You can either go the Wadia route and put up with a ~3.5-dB roll-off at 20 kHz, a measuring result caused by a format-inherent amplitude modulation, or you can accept a drastic ringing (the Gibbs phenomenon) as a result of the equalized amplitude modulation and response.

post #27 of 64
Quote:
Originally Posted by Leporello
AFAIK the resolution of any system is limited by its own noise. An infinite resolution system simply does not exist (contrary to what is often claimed). 65536 possible amplitudes translates to 96 db dynamic range. Any analog system has more noise, i.e. lower resolution.
Not really - to stay in the picture analogy: Imagine two pictures with alternating black and dark blue as well as black and bright red dots, each one pixel in width on your monitor. number of dpi will be the same, but contrast will differ a lot. SNR and dynamic range in audio are more like absence of possible dirt on your screen and contrast in video. Resolution is more like pixel bandwith...

Greetings from Hannover!

Manfred / lini
post #28 of 64
Quote:
Originally Posted by JaZZ
Supposed a well-recorded analog master tape with good bandwidth is used -- which of the following formats will come closest in sound: vinyl, redbook CD, DVD-audio or SACD?
One thing goes without saying, CD for sure is not the closest to the original. For the simple fact that DVD-audio and SACD have both several times over the resolution of CD, CD can't be better than either one, at least format-wise. Now, from the format to the quality of the playback there's a lot of things going on, and well played CD can be pretty darn good. Top-notch Cd players can allegedly sound better than entry level universal players playing high res formats.

Whether high res digital formats played back in top-notch systems can sound better/more realistic than a top-notch analog rig, or viceversa, well, I just don't know
post #29 of 64
Thread Starter 
Quote:
Originally Posted by recephasan
There was an article in the September 2004 issue of AbsoluteSound.

...basically all of them agreed that:

-2" analog tape is superb
-vinyl is beautiful
-DSD (SACD) mostly sounds very good
-PCM at 24 bits (DVD Audio) sounds very good if it is WELL CLOCKED
-CD - I don't think they even talked about it
Interesting statements. But still too subjective («superb», «beautiful») to really help to find out which format is most true to the original signal. I know that for many people beautiful and superb is enough, and I guess these attributes are often associated with accuracy. That's not my opinion at all. I tend to think that analog tape could be the most accurate data format -- but it has no meaning as a music data carrier. And I'm rather sure that vinyl is more euphonic than accurate.

A question to the vinyl advocates: Why do you think vinyl is more true to the signal than DVD-A or SACD -- from a technical perspective?

post #30 of 64
Quote:
Originally Posted by JaZZ
A question to the vinyl advocates: Why do you think vinyl is more true to the signal than DVD-A or SACD -- from a technical perspective?

Well duh! Because it's superb and beautiful (j/k)
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Dedicated Source Components
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › Which format is closest to the master: vinyl, CD, DVD-A, or SACD?