USB audio cracked... finally!

[BigAlMc]

Just now, PeterSt said:

 

Maybe it's air only in there and we forgot something. 

 

haha

 

Brilliant! Air - the ultimate in galvanic isolation. Much better than old school copper 

 

The 'air' is sounding pretty good already. But I'll hold off and give it some time to settle in before commenting further.

[PeterSt]

17 minutes ago, semente said:

Why hasn't anyone come up with a more effective alternative for connecting a Computer to a DAC?

 

A small bit of history, in case you don't know :

 

1.

I was working with a smart guy from Germany on world's first Isochronous USB DAC. Well, I helped him a bit with thinking about it and I helped with listening tests. The DAC was named Crazy-T and it took so long to get it working in good fashion that by the time it hit the market, it was caught by time itself and the 16/48 it could do max, was not sufficient.

 

2.

Someone named Gordon Rankin did a slightly better job, and USB started to do 24/96. Everybody was happy and I said that USB carried too much noise (this is again close to 10 years ago IIRC) and I said so on CA and nobody agreed.

 

3.

Of course I said so because I was ignorant and used a completely different interface for the first incarnation of the Phasure NOS1 DAC. It was a PCIe interface (don't tell). People maybe recall this photo (has been on CA in a larger series of other photos about the interface) :

 

It was the first interface doing 24/384.

 

4.

That Interface (some still use this first NOS1 version) gave us nightmares because Windows 7 of te time was not up to the task and it was always praying whether the system would boot. Besides that, it wasn't particularly plug and play (it required a PCIe card in the PC) and USB was the thing. So already within 6 months the NOS1a could be supplied with USB interface. I recall that it even sounded better, but not after a few months of fighting to get it right.

Because all is about numbers, it was the first (USB) interface capable of 24/768.

 

5.

When a first NOS1-USB went to the USA, it appeared that very strange DC Offset problems emerged, and from there on I knew that even the mains interferes with our precious audio interface.

 

I will stop here because it will be OT. But message is : I tried.

More in 7 months ... 

 

 

[manisandher]

36 minutes ago, PeterSt said:

3.

Of course I said so because I was ignorant and used a completely different interface for the first incarnation of the Phasure NOS1 DAC. It was a PCIe interface (don't tell).

 

Aah, memories. I've had (both) my NOS1s for over 6 years now, and in that time they've gone from using PCIe 24/384, to using USB 24/384, and to using USB 24/768 now. The one in my office works very nicely with Roon -> HQP, the latter upsampling to 24/705.6 (or 24/768) before feeding the NOS1.

 

36 minutes ago, PeterSt said:

More in 7 months ... 

 

Hmm... So our Lushes will become obsolete? 

 

Mani.

[marce]

4 hours ago, PeterSt said:

 

I propose that any waveform which is used to trigger on its slope(s) does not result in "switching noise" as such. And otherwise it should be defined what is the "switching" phenomenon.

 

Switching noise is a quite formal term for a digital switch and one can spend a year in taming such devices, were it about utilizing them with the least amount of "noise". I am talking about high(er) frequency switches, like 45MHz. 

Of course this is a very low frequency compared to what exists in other worlds, but the noise coming from those switches is baaaad (for audio).

Btw, they carry jitter - but now I am trying to be funny.

 

Look up simultaneous switching noise... a problem with PC's and memory switching. A few gates should hardly generate any, but a bank of DDR is another thing. Its all down to instantaneous power requirements when switching... Once had someone insist on a 100A supply to the inner pins of a BGA to cope with the switching current... Told him it was impossible, so he sulked off and a few weeks later we laid his board out (5 more people repeated my comment). It can be reduced, but it isn't going to happen in an off the shelf PC, you need capacitive planes, expensive dialectrics and many layers, all adds cost so Mr PC of the shelf will be noisy.

[marce]

1 hour ago, semente said:

 

Why hasn't anyone come up with a more effective alternative for connecting a Computer to a DAC?

 

Do you think this kind of problem is inevitable and as such may as well be tackled in the USB connection? 

I worked on a real time network as part of a  many many years ago... it eventually matured into a used industrial bus...

Look at what a phy does...

Then start to work out your protocol, your voltage levels, current, design the transceivers, write the firmware, write the software, write the protocols, write device drivers, test it, test it again... etc. etc. Takes a team a long time just to come up with a basic working interface.

[lmitche]

1 hour ago, PeterSt said:

 

A small bit of history, in case you don't know :

 

1.

I was working with a smart guy from Germany on world's first Isochronous USB DAC. Well, I helped him a bit with thinking about it and I helped with listening tests. The DAC was named Crazy-T and it took so long to get it working in good fashion that by the time it hit the market, it was caught by time itself and the 16/48 it could do max, was not sufficient.

 

2.

Someone named Gordon Rankin did a slightly better job, and USB started to do 24/96. Everybody was happy and I said that USB carried too much noise (this is again close to 10 years ago IIRC) and I said so on CA and nobody agreed.

 

3.

Of course I said so because I was ignorant and used a completely different interface for the first incarnation of the Phasure NOS1 DAC. It was a PCIe interface (don't tell). People maybe recall this photo (has been on CA in a larger series of other photos about the interface) :

 

It was the first interface doing 24/384.

 

4.

That Interface (some still use this first NOS1 version) gave us nightmares because Windows 7 of te time was not up to the task and it was always praying whether the system would boot. Besides that, it wasn't particularly plug and play (it required a PCIe card in the PC) and USB was the thing. So already within 6 months the NOS1a could be supplied with USB interface. I recall that it even sounded better, but not after a few months of fighting to get it right.

Because all is about numbers, it was the first (USB) interface capable of 24/768.

 

5.

When a first NOS1-USB went to the USA, it appeared that very strange DC Offset problems emerged, and from there on I knew that even the mains interferes with our precious audio interface.

 

I will stop here because it will be OT. But message is : I tried.

More in 7 months ... 

 

 

Quantum entanglement coming in 7 months, cool!

[jabbr]

5 hours ago, PeterSt said:

Switching noise is a quite formal term for a digital switch and one can spend a year in taming such devices, were it about utilizing them with the least amount of "noise". I am talking about high(er) frequency switches, like 45MHz. 

 

Here is a very common issue. Generally .1 uF is specified as a bypass capacitor.

Which has low impedance at 10 Mhz but not so good at 100+ Mhz. Digital switching noise from a fundamental 45 Mhz clock will have significant high frequency harmonics. That's why bypass also at 1nF, 10pF etc. . The caps need to be laid out properly also. How often is this done in audio equipment? 

[PeterSt]

34 minutes ago, jabbr said:

How often is this done in audio equipment? 

 

 

 

At least one time. 

[Jud]

2 hours ago, manisandher said:

Hmm... So our Lushes will become obsolete? 

 

Mani.

 

Hopefully!

[Jud]

2 hours ago, marce said:

I worked on a real time network as part of a  many many years ago... it eventually matured into a used industrial bus...

Look at what a phy does...

Then start to work out your protocol, your voltage levels, current, design the transceivers, write the firmware, write the software, write the protocols, write device drivers, test it, test it again... etc. etc. Takes a team a long time just to come up with a basic working interface.

 

Probably a working Ethernet interface at the resolutions many people like to run is not terribly far off.  I'm sure it will have its own problems, and people will say, years from now, "Why do we use the Ethernet interface instead of good old USB?", the same way they talk about SPDIF interfaces today.

 

BTW, while I didn't have Peter's DAC, I did have the card he used, the very nice Juli@.  Eventually sold it, because it didn't work with FreeBSD, which I was running a lot at the time.

[PeterSt]

13 minutes ago, Jud said:

BTW, while I didn't have Peter's DAC, I did have the card he used, the very nice Juli@.  Eventually sold it, because it didn't work with FreeBSD, which I was running a lot at the time.

 

Haha. Anyone wants one ? I still have 20-25 all new in box.

For 20 euros + shipping (and VAT for those liable to VAT).

Also does 24/192 recording (am I right, I think so ...).

[k-man]

27 minutes ago, Jud said:

 

Hopefully!

Given the thread title, that is an ironic response!

[austinpop]

4 hours ago, BigAlMc said:

 

@austinpop - I'm going to let it burn in a bit before doing any serious listening or A-B-ing with the Curious cable.

 

Initial impressions. It's spongy!! 

 

Spongy? I thought it was cracked... finally!

 

[plissken]

On 7/18/2017 at 1:01 PM, Jud said:

 

The following example from John Swenson (in an old Computer Audio Asylum thread) talks about altering buffer size and its effects (long, but I think worthwhile):
 

 

 

Hi Jud.

 

How can this be taken in a vacuum of all the other processes that go on in a computer? Do you have direct link? Does John provide any output from the DAC measurements showing the analog is affected?

 

If you are a Windows or Linux user go into either PowerShell or Terminal and issue a PS command:

 

Powershell: ps | out-file -filepath c:\users\winlogonnamehere\documents\ps.csv

 

Take a look at your output. My system has 197 processes running on it and that's my work-a-day. I just did a clean Win10 Pro box and there are 138 processes.

 

Caching is such an integral part of OS's and you can measure this with built in reporting:

 

 

 

The above is all going on the fresh install.

 

My other take aways from reading this thread:

 

1. Don't use USB. Get a Networked solution

2. As Alex mentioned a chipset and a manufacturer, then take Chris's and others approach:

   

Get a DAC that isn't compromised by how USB was implemented on it.

 

IMO if you have to stick $700 worth of stuff in front of your USB DAC to get it to sound proper then you've fundamentally erred.

 

[Jud]

1 hour ago, plissken said:

 

Hi Jud.

 

How can this be taken in a vacuum of all the other processes that go on in a computer? Do you have direct link? Does John provide any output from the DAC measurements showing the analog is affected?

 

John was speaking of an illustrative hypothetical example:

 

Quote

Now for a concrete example. Let's take a simple program that is just coppying audio data from a file to a buffer and then to an simple output port. It has two threads, one reading the file and putting the data in the buffer, and one taking data out of the buffer and putting it on the out port using an external clock to time the opperation. The first thread waits until the buffer is empty then fills it up and goes back to sleep. (in reality there would be two buffers used in a ping pong arrangement, but that is irrelevant to the issue at hand).

 

I doubt he would have made actual measurements of a hypothetical example. 

 

1 hour ago, plissken said:

 

If you are a Windows or Linux user go into either PowerShell or Terminal and issue a PS command:

 

Powershell: ps | out-file -filepath c:\users\winlogonnamehere\documents\ps.csv

 

Take a look at your output. My system has 197 processes running on it and that's my work-a-day. I just did a clean Win10 Pro box and there are 138 processes.

 

Caching is such an integral part of OS's and you can measure this with built in reporting:

 

 

 

The above is all going on the fresh install.

 

 

Beyond noise, what might conceivably be relevant is whether the program's power draw is essentially random with regard to the music playback, or is correlated with it.  See http://www.amr-audio.co.uk/large_image/MAC OSX audio players & Integer Mode.pdf

[Jud]

Here's the start of John Swenson's contributions to the thread in which the quotes I took from him appeared:

 

https://www.audioasylum.com/forums/pcaudio/messages/12/126159.html

[lucretius]

2 hours ago, Jud said:

BTW, while I didn't have Peter's DAC, I did have the card he used, the very nice Juli@.  Eventually sold it, because it didn't work with FreeBSD, which I was running a lot at the time.

 

I still have a Juli@xte card lying around somewhere.  I used the analog RCA connections and I could hear noise transmitted via the interconnects to the powered speakers, when no music was playing.

[Jud]

2 minutes ago, lucretius said:

 

I still have a Juli@xte card lying around somewhere.  I used the analog RCA connections and I could here noise transmitted via the interconnects to the powered speakers, when no music was playing.

 

Well then don't use the analog RCA connections. 

[lmitche]

2 hours ago, plissken said:

 

Hi Jud.

 

How can this be taken in a vacuum of all the other processes that go on in a computer? Do you have direct link? Does John provide any output from the DAC measurements showing the analog is affected?

 

If you are a Windows or Linux user go into either PowerShell or Terminal and issue a PS command:

 

Powershell: ps | out-file -filepath c:\users\winlogonnamehere\documents\ps.csv

 

Take a look at your output. My system has 197 processes running on it and that's my work-a-day. I just did a clean Win10 Pro box and there are 138 processes.

 

Caching is such an integral part of OS's and you can measure this with built in reporting:

 

 

 

The above is all going on the fresh install.

 

My other take aways from reading this thread:

 

1. Don't use USB. Get a Networked solution

2. As Alex mentioned a chipset and a manufacturer, then take Chris's and others approach:

   

Get a DAC that isn't compromised by how USB was implemented on it.

 

IMO if you have to stick $700 worth of stuff in front of your USB DAC to get it to sound proper then you've fundamentally erred.

 

You can of course tune Windows for music playback with a variety of tools like AO and by hand. My machine has 22 running process with Hqplayer upsampling to dsd512, along with Roon local storage and Tidal playback capability. Optane memory as boot disk services the many random io cache file activities with record low latency response times.

 

Sound quality is awesome.

[plissken]

15 minutes ago, Jud said:

 

John was speaking of an illustrative hypothetical example:

 

 

I doubt he would have made actual measurements of a hypothetical example. 

 

So what good does mere conjecture do?

15 minutes ago, Jud said:

 

Beyond noise, what might conceivably be relevant is whether the program's power draw is essentially random with regard to the music playback, or is correlated with it.  See http://www.amr-audio.co.uk/large_image/MAC OSX audio players & Integer Mode.pdf

 

I read the PDF. Thanks. So we just want some exclusive mode for the audio device... Track preload.

 

Great. I use Wasapi in exclusive mode and JRiver set to preload the entire track. Jriver admits they are unaware of any study showing any actual SQ improvement. But it's there by customer demand and 'why not'.

 

 

[plissken]

3 minutes ago, lmitche said:

You can of course tune Windows for music playback with a variety of tools like AO. My machine has 22 running process with Hqplayer upsampling to dsd512, along with Roon local storage and Tidal playback capability. Optane memory as boot disk services the many random io cache file activities with record low latency response times.

 

Sound quality is awesome.

 

I hammered one of my systems down to next to nothing (both N3150M with 8GB RAM, 120GB SSD, Win10 Pro).

 

You would be hard pressed to pick the "AO'd" system.

 

The fact is that pagefile's, interrupts, cache operations. All sorts of I/O are STILL going on. My bog standard fresh Windows 10 installation is running at only 6% cpu utilization and I have a few items open: Powershell, IE, Control Panel, Notepad.

[lmitche]

8 minutes ago, plissken said:

 

I hammered one of my systems down to next to nothing (both N3150M with 8GB RAM, 120GB SSD, Win10 Pro).

 

You would be hard pressed to pick the "AO'd" system.

There is no question you could choose to make the hundreds of changes performed by AO, all by hand after a clean install, assuming you know what to do. I got tired of doing so, and bought AO to automate the process and to take advantage of the authors expertise.

 

Regardless of how you get there, tuning makes a huge difference for sure.

[Jud]

11 minutes ago, plissken said:

 

So what good does mere conjecture do?

 

I read the PDF. Thanks. So we just want some exclusive mode for the audio device... Track preload.

 

Great. I use Wasapi in exclusive mode and JRiver set to preload the entire track. Jriver admits they are unaware of any study showing any actual SQ improvement. But it's there by customer demand and 'why not'.

 

 

 

If you look just over halfway down the page here, you'll see measurements, by a person with whom you may be familiar, of ground plane noise with two media players.

 

http://thewelltemperedcomputer.com/Intro/SQ/MediaPlayer.htm

[Boleary3]

7 hours ago, PeterSt said:

I must note that such a path is the most dangerous because it depends on my ears (and the ears of only a handful of others in the same "field"). So what happens is, that I judge something as sounding better (which already takes me a week to be sure) to next reason out what actually happened at the technical level. And well, as we can nicely see by this topic, is how a world of "power" is not really capable to draw firm conclusions.

Quote

On 7/16/2017 at 7:38 AM, PeterSt said:

On 7/16/2017 at 7:33 AM, semente said:

So what have you achieved by not complying with the USB spec: a more accurate audio signal transfer or "better" sound?

 

Well observed. See previous post.

The latter.

 

It is also the very first time that I worked with a subjective parameter. Really.

 

 

On 7/16/2017 at 6:56 AM, PeterSt said:

But as we will see further down the line (next posts as far as time permits) there is no voodoo going on and nothing is coincidental. What is quite crucial though is that I happened to start out with the Clairixa which the very best as I could complied to the USB2 specs. So with that as a firm base for specs which coincidentally also implied the best sound (confirmed by all of its owners) I could start changing the now well known parameters

 

This is really one of the very best threads I have read on CA, despite its unfortunate "USB Cracked" title, coming from Mani, Mr. Measurement, of all people. Peter has provided the CA community with his vision, reasoned thinking, humble admissions of limitations, and use of a subjective parameter in creating the Lush. Thankfully, Peter would be the last person to say he has "cracked" usb audio. What he has created, given his ears, his system (118 db sensitive horn speakers, I think) and the settings he, and others using XXHighend software, changed when going from the Clairixa to the Lush, is a fantastic house sound that others, likely with similar systems, will find equally compelling.

 

I have both the Clairixa and Lush usb cables ( I use XXHighend too) and to my ears, in my system (92 db sensitive speakers, Shiit Yggy DAC) , in my house, I prefer the Clairixa. (I replaced my NOS1a Phasure DAC with the Yggy, finding the Yggy more "musical" and coherent sounding. I haven't heard the G3, but I'm sure it's fantastic).Though the Lush has a very, very pleasing sound it loses the "air" of the Clairixa. To my ears, in my system, in my house the Lush imparts a smooth, rounded and detailed sonic signature, but I have to turn up the volume to get the  bloom I so love with the Clairixa. Though the sound at higher volume is good, its just too damn loud for me. I am used to what is for me better sound, at a bit lower volume.

 

(Having mentioned my Yggy, lets hope Mani doesn't troll more charts and graphs on Yggy measurements in this thread as he recently did here:  http://www.whatsbestforum.com/showthread.php?22244-Schiit-interesting-name-more-interesting-products!/page21 But, in case he does try to troll some spec graphs here, and perhaps to get him to stop his biased attacks on Schiit, let me share with you Mani's April 2016 words to me regarding letting go of the Yggy:

"How are you getting on with things? I ask because I’ll be returning my demo Yggy back to the distributor next week… and I’m already starting to feel sad! I really think I’m going to miss its sound. I’d never say this to Peter (so please keep this between you and me) but I’m almost inclined to say to hell with its measurements. There’s something about its sound that I’ve never heard before and that’s so alluring.")

 

Sorry Mani, but my mentioning the Yggy here risks your bias and lack of humility "showing up." There is certainly something more to good sound than measurements have been able to capture.

 

For any folks reading this whose sole interest is SQ, I highly recommend you give Peter's XXHighend software a try. Though it's a bit of a challenge, Peter provides fantastic support (last week from Holland, via teamviewer, he was helping me trouble shoot an issue here in Philly) and forum members post their respective XXHighend settings, so you will have a place to start. It's amazing software that permits a kind of limited individualized tuning of the sound to the components in your system.

 

P.S. I have nothing more to say.

 

Brian

 

 

 

[plissken]

18 minutes ago, Jud said:

 

If you look just over halfway down the page here, you'll see measurements, by a person with whom you may be familiar, of ground plane noise with two media players.

 

http://thewelltemperedcomputer.com/Intro/SQ/MediaPlayer.htm

Read more  

 

I'm quite familiar with Amir's article. Unfortunately, as now is the norm with shills that write, totally out of context.

 

Let me help (linky):

 

"All the jitter is gone!!!

We went back to WMP player and let it read the file for a bit and it too settled down to the same performance as Media Player Classic. Clearly system activity disturbed this DAC.

Its brochure brags about "isolation" but the word "galvanic" is not there. This means that it is not isolated from the PC ground and that is feeding into the oscillator causing those correlated jitter spikes.

He sent the DAC back to get his money. When the check arrives, I will mention which DAC it is "

 

and

 

"Yes, he got his money back. The DAC was a Shiit."
 

 

Which means another DAC was properly engineered and didn't allow for this behavior