Avoid the snake oil

Because they were sending the identical digital sound files to a competently designed DAC (Benchmark).

(If digital signals didn't work reliably in that fashion, my industry - digital post production sound - would be sh*t out of luck as we are constantly sending our digital sound files via all sorts of servers).

You are leaving out all sorts of variables that can affect the SQ of servers. Normal PC acting as a server meaning the MB has a built- in video and audio card? Running Windows or Linux? SMPS or linear power supply?
 
You are leaving out all sorts of variables that can affect the SQ of servers. Normal PC acting as a server meaning the MB has a built- in video and audio card? Running Windows or Linux? SMPS or linear power supply?

I went from using the Apple Airport Express as my music server to the Raspberry Pi, both via S/PDIF to my Benchmark DAC 2L.

Strangely this forum won't allow links that I supplied. But you can look up measurements of the Apple Airport Express by kenrockwell and Stereophile, as well as the Raspberry Pi as streamer measured at Archimago's Musings.

But the upshot is, as I mentioned, both are sending identical digital sound files via a S/PDIF cable to a DAC competently designed to deal with any possible spuria, e.g. jitter.

There is no reason to expect audible differences.

As to the other scenarios you mentioned, I'm aware plenty of audiophiles think practically anything affects sound quality. I'm also aware that we are very good at imagining things. I have personal experience of this :-)

And, as I mentioned, in the pro audio world our files are flying around the world via the internet, sucked up by all sorts of different servers and set ups, gawd-knows-what-cables, people are using internal drives, external drives, you name it, and there are non of these fussy audiophile problems. Because the sonic quality is found in the sound files - digital audio works reliably (in this sense) by it's very nature.
 
Your last paragraph was all I needed to read to understand your simplistic outlook on the role servers play in digital audio.
 
Your last paragraph was all I needed to read to understand your simplistic outlook on the role servers play in digital audio.

Can you explain why one would expect audible differences to occur between digital audio files sent by the two servers I mentioned, used how I described?

That would be a start, in showing where I've gone so wrong.
 
I went from using the Apple Airport Express as my music server to the Raspberry Pi, both via S/PDIF to my Benchmark DAC 2L.

Strangely this forum won't allow links that I supplied. But you can look up measurements of the Apple Airport Express by kenrockwell and Stereophile, as well as the Raspberry Pi as streamer measured at Archimago's Musings.

But the upshot is, as I mentioned, both are sending identical digital sound files via a S/PDIF cable to a DAC competently designed to deal with any possible spuria, e.g. jitter.

There is no reason to expect audible differences.

As to the other scenarios you mentioned, I'm aware plenty of audiophiles think practically anything affects sound quality. I'm also aware that we are very good at imagining things. I have personal experience of this :-)

And, as I mentioned, in the pro audio world our files are flying around the world via the internet, sucked up by all sorts of different servers and set ups, gawd-knows-what-cables, people are using internal drives, external drives, you name it, and there are non of these fussy audiophile problems. Because the sonic quality is found in the sound files - digital audio works reliably (in this sense) by it's very nature.

You need 10 total posts to post links.

Servers can sound dramatically different using the same file. Power supplies, designs which limit noise and even the software itself, all have an impact.


Sent from my iPhone using Tapatalk Pro
 
You need 10 total posts to post links.

Thanks for that info Mike!

(If the software alters the sound file in any way, sure that can affect the sound, but aside from that...)

Servers can sound dramatically different using the same file. Power supplies, designs which limit noise and even the software itself, all have an impact.

If that were the case this should be a huge problem in my industry: there is little consistency in the way the files are being transferred in the sense that many audiophiles would demand - all sorts of bog standard cabling, different CPUs/Hardware, different peer to peer file sharing, the "problems" which would add up to alterations of the sound of the files would go on and on.

And yet there are NOT these "dramatically different" changes in sound when we do our work this way. If there were the system just wouldn't work. I do sound design and I am often meticulously balancing a great number of tracks so you can *just barely hear* X or Y elements vs others. I literally have to match the sound of the "air" recorded in a room for dialogue tracks. If this timbre changes, it's a problem.

But despite the relatively haphazard combinations of technology (to the audiophile anyway) that all these sounds fly through, we RELIABLY get the same sound. In the mixing theater it sounds just as I created it. If another editor or mixer and I are sending various edited versions of sessions and audio files back and forth to our work stations...they sound the same. Because digital audio works reliably (so long as some minimum technological criteria are satisfied - e.g. to spec equipment).

You may indeed know something I don't know on the issue. However I have to factor in to any such claims of "dramatic differences" that audiophiles, like any other human being, are subject to biases and imagination. Therefore I'd like to see more than anecdote or speculation. In regard to the elements you mentioned, what measurable changes in the audio files are we talking about, and how was it determined they are audible (hopefully controlled for bias).

Or just take the example I gave of my own situation. Can you explain why there could have been audible differences between the files sent via the Apple Airport Express vs the Raspberry Pi, via S/PDIF, to a Benchmark DAC 2L?
 
What's the total cost of the two servers you compared?
 
Thanks for that info Mike!

(If the software alters the sound file in any way, sure that can affect the sound, but aside from that...)



If that were the case this should be a huge problem in my industry: there is little consistency in the way the files are being transferred in the sense that many audiophiles would demand - all sorts of bog standard cabling, different CPUs/Hardware, different peer to peer file sharing, the "problems" which would add up to alterations of the sound of the files would go on and on.

And yet there are NOT these "dramatically different" changes in sound when we do our work this way. If there were the system just wouldn't work. I do sound design and I am often meticulously balancing a great number of tracks so you can *just barely hear* X or Y elements vs others. I literally have to match the sound of the "air" recorded in a room for dialogue tracks. If this timbre changes, it's a problem.

But despite the relatively haphazard combinations of technology (to the audiophile anyway) that all these sounds fly through, we RELIABLY get the same sound. In the mixing theater it sounds just as I created it. If another editor or mixer and I are sending various edited versions of sessions and audio files back and forth to our work stations...they sound the same. Because digital audio works reliably (so long as some minimum technological criteria are satisfied - e.g. to spec equipment).

You may indeed know something I don't know on the issue. However I have to factor in to any such claims of "dramatic differences" that audiophiles, like any other human being, are subject to biases and imagination. Therefore I'd like to see more than anecdote or speculation. In regard to the elements you mentioned, what measurable changes in the audio files are we talking about, and how was it determined they are audible (hopefully controlled for bias).

Or just take the example I gave of my own situation. Can you explain why there could have been audible differences between the files sent via the Apple Airport Express vs the Raspberry Pi, via S/PDIF, to a Benchmark DAC 2L?

I’ve just finished a late night install. I’ll respond later, but for now, I wouldn’t expect there to be much difference between a Raspberry Pi and an Airport Express. But no serious audiophile uses either of those. Look up Taiko Extreme as an example.

Heavily graphic intensive software like Roon which is built on a P2P network needs a powerful CPU (and therefore noisy) and that degrades the sound.

I’m glad you’re asking these questions. It shows that someone on the other side is interested in learning. That’s awesome!

I would ask you, assuming you’re not using tape (which is disappointing if you’re not), what A2D are you using? Do you use compression sparingly or quite liberally? What other modifications in the song are you doing using ProTools?

I would love to learn how the production and mastering side could be improved (in your opinion). If the song is badly recorded, admittedly, there is only so much we can do on the other end. Lipstick on a pig if you will.


Sent from my iPhone using Tapatalk Pro
 
I used to use a high-end laptop for my digital music. I then built a custom desktop with a low ripple power supply and other handpicked components and no video card. The difference is sound was very noticeable with the desktop sounding much better. It was like a veil was lifted. I would expect that different servers can sound different.
 
Thanks for that info Mike!

(If the software alters the sound file in any way, sure that can affect the sound, but aside from that...)

If that were the case this should be a huge problem in my industry: there is little consistency in the way the files are being transferred ...

I don't think anyone here is questioning the integrity (i.e., bit for bit identical) of the music files; it is how those files are then transferred to the DAC and the resulting analog sound, with oh so many variables on that path.

Although a year or two ago Kuoppis posted a link to a publication from CERN that did indeed question the integrity of digital data transfer for their huge files; although it was very small and not likely to occur with files the size of audio (or even typical home video), when the file size is hundreds of TB there may be a handful of errors that occur during transfers.
 
I’ve just finished a late night install. I’ll respond later, but for now, I wouldn’t expect there to be much difference between a Raspberry Pi and an Airport Express. But no serious audiophile uses either of those.

I'm a serious audiophile and I used them, so I'm afraid that claim is incorrect. (Unless you want to go full No True Scotsman fallacy).

(BTW I see the MBLs in your posts. I'm a big fan of MBLs, even owned the MBL 121s for many years!)

Look up Taiko Extreme as an example.

Am I too assume that "Much More Expensive Is Better?" Should I simply buy in to the marketting claims of such an outrageously priced device?

There have been numerous very highly priced items, including digital devices, measured by Stereophile and others that have been either embarrassing given the claims made for the device, or which have been handily beaten by much cheaper gear. Take for instance the recently reviewed (on Stereophile) CH Precision D1.5 SACD/CD player/transport, which purports to go to heroic efforts to reduce any forms of distortion, yet which is handily outperformed by vastly cheaper Benchmark gear.

So, again, I'm not looking for marketing speak or anecdotes. Can you point me to where the Taiko Extreme has been shown to be measurably superior to cheaper, competently built streamers? Then we could address whether any audible claims are even credible.


Heavily graphic intensive software like Roon which is built on a P2P network needs a powerful CPU (and therefore noisy) and that degrades the sound.

What's the evidence for that claim? Does it require more CPU than for, say, a Pro Tools system, either desktop or used in mixing theaters? In decades of using such systems I've yet to meet anyone in post production sound, nor anyone from Pro Tools, express the claim you have made about a powerful CPU "degrading sound." Can you understand why I might have grounds for skepticism before accepting these claims?

I’m glad you’re asking these questions. It shows that someone on the other side is interested in learning.

I'm always interested in learning, especially that I may be wrong or may be misinformed. However, I also have justified skepticism about claims in the audiophile world because there has been a tremendous amount of pseudo-science promoted and believed.

I would ask you, assuming you’re not using tape (which is disappointing if you’re not),

???

I started in the business when tape was the norm. I've never been more glad to leave a technology behind.

what A2D are you using? Do you use compression sparingly or quite liberally? What other modifications in the song are you doing using ProTools?

I would love to learn how the production and mastering side could be improved (in your opinion). If the song is badly recorded, admittedly, there is only so much we can do on the other end. Lipstick on a pig if you will.

The questions you are asking seem to assume music production, and would be a whole other topic (and one better answered by someone in music production). I work in film sound post production. And we certainly care about sound quality. The point being that if all the things you seem to believe affected sound quality were true, it would imply we'd be having problems in digital sound production that we are simply not having.
 
What's the total cost of the two servers you compared?

What a strange question, IMO. It doesn't matter what a device cost: what matters is how well it does it's job. And there are functions for which less expensive devices are perfectly adequate (e.g. sending digital information reliably).

I mentioned the devices - Apple Airport Express as my music server to the Raspberry Pi (3), both via S/PDIF to my Benchmark DAC 2L.

Again: can you give a reason beyond marketing claims, speculation or anecdote for why I might expect one to have sounded different from the other? In other words, any objective evidence?

Or for instance why your music server (not your DAC) would produce a measurably/audibly different signal from the same digital files, vs say my Raspberry Pi?
 
I don't think anyone here is questioning the integrity (i.e., bit for bit identical) of the music files; it is how those files are then transferred to the DAC and the resulting analog sound, with oh so many variables on that path.

If they play the same digital files, it's the same file. The sound is encoded in the file.

Again, I can send a sound file over Ethernet cables, internet provider cables, through the air, to any number of FTP servers, which go on to other work stations through all sorts of "variables" and it can be sent back to me through all those "variables" and it remains the same sound file (sonically identical and zero change in waveform, for instance). It plays back the same. Because that's basically how digital audio has been designed to work. If there were truly the amount of corruption happening that many audiophiles believe, we'd be having all sorts of problems that we do not have. Think of Tidal - the sound files are tiny pieces, coming in packets, flying all over the internet, sometimes even arriving at different times, yet because tcp ip error correction WORKS it ends up being the same file to be played via your DAC. A DAC can in principle change the sound depending on the design, but it's being fed the same files. It would seem Tidal or other streaming services couldn't even work if the amount of "variables" being suggested alter the integrity/sound of the files.

If my Benchmark DAC is sent a digital sound file from my desktop computer, or my raspberry pi or whatever, it has received and is therefore playing the same sound file, hence the same sound.

Although a year or two ago Kuoppis posted a link to a publication from CERN that did indeed question the integrity of digital data transfer for their huge files; although it was very small and not likely to occur with files the size of audio (or even typical home video), when the file size is hundreds of TB there may be a handful of errors that occur during transfers.

Yet another reason why I await some actual objective evidence for the type of audible claims being made here.

Cheers.
 
Oh dear Matt, you have so much to learn young grasshopper. You can hear what you can measure, but you can’t measure what you can hear (tone, depth of soundstage, instrument separation, etc).

Feel free to drop by the store with your Apple device and I’ll use the Taiko for comparison purposes. I’ll let you hear the difference.

Im far too busy to argue with meter readers.


Sent from my iPhone using Tapatalk Pro
 
An interesting topic is the rapid rise in “vloggers”. Are they a threat to print media? Will we see print media have to jump on board into the vlogger world?

Someone sent me this video the other day and my reaction was “who the F are these guys?”

Danny at GR-Research Is Elitist and Delusional - YouTube

60,000 subscribers.


Sent from my iPhone using Tapatalk Pro

Yea Danny is bigger and more influential than PF ..! He actually measures and listen to audio gear ..!
 
Oh dear Matt, you have so much to learn young grasshopper. You can hear what you can measure, but you can’t measure what you can hear (tone, depth of soundstage, instrument separation, etc).

Feel free to drop by the store with your Apple device and I’ll use the Taiko for comparison purposes. I’ll let you hear the difference.

Im far too busy to argue with meter readers.


Sent from my iPhone using Tapatalk Pro

While I appreciate your kind offer, given you are eschewing any objective evidence, your reply seems to fall in to the usual Golden Ears response. "You can't measure what I can hear!"

The Golden Ears approach to audio (where we have utter confidence in our subjective inferences over any objective verification) doesn't allow any way to find out you are wrong. After all, if I come over and compared an Apple Server to the Taiko and detected no difference, you can always claim to hear a difference. Hence you would be likely to ignore my negative result as suggesting you may have been fooling yourself as a data point and just put it down to your ears being better than mine.

As to:

You can hear what you can measure, but you can’t measure what you can hear (tone, depth of soundstage, instrument separation, etc).

That is too vague and hence misleading. We can hear *some* things that we can measure, but not all. The reason why we often invent and use measurement devices is to EXTEND beyond what our senses can detect.
I hope you have a CO2 monitor in your house! You won't be seeing X-rays any time soon, but instruments can detect those. You also can't hear certain levels of distortion that can be measured (and depending how old you are, you likely can't hear above a certain frequency response that can easily be detected with instruments).

It's very important to acknowledge that we can measure things we can not hear. Because sometimes...very rarely in high end...a company will say 'look, our product made a measurable difference' but that doesn't automatically entail it is an AUDIBLE difference.

As to things we can't measure: first of all, we certainly can tell you that if two sound files are identical and the output signal measures precisely the same you are NOT in all likelihood hearing any differences - timbre, soundstage or otherwise. Measurements matter.

Further, of course tone can be measured! It's one of the most salient characteristics of sound - the frequency response of a loudspeaker (or EQ settings etc) will reliably affect the tone. I manipulate tone and timbre of sounds all day long in my work! Further: The overtones and harmonics that distinguish instruments and their timbre are well documented and measurable!
(Ever wonder how synth designers emulate different instruments if the character of tone/timbre weren't understood in measurable terms?)

As for soundstage depth and instrument separation, there is plenty known about what measurable aspects can affect those impressions - e.g. room acoustics, speaker design, etc. (I play with soundstage depth and width all the time in my listening room, via altering the acoustics). And in terms of the actual sound source: Those very properties are manipulated all the time by people creating sound (like me) and by mixers. Tons of plug-ins, EQ, reverbs, transient shapers etc work to modulate the depth and width of imaging in a mix - all of course had to be built upon measurable, predictable sonic characteristics to even work. It's not Voodoo. (There are even tutorials from mixers on youtube explaining how you can manipulate the soundstage and depth with various tools).

When people say things like "science can't measure X" often what they really mean is "I have no idea how X is measured." It's a projection of one's ignorance in that sense. I don't know about your technical knowledge so I'm not automatically assuming this of you, but...perhaps you may have a lot to learn as well on the topics you are mentioning? ;-)

I am not, btw, claiming to be anything like a technical expert myself on all these matters. Certainly not!

But I do know enough to recognize legitimate reasons for skepticism, and when I'm seeing the type of responses that are more-of-the-same in terms of a Golden Ear type response. (Assertions and anecdote about "what people hear" without additional technical/measurable corroborative evidence, or a dismissal of the relevance of measurements to one's claim).
 
While I appreciate your kind offer, given you are eschewing any objective evidence, your reply seems to fall in to the usual Golden Ears response. "You can't measure what I can hear!"

The Golden Ears approach to audio (where we have utter confidence in our subjective inferences over any objective verification) doesn't allow any way to find out you are wrong. After all, if I come over and compared an Apple Server to the Taiko and detected no difference, you can always claim to hear a difference. Hence you would be likely to ignore my negative result as suggesting you may have been fooling yourself as a data point and just put it down to your ears being better than mine.

As to:

You can hear what you can measure, but you can’t measure what you can hear (tone, depth of soundstage, instrument separation, etc).

That is too vague and hence misleading. We can hear *some* things that we can measure, but not all. The reason why we often invent and use measurement devices is to EXTEND beyond what our senses can detect.
I hope you have a CO2 monitor in your house! You won't be seeing X-rays any time soon, but instruments can detect those. You also can't hear certain levels of distortion that can be measured (and depending how old you are, you likely can't hear above a certain frequency response that can easily be detected with instruments).

It's very important to acknowledge that we can measure things we can not hear. Because sometimes...very rarely in high end...a company will say 'look, our product made a measurable difference' but that doesn't automatically entail it is an AUDIBLE difference.

As to things we can't measure: first of all, we certainly can tell you that if two sound files are identical and the output signal measures precisely the same you are NOT in all likelihood hearing any differences - timbre, soundstage or otherwise. Measurements matter.

Further, of course tone can be measured! It's one of the most salient characteristics of sound - the frequency response of a loudspeaker (or EQ settings etc) will reliably affect the tone. I manipulate tone and timbre of sounds all day long in my work! Further: The overtones and harmonics that distinguish instruments and their timbre are well documented and measurable!
(Ever wonder how synth designers emulate different instruments if the character of tone/timbre weren't understood in measurable terms?)

As for soundstage depth and instrument separation, there is plenty known about what measurable aspects can affect those impressions - e.g. room acoustics, speaker design, etc. (I play with soundstage depth and width all the time in my listening room, via altering the acoustics). And in terms of the actual sound source: Those very properties are manipulated all the time by people creating sound (like me) and by mixers. Tons of plug-ins, EQ, reverbs, transient shapers etc work to modulate the depth and width of imaging in a mix - all of course had to be built upon measurable, predictable sonic characteristics to even work. It's not Voodoo. (There are even tutorials from mixers on youtube explaining how you can manipulate the soundstage and depth with various tools).

When people say things like "science can't measure X" often what they really mean is "I have no idea how X is measured." It's a projection of one's ignorance in that sense. I don't know about your technical knowledge so I'm not automatically assuming this of you, but...perhaps you may have a lot to learn as well on the topics you are mentioning? ;-)

I am not, btw, claiming to be anything like a technical expert myself on all these matters. Certainly not!

But I do know enough to recognize legitimate reasons for skepticism, and when I'm seeing the type of responses that are more-of-the-same in terms of a Golden Ear type response. (Assertions and anecdote about "what people hear" without additional technical/measurable corroborative evidence, or a dismissal of the relevance of measurements to one's claim).

If you prefer to look at meters which only tell part of the story (see Nelson Pass’s comments) rather than trust your ears, than may I suggest:

Audio Science Review (ASR) Forum

For what actual engineers say:

“If it measures good and sounds bad, — it is bad. If it sounds good and measures bad, — you've measured the wrong thing.”

Here’s how good relying on measurements can be (they measured incorrectly and slammed a product for their bad measurements!!!):

EarMen Tradutto DAC - Good Measurements Gone Wrong - SoundStage! Real Hi-Fi (Ep:29) - YouTube

Measurements vs listening - YouTube

Podcast: Beware the measurements! | Darko.Audio

Meet the Stellar Phono Preamp – PS Audio

If it quacks like a duck… – PS Audio


Sent from my iPhone using Tapatalk Pro
 
Back
Top