An interesting argument against sampling frequencies above 96kHz

[gvl]

...In any event, if only the most trained listeners have tremendous difficulty being able to name which is the 96k and which is the 44.1k then what about telling the difference between 96 and 192? There is no evidence that anyone can do this at all...

The point I was trying to make that was discarded based on "experiences" that most modern DA converters will upsample/oversample (call it however you want) the material one way and another blurring the differences between the lower and higher resolutions, which makes it more difficult to detect the differences. The kind of feedback I received to my comment clearly indicates that many don't have a slightest clue what happens inside their DACs and their "experiences" are not what they think they are. It would be interesting to see a study that is performed on a non-oversampling multi-bit DAC that doesn't have any digital filtering built-in.

 

[E-Stat]

I continue to look at high resolution designed for computer audio playback - and will continue to do so - but so far I have not been overly impressed by the sound - perhaps - again - it is their analog output stage that is the culprit and not the technology. I have certainly not heard all the Computer Audio DACs out there. It only takes one.

Audio shows are less than optimum for multiple reasons: hotel room acoustics, lack of fine tuned room treatments, AC quality and the inability to use isolated network computer connectivity for computer based playback. The last thing I want in my listening room is a general purpose computer - even a laptop which today typically uses a high powered quad core Intel i7 processor.

I've found the ideal arrangement is to place the music server elsewhere - in my case the office on the main floor. It's RFI/EMI emissions are completely isolated from the listening room located upstairs - driven by dedicated AC lines sourced from a different circuit box than the main floor. All you find in the listening room is the tiny microRendu player connected via CAT7 and sourced by a power supply that is galvanically isolated from the AC.

 

[RGA]

Audio shows are less than optimum for multiple reasons: hotel room acoustics, lack of fine tuned room treatments, AC quality and the inability to use isolated network computer connectivity for computer based playback. The last thing I want in my listening room is a general purpose computer - even a laptop which today typically uses a high powered quad core Intel i7 processor.

I've found the ideal arrangement is to place the music server elsewhere - in my case the office on the main floor. It's RFI/EMI emissions are completely isolated from the listening room located upstairs - driven by dedicated AC lines sourced from a different circuit box than the main floor. All you find in the listening room is the tiny microRendu player connected via CAT7 and sourced by a power supply that is galvanically isolated from the AC.

You've mentioned this a few times - but again - the CD transport has to sit right beside the amplifier and or DAC and suffers no issues - AND the problem in the playback is NOT RF noise in the line - it's a lack of tone dimension and sounding 'fake" - if it was noise and or hum I could "listen that out" - and I am not thrilled with the excuse because if this is a problem the dealer should put a massive shield around the computer - or place it at the other end of the room. It is their job to demonstrate not convince you to spend the money first and then have you "experiment" on their "word".

And I agree that shows are not ideal. Nevertheless all the rooms are in the same hotel in an even playing field. I merely cover the room at the particular show - I have had rooms score bottom 5 at one show and wound up being very best at another - quite the difference.

I have zero "noise" issues during computer audio playback. Perhaps try an ARC CD6 (as ARC uses the same Philips Pro 2 transport mechanism as the AN CD Two/II and up) as a transport with an Audio Note DAC or Border Patrol (as AN rubs you the wrong way) and compare to your CA and see what happens.

 

[RGA]

Just taking your argument further, Just because a designer can build a toob amp, does not mean they can build an analog output stage. Both can happen and thus your point is?

Generalizations always end up being one offs and not true in actuality. That is the nature of this hobby.

The bolded point, in your post, is not objective but subjective and your whole post is that way. Nothing wrong with that, mind you, but we are trying to discuss objectively. Jes saying.

There is plenty of information out on the idea that tones, beyond out hearing range, CAN in fact affect the way we hear things.

Well I am not a big fan of toobs or SS or anything else - I am a fan of what sounds good. Ah Tjoeb! was one of the first tube CD players I heard and I kinda hated it. Easter Electric made a SS and tube DAC and many/most preferred the SS section including me.

Most tube CD players simply stick some tubes after some OPamps and the tube serves as a buffer to basically add a bunch of distortion to the signal. None of this applies to Audio Note or Border Patrol CD players which are rather entirely different animals.

Objectively we can run a DBT and all one needs to do is go to a local university with a music department and gather up 30 classical and or jazz trained musicians under the age of 40 (likely to still have good hearing) and take out a few hotel rooms and give all listeners a card - they listen to two or three rooms and drop their card in a box of the room they felt recreated music properly and add up the scores. The systems are blacked out and both rooms played at the same volume.

It's why when faced with measurements arguments I point out that no feedback amplifiers tend to ALWAYS win. Heck Martin Colloms had top SS designers bring their top of the line SS amplifiers (like Meridian and Naim etc) and they included a low feedback tube amp and ALL the SS designers chose the tube amp over ALL the SS amps including the ones they designed and were selling!

And it's not just about tubes - the Sugden A21a has also been in several such tests and came out on top and it measures similar to a SET. Class A.

The point is - and John Atkinson makes it all the time - the measurements (of the best kind) don't equate correlationally as sounding better to gear of the poor measuring variety.

That's why I tend to always bring up Audio Note CD players because they are typically the worst measuring in the entire audio industry (try to find worse). No jitter reduction of any kind, no error correction, no clock matching, fairly high distortion, no over/upsampling etc etc. And consistently picked (even by engineers and people who worked to design SACD) and Recording and Mastering engineers like Steve Hoffman (Beatles, Doors, Dylan, Pink Floyd, Eva Cassidy, Miles Davis, John Coltrane, Frank Sinatra --- ahh not a no name hack) has an AN DAC. (and a PS Audio for that matter).

This is why it is fascinating - these two would be "completely" at opposite ends of the audio spectrum in terms of design and measurements. So I go back to the listening audition because if AN wins award after award after award and wins blind level matched session all the time with the WORST measurements then it indicates that measurements are not giving you jack squat knowledge as to whether it will sound better.

And on top of those crappy CD measurements - you feed that into a SET amp for even more worse measurements into "mediocre" speaker measurements.

PS I started out as a Bryston/Krell/PMC/Genelec/B&W sorta guy. Bomb proof measurements. Meh.

 

[E-Stat]

You've mentioned this a few times - but again - the CD transport has to sit right beside the amplifier and or DAC and suffers no issues - AND the problem in the playback is NOT RF noise in the line - it's a lack of tone dimension and sounding 'fake"

I disagree with your assessment. There is most certainly FAR more RF noise generated by a general purpose computer than a CD-ROM transport.

and I am not thrilled with the excuse because if this is a problem the dealer should put a massive shield around the computer

Which would have no effect on the digital noise spewed back into the mains.

I have zero "noise" issues during computer audio playback.

In my experience, RFI/EMI noise affects the performance of connected gear. It's not a matter of hearing "hiss". In my system, just replacing an otherwise nice ($400 ) linear power supply with an Uptone LPS-1 ultra capacitor unit which is galvanically isolated from the AC improved focus and dynamics markedly in my system.

Maybe someday, dealers will use truly isolated, networked systems as you find in professional circles using DAWs.

 

[RGA]

This is all well and good E-stat - but it's the manufacturer's JOB to demonstrate their gear properly then isn't it? It's pretty stupid to expect someone to spend $30,000 on a Computer Audio DAC and then tell them ÿes sir we know this sounds utterly shitty but if you just put your computer 1000 feet from the stereo system then "we promise" it will sound better than the competitor's DAC that was actually demonstrated to sound better for 1/10th the price.

It's not unlike students who tell me their homework was done perfectly but it is at home - maybe but I can't take their word for it.

Do you know how many times a Magnepan fan will tell me that "oh you only heard the stock Magnepan - you need to hear a modified one done by company XYZ" - wait what? I have to buy a speaker that comes to market as an inferior pile of rubbish and the only way to make it sound good is to ship it to a company in Canada specializing in mods to then FIX the thing. And these people get irritated at me for not doing all of this in order to evaluate the thing.

Here's a plan and advice to computer audio DAC makers who believe the computer is shoddy - BUILD your own computer dedicated entirely to computer audio playback that you can sit beside your stereo because this is how it "oughta" be. Your stereo and you should be in the same room and achieve quality sound. And if you really must have the computer 1000 feet from your stereo then pony up the cash and take out a second hotel room at an audio show and put your computer in the hotel room at the other end.

Otherwise all you have is a promise from a company advertising agent and a few internet posters who can make unsubstantiated claims.

At the end of the day the manufacturer and dealer have to be able to demonstrate their quality first. They do that through audio shows, dealers, directing people to owners so I can hear it at their homes (which I have done many times).

Maybe I am too old school in wanting companies to demonstrate their product sounds excellent rather than selling me words "buy first and we'll tell you how to make it sound good - just trust us."

 

[E-Stat]

This is all well and good E-stat - but it's the manufacturer's JOB to demonstrate their gear properly then isn't it? It's pretty stupid to expect someone to spend $30,000 on a Computer Audio DAC

Question if you please - what exactly is a "computer audio DAC"? Do you mean USB DAC? Like one of those tiny Dragonfly units made by Audioquest that derives power from the USB? They run under $200.

As for me, I use "plain old" DACs. Those which connect to whatever transport you want using standard digital inputs like you find on your beloved brand Audionote DAC5. I can use their transport with my DACs or - use their DAC with my streamers. Having trouble understanding the distinction you're drawing.

The only DAC I know of in the $30k range is from dcs called the Vivaldi. Is that a "computer audio DAC"? Like that of mine or those from Audionote, it offers multiple digital inputs. They make no mention of it being a "computer audio DAC". Would you care to illustrate what a $30,000 Computer Audio DAC" looks like?

The only computers I've seen in show pictures are laptops from companies like Apple and HP. Those providers couldn't care less about high end audio.

...but if you just put your computer 1000 feet from the stereo system

Gee, I don't know anyone whose home is two-tenths of a mile long. Most certainly not mine!

Here's a plan and advice to computer audio DAC makers who believe the computer is shoddy - BUILD your own computer dedicated entirely to computer audio playback that you can sit beside your stereo because this is how it "oughta" be.

The highest performance computers aren't "shoddy" - they just radiate lots of RFI/EMI. There are companies like Aurender who make expensive "music players" with built in disk drives, but those still have the noisy bits in the same box. And I find that approach unnecessary when you apply common sense.

Your stereo and you should be in the same room and achieve quality sound

My primary audio system resides upstairs - only the media lives downstairs in the office on the media server. If you refer to using remotely networked content, why the arbitrary limitation? My now unused CD and SACD collection lives in a closet off the listening room. And I'm most certainly in the room while listening to it!

Otherwise all you have is a promise from a company advertising agent and a few internet posters who can make unsubstantiated claims.

The world of audio does not exist solely at shows in hotel rooms. Perhaps you should get out and listen more to systems used in actual environments using room treatments, dedicated power lines, lack of fluorescent fixtures, dimmers, etc.

 

[RGA]

E-STAT I have thought more about what you have said and my beef may have nothing to do with the computer but with the actual DACs. I am referring to the full size DA converters like dCS, MSB, Ayre, PS Audio - the DACs the size of a CD player not the little thumb size things like M2Tech.

Of course if I was not a fan of the DAC with a CD transport connected up to them via SPDIF then there is likely a good chance that I would be no more a fan of said DAC with a USB cable connected up to them. IE; if I don't like the sound of the DAC then what is fed to it whether the best CD on the planet of 24/192 files from a computer the sound is poor.

And of course it will be poorer STILL if as you note there is an abundance of RFI signals making it worse.

So what in fact one should do at an Audio Show is bring their best CDs and judge the sound quality of the DAC - if it sounds good - then purchase whatever that DAC of choice is and if need be buy a USB to S/PDIF converter from the likes of M2 Tech. I know some folks who do this even if their modern DAC has USB inputs because they feel the sound is better with the converter and into the SPDIF input because they feel the SPDIF sounds superior (though this appears to be DAC maker depoendent). The M2 Tech is like $180 so it's not a killer.

 

[E-Stat]

Of course if I was not a fan of the DAC with a CD transport connected up to them via SPDIF then there is likely a good chance that I would be no more a fan of said DAC with a USB cable connected up to them.

One of the challenges with the Universal Serial Bus is that it was not really designed for music. Using your MacAir to feed a DAC uses the onboard USB bus which is also tied to others on the same computer and is inherently noisy. Realizing this, there are reclockers like the Uptone Regen which are said to improve that situation. The designer, John Swenson, took that concept a step further and designed a tiny digital player with an equally small RFI footprint with a single mission - music playback. It incorporates only what was needed and included his Regen circuitry.

He also attacked noise from the AC side as well with the LPS-1 Power supply which generates the nominal 1A required. It is effectively a battery supply with ultra low impedance drive that is galvanically isolated from the AC. You should audition the microRendu / LPS-1 combo. There's actually a further refined unit called the ultraRendu. In this case, it is the DC power cable that can make a difference - for not much money either ($35). Swenson found that using a starquad design like Canare with lower inductance made improvements. For me, it enhanced low level dynamic expression.

I've found that the predominant sonic signature of a DAC has more to do with factors other than the DAC chip(s) themselves: power supply, analog output stage design and implementation. I chose an Audio Research for its class A, zero feedback discrete, balanced JFET output stage with multiple stiff power supplies (R-core tranny for audio) - as opposed to the more common use of high feedback op amps. While it has a USB input which the microRendu requires, I needed a USB to S/PDIF converter and use the 75 ohm BNC input because it does not natively support the stripped down Linux OS of the uRendu.

 

[nyhifihead]

with the technology put into mutec, singxer, and focusrite digital interfaces you can essentially bypass the USB aspect of computer audio. Even a traditional usb - spdif is a great option

audio over IP solutions like the focusrite rednet 3 have been reported to have superior sound than usb based digital interfaces, and I'd believe it as USB over cat5(RJ45) is becoming more popular too with ps audio even offering a model. IP transmission with modern day NICs are actually very noise resistant and galvanically isolated

I use a gustard u12 using AES/spdif coax depending on my dac but I recently had my hands on a singxer su-1 and it is extremely impressive for its price, transparent and black background

 

[doctor fuse]

First of all, understand that your title is not worded correctly. Sampling at 96 kHz does not mean the goal is to reproduce frequencies there. It allows for using a more gradual filter in order to minimize phase distortion and ringing within and just above the audible band.

The value of oversampling is debatable with potential benefits and tradeoffs. One of my DACs allows for choosing to play native or to always upsample to 192 kHz. I choose the native option.

I understand it is not audible frequency, but sampling rates are still "frequency" of a signal, are they not?

 

[E-Stat]

I understand it is not audible frequency, but sampling rates are still "frequency" of a signal, are they not?

At the lower part of the spectrum, yes. Due to Nyquist, however, the highest reproducible frequency is but half of the sample rate.

Hence, the goal of a 96/24 is NOT to reproduce frequencies out to 96 kHz. It is to preserve the full range of instruments using a more gradual "brickwall" filter slope so as to preserve integrity in the time domain.

 

[doctor fuse]

At the lower part of the spectrum, yes. Due to Nyquist, however, the highest reproducible frequency is but half of the sample rate.

Hence, the goal of a 96/24 is NOT to reproduce frequencies out to 96 kHz. It is to preserve the full range of instruments using a more gradual "brickwall" filter slope so as to preserve integrity in the time domain.

Yes, that jives with my [limited] understanding. Putting the upper audible frequency range in the 48kHz range - similar to the upper range of some higher-end phono cartridges, amps, and speakers (my Tannoy dual concentric horn tweeters are rated at 30kHz, at less than +/-3 dB, I believe).
I am probably trying to rationalize my subjective preference for good analog recordings and playback, by using the limited amount of science I actually know (but may not properly understand).

 

[gvl]

Speakers and cartridges is a completely different story, it has little to do with DACs.

 

[E-Stat]

I am probably trying to rationalize my subjective preference for good analog recordings and playback, by using the limited amount of science I actually know (but may not properly understand).

I find that it takes a pretty darn good digital playback system to match some of the virtues of the best analog gear. For me, 96/24 (and higher) combines the strengths of both analog and Redbook digital.