Is this guy for real?

[Bruce B]

What I was saying is that adoption that Mike is talking about for 4x DSD in my opinion is purely based on the assumption that "4X" anything is better than that anything by itself. If tomorrow an 8X comes, that will be the next to have. No one is adopting it because of the reason you mention. After all, that is an objective observation, not subjective.

Well I've done this dozens of times, but I swear I can't tell the difference between 2xDSD and 4xDSD.
From what I've researched and read, 4xDSD was just created so multi-tracking can be done without UHF noise build up.
2xDSD pushes the noise out to 50k and 4xDSD pushes it out to 100k, both well away from the audible band.
2xDSD is all you need, but if you're going to pay me more to transfer at 4xDSD, then I'm not going to argue. )

 

[amirm]

Well I've done this dozens of times, but I swear I can't tell the difference between 2xDSD and 4xDSD.
From what I've researched and read, 4xDSD was just created so multi-tracking can be done without UHF noise build up.
2xDSD pushes the noise out to 50k and 4xDSD pushes it out to 100k, both well away from the audible band.
2xDSD is all you need, but if you're going to pay me more to transfer at 4xDSD, then I'm not going to argue. )

Now that is a reasonable response I like to see .

 

[Fitzcaraldo215]

Well I've done this dozens of times, but I swear I can't tell the difference between 2xDSD and 4xDSD.
From what I've researched and read, 4xDSD was just created so multi-tracking can be done without UHF noise build up.
2xDSD pushes the noise out to 50k and 4xDSD pushes it out to 100k, both well away from the audible band.
2xDSD is all you need, but if you're going to pay me more to transfer at 4xDSD, then I'm not going to argue. )

I agree. While I have not done it frequently, a friend and I had occasion to do some further listening a short time ago. He, a noted recording critic, had obtained some test remasterings of a very famous Proprius analog stereo recording in DSD, 2x DSD, 192k and 356k. Any differences through an excellent DAC and system were miniscule and essentially nonexistent in very careful listening - totally underwhelming. He did express a very slight preference for 2xDSD, but acknowledged he would never be able to identify it in blind conditions.

Actually, the very best sound that night by far was listening in my normal mode: DSD converted to 88k PCM with Dirac Room EQ and DSP bass management applied. I had long before concluded that conversion of DSD to 88k PCM was slightly preferable to 176k in my system.

We concluded that the numbers game for ever higher sampling rates is basically a marketing ploy rather than a significant sonic advance in playback. But, yes, recording, mixing, etc. at ultra sampling rates may provide better recordings by keeping more ultrasonic noise out of the recording chain before final conversion to normal playback resolution.

 

[c1ferrari]

What I was saying is that adoption that Mike is talking about for 4x DSD in my opinion is purely based on the assumption that "4X" anything is better than that anything by itself. If tomorrow an 8X comes, that will be the next to have. No one is adopting it because of the reason you mention. After all, that is an objective observation, not subjective.

Okay, thanks for the clarification.

 

[c1ferrari]

Well I've done this dozens of times, but I swear I can't tell the difference between 2xDSD and 4xDSD.
From what I've researched and read, 4xDSD was just created so multi-tracking can be done without UHF noise build up.
2xDSD pushes the noise out to 50k and 4xDSD pushes it out to 100k, both well away from the audible band.

Thanks for the info. I'm curious if there are other factors for DSD256? I'm very vaguely reminded of a 'high speed' solid state company,
which states in its literature about bandwidth into the MHz range that was a consequence of the company's interest in thermal stability of components
at specific junctions in their circuit, if recollection serves. IOW, MHz frequencies are also well-away from the audio pass band and it met design objectives.

 

[Mike Lavigne]

What I was saying is that adoption that Mike is talking about for 4x DSD in my opinion is purely based on the assumption that "4X" anything is better than that anything by itself. If tomorrow an 8X comes, that will be the next to have. No one is adopting it because of the reason you mention. After all, that is an objective observation, not subjective.

we don't yet know whether 4xdsd will become a format for consumers, and a step forward. I've not heard it yet myself. I've spoken to people who have that report some degree of improvement. it seems the step forward from 2xdsd to 4xdsd is smaller and less clear then PCM->dsd or dsd -> 2xdsd. if Bruce says he cannot hear an improvement then that's a new data point for me.

my opinion is that digital music needs to keep pushing the envelope for improvement. will the next step be higher level dsd, higher level PCM, or some currently unknown digital format?

it will take a crystal ball to know for sure; for now I am looking forward in the short term to 4xdsd and it's maturation with hardware and content to see whether it is another step forward. my mind and ears are open to it.

no doubt in my mind that digital transfers are not transparent to the source at this time. I hope that I see that change.

 

[Bruce B]

Thanks for the info. I'm curious if there are other factors for DSD256? I'm very vaguely reminded of a 'high speed' solid state company,
which states in its literature about bandwidth into the MHz range that was a consequence of the company's interest in thermal stability of components
at specific junctions in their circuit, if recollection serves. IOW, MHz frequencies are also well-away from the audio pass band and it met design objectives.

I think the only thing I've heard about that is a couple of manufacturers are up sampling and running their clocks into the 10MHz range because it's both an integer of 44.1 and 48k. This way they don't need 2 clocking systems.

 

[c1ferrari]

^^^ Reasonable and let's hope...well-implemented

 

[rbbert]

I'm not so sure we necessarily need new digital formats, what is needed is better A>D conversion and better matching of A>D (professional recording) and later D>A (at listener playback). JMO of course.

 

[Andre Marc]

we don't yet know whether 4xdsd will become a format for consumers, and a step forward. I've not heard it yet myself. I've spoken to people who have that report some degree of improvement. it seems the step forward from 2xdsd to 4xdsd is smaller and less clear then PCM->dsd or dsd -> 2xdsd. if Bruce says he cannot hear an improvement then that's a new data point for me.

my opinion is that digital music needs to keep pushing the envelope for improvement. will the next step be higher level dsd, higher level PCM, or some currently unknown digital format?

it will take a crystal ball to know for sure; for now I am looking forward in the short term to 4xdsd and it's maturation with hardware and content to see whether it is another step forward. my mind and ears are open to it.

no doubt in my mind that digital transfers are not transparent to the source at this time. I hope that I see that change.

"we don't yet know whether 4xdsd will become a format for consumers, and a step forward?"

What consumers are you referring to? What is left of the music buying public at large, where AAC files from iTunes and CDs are still the dominant format? Or
relatively small percentage of audiophiles who consume music on a regular basis? Are you applying this statement to a small number of classical recordings and
esoteric drivel often seen in newer formats?

On top of that Billboard did a recent study and those who identify them selves as audiophiles purchase far, far less music than those who don't, and attend far fewer live events.

And let's get real. 99% of all popular recordings are done 24/44/1, 24/48, or 24/96.

I have heard DSD128 and DSD256 on Uber DACs and there is utterly none or a miniscule difference where any phantom advantage is wiped out by cost and file size.

I have also spoken extensively to a few label owners who offer files in PCM, DSD64/128/256 and they said 128 was absolutely the sweet spot with quad DSD oferring no
benefit they found audible. Bruce is spot on.

 

[amirm]

my opinion is that digital music needs to keep pushing the envelope for improvement. will the next step be higher level dsd, higher level PCM, or some currently unknown digital format?

Here is the thing about sampling rates Mike. They need to be high enough but no more. High speed electronics is the enemy of ultimate fidelity. We want the digital aspects of the design to run as quietly as they can. Higher speeds require higher slew rates and that in turn requires higher current pulses. These more powerful pulses transmit through the air, over the power line, etc. and can bleed into other parts of the circuit. And all else being equal a slower DAC will always be more accurate than a faster one. My oscilloscope has 4 Ghz Analog to digital converter. 44.1 Khz audio runs at 0.00004 Ghz. The price for that is that the ADC in my scope runs at just 8 bits. Not 16 bits and forget about 24 bits.

it will take a crystal ball to know for sure; for now I am looking forward in the short term to 4xdsd and it's maturation with hardware and content to see whether it is another step forward. my mind and ears are open to it.

no doubt in my mind that digital transfers are not transparent to the source at this time. I hope that I see that change.

Thanks for saying "in my mind."

 

[DaveyF]

Here is the thing about sampling rates Mike. They need to be high enough but no more. High speed electronics is the enemy of ultimate fidelity. We want the digital aspects of the design to run as quietly as they can. Higher speeds require higher slew rates and that in turn requires higher current pulses. These more powerful pulses transmit through the air, over the power line, etc. and can bleed into other parts of the circuit. And all else being equal a slower DAC will always be more accurate than a faster one. My oscilloscope has 4 Ghz Analog to digital converter. 44.1 Khz audio runs at 0.00004 Ghz. The price for that is that the ADC in my scope runs at just 8 bits. Not 16 bits and forget about 24 bits.

Amir, are you saying that the lower sampling DAC's can sound better? My old EAD 7000Mk3 samples at 8x only...and yet it still sounds pretty darn good ( Particularly when in the HDCD decode mode). Many have commented that it has a sort of analog way about it. Perhaps NOT as revealing as some of the newer DAC's but still sounds incredibly decent given its age.

 

[c1ferrari]

And all else being equal a slower DAC will always be more accurate than a faster one.

Really, Amir

 

[Bruce B]

I'm not so sure we necessarily need new digital formats, what is needed is better A>D conversion and better matching of A>D (professional recording) and later D>A (at listener playback). JMO of course.

+1. Totally agree!

 

[Phelonious Ponk]

What we need is better mastering. Then we can talk about formats. Until then, it's a moot point. A well-mastered medium-resolution MP3 will beat a crap (pretty much the standard) master formatted at hi-res (or vinyl or tape) all day, every day.

Tim

 

[rockitman]

I think I have said before mastering is probably 99% of what makes any recording sound good in any genre or format, mastering is where we make that all important illusion attempt.

Problem is, that can't be fixed unless the record co.'s see or acknowledge that and redo all the classics w/o compession, ect...thus if you like to listen to music from the 50's-90's...your screwed if you listen to digital for the most part...

 

[rbbert]

Except for pop/rock (including modern country, hip-hop, etc) and more recently most jazz, digital recordings are often not excessively compressed. I have a ton of music from the last 25 years, and outside of those categories there is much that is well mastered and sounds very good. One could start with most of the ECM catalog, for example.

 

[rockitman]

Except for pop/rock (including modern country, hip-hop, etc) and more recently most jazz, digital recordings are often not excessively compressed. I have a ton of music from the last 25 years, and outside of those categories there is much that is well mastered and sounds very good. One could start with most of the ECM catalog, for example.

I agree...genre and time period dependent...

 

[Fitzcaraldo215]

...

my opinion is that digital music needs to keep pushing the envelope for improvement. will the next step be higher level dsd, higher level PCM, or some currently unknown digital format?

...

no doubt in my mind that digital transfers are not transparent to the source at this time. I hope that I see that change.

If we can believe the advance marketing campaign, then the answer might be MQA. I do expect it to be huge, if for no other reason than its efficient compression. But, the concept it embodies of correcting the filter ringing of both the original a-d and the final d-a might well bring us a higher level of transparency. Time will tell.

 

[Phelonious Ponk]

I think I have said before mastering is probably 99% of what makes any recording sound good in any genre or format, mastering is where we make that all important illusion attempt.

Oh I can't go along with that one. The recording itself, and the mixing are, if anything, more important than the mastering, though truly bad mastering can screw up both of them. Fortunately, as rbert says, it's pretty genre-dependent. The loudness wars have impacted folk, Americana, even jazz, but more often than not, they haven't ruined it. There are plenty of compressed recordings that still sound good. I'm not just looking for a de-escalation of the loudness wars, I'm looking for great recording, mixing and mastering techniques that make the most of the current technology. There are some out there, but outside of classical, they are too few and far between. I just disagree that mainstream recordings were fabulous in the vinyl age. There is plenty of mainstream rock from the 60s, 70s and 80s that doesn't sound very good, if for no other reason than that multitrack technology wasn't up to the task that many of the densely layered recordings of that day demanded of it. Most of the truly great analog recordings were made before the mid 60s.

Tim