Signal integrity is a critical issue- dropouts and glitches can cause major problems in industrial application. Audio applications as well- which is why people who actually know about this stuff have said over and over and over and over and over that the symptoms are NOT audio degradation (which would require that the wires somehow magically recalculate numerical values and reconvert them to properly formatted signals), but rather dropouts, pops, and loss of signal.
If you have any of that, change the cable. If you don't, your cable is as good as it can get for this application, comic book audio websites notwithstanding.
If I used silver cable would my bank balance improve?
As a customer, your bank balance will go down. As a manufacturer, with a clever marketing campaign, it will increase substantially
Exactly SY. If, for example, a specific cable would improve, say transient response, what is it this cable needs to do?
It needs to search for specific data bytes in the stream until it finds a data packet of several bytes that encodes for a transient signal, and replace the numerical values in these bytes with values that give a higher level for the spectral components that determine the transient response. All of this on the fly without missing a beat, so to say.
Some cable, huh?
jan
BTW I need more popcorn...
Adds at least one digit, but now if it was gold plated....If I used silver cable would my bank balance improve?
You nailed it on the head. The uninformed masses do not recognize plausible from mythical or is it mystical?
I just read through this thread last night and preface this posting with the fact that I consider myself "a cable skeptic" and strongly believe in a scientific approach to audio engineering.
I also have little respect for those who attempt to profit from the ignorant and "unwilling to learn" audiophool brigade.
With that out of the way, I have little regard for the flat earth audio society who dismiss other people's findings out of hand and and take a closed minded approach to audio engineering.
A sound inquiring mind ( no pun intended) would attempt to replicate the cable described here and attempt to verify or refute the protagonist's claims with good science. What I see instead is a number of destructive posts with no attempt to suck it and see.
It is often said that the ear is the "finest instrument" and I clearly remember the days when the flat earth brigade claimed that all amplifiers with similar measurements sounded the same. Some even claimed that record players using the same arm and cartridge would also sound the same. I often wonder if those folks ever attempted to verify their claims without bias.
Lets accept that some folks here have stumbled upon a yet unexplained phenoma that results in a perceived improvement in sound and build from this observation with solid science.
I also have little respect for those who attempt to profit from the ignorant and "unwilling to learn" audiophool brigade.
With that out of the way, I have little regard for the flat earth audio society who dismiss other people's findings out of hand and and take a closed minded approach to audio engineering.
A sound inquiring mind ( no pun intended) would attempt to replicate the cable described here and attempt to verify or refute the protagonist's claims with good science. What I see instead is a number of destructive posts with no attempt to suck it and see.
It is often said that the ear is the "finest instrument" and I clearly remember the days when the flat earth brigade claimed that all amplifiers with similar measurements sounded the same. Some even claimed that record players using the same arm and cartridge would also sound the same. I often wonder if those folks ever attempted to verify their claims without bias.
Lets accept that some folks here have stumbled upon a yet unexplained phenoma that results in a perceived improvement in sound and build from this observation with solid science.
Last edited:
Good science does not have to be an experimental approach under all circumstances. The fact that a digital transmission line without timing information does *not* change the signal if it is bit-perfect does not need experimental verification. Applying well-known and understood laws of physics is enough in this case.
Imagine someone writing poems on a typewriter. Then he puts the sheets in an appropriately sized cardboard box and sends them to someone else. Would you really pay attention to someone who argues that the poems sound more impressive on being read out loudly when they have been sent in a cardboard box of different construction? As long as the sheets arrive undamaged and readable in both cases?
Rundmaus
There is absolutely no reason to draw such conclusion without objective (i.e. blind) listening tests first. No such have been presented, never before, not in this case either. Once these are on the table, we can start talking about solid science. Until then it is just about excited feelings, nothing to pay serious attention to.
Let us assume for the sake of argument that a properly conducted double-blind trial confirmed that listeners could distinguish between different USB-compliant cables, used in a setup where the USB carried data only. 'USB carrying timing' would have to be excluded as USB (and most personal computer software) was never intended to carry timing - although some seem tolerant of the resultant sound. Non-compliant cables would have to be excluded because of course they could sound different through simple signal corruption - although some may prefer this sound.
What would I conclude? That some USB cables are 'better' for carrying audio than others? No, that conclusion is not open to me as it defies well-established science. My conclusion would have to be based on things like ground current loops, and RF interfence injection into DAC grounds etc. These are all equipment weaknesses, which different cables may expose. I return to the point I keep making in various contexts: cable fussiness is a sign of poor equipment design.
What would I conclude? That some USB cables are 'better' for carrying audio than others? No, that conclusion is not open to me as it defies well-established science. My conclusion would have to be based on things like ground current loops, and RF interfence injection into DAC grounds etc. These are all equipment weaknesses, which different cables may expose. I return to the point I keep making in various contexts: cable fussiness is a sign of poor equipment design.
Last edited:
This may back up the common statement that with commercial hifi kit it may be a good idea to have all components made by the same company.
Not really an option for us diy fanatics though is it.
With so many endless combinations of different kit, some units better than others it MAY be worthwhile trying different cables to see if it makes any difference. I certainly wouldn't encourage people to spend silly money on them though.
I'm keeping an open mind on cable differences from now on and wont be commenting any further on the perceived benefits of trying them or not. Everyone's different with different hearing and kit, what makes you happy is what actually counts at the end of the day.
Not really an option for us diy fanatics though is it.
With so many endless combinations of different kit, some units better than others it MAY be worthwhile trying different cables to see if it makes any difference. I certainly wouldn't encourage people to spend silly money on them though.
I'm keeping an open mind on cable differences from now on and wont be commenting any further on the perceived benefits of trying them or not. Everyone's different with different hearing and kit, what makes you happy is what actually counts at the end of the day.
Great, seems we are settting off in stride with each other.
And... you lost me. It is the people who dismiss outlandish claims who are closed minded? To me, "other people's findings" are nonsense to anyone but the person making them, except when carried out in a proper scientific manner (which in the present context is essentially never).
Good science can never prove that a particular USB cable definitely sounds better, but it sure can prove that a USB cable can make no difference whatsoever. Anyone with a moderately detailed knowledge of how USB (and the computer as a whole) works will understand why a different/better/blessed USB cable can never make any difference whatsoever.
If the USB cable could make that difference when used with a DAC, then it must have the same effect on the audio files when copying them to or from a USB hard drive. The file copied with the Ultra Super Such Wow USB cable will be permanently improved by the cable. But of course you don't believe that. It's nonsense. When you copy a file to or from your USB drive, it is a perfect copy. It's the same data on both sides. Hmm.
Oh let's definitely not do that.
Why? USB isochronous synchronous and adaptive modes are regular USB modes designed by the USB consortium.
There are computers and interfaces which are intended to carry time-sensitive information; they are used by industry and science. Modern personal computers, their software and USB do not fit this category as they are intended for IT, not real-time processing. That is why decent audio will always use USB to carry data only. The fact that the USB spec defines 'audio' modes does not mean that USB can carry high quality audio timing data - I suspect they were thinking of 'computer quality' audio.
It seems to me that people who play with USB audio cables mainly do not realise that they are probably already listening to a degraded system, which works surprisingly well given its inadequacies. Under those circumstances a cable could affect the sound, but then the sound is already affected by the high jitter caused by the weak system. If I tried transmitting high quality audio via a wet string then I might find that string quality and water quality affect the sound, but that is not an argument for better strings but an argument for finding a better audio system than wet strings.
It seems to me that people who play with USB audio cables mainly do not realise that they are probably already listening to a degraded system, which works surprisingly well given its inadequacies. Under those circumstances a cable could affect the sound, but then the sound is already affected by the high jitter caused by the weak system. If I tried transmitting high quality audio via a wet string then I might find that string quality and water quality affect the sound, but that is not an argument for better strings but an argument for finding a better audio system than wet strings.
Pano's mud experiments demonstrated that even the string quality is likely to be unimportant.
USB bus is timed by hardware clock of the USB controller. It does not relate to CPU at all, hence no point in comparing with realtime processing. The model is actually fairly similar to PCI soundcards, just the clock must be PLLed from the 1ms incoming clock, not the 24.xxxMHz internal card clock. Of course the recovered clock quality can hardly compare to precise crystal clock of PCI cards but many of those use PLL for master clock generation too.
I am not saying USB adaptive is ideal for audio, but the transmission clock has no tie to CPU processing and it does use HW-generated clock. Even system with highly loaded CPU provides decent USB clock with either equal (for 48/96/192 samplerates) or deterministic (for samplerates not divisible by 1ms) number of samples in each 1ms USB frame. I did test that, monitoring the number of samples in each usb frame for a lightly as well as heavily loaded linux system.
I am not saying USB adaptive is ideal for audio, but the transmission clock has no tie to CPU processing and it does use HW-generated clock. Even system with highly loaded CPU provides decent USB clock with either equal (for 48/96/192 samplerates) or deterministic (for samplerates not divisible by 1ms) number of samples in each 1ms USB frame. I did test that, monitoring the number of samples in each usb frame for a lightly as well as heavily loaded linux system.
Honestly, I do not see much of valuable information in that writeup - just claims without proper support with measurements/examples/experimental results. Why "A computer sadly cannot maintain perfect timing of the data sent via USB"? E.g. USB controllers in add-on USB cards use their own crystals for USB frame clock generation. I am not saying it is perfect by any means, the 1kHz clock still has to be PLLed to much higher values, but it is not inherently tied to computers (in term of timing of processing operations).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.