If I remember the specs correctly both the clock in the USB transmitter and the clock in the USB receiver allows for use of "bad" clocks with huge amounts of jitter and drift. I replace those clocks with clocks with better than 7ppm and 0.3ps jitter in my systems. Then I use galvanic isolation between the USB transmitter and USB receiver. Even then the USB cable are affecting the fidelity.
One of the comments appearing more and more regarding digital signals and digital engineering is that it is analogue. I believe this is so analogue problems and phenomena can be attributed to digital signal transmission to back up some of the claims regarding differences perceived when digital sourced playback is used.
Signal integrity issues are different between analogue and digital. For example high frequency attenuation on an analogue waveform will reduce the treble output and depending on the level could be quite noticeable, whereas high frequency attenuation of the digital waveform will round the corners of the wave (and as long as not to severe is something to be desired) but the data will still be the same.
The main difference though between the two is that any change to the analogue waveform in transit changes the resultant sound, changes to digital waveforms up to a certain level will not change the resultant sound, the data will still be the same. Where the digital waveform is corrupted so much that the bits cannot be read consistently will become instantly apparent as the whole system will have problems, and drop outs or worse will be very apparent.
That said USB audio is not that taxing speed wise or signal wise and the use of LVDF means that the signal is even more resilient against noise and other signal integrity problems.
As to screening, a cable either does it to the required level or it doesn’t, if it doesn’t it is flawed and should not be used, and also should not have passed CE or FCC qualification and should not be on the market. No point discussing such a cable.
If you are worried about noise from your source affecting your DAC via the cable, measure it, if it is there sort it through filters or isolation, and here Henry Ott can be your friend and design guru.
At the end of the day all the problems that are real can have solutions engineered to solve them.
Signal integrity issues are different between analogue and digital. For example high frequency attenuation on an analogue waveform will reduce the treble output and depending on the level could be quite noticeable, whereas high frequency attenuation of the digital waveform will round the corners of the wave (and as long as not to severe is something to be desired) but the data will still be the same.
The main difference though between the two is that any change to the analogue waveform in transit changes the resultant sound, changes to digital waveforms up to a certain level will not change the resultant sound, the data will still be the same. Where the digital waveform is corrupted so much that the bits cannot be read consistently will become instantly apparent as the whole system will have problems, and drop outs or worse will be very apparent.
That said USB audio is not that taxing speed wise or signal wise and the use of LVDF means that the signal is even more resilient against noise and other signal integrity problems.
As to screening, a cable either does it to the required level or it doesn’t, if it doesn’t it is flawed and should not be used, and also should not have passed CE or FCC qualification and should not be on the market. No point discussing such a cable.
If you are worried about noise from your source affecting your DAC via the cable, measure it, if it is there sort it through filters or isolation, and here Henry Ott can be your friend and design guru.
At the end of the day all the problems that are real can have solutions engineered to solve them.
What type of USB data transmission are you using sync or async?
By the way:
USB Audio Synchronous does not have error check => cannot be seen as bit perfect (only in exceptional cases), anything can happend with the data stream.
USB Audio Asynchronous has error check at USB packet level => can be bit perfect if no packets are dropped. at the spped that a 44.1 16b pcm stream is transmitted between 2 hosts it had to be a verey very very bad cable to affect pcm stream
All the data that flows between 2 USB host is prepacked in USB packets with CRC checksum ..... so you have an additional layer between cable and pcm stream, that assures bit perfect data transmission.
From Universal Serial Bus Specification Revision 2.0 April 27, 2000
Sync asyn Audio USB are not directly related to error correction so the implementation may differ. Syn asyn refers onlty to
And from Universal Serial Bus Device Class Definition for Audio Devices Release 2.0 May 31, 2006 May 31, 2006
Last edited:
We are back to the roots - do you have credible blind test results to back up the claim?
I added 40dB filtration to the USB receiver PSU and still there are audible differences between USB cables.
So regardless of improvements of PSU´s, galvanic isolation, clocks etc. etc. there are audible differences between USB cables.
So your HOW! needs more investigation and research to be answered,
but WHY! are a even better question because then the HOW! could have been answered.
I have stated previously that blind tests are not a tool that will work.
But I understand that it is brought up.
Waiving the "blind test" flag are as I see it more than 100% safe as it is a tool that do not work and not give any result..
OK. So what other objective confirmation of the effect do you offer? You want to discuss technical causes of the phenomen, we want to make sure it actually exists first. Otherwise such discussion makes no sense.
You've made a claim. It is an extraordinary claim. You admit that you have no evidence. Classic onus probandi fallacy, and not for the first time.
As most audiophile misconceptions, it is based on some truth, taken out of its original context. The origin of that saying is probably that, when the clock is embedded in the digital stream (as in spdif or the worse kind of usb transfer), clock recovery is mostly an analog process (PLL).
Nothing to do with data transfer though, as you pointed out.
For the rest, I completely agree with you.
You misunderstand me.
A cable will either send the data correctly or fail and have to send it again. The fact that one data protocol allows this resending and another does not will have no bearing on whether the cable fails in the first place.
So as I said a cable either provides enough data to be entirely recoverable, or it does not. Something can't be 50% recoverable, either there is enough data to error correct and rebuild from or there is not.
a cable either works or it does not. There is no third state.
A cable will either send the data correctly or fail and have to send it again. The fact that one data protocol allows this resending and another does not will have no bearing on whether the cable fails in the first place.
So as I said a cable either provides enough data to be entirely recoverable, or it does not. Something can't be 50% recoverable, either there is enough data to error correct and rebuild from or there is not.
a cable either works or it does not. There is no third state.
This argument will never end while:
1. people seem to ignore or forget that different USB audio transfer methods will have very different properties - the weaker methods (which should never be used, but are used) may suffer from clock jitter which could (just) be cable-dependent, but much more operating system driver dependent.
2. people make unsubstantiated claims of audibility, perhaps with commercial considerations in mind.
USB was designed for delivering data. It is reliable for that, while perhaps not as good as some other methods. Higher level protocols can ensure that the data is correctly delivered. Then timing is in the hands of the receiver. Under these circumstances cables can have no effect, unless there are design flaws in the electronics. So cable sensitivity means one or both of:
1. the wrong USB method is being used - designer error.
2. the endpoints are poorly designed - designer error.
There should be a stigma attached to admitting that your equipment is cable-sensitive; at present in some ignorant quarters it is almost a badge of respect!
1. people seem to ignore or forget that different USB audio transfer methods will have very different properties - the weaker methods (which should never be used, but are used) may suffer from clock jitter which could (just) be cable-dependent, but much more operating system driver dependent.
2. people make unsubstantiated claims of audibility, perhaps with commercial considerations in mind.
USB was designed for delivering data. It is reliable for that, while perhaps not as good as some other methods. Higher level protocols can ensure that the data is correctly delivered. Then timing is in the hands of the receiver. Under these circumstances cables can have no effect, unless there are design flaws in the electronics. So cable sensitivity means one or both of:
1. the wrong USB method is being used - designer error.
2. the endpoints are poorly designed - designer error.
There should be a stigma attached to admitting that your equipment is cable-sensitive; at present in some ignorant quarters it is almost a badge of respect!
No an ordinary claim as a huge percentage of those that have tried to change USB cables or cables in general can hear differences. I do not know the exact % number, but based on the people I know that have checked this I would say maybe 50 to 80% can hear differences and 20% can evaluate the differences to be either positive or negative.
If you are among those that do not have the ability to hear any difference - that is OK.
Please reveal what evidence do you have that can make such statements?
And again not for the first time...
Some measurements to show us what's going on! Cos if your going to add filtration you need to know what frequencies to filter...
Anecdote is not evidence; what's worse is that it isn't even a good anecdote!
I think we can safely group this with your other wild and unsupported claims. Lame retry on the onus probandi; please assume that people here aren't stupid.
If that's how you prove things, you have just proved UFOs exist and dinosaurs and humans lived together. Show us some proof or throw your come td into the opinions pile, that's all they are, and as such are useless.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.