[crazychile]

*sigh*

Digital technology is predicated on the fact that a bit is a fundamental unit of data having precisely two states, with no scope for grey areas. Some apparently prefer to believe in ghosts in the machine rather than accept psychology plays any part in what they hear.

I understand digital technology perfectly well. Most of what you have stated is correct on a very basic level. Unfortunately your premise is flawed because it only looks at one ingredient in the soup. You've completely ignored the impact of power management, oversampling and digital filtering effects, how data integrity is maintained pre and post conversion, and many other variables. It's the combined effects from all of these that impact sound quality.

There is also a big difference between static and rapidly changing data. Yes, if something is corrupted or missing it will either not work at all, or work intermittently. But digital music is not the same as opening a Word document. If you understand how the entire digital conversion process really works with audio, then you realize how there is effectively guess work being performed that fills in the gaps between the points that are known. This is an oversimplification, but when you understand digital audio, you understand the impact that seemingly insignificant variables have on effecting the amount of guess work or filling in of the gaps that takes place when converting the 1s and 0s.

Another way of looking at it would be similar to the differences between a lossy and lossless file of the same recording. The MSB or Macro info is going to be comparable between the two, but the LSB or micro details won't. In my experience, the micro details are what make the music sound "live" rather than just a pattern of notes.

I still hold firm on my original position that USB as it relates to audio is not fully understood. USB was never originally designed as an interface for digital audio. More than a few manufacturers have openly admitted that they have made refinements to their designs to bring the performance closer to other more established means of data transmission. This is a learning process that evolves over time.

I am a skeptic of marketing claims and don't own any high priced cables (digital or analog), or any magic tuning pucks to place on my gear. However, i can often hear subtle, repeatable differences when variables are changed. I can't always explain them, but try to keep an open mind rather than saying they don't exist, because they can't be explained from the knowledge we currently have.

 

[WraithApe]

I understand digital technology perfectly well. Most of what you have stated is correct on a very basic level. Unfortunately your premise is flawed because it only looks at one ingredient in the soup. You've completely ignored the impact of power management, oversampling and digital filtering effects, how data integrity is maintained pre and post conversion, and many other variables. It's the combined effects from all of these that impact sound quality.

There is also a big difference between static and rapidly changing data. Yes, if something is corrupted or missing it will either not work at all, or work intermittently. But digital music is not the same as opening a Word document. If you understand how the entire digital conversion process really works with audio, then you realize how there is effectively guess work being performed that fills in the gaps between the points that are known. This is an oversimplification, but when you understand digital audio, you understand the impact that seemingly insignificant variables have on effecting the amount of guess work or filling in of the gaps that takes place when converting the 1s and 0s.

Another way of looking at it would be similar to the differences between a lossy and lossless file of the same recording. The MSB or Macro info is going to be comparable between the two, but the LSB or micro details won't. In my experience, the micro details are what make the music sound "live" rather than just a pattern of notes.

I still hold firm on my original position that USB as it relates to audio is not fully understood. USB was never originally designed as an interface for digital audio. More than a few manufacturers have openly admitted that they have made refinements to their designs to bring the performance closer to other more established means of data transmission. This is a learning process that evolves over time.

I am a skeptic of marketing claims and don't own any high priced cables (digital or analog), or any magic tuning pucks to place on my gear. However, i can often hear subtle, repeatable differences when variables are changed. I can't always explain them, but try to keep an open mind rather than saying they don't exist, because they can't be explained from the knowledge we currently have.

 
Perhaps you could point out where in the USB specification it says that the standard is insufficient for the transfer of audio data?
 
Despite these attempts at obfuscation, the fact remains that digital information transmission is content agnostic - it doesn't matter whether the information encoded prior to transmission is audio, visual or textual; at the point of transmission it just comes down to packets of data. So yes, there is no difference between a digital music file and a Word document in that respect. I'm well aware that A/D and D/A conversion + perceptual coding are complex issues but that's muddying the water when the focus is purely on data transfer, not pre- or post-transfer conversion and processing. Native USB CRC enables a virtually bit-perfect transfer via cables conforming to the USB standard; what's done with that data on receipt, or prior to being transmitted, really isn't the issue here.

 

[Jabbah]

I understand digital technology perfectly well. Most of what you have stated is correct on a very basic level. Unfortunately your premise is flawed because it only looks at one ingredient in the soup. You've completely ignored the impact of power management, oversampling and digital filtering effects, how data integrity is maintained pre and post conversion, and many other variables. It's the combined effects from all of these that impact sound quality.

And exactly how are any of those relevant when discussing data transmission cables?

There is also a big difference between static and rapidly changing data.

What is "static data" in the context of data transfers?

But digital music is not the same as opening a Word document.

From the point of view of the cable, what exactly is the difference between a stream of bits representing audio and a stream of bits representing other data?

If you understand how the entire digital conversion process really works with audio, then you realize how there is effectively guess work being performed that fills in the gaps between the points that are known. This is an oversimplification, but when you understand digital audio, you understand the impact that seemingly insignificant variables have on effecting the amount of guess work or filling in of the gaps that takes place when converting the 1s and 0s.

Wow, could you fit any more condescension in a single paragraph? I think you would benefit from reading up on Nyquist-Shannon sampling theorem. This video may also help:

https://youtu.be/cIQ9IXSUzuM

Hint: There is no guess work, for a given set of samples there is a unique solution to reconstruct the wave form.

Another way of looking at it would be similar to the differences between a lossy and lossless file of the same recording. The MSB or Macro info is going to be comparable between the two, but the LSB or micro details won't. In my experience, the micro details are what make the music sound "live" rather than just a pattern of notes.

All bits in a bit stream have equal probability of being flipped. The least significant bit doesn't magically become more susceptible because we call it the LSB. It is impossible for a cable to affect the "micro details" and not the "macro details".

I still hold firm on my original position that USB as it relates to audio is not fully understood. USB was never originally designed as an interface for digital audio.

USB audio is very well understood and it most certainly was designed with the inclusion of streaming audio interfaces, that's exactly what isochronous endpoints are in the spec for.

More than a few manufacturers have openly admitted that they have made refinements to their designs to bring the performance closer to other more established means of data transmission. This is a learning process that evolves over time.

News flash: company claims something is really difficult but that they have refined their offering to make it better than cheaper rivals.

I am a skeptic of marketing claims and don't own any high priced cables (digital or analog), or any magic tuning pucks to place on my gear. However, i can often hear subtle, repeatable differences when variables are changed. I can't always explain them, but try to keep an open mind rather than saying they don't exist, because they can't be explained from the knowledge we currently have.

We can explain it from the knowledge we currently have; it's called expectation bias.

 

[crazychile]

Hey guys, my apologies if what I wrote came off as condescending.  This is supposed to be in the spirit of healthy debate.  I’m not trying to start a p!ss!ng contest here. My only agenda that anyone reading the thread might CONSIDER that maybe there’s more to this than just  1s and 0s and the promise of perfect sound forever.  Plenty have stated their opinion on one side of the issue, I just wanted to present the other side.
 
My final response in this thread  for anyone who cares is this… take a look at the white papers from http://www.usb.org/developers/whitepapers/ :
 
“USB 3.0 Jitter Budgeting”: If usb.org has taken the time to write a paper about jitter and the methodology used to model the data that is used to improve the jitter margin, then maybe this has an effect on system performance?
 
“Managing Connector and Cable Assembly Performance for USB SuperSpeed”  is an interesting read that covers EMI and RFI in addition to other topics. Is it reasonable then to conclude that how a USB cable is constructed (and the materials used) might have an effect on EMI and RFI?
 
This paper also refers to the USB specification that defines a  90 ohm nominal impedance. How many cables on the market are at or near this spec (or contribute to the total load that is generated by the  I/O receivers + cable). How many companies even specify what the impedance of their cables are? Maybe then if a cable deviates from 90 ohms it could technically perform it’s basic functions yet still be a variable regarding  the overall performance?
 
And just maybe if someone were to come out with a gadget (like the one I mentioned in my first post) where one of the objectives were to bring the incoming signal closer to this 90 ohm ideal impedance, then is it possible that this might actually  make a verifiable difference in the sound?

 

[Jabbah]

 
“USB 3.0 Jitter Budgeting”: If usb.org has taken the time to write a paper about jitter and the methodology used to model the data that is used to improve the jitter margin, then maybe this has an effect on system performance?  
“Managing Connector and Cable Assembly Performance for USB SuperSpeed”  is an interesting read that covers EMI and RFI in addition to other topics. Is it reasonable then to conclude that how a USB cable is constructed (and the materials used) might have an effect on EMI and RFI?

 
You do realise that those papers are for generic data transfers at high speeds eg 5Gbps? Jitter in a data stream will not lead to jitter in audio, it will lead to a bad CRC and the packet being rejected. Same with cable performance, any issues will lead to packet rejection. Digital data transfer is incredibly robust, it cannot lead to corruption of the data as any packets that do not match the CRC will be rejected. This goes as much for audio data as any other data.
 

 And just maybe if someone were to come out with a gadget (like the one I mentioned in my first post) where one of the objectives were to bring the incoming signal closer to this 90 ohm ideal impedance, then is it possible that this might actually  make a verifiable difference in the sound?

 
No, all the bits in an data stream (audio or otherwise) are equal, it is impossible for digital data streams to be affected in a way that would change the dynamics of the audio. Digital corruption is very well understood and the only outcome with a very poor cable with digital streaming data (isochronus endpoint) is that the odd data packet will be dropped. Streaming audio is at such a low data rate compared to the rates that USB is capable of, that if you are suffering from dropped packets then there is likely a catastrophic failure in the cable.

 

[CanadianMaestro]

A USB data cable is not a good fix for jitter issues, even over longer distances. So save the $$ and instead explore DACs with vanishingly low jitter values.

 

[CanadianMaestro]

I found that DH Labs or the Furutech ADL does well for computer USB-to-DAC/HP amp connection. And relatively cheap.

 

[cel4145]

No, all the bits in an data stream (audio or otherwise) are equal, it is impossible for digital data streams to be affected in a way that would change the dynamics of the audio. Digital corruption is very well understood and the only outcome with a very poor cable with digital streaming data (isochronus endpoint) is that the odd data packet will be dropped.

Crazychile, you need to listen to what he's saying here. And it's the same thing that others have said before. *If* data corruption happens, it CANNOT result in the types of audible differences that people attribute to better USB cables. This is very flawed cause and effect reasoning.

 

[Tablix]

The only answer is placebo affect, nothing more.  You think recording studios spend more than the absolute minimum for a USB cable? This whole topic of the thread is pointless.  The ONLY reason to spend more than $10 is cable length or durability, and considering 99% of users never move their cables, durability doesn't even enter the equation either.

 

[CanadianMaestro]

Fck placebo and anyone's preaching about how much one should spend. If a component makes my system damn musical, i don't give a schiit about the science behind that gear. I just enjoy my music. And nobody tells me how much I should spend on any of my gear.

 

[Tablix]

Your money, spend it how you like. I am just saying a recording studio isnt going to spend $100 on a USB cable, so why should you.  After all your USB cable is going to improve the data you bought from them,  /sarcam.off

 

[CanadianMaestro]

The goal of pro rec studios is different than the goal of hobbyist home audio, especially computer audio, where USB is most relevant. Pro studios, or the ones I am familiar with, spend tons of dough on getting an unadulterated, uncolored signal. Power conds, analog ic's.

Sorry for my inflammatory remark.

 

[StanD]

The goal of pro rec studios is different than the goal of hobbyist home audio, especially computer audio, where USB is most relevant. Pro studios, or the ones I am familiar with, spend tons of dough on getting an unadulterated, uncolored signal. Power conds, analog ic's.

Sorry for my inflammatory remark.

A pro studio is a business and has a budget. They do not waste money on cables made from unobtainium. Most of their wiring is of standard lines/impedances, many times 600 Ohms. They use good quality affordable cable.The total lengths of cable they use would put them out of business if they used some of the overblown hype ridden overpriced cable that some audiophiles hyperventilate about.

 

[CanadianMaestro]

A pro studio is a business and has a budget. They do not waste money on cables made from unobtainium. Most of their wiring is of standard lines/impedances, many times 600 Ohms. They use good quality affordable cable.The total lengths of cable they use would put them out of business if they used some of the overblown hype ridden overpriced cable that some audiophiles hyperventilate about.

You should email exactly what you wrote, to Astoria Studios, DG, and others, run by very competent engineers responsible for some of the most beloved recordings ever put on media, like Dark Side of the Moon, countless jazz classics, etc. then wait for a response from them, should be very eye opening. The idea of using $10 cabling or whatever behind and in front of electronics costing hundreds of thousands, is really sensible....

 

[Tablix]

USB is not an analogue cable, the end. I have worked in maybe 20 studios, and data cables are data cables.  When it comes to an XLR cable they will spend the money, but when it comes to data they understand that data either arrives or it doesn't.  USB is transferring data, the weak link is the DAC or the ADC, and not the cable.
EDIT: I am not claiming to have worked with Pink Flyod, but I have worked with numerous local bands and recorded my own tracks.... and at no time have I used anything but a generic USB cable.