What sampling / bit rate would equal vinyl?

[RUR]

I know they have been.... but only two ICs in a DAC or a Transport ... TWO ICs !?!?

Two in the DAC, anyway...

The 4715 is surely among the smallest DACs in high-end audio, and the main reason for that is Kimura's desire to keep signal paths as short as possible. Mission accomplished, I'd say. The converter is the more radical of these two products: It has no digital filter chip—so it can run neither resolution-enhancement schemes à la HDCD nor bog-standard oversampling—and it doesn't even contain an analog filter. What it does contain is a total of only 20 parts: 7 resistors, 3 film capacitors, 6 electrolytic capacitors, 2 voltage regulators, and 2 chips. Of the last, one is a Philips TDA1543T 16-bit dual DAC, the other a Crystal CD8412 input receiver. That's it—that and a tiny circuit board, a few jacks, and a neat-looking ceramic box. There isn't even an op-amp for converting current to voltage at the output, because that's done passively, with two resistors.

http://www.stereophile.com/digitalprocessors/800/index.html

 

[FrantzM]

The Voltage regulators are " chips"

 

[RUR]

Four, then. Not including transport and external power supplies.

 

[DonH50]

Interesting article. The jitter description is wrong as jitter depends upon the signal, not the clock. Measurements are not very good, and it is obvioussly quite load-sensitive, and without reclocking it is at the mercy of the data stream's clock. However, seems like it provides the magic in the right system.

 

[Orb]

Highly-integrated chips have been around a while...

I would bet that does not include the laser receiver, however, and I wonder if there is one or more in the power supply.

Yeah, plus the board with the servo code, usually this is one of the largest circuit boards in a CD player (and most times the largest chip as well).
Cheers
Orb

 

[Orb]

Interesting article. The jitter description is wrong as jitter depends upon the signal, not the clock. Measurements are not very good, and it is obvioussly quite load-sensitive, and without reclocking it is at the mercy of the data stream's clock. However, seems like it provides the magic in the right system.

I have not looked so take my comments with a pinch of salt as I may be missing the context.
However it may be semantics Don, even Dunn states there is converter clock jitter relating to the timing of sample conversions; I appreciate that technically it could be argued that this still can be associated with the signal.
The sampling jitter comes back to oversampling techniques for its elimination, so I am assuming they mention this and are a NOS.
Not necessarily disagreeing just showing what may be their POV and possibly fits with Dunn's description of Converter Clock Jitter.

Thanks
Orb

 

[DonH50]

Jitter on the clock causes problems, yes, but changing the clock frequency does not impact the level of jitter that causes problems with the signal. At the top of the second page he implies the jitter threshold scales with oversampling rate and resolution; it only scales with resolution and signal frequency. Going from 16 bits and 44.1 kS/s to 20 bits and 8x OSR only reduces the threshold by 16 (the extra four bits going from 16 bits to 20); the clock rate does not play into it. The 1.35 ps quoted should be "only" 10.8 ps.

I have only skimmed the article; interesting they ony use a single dual-channel DAC. Also, IIRC, a schematic was posted a while back in another thread, and in fact there is some output filtering as part of the output circuit (aside from the usual RC that always happens in the real world), but no distinct high-order filter as is usually found in NOS DACs.

All in all, an interesting "minimalistic" approach to the problem, akin to the old "straight wire with gain" raised as an ideal for preamps and amps. I do tend to think throwing more transistors at a problem is not always the best solution...

 

[Orb]

I guess I should read it sometime, however changing the sampling rate may impact jitter in that it increases the noise floor (talking about sampling jitter being timing inaccuracy in the D/A conversion).
Being lazy here so just quoting part of Dunn with regards to where I am coming from - but appreciate as I say the context may not match their narrative for the product design:

Heavily oversampled noise-shaped Delta Sigma Modulators produce ultrasonic noise at high amplitude. As this noise signal is concentrated at high frequencies, model 1 indicates that jitter sidebands will have much higher amplitude than for equivalent signals in the audio band, and so sampling jitter will have a greater impact on the noise floor.
As the model 2 DAC sideband amplitudes scale with sideband frequency rather than input frequency, so the effect of jitter on the ultrasonic noise from the modulator output is reduced by the same proportion. In a bad case this produces a noise floor that rises with frequency.

However when looking at modern well designed products I cannot see any issues relating to oversampling-sampling jitter, but maybe they are justifying their position from a theory technical standpoint (but as you say it needs to be looked at carefully).

Cheers
Orb

 

[DonH50]

I did not catch what paper you quoted? And what the heck does "heavily oversampled" mean?

In any event, oversampling DS modulators do indeed have high-amplitude ultrasonic noise and a noise floor that rises with frequency; that's what they are designed to do! They shape the in-band noise out of band, into the HF regions, then filter it off. That has nothing to do with jitter. I have a feeling I am missing the context of the paper.

The theory was presented (loosely) in one of my fundamental jitter threads, and says that clock rate does not matter for jitter, just resolution (number of bits) and signal frequency. I do agree wider bandwidth may lead to greater noise, but that is not the same as increased jitter on the clock. Higher clock frequencies may have higher jitter, or not; most are in a controlled loop (DLL/PLL) so it is the loop bandwidth that sets the jitter, not the carrier frequency (to first order).

HTH - Don

p.s. How did we get here from the bit rate needed to sample vinyl? I wasn't intending to post here as I don't really know -- too much depends upon the listener and what bandwidth is really needed, which is pretty subjective. Alas, I suspect very high rates are needed to capture the full bandwidth of the ticks and pops! (It's a joke, guys, don't crucify me over it...)

 

[garylkoh]

HTH - Don

p.s. How did we get here from the bit rate needed to sample vinyl? I wasn't intending to post here as I don't really know -- too much depends upon the listener and what bandwidth is really needed, which is pretty subjective. Alas, I suspect very high rates are needed to capture the full bandwidth of the ticks and pops! (It's a joke, guys, don't crucify me over it...)

I think that it also depends on the type of music. One of the findings during the digital/vinyl event we did that surprised me was that large orchestral works might need lower rates than simple, solo instrument of voices. At 24/96 we could not pick out the difference between vinyl and digitized vinyl (not even the trained listeners) on orchestral works. However, many of the trained listeners could discern differences with small groups/solo piano. I did my own tests with myself at 24/48, and didn't hear a difference again on orchestral works. I'm doing some digitization of choral music with 24/44.1 resolution, and it's still very difficult to tell the difference.

 

[Vincent Kars]

orchestral works might need lower rates than simple, solo instrument of voices

masking?
At least when testing codecs for transparancy, small settings are recommanded.

 

[FrantzM]

What I find bothersome with the title of the thread itself is the implication that Vinyl is what Digital should aspire to ... I am not sure this is currently the case objectively ..Preferences are another thing ...

 

[garylkoh]

masking?
At least when testing codecs for transparancy, small settings are recommanded.

I recall Amir mentioning that when testing CODECs that orchestral harmonic content was easier to compress than single instruments. May be the same thing??

 

[DonH50]

Could also be mic'ing and recording technique...

 

[Orb]

Don,
I have not been ignoring you bud, just been feeling very lazy and to respond in a useful way to you would need me to kick my backside and go up a gear
Yeah this has digressed in a major way, but I blame others as I was responding to you, while you wereI w responding to someone else
I will send you a PM with the link to the papers, unfortunately there are other ESS related articles and papers I had that were of interest relating to oversampling and a related interpolation patent they have, if I ever find these I will let you know but spent awhile already trawling my stuff trying to find them bah.
Relating to the oversample and how much is too much, well on Stereophile they have an example of the noise floor issue from 384khz with one product (think was Cary), but as I mentioned it should not really be an issue with the better implemented products IMO.
Thanks
Orb