Computer Audio Asylum

Music servers and other computer based digital audio technologies.

Return to Computer Audio Asylum


Message Sort: Post Order or Asylum Reverse Threaded

USB cable errors - NOT

66.220.116.223

Posted on August 25, 2014 at 11:54:40
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
I have been running 24/192 to an XMOS reference board with a variety of USB cables for days now and monitoring errors with TUSBAudio SPY program. There have been zero errors of any kind with any of the cables I have tried. One is a 16-foot Belkin Gold. Another is a cheap Emerson 16-footer. The XMOS board is powered from the cable, nothing fancy. I am using KS with Foobar2000 on a Win7 hp laptop. Based on this, I'm not convinced that errors ever occur with an impedance-matched cable.

Maybe if the system has noisy ground-loops you could have errors???

 

Hide full thread outline!
    ...
RE: USB cable errors - NOT, posted on August 25, 2014 at 12:26:10
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002
Maybe if the system has noisy ground-loops you could have errors???

It was fmak who was reporting USB errors in his setup, no one else that I know of had reported any.



 

You raise an interesting point, posted on August 25, 2014 at 12:42:27
Joe Appierto
Audiophile

Posts: 1050
Location: Central NJ
Joined: January 3, 2004
Are errors in reproduction the only criterion why one cable might "sound" different from another?
Joe

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 12:45:17
Roseval
Audiophile

Posts: 1843
Joined: March 31, 2008
Sometimes users complain they can hear e.g. their mouse when using a USB DAC
I don’t think you can rule out CRC errors
Lindemann offers a couple of explanation why you might get errors.


The Well Tempered Computer

 

I bet the system still sounds different with different cables, despite that., posted on August 25, 2014 at 12:50:57
carcass93
Audiophile

Posts: 7181
Location: NJ
Joined: September 20, 2006
The obvious conclusion would be that it's not, or most likely not only, those CRC errors, or whatever that software registers, that affect sound quality.

 

RE: I bet the system still sounds different with different cables, despite that., posted on August 25, 2014 at 13:31:11
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
it does sound different with these different cables, but no errors reported...

 

RE: You raise an interesting point, posted on August 25, 2014 at 13:34:49
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
I know of other factors that might make cables sound different, including:

1) Common-mode noise - due to ground-loop
2) RFI
3) jitter
4) Cable +5V quality

The problem is all of these are effectively eliminated when using a galvanically isolated async interface with an external USB cable power supply. These different cables still sound different in this environment.

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 13:38:19
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
I believe CRC errors would be reported by USB SPY.

 

I agree (nt), posted on August 25, 2014 at 13:54:32
Mercman
Audiophile

Posts: 6580
Location: So. CA
Joined: October 20, 2002
.

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 14:02:40
Roseval
Audiophile

Posts: 1843
Joined: March 31, 2008
Can you provoke them e.g. remove shielding, wrap around a PSU, voltage regulator?
The Well Tempered Computer

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 19:00:22
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000
"It was fmak who was reporting USB errors in his setup, no one else that I know of had reported any."

Because nobody ever bothered checking. I guess his investigation prompted Steve to try it himself.

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 19:36:59
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002
If I'm not mistaken, a couple others tried testing for USB errors, possibly Tony. In any case, it still holds true that the only one I recall reporting errors was fmak. If he's experiencing errors, that might be one reason for his obsession with USB tweaking and hearing a difference with every tweak. Fmak also told me that placing a ferrite core around a USB cable would cause data errors. Some high-end USB cable makers do in fact use ferrites around their cables so go figure.



 

RE: USB cable errors - NOT, posted on August 25, 2014 at 20:41:53
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
"If I'm not mistaken, a couple others tried testing for USB errors, possibly Tony."

Correct, I have always counted the errors when possible. As a matter of principle, when ever I get a new piece of hardware or software I try to break it, or otherwise test its limits, unless the nature of the product is such that it might be permanently damaged. I have seen all the reported errors for my playback chain, those reported by HQplayer and those reported by the USBPAL device driver for my Mytek. However, from the Mytek documentation it was not exactly clear what the cause was for USB errors. I generate all the underrun errors by load testing the PC with non-audio tasks. Among other benefits this vets that the error counters do actually count...

Note that USB data errors will be detected by the CRC and the resulting frame dropped. Thus the audible effect of USB data errors is likely to be similar from the audible effect of buffer underruns where only a single buffer is dropped. (This is different from what one hears if there are data errors on an SPDIF connection where there will be audible clicks if they hit a high order bit at a time when the music is quiet.)

With the Mytek firmware and driver I have noted that using smaller buffers increases the chance of various error counters counting, but decreases the audible effects when they occur, such that often they are inaudible. When grossly overloading the computer and getting underrun errors with large buffers the sonic glitches will be obvious.


Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: You raise an interesting point, posted on August 25, 2014 at 21:04:53
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
Is there different gauge wire's on the 5V power lines?

What's the nature of the difference you got?

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 21:05:36
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000
"Fmak also told me that placing a ferrite core around a USB cable would cause data errors. Some high-end USB cable makers do in fact use ferrites around their cables so go figure."

I believe Gordon also recommended not using ferrites. Maybe Fred knows more than you think... The operable word here is "think".

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 21:34:37
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

While it's possible that using a ferrite choke -might- cause audible differences to some, I highly doubt that they outright cause data errors. They are commonly used on many general purpose USB cables supplied with disk drives and printers, as well as on some 'high-end' audiophile USB cables. If they -caused- data errors I don't think we'd see them used for so many applications.



 

RE: USB cable errors - NOT, posted on August 25, 2014 at 21:56:55
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000
"While it's possible that using a ferrite choke -might- cause audible differences to some, I highly doubt that they outright cause data errors. They are commonly used on many general purpose USB cables supplied with disk drives and printers, as well as on some 'high-end' audiophile USB cables. If they -caused- data errors I don't think we'd see them used for so many applications."

I thought we discuss audio here... Errors are not an issues with everything else unless they are extreme. This is basic computing, but we are not talking basic computing here, are we?

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 23:08:54
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002
I thought we discuss audio here...

See the OP. We're talking about USB errors.... and if extreme would certainly cause audio problems.






 

RE: USB cable errors - NOT, posted on August 25, 2014 at 23:13:28
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000
"See the OP. We're talking about USB errors.... and if extreme would certainly cause audio problems."

I am sorry, I forgot Steve builds printers...

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 23:23:06
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

I am sorry, but did you not see my reference to USB audio cables?



 

RE: USB cable errors - NOT, posted on August 25, 2014 at 23:36:02
soundchekk
Audiophile

Posts: 2424
Joined: July 11, 2007

I don't think that data corruption is a main concern here.

It's been Gordon who came up with the subject not long ago.
I think he was/is still fishing for a reason why the hell
he and a million other users still experiencee differences
on transport tuning measures (SW and HW).

For sure it's not all "common mode noise". ;)

Anyhow. Thx for letting us now.


-----------------------------------------------------------------

blog latest >> The Audio Streaming Series - tuning kit pCP

 

RE: USB cable errors - NOT, posted on August 25, 2014 at 23:44:32
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002


If you're not seeing USB errors but still hearing the differences among the various USB cables, do you think that you would likely hear even bigger audible differences or inconsistencies if USB errors were present?


 

TUSBAudio SPY , posted on August 26, 2014 at 03:40:10
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
The issue for me is:

What is the spy program detecting?

I am unable to get any information from those who hold Thyscon licenses and who should be able to get sensible answers from them.

I do know that there are errors from my desktop with front cable connected
usb ports to the motherboard.

 

RE: TUSBAudio SPY , posted on August 26, 2014 at 03:58:07
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
I am unable to get any information from those who hold Thyscon licenses and who should be able to get sensible answers from them.

WaveIO designer Lorien forwarded your query to Thesycon. Its reply, which is linked below, notes inter alia that ". . . some events show up as errors but are not really error conditions".

 

RE: TUSBAudio SPY - Answer from Theyscon, posted on August 26, 2014 at 06:31:55
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
I wait for the answer from Thesycon and here it is:

[...]Regarding the issue: This error indicates that the application (the player, e.g. foobar) was not able to provide audio buffers in time. But the correct interpretation depends on when the error happens: a) If the error is counting up while a song is playing then this indicates that there is a problem. b) If the error is counting up while the player switches from one song to the next, or during user-initiated start/stop transitions then this is normal behavior. Typically, during such transitions the player is not able to provide buffers in time. One should use Options - Clear in the menu to reset the counters and then watch carefully when the counter increments.

Note: [...]Spy Utility is intended to be used by developers and support staff only. Correct interpretation of the counters is not trivial. Due to internal implementation details some events show up as errors but are not really error conditions.

Maybe making the buffer in TUSBAudio Control Panel a little bit large could solve the issue (hoping it's indeed a real problem and not a 'hidden' code showed as an error, as Udo said above)?

http://www.diyaudio.com/forums/digital-source/188902-xmos-based-asynchronous-usb-i2s-interface-212.html#post3987062

post 2111

It seems that a keen vendor like Lucian was able to get an answer from Theyscon whereas bigger players just keep their silence.

This goes someway in settling the matter when we hear no issues.

 

RE: TUSBAudio SPY , posted on August 26, 2014 at 06:48:42
Ivan303
Audiophile

Posts: 48887
Location: Cadiere d'azur FRANCE - Santa Fe, NM
Joined: February 26, 2001
"Spy Utility is intended to be used by developers and support staff only. Correct interpretation of the counters is not trivial."

Kinda says it all, huh?


First they came for the dumb-asses
And I did not speak out
Because I was not a dumb-ass

 

RE: TUSBAudio SPY , posted on August 26, 2014 at 10:49:43
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
SPY tool is detecting for me:
Restartinitiated
Haltstreamfailed
Packetsizeinvalid
Packettoolong
ISObuffertoolate
ISObufferfailed
ISOpacketfailed

I think the important one is ISOpacketfailed.

I did get a single even of packettoolong when running 44.1, but never happened again.

You might try a laptop to see if its the ground-loop, or power the device from an external USB supply. Might also be noisy AC power....

Steve N.



 

RE: You raise an interesting point, posted on August 26, 2014 at 10:56:02
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
I'm using an external supply for +5V. The wires are silver-plated 20 gauge copper with Teflon ins, twisted together to reduce inductance.

The differences I hear in cables are primarily better focus and imaging with the better or shorter ones.

 

RE: USB cable errors - NOT, posted on August 26, 2014 at 10:57:43
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
Ferrites on the entire USB cable is a bad idea. Slows the edge-rates and messes with the impedance matching. Better to only put it on the ground wire.

Steve N.

 

RE: USB cable errors - NOT, posted on August 26, 2014 at 10:58:59
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
Disk drives and printer protocols allow for retry on an error. Streaming audio does not.

 

RE: USB cable errors - NOT, posted on August 26, 2014 at 11:01:16
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
I'll try some things and report back.

 

RE: You raise an interesting point, posted on August 26, 2014 at 11:38:59
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
Since it sounds better with the shorter cables might point to Em radiation off the USB cable finding it's way into your analog somehow. Perhaps the key is finding impedance matched cables also proven to be well shielded.

It sounds like you have really gone the extra mile to try and reduce influence from the pc's noise.

 

RE: USB cable errors - NOT, posted on August 26, 2014 at 12:02:18
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002
Disk drives and printer protocols allow for retry on an error. Streaming audio does not.

How would hard data errors and the lack of retries manifest themselves audibly in the audio playback chain?




 

RE: USB cable errors - NOT, posted on August 26, 2014 at 13:36:35
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
"How would hard data errors and the lack of retries manifest themselves audibly in the audio playback chain?"

There are two possibilities:

1. process the erroneous data. What you get is what you hear.
2. ignore the erroneous data and attempt to guess what it was. What you hear depends on the algorithm.

In the case of SPDIF there is no further error detection, so it's case 1. You will hear clicks, amplitude depending on which bits in the PCM word get corrupted.

In the case of CDA (Red Book coding) there is a strategy to correct errors and to guess at single sample uncorrectable errors by interpolation, otherwise to mute. Most single errors are successfully interpolated without clicks. There are test tools that CD mastering engineers run to evaluate the quality of the raw bits on a disk to see what margins for error correction remain.

With USB audio, the DAC would implements some kind of strategy. I suspect this includes discarding any packets that have CRC errors. Here error detection is done on a packet basis which contains roughly 1 millisecond of audio. One can simply duplicate the missing sample or one can drop it and slip the tempo one millisecond, etc... If this doesn't happen too often it won't be noticeable on most music to untrained listeners. (The details will depend on the particular USB protocol the DAC uses, which may be proprietary.) The aural effects are likely to be similar to buffer underruns due to computer system overload. One strong reason for discarding packets that fail the CRC concerns the USB frame format. Bit stuffing is used to delimit the data packets, which means that a single bit in error can turn into a different number of bits in the output, resulting in loss of PCM word framing. If PCM framing is lost the results can be a horrible (if not dangerous) burst of noise.


Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: USB cable errors - NOT, posted on August 26, 2014 at 16:16:23
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002
OK. But 'errors' are not the same as a slight rounding of the leading edge of a data stream.

I was told by an inmate that a ferrite choke on a USB cable will cause errors. I do not believe this to true and I do not believe anyone has proven it.

I would think that if such errors occurred they would have a greater chance of being distinctly audible vs a slight 'rounding of the leading edge' of a data stream. Possibly clicks, pops, and bursts of noise.

 

RE: USB cable errors - NOT, posted on August 26, 2014 at 19:16:51
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000
""Disk drives and printer protocols allow for retry on an error. Streaming audio does not.""

"How would hard data errors and the lack of retries manifest themselves audibly in the audio playback chain?"

He was not talking about audio. He was explaining the difference,just like I tried to. You can keep the ferrets for your printer cable.


 

RE: USB cable errors - NOT, posted on August 26, 2014 at 19:47:11
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002
So he explained the differences but has not yet answered my question. And if you read the OP he is in fact talking about audio and so am I.

I have been running 24/192 to an XMOS reference board with a variety of USB cables for days now and monitoring errors with TUSBAudio SPY program.

Hence my question: How would hard data errors and the lack of retries manifest themselves audibly in the audio playback chain?

Is this question not relevant? If not, please tell me why.


 

RE: USB cable errors - NOT, posted on August 27, 2014 at 06:27:34
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
I think I'd worry adding the bead changes the cable impedance too much when attempting adding one to a design already compliant to some controlled impedance tolerance range. If you change the resultant impedance too much, data errors are inevitable. I never played with this very much and so don't have a good feel for how much various beads alter the curves.

When the right test gear is available to a designer it is nothing to verify cables in whatever configuration you can come up with and then compensate until they meet spec. Without the gear it's much more difficult to get it right. Unfortunately the price of the necessary gear to make this task easy is way out of range for it to seem interesting to most of us, at least for now.

 

RE: USB cable errors - NOT, posted on August 27, 2014 at 07:54:51
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

It seems to me that since some high-end USB audio cable makers do use ferrite chokes on their cables, they are highly unlikely to cause outright 'errors'.

They may slightly alter the sharpness of a leading edge but that is not the same as an 'error', and the overall benefit of the choke may outweigh any slight rounding.

You're right, w/o test equipment I don't think anyone here is qualified to definitively state that a ferrite choke will cause errors.


 

RE: USB cable errors - NOT, posted on August 27, 2014 at 09:30:16
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
My point, though maybe not entirely clear, manufacturers with the right test gear can verify the effect adding beads has and properly compensate for the negative side effects to ensure the resulting assembly with the bead still meets spec. Without the test gear one would have no way of knowing what affect the bead had or how potential compensations for its affect behaves. As I said I'm just guessing here and you may be 100% correct that for USB band signals the effect is negligible.

 

RE: USB cable errors - NOT, posted on August 27, 2014 at 10:32:51
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
Abe - not all ferrites are the same. They are comprised of different materials with different properties. Some may have the potential of causing errors. Like most things in life, its not black and white.

 

RE: USB cable errors - NOT, posted on August 27, 2014 at 12:03:35
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002
Understand. Of course not all ferrite chokes are the same and not all will cause errors.


 

Understand., posted on August 27, 2014 at 12:06:40
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

I'm just a little surprised that an inmate here, without the necessary test equipment, would make a blanket statement that ferrite chokes on USB cables cause errors.



 

this is a protracted discussion of no substance., posted on August 29, 2014 at 01:13:36
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
I never said anything about ferrites causing errors. It was Gordon who posted and he has more equipment to measure than anyone else who insists on posting pointless speculations.

 

It is clearly the hand of God typing at your keyboard and no mere mortal., posted on August 29, 2014 at 06:42:11
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
I suppose the guy who invented Gordons gear had to just guess about how to make it work until a miracle occurred and it started working? Hint: the answer is no. Modern physical theory adequately covers design and understanding of controlled impedance cables for usb even if it is too much a pain for most to bother to pull off manually.

 

RE: Understand., posted on August 29, 2014 at 08:37:54
rick_m
Audiophile

Posts: 6230
Location: Oregon
Joined: August 11, 2005
"I'm just a little surprised that an inmate here, without the necessary test equipment, would make a blanket statement that ferrite chokes on USB cables cause errors."

Why on earth would that surprise you? That is the epitome of what we are all about. If you can't proudly state that one listen to your new amplifier was enough to know that the sound was being squimmied by the cheap chinese ink used to label the powercord then you just don't have the ears and perception to be an audiophile.

Not only are our men strong, women beautiful and kids above average, our perceptions are honed to the point where we know how good something will sound just by evaluating the data on it's pricetag...

And loving every minute of it! However nutty we may seem, we are the epitomy of sanity compared to football fans...

Rick

 

RE: Understand., posted on August 29, 2014 at 09:24:42
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
Why on earth would that surprise you?

Part of the surprise might be because the inmate in question protested that "I never said anything about ferrites causing errors". If in fact he had, I at least couldn't find where so I concluded, not unreasonably, that he hadn't said anything about ferrites causing data errors.

What he did say was that an Xmos test program was reporting errors (not a ferrite in sight) and that his efforts to find out what the reports meant had not been terribly successful.

A crowd than gathered round to poke sticks through the bars of his (metaphorical) cage and demand loudly what the effect of those data errors on sound quality might be.

How could one possibly answer that? It was all pretty baffling and teeeedious both but it was not, as you rightly suggest, very surprising.

 

Guilty I am!, posted on August 29, 2014 at 11:05:32
rick_m
Audiophile

Posts: 6230
Location: Oregon
Joined: August 11, 2005
I wasn't really responding to the thread, just Abe's post. I should have looked more carefully at the context...

Rick


 

RE: Understand., posted on August 29, 2014 at 22:41:12
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

You weren't looking in the right thread. Quite some time ago when I posted about having a ferrite choke on one of my USB cables, the inmate made the blanket statement that they cause USB errors. Not true. This was a while back about the same time said inmate was suggesting defragmenting SSDs was a good thing. And he also would partition his SSD so one partition would wear less than the other (doesn't work that way with SSDs).

Sorry but this inmate has zero credibility with me. It's one thing to tweak and theorize while criticizing "IT guys". It's another matter completely when fmak'ing total bogus misinformation.


 

RE: Understand., posted on August 30, 2014 at 01:43:13
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
You weren't looking in the right thread.

Not convinced the fault's mine - why should I have to search the archive for a point made "quite some time ago" merely to understand your post? Given that you're belittling an inmate, why not provide a link to the pertinent thread so we can judge your claim for ourselves?

I say that because it is clear from this and other threads that many inmates still do not understand the (very basic) difference between data errors and timing errors, i.e. why the performance of a computer driving printers, scanners, storage devices and the rest does not inform about how it fares when driving an audio device.

The confusion leads to claims such as the one that error-free data transfers over USB "prove" that USB cable quality is, within obvious limits, irrelevant to USB audio.

As it goes, for reasons not relevant here, I don't hear much difference between USB cables in my dedicated-to-audio PC setup but I do understand why competent engineers are adamant that simply ensuring that a cable complies with spec is by no means the whole story for audio. I do not ridicule those who do report differences.

(Of course, many USB cables and interfaces that seem nonetheless to work reasonably well in general computing are waaaay out of spec but that's a different topic.)

In short, absent demonstration that fmak was unequivocally suggesting that ferrite beads on a USB cable cause data errors rather than exacerbate timing errors, the "zero credibility" thing works both ways. I doubt fmak would make such a claim though, of course, some might misinterpet his point.

Especially someone who asks, apparently in all seriousness, "How would hard data errors and the lack of retries manifest themselves audibly in the audio playback chain?"

 

RE: Understand., posted on August 30, 2014 at 07:15:00
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002


Not convinced the fault's mine - why should I have to search the archive for a point made "quite some time ago" merely to understand your post?

No one is placing fault or making you search the archives. I'm not making this stuff up.

Especially someone who asks, apparently in all seriousness, "How would hard data errors and the lack of retries manifest themselves audibly in the audio playback chain?"

A fair question, no?


 

RE: Understand., posted on August 30, 2014 at 07:47:59
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
I'm not making this stuff up.

Hadn't suggested you were but whether you're correctly interpreting what was written is a different matter.

A fair question, no?

Maybe I'm missing something but it struck me as a red herring, esp in the light of Steve's report.

 

RE: Understand., posted on August 30, 2014 at 08:26:29
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

Maybe I'm missing something but it struck me as a red herring, esp in the light of Steve's report.

No, not at all.

I'd be curious to know if 'hard errors' are observed over the USB audio interface as Steve witnessed, what do they sound like out the DAC?

I'm not talking about so-called slight timing variation (jitter) possibly due to differences in USB cables or any number of other reasons, but an actual distinctly audible event correlated against observed 'hard errors' over USB. Clicks, pops, scratches, bursts, etc. ??

Has anyone done this?





 

RE: Understand., posted on August 30, 2014 at 12:10:37
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
"This was a while back about the same time said inmate was suggesting defragmenting SSDs was a good thing"

That was not my recollection, so I went back and checked. My recollection was correct.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

So, in other words, without trying to be needlessly polite, it - ..., posted on August 30, 2014 at 12:28:15
carcass93
Audiophile

Posts: 7181
Location: NJ
Joined: September 20, 2006
... "inmate was suggesting defragmenting SSDs was a good thing" - is a BS and a lie, disseminated by a notorious BS'er and a liar, in other do discredit an inmate who forgot more about audio-related matters, than nasty Apple-fixated coward ever knew.

The fact that he also absolutely can't remain focused on audio-related topics - SSD defragmenting, value of Apple stocks etc. - while posting in Computer Audio forum, just adds an insult to (apparent brain) injury.

 

Here you go......, posted on August 30, 2014 at 13:10:42
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

I just don't have ANY problems with XP on SSDs. Even w/o dual boot on XP, you can Trim as an external drive or run Diskeeper SSD Free Space Optimizer.

8: Trim (4.11) Open this result in new window
Posted by fmak on 2014-04-04, 22:23:22 (93.109.161.60)
This is why I have dual boot. On XP I run programs that W8 won't plus reliable and easier/more informative Acronis and Diskeeper. On W8 Trim works. On some earlier SSD, you can also run Garbage collec .......


Diskeeper SSD Free Space Optimizer IS a defragmenter. No benefit on SSD. Look it up.

In other posts, he suggests partitioning SSDs to minimize wear in one of the partitions. Won't work as SSDs are not 'electro-mechanical' in nature like HDDs. The controller will write to whatever NAND flash cells it chooses regardless of how the SSD is partitioned.

Like I said, tweaking and theorizing is one thing but spewing factually inaccurate misinformation is another.



 

RE: Here you go......, posted on August 30, 2014 at 13:34:00
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
I think there may be some confusion regarding software product names and versions. Here is fmak's post that I was working off:

"It is not defrag; read it up. Hyperfast
It works well."


I suggest reading the entire thread for more clarity.

My only experience with Diskeeper comes from using an early version to attempt to clean up a huge MFT. The only effect of the product was to trash the entire volume. As a result of this experience, I will never remotely consider purchasing any product from that company.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: Here you go......, posted on August 30, 2014 at 13:59:52
Ryelands
Audiophile

Posts: 1867
Location: Scotland
Joined: January 9, 2009
I suggest reading the entire thread for more clarity.

Ye Gods! Where's the fun in that? All we need do is pull quotes at will and create mayhem where there used to be clarity. What kind of forum are you after?

Seriously, assuming (reasonably) that the link is the quote AC had in mind, it strikes me that it's a bit rich to "quote" out of context and without source a four-year-old remark on a different topic to discredit by association.

Still, what do we simple rednecks know about city folks' serrfisticated diskussions and flosifcal ways? Yeeeeeeeeehaw!

 

RE: Here you go......, posted on August 30, 2014 at 18:33:59
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

He said he used that defragger on his SSD. Lost credibility. Not sure how you see that comment of his being quoted 'out of context'.

Same inmate suggested partitioning a SSD to localize wear to the partition being written to. That doesn't work. Lost credibility again. Shall I dig that one up too?

Here's a guy who criticizes what he calls "IT guys" yet he spews factually incorrect computing nonsense. Believe what you will.



 

RE: Here you go......, posted on August 30, 2014 at 18:58:14
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
"Same inmate suggested partitioning a SSD to localize wear to the partition being written to. That doesn't work. Lost credibility again. Shall I dig that one up too?"

You may be right, but the verification would depend on details of SSD firmware, operating system drivers and file system software. At the time, I didn't particularly care, as I had reached the conclusion that SSDs were not (yet) mature technology and therefore not to be trusted with my precious data....

If you were arguing that most people tweaking computer audio systems don't have a clue as to what is going on, then I would agree with you. I would put myself in the same category since it has been decades since I had direct access to hardware and software designers and access to hardware logic diagrams and software source code. This is why I believe that tweaking operating systems for better sound quality will ultimately prove to be a futile effort, since it is vastly more difficult than simply designing and building proper isolation into DACs and downstream analog equipment. (If you are unsatisfied with the sound you get with your existing hardware, then these tweaks may be the best available alternative.)



Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: Understand., posted on August 30, 2014 at 18:58:49
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000
"I'd be curious to know if 'hard errors' are observed over the USB audio interface as Steve witnessed, what do they sound like out the DAC?

I'm not talking about so-called slight timing variation (jitter) possibly due to differences in USB cables or any number of other reasons, but an actual distinctly audible event correlated against observed 'hard errors' over USB. Clicks, pops, scratches, bursts, etc. ??"

You are really obsessed. If there are errors how can one correlate in real time what "AN ERROR" might sound like??? If you keep going back in time you might be able to find some hieroglyphics to try and make you stupid points!!!

 

Huh?, posted on August 30, 2014 at 19:26:25
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002
Making your usual helpful contribution again huh Bob?

Why can't one observe 'hard errors' over USB audio in realtime while listening to the output of a DAC in realtime?

Is that not possible? Since you're so bright and feeling so helpful today, please explain.


 

RE: Understand., posted on August 30, 2014 at 19:29:25
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
If there are data errors on a USB cable, then sooner or later they will be audible if you listen carefully. You might not hear each individual data error, but if 10 occur a careful listener will likely notice at least one problem.

The sound of a USB error will depend on the particular music being played ant the particular algorithms in the DAC that respond to the error indication. The situation is more complex with USB errors than it is with SPDIF errors. SPDIF errors will be audible when they happen to hit high order bits of a PCM word, particularly if the music is quiet at that point. Because of the design of SPDIF there is no way the DAC can detect an error and perform muting, hence there will be clicks. In the case of USB errors the DAC will definitely be able to detect data errors, but a single bit error will affect not just one audio sample, but potentially all the data left in the remainder of a 1 msec block of audio.

There should be no data errors on USB audio. They should be at the level of people being killed by errant meteors. If there are errors the likelihood is that some piece of equipment is broken or that some fool paid big bucks for an expensive cable "designed" by an idiot who didn't know what he was doing.




Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: Huh?, posted on August 30, 2014 at 19:35:57
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000
How about this... They may sound like one hand clapping, and then again they may not... Maybe someone can explain this to you...

We need to start calling you Glenn Close...

 

RE: Understand., posted on August 30, 2014 at 19:40:41
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000
Thank you so much Tony (Saint Anthony) . You have much more patients than I do...

 

RE: Huh?, posted on August 30, 2014 at 21:33:24
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

You're full of hot air Bob. Nothing more.

 

RE: Huh?, posted on August 30, 2014 at 21:43:17
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000

 

RE: Understand., posted on August 30, 2014 at 21:51:37
AbeCollins
Audiophile

Posts: 46200
Location: USA
Joined: June 22, 2001
Contributor
  Since:
February 2, 2002

There should be no data errors on USB audio. They should be at the level of people being killed by errant meteors. If there are errors the likelihood is that some piece of equipment is broken or that some fool paid big bucks for an expensive cable "designed" by an idiot who didn't know what he was doing.


Thanks Tony. Yet at least one inmate has reported USB errors which by your comments would be a good indication that his equipment or cable is broken. It wouldn't make sense at that point to pursue other audio tweaks if the fundamental functionality of his USB interface is flawed.

I'd still be curious to know if anyone has observed these 'hard errors' using software or test equipment while listening for them, perhaps against a continuous tone to make them easier to hear.


 

Hey, that's not fair - how come I don't get called Saint something?, posted on August 31, 2014 at 08:54:33
carcass93
Audiophile

Posts: 7181
Location: NJ
Joined: September 20, 2006
I'm not too pretentious - if all good ones are already taken, I'll be happy with something less traditional.

How about St. Baal?

 

RE: Understand., posted on August 31, 2014 at 09:05:59
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
He reported USB errors, but it may have been the result of a bug in the error counting software. I can get reported USB errors from my USBPAL driver if I try running with 8 msec buffers and none with 4 msec buffers. Also, there are a horrendous number of them even if no audio application is running. With either buffer size audio seems to play without obvious glitches.

If I load my system up heavily then sometimes another error "ASIO errors" will get counted. This is particularly true when I do room correction of a DSD file as CPU load gets to 75 percent (all cores). When these happen usually errors show up in pairs and they are usually audible. Typically, something like using the scroll wheel on the mouse while looking at a complex web page will provoke these errors. These errors mean that the application is not supplying buffers to the driver fast enough.


Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: LMAO!!! nt, posted on August 31, 2014 at 09:39:08
Bob_C
Audiophile

Posts: 2667
Location: NY
Joined: July 31, 2000
nt

 

Better setup needed for errors, posted on September 1, 2014 at 13:21:05
Gordon Rankin
Manufacturer

Posts: 2928
Joined: June 9, 2000
Steve,

I was doing the same thing, using my TotalPhase analyzer to look at frames and see if there was any errors. Last year I needed to spend some money on stuff so I bought a TEK MSO scope with their USB package and a differential probe. This setup actually can do full USB acceptance testing.

Man I was not ready for what I saw. I designed up a small board that would break out the D+D- signals on either end of a cable. Depending on computer, dac, cable, direction of the wind. I got all kinds of interesting data.

Also I reprogrammed the XMOS to show when the PHY received errors. From all this the data was not as good as what you are seeing.

First looking at EYE patterns and especially the feedback pipe from the DAC back to the computer there was a lot of cable issues. Many times the turnaround from the IN passed packet to the DAC sending the feedback data was many times in error. You could see this on ports from several computers that shared peripherals on the same USB buss and from cheap computers and single board Linux systems. I have checked about 12 Windows machines and several Apple and at least 18 Linux boxes and the results for cables are all over the map.

I think there is allot more needed from some of these cable companies. I don't think any of them are testing at this level.... AND THEY SHOULD BE!

Thanks,
Gordon
J. Gordon Rankin

 

RE: Better setup needed for errors, posted on September 1, 2014 at 13:58:30
Tony Lauck
Audiophile

Posts: 13629
Location: Vermont
Joined: November 12, 2007
Given that you have the necessary tools to determine conformance, perhaps you might publish some data that exposes the cheats who put out cables that do not meet USB specifications, but which command high prices because of marketing designed to convince golden ears that they sound good. It would suffice to list a few conforming products by brand, model and price. And then to list some non-conforming products that cause data errors because they transform a legal signal starting out at one end into an illegal signal at the other.

If you have the stones, it would be even better to name names, but I can see reasons why not to do this.


Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

 

RE: Better setup needed for errors, posted on September 3, 2014 at 11:38:18
audioengr
Manufacturer

Posts: 6017
Location: Oregon
Joined: April 12, 2001
The USB3318 PHY seems to be more sensitive than other PHYs I have used. Does not work with cables that work fine with my M2Tech interface. Other developers have also reported this sensitivity.

Have you tested for the same errors using your prior non-XMOS interface?

Steve N.

 

Page processed in 0.046 seconds.