DSD downloads worth the trouble?

[spdif-usb]

Of course DSD has bits, I never said it didn't. What it doesn't have is digital values.

 

It's absolutely store-able and retrievable on a digital computer, It's just not process-able, for it contains no process-able values. In fact, it contains no digitally expressed values at all. It's a density modulated pulse train. It must be converted into another format that can contains values, like PCM or multi-bit PDM.

 

DSD (1-bit two level PDM) is the product of a Delta-Sigma Modulator, where the instantaneous signal level modulates the bit density of the resulting bit stream. To realistically operate, the bit stream must be several orders of magnitude higher in frequency than the highest frequency to be modulated. That's a completely different concept to a PCM sample based digital word stream that's in fact a list in time of 2's compliment binary digital word values.

 

The reason I say DSD is an analog signal (a simplification I admit, however true), is the original modulating signal can be retrieved using a simple analog integrator. I realize it's a concept with which you're uncomfortable.

Has it ever occurred to you the fact a bit IS a digitally expressed value? The only true reason you keep saying DSD is an analog signal is because you consistently utterly fail to understand the difference between an analog signal and a digital one.

[spdif-usb]

Ah, now the ultrasonic noise filtering doesn't need to be "after the DAC" (unlike what you said at first), it just needs to be "at some point." I agree.

 

Of course PCM needs filtering as well. Not a big deal, some folks like PCM, some folks like DSD, some folks like vinyl, some folks (myself included) like all of them.

 

Got anything to contribute to the thread from a music appreciation point of view?

I am merely trying to point out that urging other people to put their head in the sand when it comes to known facts about DSD is not being very helpful, especially not with all the marketing B.S. that goes on. To be fair, this also applies to the old FAD that still currently surrounds the Redbook CD format so, wake up, please.

[spdif-usb]

Valve Audio Devices are employing a similar approach in their upcoming DSD DAC 12. Here's how they describe it: direct filtered DSD64/128 stereo playback platform no decimation no processing as close as one can get to analogue with `files`

...Except that it does NOT get as close as one can get to analogue with `files`.

[spdif-usb]

Humm, DSD could reveal too much of the distortion embedded in Progressive rock recordings

It could, but doesn't...

[PAP]

Hi Everybody

DSD what does it SOUND like, if I may repeat myself.

 

[Hiro]

I am merely trying to point out that urging other people to put their head in the sand when it comes to known facts about DSD is not being very helpful, especially not with all the marketing B.S. that goes on. To be fair, this also applies to the old FAD that still currently surrounds the Redbook CD format so, wake up, please.

 

The biggest marketing BS of all is the so-called "24-bit" audio. With most 24-bit PCM converters maxing out at 17-bit THD+N... and let's not forget about the often-repeated vinyl marketing BS that "analog" is superior to anything digital...

[spdif-usb]

One can say about DSD what one wants (and my own view of it is pretty dim)

It would appear to me that this slightly depends on who it is you're talking to........

, it is a fact that it can be seen as digital and analogue at the same time.

It is also a fact that the quantization error that is present in DSD as a result of it needing a 1-bit quantizer can (again, mistakenly...) be seen as an error that is an analogue error.

 

Last time I checked, it was also still a fact that this error is present in DSD regardless of whether any further digital processing is applied between the ADC and DAC.

[wisnon]

Has it ever occurred to you the fact a bit IS a digitally expressed value? The only true reason you keep saying DSD is an analog signal is because you consistently utterly fail to understand the difference between an analog signal and a digital one.

 

 

If its a signal, then its ANALOG as it exists in the real not virtual world. The information that electricity carries is digital info either encoded voltage measure (PCM) or pulse density info (DSD). However, the density is such a simple representation that physically resembles a sine wave, and as such is very close to the original music sine wave.

[spdif-usb]

DSD is indeed a direct output of the delta sigma modulator.

In theory, yes... BUT... unfortunately not in practice.

[wisnon]

I am merely trying to point out that urging other people to put their head in the sand when it comes to known facts about DSD is not being very helpful, especially not with all the marketing B.S. that goes on. To be fair, this also applies to the old FAD that still currently surrounds the Redbook CD format so, wake up, please.

So why not answer his question...where exactly is this theoretical inferiority of DSD and why do many prefer it?

 

Not being snarky, I really want to see this. I do appreciate your contribution and keep an open mind. The only person so far that gve a half decent explanation was PeterST and even there I disagreed on a fundamental point that tore down the bastion of his otherwise impressive argument (centred around dynamics).

[spdif-usb]

One thing that I will say is that I have not seen from the DSD-negative side purely technical arguments that I am able to understand.

Perhaps the reason why you haven't seen them is just because you haven't very much looked for them.

[spdif-usb]

So the PCM pundits are NOT obnoxious???

Most of them are IMO, but that is mainly because they cannot understand the theories behind and the proven facts about digital audio as a whole. Even today, even the Nyquist theorem is still being misinterpreted by many, including many experienced electronics engineers, very unfortunately.

[spdif-usb]

, or did those grapes come from the north or south side of the hill, it's all a preference.

Usually I prefer beer just so I don't need to worry about that problem! :grin:

[spdif-usb]

The biggest marketing BS of all is the so-called "24-bit" audio. With most 24-bit PCM converters maxing out at 17-bit THD+N... and let's not forget about the often-repeated vinyl marketing BS that "analog" is superior to anything digital...

24-bit was chosen from a practical convenience standpoint because it's three bytes. Despite that, I agree many commercial 24-bit PCM music releases haven't even been recorded using a mic that can output a signal with a noise floor better than the equivalent of 16-bit, and, anything that can go wrong will go wrong... Murphy's law.

[Hiro]

24-bit was chosen from a practical convenience standpoint because it's three bytes. Despite that, I agree many commercial 24-bit PCM music releases haven't even been recorded using a mic that can output a signal with a noise floor better than the equivalent of 16-bit, and, anything that can go wrong will go wrong... Murphy's law.

 

I'm not talking about mics. I'm simply pointing out the fact that most PCM converters have THD+N at 17-bit level.

 

24-bit audio?

 

Please.....

[spdif-usb]

If its a signal, then its ANALOG as it exists in the real not virtual world. The information that electricity carries is digital info either encoded voltage measure (PCM) or pulse density info (DSD). However, the density is such a simple representation that physically resembles a sine wave, and as such is very close to the original music sine wave.

It is not very close to the original music sine wave even by far.

 

[Jud]

Looks like this thread is pretty well tapped out.

 

 

Sent from my iPhone using Tapatalk

[spdif-usb]

I'm not talking about mics. I'm simply pointing out the fact that most PCM converters have THD+N at 17-bit level.

 

24-bit audio?

 

Please.....

If the noise floor is the equivalent of 17-bit, you can convert 24-bit to e.g. 20-bit or 18-bit without any audible loss in quality. Adding a few extra bits doesn't hurt performance given today's standards, but you are right about the fact there will always be a bunch of charlatans who will take into abuse every chance that they see fit.

[Hiro]

Adding a few extra bits doesn't hurt performance

 

And sales, I would think... but 24-bit PCM audio, it is not.

[Paul R]

Perhaps the reason why you haven't seen them is just because you haven't very much looked for them.

 

Or, perhaps, such a deficiency simply does not exist. In all seriousness, the only deficiency I can see with DSD, in comparison to PCM that is, is that DSD currently does not have as wide a set of editing tools. And that is a deficiency that will be remedied with time.

 

In regard to quantization error - any recording system with less than an infinite number of discrete sampling periods per time unit will have this class of error. However, not only does DSD have less because of high sample rates, since DSD is intrinsically different in the way it represents recorded sound than PCM, it has at least an order of magnitude less error than PCM.

 

DSD is, however, far more susceptible to timing errors than is PCM, which we all know is vulnerable to jitter. The quality of the computer, software, and DAC appears to me to be the controls on this problem both in PCM and DSD systems. Nothing new there!

[Paul R]

It is not very close to the original music sine wave even by far.

 

[ATTACH=CONFIG]14301[/ATTACH]

 

Where did you find this diagram? It does not look at all close to what the real representation OS a pure sign wave in DSD would look like. I think this is a rather misleading representation, as it is shown.

 

But even assuming it is accurate, the sine wave you show generated from the signal is indeed very smooth. Which is one of the fundamental differences in PCM and DSD.

[Hiro]

Or, perhaps, such a deficiency simply does not exist. In all seriousness, the only deficiency I can see with DSD, in comparison to PCM that is, is that DSD currently does not have as wide a set of editing tools. And that is a deficiency that will be remedied with time.

 

And it's a deficiency that needs to be addressed in the future, given that the current audio world is dominated by SDM, and I mean "99.9% dominated" by SDM. It is only logical that native SDM production tools be developed.

[spdif-usb]

And sales, I would think... but 24-bit PCM audio, it is not.

Yes, like I already said twice. If you keep repeating this enough times, the quantization error of a 1-bit quantizer might perhaps one day magically fully disappear. :grin:

[spdif-usb]

Or, perhaps, such a deficiency simply does not exist.

Or, perhaps, 1 + 1 = 3.

[Hiro]

Yes, like I already said twice. If you keep repeating this enough times, the quantization error of a 1-bit quantizer might perhaps one day magically fully disappear. :grin:

 

In the meantime you may keep believing that PCM decimation filters, reconstruction filters, oversampling filters, and brick wall filters are perfect. Um, sure.