Audiophile USB Cable?

[Ammonite Audio]

Now that my HiFace Evo has arrived, I do need a suitable USB cable of around 1.5m to connect it up to the computer. So, do I go for an audiophile cable, or a boggo cable for a few pounds, or something in-between? I admit to being a cable believer, and I have even experienced (although very briefly) a Cryoparts (http://www.itemaudio.co.uk/cryoparts_usb.html) USB cable - the difference between that cable and another was audible, but whether it was better is not clear in my memory. That Cryoparts cable strikes me as being sanely priced.

So, who has any experience of USB cables for audiophile applications? Is it necessary to use a cable better than the cheapo ones; and if so, how much extra expenditure is warranted? Is there a point beyond which no further benefits accrue?

As this is a largely subjectivist forum, I do hope that I am not about to ignite a flame war here!

[Alex_UK]

Our good friends Belkin came up trumps for me - I doubt it makes much difference to the sound to be honest, but you get a chunky quality cable, gold plated connectors, and pretty blue LEDs! All for under three quid delivered! :eek:

http://cgi.ebay.co.uk/BELKIN-GOLD-USB-A-B-MINI-CABLE-1-8m-LED-STATUS-/270591754749?pt=LH_DefaultDomain_3&hash=item3f008681fd#ht_1341wt_905

[WAD62]

I suppose it must depend on whether you're playing synchronously or asynchronously via the USB.

[Ammonite Audio]

I suppose it must depend on whether you're playing synchronously or asynchronously via the USB.

The HiFace is an asynchronous interface.

[WAD62]

Then I wouldn't have thought the USB cable quality would matter that much, as it's really just performing an FTP to the DAC. Does the quality of your printer cable affect the quality of the print etc?

...I'll stand back whilst someone corrects me, as this is speculation. ;)

[Vincent Kars]

Sending bits over a wire is sending a block pulse.
This square wave will degrade with the length of the cable.
The higher the frequency, the more critical the cable quality.

This is a nice test.
Take your audiophile USB cable.
Connect it to your DAC.
Almost all USB DAC's use USB class 1 audio.
The USB will run at 12 Mbps
Connect the same cable to a high speed device e,g, a USB HD
Now the USB will run as 480 Mbps
If this fails you know that this cable is a very bad digital cable not even USB 2 compliant.
This might explain why some of these audiophile cables do have a sonic signature.
Constructed so badly, they probably create a hell of a lot input jitter even at low speed.

[WAD62]

...and do asynchronous DACs use class 1 USB?

Surely not, time for some digging!

[WAD62]

Some good info here...

http://www.thewelltemperedcomputer.com/HW/USB_DAC.htm

[Vincent Kars]

>>and do asynchronous DACs use class 1 USB? <<
Most of them do.
Until recently, any DAC not requiring installing a driver at the computer side uses the USB audio driver of the computer. This is a USB class 1 audio driver.
Today OSX and recent Linux distro's come with a native mode USB audio class 2 driver.
The higher sample rates (176/192) supported by class 2 requires high speed so 480.

The mode (adaptive or asynchronous) can be used in both class 1 and 2.

[Mark Grant]

Another example.

At home I have a USB mobile phone dongle as backup to normal wired broadband.
This needs to be on the end of a 5 metre USB cable so it can be in the loft to get a decent mobile phone signal.

I tried 4 different 5 metre USB cables before I found one that would work with the dongle.

All the ones that did not work with the dongle worked perfectly when tested with a laser printer and continuity tested with a multimeter.

Probably because the dongle was drawing power along the cable and some where too thin to work reliably, but shows that not all USB cables are the same, regardless of what you read on the internet.:lol:

Not saying the sound will be different, but all cables not made the same.

A DIY job would be possible with some connectors and suitable cable, not many wires needed so should not be at all difficult.

Mark.

[Vincent Kars]

Any experience with one of these? http://www.wireworldcable.com/categories/usb_cables.html

They claim USB up to 7 m!

[Welder]

“So, who has any experience of USB cables for audiophile applications?”

It all started to go wrong here; unless you can hear digital data transfer that is.

It’s that word “audiophile”. If your one of these then of course there’s a difference; night and day, huge, miles better, and a queue of vendors more than happy to take as many hundreds of pounds off you as you are prepared to spend.

Of course you will hear the difference; well, I mean, you would look a bit of a pratt if you had spent £200 on a USB cable when you could have spent £5 and you couldn’t tell the difference wouldn’t you :rolleyes:

@vincent.
Yep tried wire world ultra violet. Belkin gold 1.5m "sounds" just as good at a fraction of the price. As you suggest, its the construction quality and of course the length that makes a difference.

[WAD62]

FYI. Another nice article here, which outlines the difference between 'Adaptive' and 'Asynchronous';

http://www.computeraudiophile.com/content/Asynchronicity-USB-Audio-Primer

[WAD62]

...and finally Cyril

How network players achieve the same asynchronous end...

http://www.computeraudiophile.com/content/Asynchronous-USB-vs-Network-data-transfers

P.S. Hiface recommend their own cable at £5, so I think the answer is there, I'd say if it works at all, it's working perfectly, it's just FTP after all.

[Welder]

Oi Vincent, are you paying this Will chap or something :lolsign:

[WAD62]

Oi Vincent, are you paying this Will chap or something :lolsign:

Oops must learn to read, better put my specs on

[lovejoy]

I spent £25 from Ebay on a Kimber USB. I also spent £5 on a 0.5 metre Belkin Pro cable. I've also tried a standard USB2 certified cable.

I have an Arcam rDAC with async USB. The Kimber sounds far better than the Belkin which has a rather hard sounding top end. The Kimber gets rid of all of the digital glare in the system and sounds really natural. The Belkin is marginally better than the standard cable.

Ridicule these statements all you like. I thought the same thing before I actually sat down and listened for myself. I've also demonstrated this to non-believing friends who were not told which cable was which and they all agreed the Kimber was a marked improvement.

[Vincent Kars]

Oi Vincent, are you paying this Will chap or something :lolsign:

Because he refers to ComputerAudiophile instead of the Well Tempered Computer?:steam:

[WAD62]

I've just noticed Vincent's signature honest!!!

God you have to tread carefully round here...:)

[WAD62]

I spent £25 from Ebay on a Kimber USB. I also spent £5 on a 0.5 metre Belkin Pro cable. I've also tried a standard USB2 certified cable.

I have an Arcam rDAC with async USB. The Kimber sounds far better than the Belkin which has a rather hard sounding top end. The Kimber gets rid of all of the digital glare in the system and sounds really natural. The Belkin is marginally better than the standard cable.

Ridicule these statements all you like. I thought the same thing before I actually sat down and listened for myself. I've also demonstrated this to non-believing friends who were not told which cable was which and they all agreed the Kimber was a marked improvement.

Observation is better than speculation...

Does the rDAC use the USB as a power supply, or does it have its own?

The only other option would be that the rDAC is not truly asynchronous, IMHO ;)

[lovejoy]

The rDac has it's own power supply. The async USB is licensed from dCS. From the Arcam website:

* dCS Asynchronous USB technology used under license. Patents pending GB0817141.5 and US 12/63804.

I guess a look at the patents will tell for sure, but I believe it is true asynchronous transfer.

[Welder]

“I spent £25 from Ebay on a Kimber USB. I also spent £5 on a 0.5 metre Belkin Pro cable. I've also tried a standard USB2 certified cable.

I have an Arcam rDAC with async USB. The Kimber sounds far better than the Belkin which has a rather hard sounding top end. The Kimber gets rid of all of the digital glare in the system and sounds really natural. The Belkin is marginally better than the standard cable.

Ridicule these statements all you like. I thought the same thing before I actually sat down and listened for myself. I've also demonstrated this to non-believing friends who were not told which cable was which and they all agreed the Kimber was a marked improvement.”

http://www.mediafire.com/imgbnc.php/170346ca7b8039d3db4180efa76de7389605d6a5de6737bc00 0b769a32a350fb2g.jpg (http://www.mediafire.com/imageview.php?quickkey=7bxmwgupcyajdgd&thumb=5)

[WAD62]

John,

I wouldn't dream of ridiculing tested observations, I've heard my Chord Chamaeleon interconnect go from tinny hell to a beautiful rich sound just from running it in over a week...how the hell does that work? I don't know, but it does...

Perhaps I'm at fault for applying an IT term to the audio market, in too simplistic a manner.

If it were truly asynchronous the DAC would either get the file as is (or segments of), into its buffer (or not at all), and then the DAC would feed from there.

For there to be any degradation in the sound quality due to the USB cable quality, something else must be occurring 'real time' along the cable...:scratch:

[lovejoy]

For there to be any degradation in the sound quality due to the USB cable quality, something else must be occurring 'real time' along the cable...:scratch:

Who says that there is anything affecting the digital signal? The signal between the computer and the DAC may be digital pulses, but it's still a series of analogue square waves and subject to the same laws of all analogue signals, so RF for one is going to be a major problem. The computer is generating loads of it and there's no galvanic isolation between the computer and DAC so the DAC will be picking it up. Different cables are going to be affected by this noise in different ways. There's probably a million and one other reasons that we neither know about or have the capacity to measure yet. I don't see this as a suspension of disbelief.

[WAD62]

...and therefore it's not the asynchronous data transfer that's affected, but that the lower quality cable could expose the DAC to influences that will adversely affect its conversion performance?

Oh dear I'm out of the shallow end...

Ah I think I see your point...:eyebrows:

[Welder]

Who says that there is anything affecting the digital signal? The signal between the computer and the DAC may be digital pulses, but it's still a series of analogue square waves and subject to the same laws of all analogue signals, so RF for one is going to be a major problem. The computer is generating loads of it and there's no galvanic isolation between the computer and DAC so the DAC will be picking it up. Different cables are going to be affected by this noise in different ways. There's probably a million and one other reasons that we neither know about or have the capacity to measure yet. I don't see this as a suspension of disbelief.


Ah well, that must be it then.
You see, I though asynchronous USB sent discrete data packets which are nothing like an analogue signal.

Well my Dac has got galvanic isolation between computer and analogue stage and so I believe have other USB Dacs.

One would have thought that if all this noise was getting picked up by the cable then it might interfere with the data transfer; or does all that noise get carried through the up sampling process, through the analogue stage and then is still audible after its gone through a shed load of electronics and a speaker crossover

[Vincent Kars]

Who says that there is anything affecting the digital signal? The signal between the computer and the DAC may be digital pulses, but it's still a series of analogue square waves and subject to the same laws of all analogue signals, so RF for one is going to be a major problem. The computer is generating loads of it and there's no galvanic isolation between the computer and DAC so the DAC will be picking it up.

There are DAC's like the Ayre QB9 using optocouplers between the USB receiver and the DAC (the PCB) to avoid this.

[Sketchy]

guys, all digital cables need to do is produce 1's and 0's

its not like analogue cables where they can transmit more sound, all digital cables need to do is produce the 1's and 0's and regardless of brand or if the manufacturer tells you their is higher quality or what not, the signal will be the exact same 1's and 0's

anyone who tries to tell you any different is talking out of what they sit on all day.

so to answer youyr question, there is no better sounding USB cable than another, there is better looking but even gold plated ends makes no difference, its all 1's and 0's any difference people say they find in the sound is 100% theoretical bull that doesnt make the sound any better.

sorry for the rant, I kinda know a fair bit about this kinda thing, having worked for foxconn (hon hai precision company T.W.) for 5 years in the past and they making 99% of the parts of all USB cables on sale in the world today.

[WAD62]

There are DAC's like the Ayre QB9 using optocouplers between the USB receiver and the DAC (the PCB) to avoid this.

A question therefore for you Vincent, I'm running a squeezebox receiver into an Audiolab 8000DAX via a BNC connection.

The asynchronous part is therefore sorted, however as the squeezebox receiver itself is a mini linux computer (well to some extent anyway), would I be better off using a good quality 'optical' connection into the DAC, replicating the effect of the optocoupler?

Or would I be robbing 'peter to pay for paul' as it were?

[Sketchy]

Not saying the sound will be different, but all cables not made the same.

Mark.

using a powered hub instead of the port on your computer will have had the exact same effect.

all a digital connection does is send 1's and 0's to the DAC (or whatever you are plugged in to) the DAC (as a example) then turns them into audio based on the digital information sent to it.

digital is linear, unlike analogue, the only difference in digital cables is how they look from a sound point of view, if the 1's and 0's get through then the sound will always be the same.

[WAD62]

digital is linear, unlike analogue, the only difference in digital cables is how they look from a sound point of view, if the 1's and 0's get through then the sound will always be the same.

Hi Gary,

The data transfer part is not what's in question here, see my earlier posts, what I believe the other chaps are highlighting is the fact that a rubbish piece of wire connecting your DAC to your computer might introduce some effects that may adversely affect the DAC chip's performance when doing the actual conversion to analogue (RF and the like).

[Mark Grant]

using a powered hub instead of the port on your computer will have had the exact same effect.


I was talking about a USB mobile data dongle :) , with various different cables that are advertised as ' certified' some of them dont work with the dongle, one of them does.
Tried it with a powerd hub, just the same, most likely some of the 'certified' cables are made from thin wire so volt drop is too high to fully power the dongle or poor build quality of the cable.

'Certified' does not mean very much these days, too many false claims are made.

[DSJR]

As I remember, the square-wave pulses of a digital signal can become corrupted by interference on the line, causing ringing of the flat-top of the wave and possibly errors in the accuracy of the "leading edge." These issues may have been dealt with these days for all I know, but a computer buff friend of mine needed a lot of convincing where digital audio streaming is concerned, as timing is an issue, unlike computer data, where almost random chunks are sent from one box to another and, as long as the info gets to the far end intact, timing (to a degree) isn't an issue.

[WAD62]

As I remember, the square-wave pulses of a digital signal can become corrupted by interference on the line, causing ringing of the flat-top of the wave and possibly errors in the accuracy of the "leading edge." These issues may have been dealt with these days for all I know, but a computer buff friend of mine needed a lot of convincing where digital audio streaming is concerned, as timing is an issue, unlike computer data, where almost random chunks are sent from one box to another and, as long as the info gets to the far end intact, timing (to a degree) isn't an issue.


Hi Dave,

Read the earlier bits about synchronous v asynchronous...

[Mark Grant]

if the 1's and 0's get through then the sound will always be the same.

In theory yes and I agree and that's why I use simple 75 ohm Belden coax most of the time.

Some systems do appear to show differences with some designs of cable.
In theory there should not be a difference but have a listen to many peoples systems and hear.

Thats the good thing about this forum, people invite each other to visit for a listen :) :cool:

[Sketchy]

Hi Mark, I can relate to your 5 meter USB signal issue as 5 mtr is the longest supported length of USB cable before you need a hub to extend it.

hence why I said about the powered hub part, you should only use a 5mtr length with another hub at the other end, or you potentially risk signal loss over distance regardless of cable type.

that same issue will not happen with cables under 5 meters, 5 meters is the max recommended for the USB standard (although the newer USB 3 is 3 meters if memory serves me correctly)


Hi Gary,

The data transfer part is not what's in question here, see my earlier posts, what I believe the other chaps are highlighting is the fact that a rubbish piece of wire connecting your DAC to your computer might introduce some effects that may adversely affect the DAC chip's performance when doing the actual conversion to analogue (RF and the like).

Hi Will, if the dac receives the numbers (1' and 0's) it converts them into audio.

digital only works if it gets the numbers, its like morse code, if the line is bad and you still hear the beeps you hear the beeps and decode the mssg.

the way any digital signal works is it will ask again for information it missed until it has the information it needs, if it lost one number it will request that number again from the DAC, the only time interference becomes an issue is if no numbers can get through or if interference is applied before the DAC via the analogue cable.

to make a USB cable you have to pay intel royalties, it has to be to a very minimum standard to get the licence to make it, as a example a cheap as chips USB cable trade from taiwan or china costs about 18p for a 1mtr length, id happily pay 100 times that retail for a cheap cable and know it will work right whilst not being the cheapest thing you can buy.

paying more than 100 times that price is just dumb as it simply will not enhance the sound (if thats what you are trying to achieve)

[Reid Malenfant]

Hi Gary, i wouldn't dismiss cables as being problematic in the digital domain ;)

In reality it all depends on what type of DAC you have imo. Digital is dependant for timing as Dave pointed out & cable capacitance can & indeed does round off the leading & following edges of any digital waveform. At this point in time the DAC doesn't know how to time a damn thing as it could well be receiving waveforms approaching sign waves. If it can't re-time things dead accurately then time related distortion will be introduced, ie jitter.

As far as i know there is only one way round this & only one manufacturer uses it. The DCS ring DAC selects delays for each sample to keep all of them in time with each other & thus gets rid of (or minimises) jitter.

I'm no expert but the difference between these rounded off square wave 1s & 0s means a crap or non existant picture via HDMI or a rather pleasant one :eyebrows:

No-one can say that audio wouldn't be the very same in all honesty. I find cables that i use between CD transport & DAC etc are much more important than analogue cables as they need to cover as vastly wider bandwidth without corrupting those all important edges ;)

[Mark Grant]

to make a USB cable you have to pay intel royalties, it has to be to a very minimum standard to get the licence to make it,

That's interesting, I did not realise intel where cashing in on every officially made USB cable. !

There will be factories that make stuff unofficially to a very low standard, which is a shame but most people want cheap.
It is the same with HDMI cables, there are rules and fees to be paid, but they are not enforced very well, which makes a mockery of the fee situation.

[Sketchy]

That's interesting, I did not realise intel where cashing in on every officially made USB cable. !

yes it was a intel employed engineer who invented it, its their proprietary technology that needs licensing.


There will be factories that make stuff unofficially to a very low standard, which is a shame but most people want cheap.

but why when it costs about 18p per 1 mtr length to buy officially there is no point, I doubt there will be many people buying anything cheaper (although there are lots of idiots out there who purchase for disties/high street retailers/etc so I wouldnt be surprised)

[Sketchy]

Hi Gary, i wouldn't dismiss cables as being problematic in the digital domain ;)

In reality it all depends on what type of DAC you have imo. Digital is dependant for timing as Dave pointed out & cable capacitance can & indeed does round off the leading & following edges of any digital waveform.

"can"

"indeed does"

im sorry but its numbers with a dac, if the number gets through its a number, its not a slightly rounded off number that sounds differently or has a accent, digital doesnt work like that, it is linear, it is either on or off, digital numbers cannot stray between 1's and 0's there is no 0.5 due to corruption or interference.

if the numbers get through the signal is output from the dac, its not like analogue at all where a high bandwith high conductivity cable WILL and DOES make certain differences, its just numbers, its not anamorphic like a phono interconnect, its either on or off, if the signal is broken before it has recieved the data it requests again, in all my years I have never ever ever heard of any USB cables "sounding" different.

but if you want to throw money about I have a rock in the garden that scares away tigers, I know this is true because there has never been a tiger in my garden, you can buy it if you want but it will cost you dearly ;)

[StanleyB]

Every time I read about rounding off of leading and trailing edges in a digital signal I can't stop grinning. PLL (Phase Locked Loop) circuits that are used to decode the bitstream are not concerned about how square or round the waveform is. It's instead looking for a gap between two waveforms, which could be triangular waveforms, sawtooth, sinewave, etc. Those are the timing period, which the PLL use to recreate the data.
The amount of pseudo-science that crop up in some allegedly informed statements can be painful reading at times.

[Ammonite Audio]

I did ask if anyone had any experiences with different USB cables:( I for one do not care about the science of the digits - I have heard enough real differences (benefits indeed) with regard to digital cables to know that something, somewhere matters, so I am completely open-minded in pretty much all areas of audio. So, has anyone (apart from Rich who has kindly related his experience) got any personal experiences, either positive, negative or neutral, to add?

[Clive]

I have 2 options for placing my laptop, one involves using a 1m cable, the other two 5m cables joined via an active repeater. There's a very noticeable difference in SQ, especially so since I did the passive mod on my 7502.

[John]

I found some differences I got a kimber and last time I said I heard a difference I got shot down but I also have one free cable that performs at the same level so not sure what going on. The other printer usb leads I got sound pretty crap

[WAD62]

Hi Will, if the dac receives the numbers (1' and 0's) it converts them into audio.

digital only works if it gets the numbers, its like morse code, if the line is bad and you still hear the beeps you hear the beeps and decode the mssg.

the way any digital signal works is it will ask again for information it missed until it has the information it needs, if it lost one number it will request that number again from the DAC, the only time interference becomes an issue is if no numbers can get through or if interference is applied before the DAC via the analogue cable.



Thanks for enlightening me...

How I ever managed to define and implement a reliable asynchronous message transfer system for IBM mainframes, via a half-duplex protocol, back in the eighties (prior to the advent of MQ) is beyond me, must have been a stroke of luck.

I suggest you read about the differences between synchronous and asynchronous USB DACs, there are some links at the start of the thread.

We're not talking about data transmission here, an asynchronous DAC will always receive the data perfectly into its buffer, it can then process it as it sees fit, it's then that we are into digital to analogue conversion stuff...and that's where I dip out, square waves, saw tooths, RF interference, I'll leave that for people better qualified than I.
;)

[Vincent Kars]

if it lost one number it will request that number again from the DAC
I’m afraid this is not true most of the time.
SPDIF or Isochronous USB are unidirectional.
These protocols don’t offer any possibility for a retry.
Even networking protocols typically used for AV as UDP don’t offer a retry.

[Peter Galbavy]

A USB cable is either a USB cable - standards compliant etc. - or it isn't. There is no such thing as "audiophile" grade.

The reason there are length limits in the standards that specify various cable types, USB included, is because the people who specified these things actually put thought and science into it and accounted for loss and smearing and recovery and all the rest.

[WAD62]

A USB cable is either a USB cable - standards compliant etc. - or it isn't. There is no such thing as "audiophile" grade.

The reason there are length limits in the standards that specify various cable types, USB included, is because the people who specified these things actually put thought and science into it and accounted for loss and smearing and recovery and all the rest.

Do you understand the difference between synchronous and asynchronous USB DACs ?

[Welder]

Are you going to explain it to us then Will?

Perhaps you could explain what influence the cable has using different protocols while your at it.

I would also like to know what degree of galvanic isolation is needed to prevent any noise at any frequency from reaching the decoder.
Wait, there’s more. Can a decoder recognize a partial zero or one and are there such things as noisy zeros and ones; ie are the power rails the only transmitter of noise?
If the timing is calculated by the gap rather than the wave form and the buffer stacks both gaps and bits in theory at least jitter (time dependant) should in theory be non existent with asynchronous data transfer yet jitter is still measurable. Is this because the Dac clock isn’t accurate enough?

I’ve read lots about this stuff and there is still loads I don’t really undertstand.

[WAD62]

John,

In the asynchronous model the DAC is feeding from its own local Buffer, the originating computer is now out of the equation, in terms of processing, and is only subsequently used as a slave file server, all timing and processing is local.

As to what the DAC then does with it I leave to more learned follows like yourselves.

A micky mouse analogy;

I have a book that I want you to read;

Asynchronous; I send you the book, onece I've received confirmation that you got the book my job's done, and you read it at your leisure.

Synchronous; I send the book to you one word at a time, and I have to keep sending you the words at a pace that allows you to read it coherently, my job is only over once you've read the book.

[Welder]

I wasn’t taking the piss Will. I really do have an awful lot of questions.
You see, I think StanleyB is right about the gaps and not the bits being used for timing.
(there you go Stan you can’t say I never agree with you ;))

If we are right then I can’t see what influence the shape of the wave has either unless there is something in the theory that not all ones and zeros are equal but are still recognized. This begs the question how unequal do they have to be before the decoder rejects them and what effect does this have on the conversion if any.

Oh arr, I also want to know why some USB ports are quite happy supplying 100mA packets to a Dac that requires continuous power of above 100mA and some it seems are not.
And, why does a USB gate close given the same instructions under USB power but if you feed it continuous 500 mA power via a hub it seems to stay open but in both cases the Dac works?

[WAD62]

I wasn’t taking the piss Will. I really do have an awful lot of questions.


No problem John, the data portion I understand, the conversion part (or the reading of the book) is a black art to me...I watch with interest. :)

[Reid Malenfant]

I wasn’t taking the piss Will. I really do have an awful lot of questions.
You see, I think StanleyB is right about the gaps and not the bits being used for timing.
(there you go Stan you can’t say I never agree with you ;))
Ok John :)

Perhaps you ought to think of these rounded off waveforms again :eyebrows: If you draw a few square waves & round off each & every corner whether they be at the leading or trailing edges you'll soon see that even if the PLL takes it's timing from the zeros you still need to establish at exactly what point constitues zero no?

Is it exactly zero volts or is it 0.1V? Or whatever, don't forget the trailing edge of that square wave will be rounded off at both the top & bottom when it approaches zero ;)

So imo it's important to keep the square wave edges as sharp as possible to reduce jitter.

Has anyone thought about asking anyone who owns a Caiman or other DAC that has USB input as well as S/PDIF etc which input sounds better if any?

[Rare Bird]

I have a brand new Kimber USB (Type 'B') cable & HRT (High Resolution Technologies) 'Streamer II'..Sound fantastic, deffo the best streamer i've ever heard regardless of the cost.The cable does sound better than my printer cable.

[Welder]

I see the analogy Mark but……
Say you have this square wave on a scope. Give it a slope on either the leading or trailing edge, the slope when it hits the mean or base line is then off and there is no wave until the next. I thought that was the beauty of digital. You can shape the wave, chop bits off even, but at zero volts if you will, the wave has passed and the zero reading starts. On and off; never mind the shape of the on :scratch:

Wait, you're saying the ones are of differing lengths because of the wave deformation and this alters the gap where zero is measured. If that was true the ones would all be different. Surely if the decoder only accepts ones there must be a definition of what one is as far as the decoder is concerned. If your getting lots of “different” ones wouldn’t the decoder reject them?
From what I can remember the zeros don’t have a dimension in digital data, they just exist. Its how they are fed that influences the timing and if they are stacked then any irregularities in the feed are ironed out. Isn’t that what asynchronous transfer is supposed to ensure?

[Reid Malenfant]

I can't remember exactly the voltage used on the square wave output on an S/PDIF. You can bet your bottom dollar though that a one will be regarded as a voltage above a certain point & a zero will be regarded as a voltage below a certain point.

Quite similar to a schmitt trigger in that respect ;)

I see what you are getting at though, but i don't think it works that way :scratch:

[WAD62]

Its how they are fed that influences the timing and if they are stacked then any irregularities in the feed are ironed out. Isn’t that what asynchronous transfer is supposed to ensure?

I believe so John, if it's truly asynchronous, it will be feeding from its own buffer pool/cache, which theoretically should have many seconds worth of data stacked up and ready for processing, hence the need for local processing capabilities, and a more complex driver in an asynchronous setup.

If I drop my router my spueezebox will still happily play for about 20 seconds after I've done it (Ok not USB but still asynchronous).

[Peter Galbavy]

Do you understand the difference between synchronous and asynchronous USB DACs ?

As much as the differences being sync and async comms systems, yep. So ? A USB cable is either a USB cable or defective/counterfeit. It either works within spec or it doesn't. Failure is not an on/off thing, it's a "does this cable meet the specification ?"

[Peter Galbavy]

The amount of pseudo-science that crop up in some allegedly informed statements can be painful reading at times.

And I pronounce it "audiofool".

[Vincent Kars]

Maybe this link is of use: http://www.audioasylum.com/forums/pcaudio/messages/7719.html

[lurcher]

John,

In the asynchronous model the DAC is feeding from its own local Buffer, the originating computer is now out of the equation, in terms of processing, and is only subsequently used as a slave file server, all timing and processing is local.

As to what the DAC then does with it I leave to more learned follows like yourselves.

A micky mouse analogy;

I have a book that I want you to read;

Asynchronous; I send you the book, onece I've received confirmation that you got the book my job's done, and you read it at your leisure.

Synchronous; I send the book to you one word at a time, and I have to keep sending you the words at a pace that allows you to read it coherently, my job is only over once you've read the book.

Not true I am afraid. USB be it async or not still is controlled by the host clocks. the host still sends a burst of data every 1ms (for USB 1) all async adds is the ability of the device to signal to the host that it should send more or less bytes in the next packet. The host can then clock the data out of each burst at its only local low jitter clock rate and by controlling the amount send every ms avoid the buffer running empty or out of space.

The is no retry or confirmation that data is received with isochronous transfer (how standard audio is handled).

[WAD62]

Not true I am afraid. USB be it async or not still is controlled by the host clocks. the host still sends a burst of data every 1ms (for USB 1) all async adds is the ability of the device to signal to the host that it should send more or less bytes in the next packet. The host can then clock the data out of each burst at its only local low jitter clock rate and by controlling the amount send every ms avoid the buffer running empty or out of space..

If the host is in control of the process then it can not be said to be asynchronous.

Please refer to this article, as posted on the first page of this thread http://www.computeraudiophile.com/content/Asynchronicity-USB-Audio-Primer

Thanks

[Vincent Kars]

The mechanism as describe by Lurcher is correct.

The PC generates the clock of the bus (12 MHz).
The PC send packets (frames) over the bus.
These packets can contain a certain amount of data.
In case of adaptive mode the PC is in control of the amount of data.
The DAC needs to keep pace with this data flow to avoid over/under run.
This is adaptive mode.
In asynchronous mode there is feedback loop.
Now the DAC is in control of the amount of data send.
This allows for a DAC with a fixed clock speed as it don't have to adapt it self to the data flow.

[Ali Tait]

Nick knows his stuff.He has designed and built his own dacs.

[Welder]

I’m getting confused here (asynchronous only)
If Vincent’s explanation and others is correct then Nick’s can’t be right, or maybe I’m still misunderstanding something here :scratch:

Two clocks, a Dac with a clock and a host (the pc or server) with a clock.

In Nicks version the host is sending a burst of data at a regular 1ms interval to the Dac buffer (irrespective of what the Dac clock says) and the Dac clock just tells the pc clock how many bytes to send.
I think that’s adaptive mode :scratch:
So what happens when the buffer is full? Which clock says stop or go? The PC can’t continue to send at regular intervals to a full buffer can it? :scratch:

If it’s the Dac clock controlling both the quantity and timing of the data of the data packets then that is what I understand to be truly asynchronous.
I thought in asynchronous mode the Dac clock not only dictated the number of bytes sent to the Dac buffer but also the timing (when to send) and that is what gave the modes reputed advantage with regard to jitter? :scratch::doh:

[StanleyB]

Unfortunately some of the stuff I have read about the subject on this forum are mere verbatim repeats of what has been claimed.

Let's correct that, shall we? I'll take as an example the PCM2902 USB chip, which I happen to use. It has its own 12MHz clock, and a data storage buffer that is loaded with music data by the PC. The audio data coming out of the PSM2902 is independent of the audio data being sent from the PC to DAC USB chip. As a matter of fact, the USB chip only starts to play music when it detects a new arrival of data from the PC into the buffer. i.e. unlike what is suggested for adaptive mode, the data from the PC is not decode on the fly into a playback stream. In the case of the PCM2902 there is a 1ms buffer time.

[Welder]

Not really very helpful Stanley.

Of course a great many comments seem to be verbatim repeats of what has been written by other “experts”, that’s how most of us learn :eyebrows:
Given not many of us are electrical engineers and haven’t designed a Dac chip, or done the programming required for the various transfer modes we are bound to rely on the information of those who have :rolleyes:

I gather from past posts that you think asynchronous USB data transfer is a load of hype but I have yet to read from you a satisfactory explanation as to why you think this.
On the other hand, there is a great deal of writing by those who have designed or been involved in the design of Dac chips and done the programming for asynchronous mode for a number of Dacs that have a completely different opinion to you :scratch:

So, what do we poor seekers of information do, believe you or them, or have we all got to go and do a degree in electrical engineering, develop a Dac chip and write our own programs to find out coz you lot who apparently know cant agree? :mental:

[technobear]

Welder,

In adaptive mode the PC is in control of the rate of sending data and the DAC adapts to it. Regardless of what method the DAC uses to generate its own playback clock, it has to be continuously adapted to the rate at which the PC is sending (usually done by using a PLL). The buffer will never overflow because the DAC will always keep up with the PC.

In asynchronous mode, the PC remains in charge of the USB bus speed and the number of packets sent but the DAC takes control of the amount of data sent in each packet. This allows the DAC to run with a (potentially very high quality) fixed clock. If the buffer starts to fill up, the DAC tells the PC to send a little less data in each packet. If the buffer starts to empty, the DAC tells the PC to send a bit more data in each packet.

So in the adaptive case, the DAC keeps up with or adapts to the PC rate of send.

In the asynchronous case, the DAC controls the rate of send from the PC and the clock in the DAC is fixed.

[Reid Malenfant]

So in the adaptive case, the DAC keeps up with or adapts to the PC rate of send.
So in this instance the quality of the audio output is determined by whatever clock is operative in the PC if i'm reading correctly :scratch:

In the asynchronous case, the DAC controls the rate of send from the PC and the clock in the DAC is fixed.
& in this case a far more accurate clock could be used in the DAC itself (something like a Trichord or similar) & audio quality should be capable of being much better due to the better clock...

If that's the case then i at least understand now & can see a definate advantage to Asynchronous conversion. Damn it's effectively what Mark Levinson as well as DCS are using in there top end DACs :eyebrows:

[Welder]

Thank you Chris :)

I realized after I posted that the Dac buffer is never full :mental:
I had it right the first time round and then confused myself; easy done at my age :rolleyes:

Mark, don’t forget HRT. They do their own asynchronous mode programming as well ;)

[Sketchy]

so we are all agreed a usb cable is a usb cable?

super, i'm of to the pub

[Ammonite Audio]

Having started this thread, I'm grateful for the handful of comments about different (or indeed better) musical reproduction when using certain certain USB cables. A question for the computer experts who have been vocal here - have any of you ever tested your science with a bit of listening? Would you allow your science to be tested by your own ears? Probably not. But I am open-minded and am happy to experiment; and to state my experiences here on AoS.

[Welder]

Hi Hugo.

Sorry old chap (being largely responsible for derailing your thread).

I’ve bought or borrowed a few USB cables and I haven’t been able to hear enough of a difference to state any one is better than another. I personally find it extremely difficult to differentiate between any cables but I still make them and still try any I can get my hands on.
I have assumed in the past the shorter the better (0.5m) but I have ended up using a 6’ Belkin Gold because I have convinced myself that if there is a difference between USB cables I’ve tried, over extended listening periods, this cable may just suit my system/ears best.
I have no rational explanation for my choice and if someone swapped cables and I was unaware of it, then I have very little confidence that I would even notice.

With cables, and other components come to that, as audiophiles I assume what most of us are listening for/aiming for is perfect reproduction of the music from our chosen media. To be able to judge if we are closer or further away from that we would need to know what that perfect reproduction sounded like. Frankly, I don’t have a clue so I settle for what I find pleasant and involving which may be nothing like the same for another.

So, if you happen to find a particular cable sounds better in your system that's great. However, that doesn’t necessarily mean it’s a better cable, or that it would suit another system or ears.

So yes, I do listen, every day, but when the listening doesn’t provide a definitive answer then I look to science to give a rational explanation of why this may be. To my mind it’s a bit pointless asking on an internet forum which component is better than another, you have to decide for yourself, by listening, or whatever method you feel most comfortable with and then any positive result is likely to only apply to you.

[technobear]

How could a fancy USB cable possibly make a... oh hang on, someone at the door :)

.
.
.

...and back. It was the postman. He just delivered something called a "Kimber USB cable" :scratch:

Oh well, the system is already warm so here goes nothing :lol:

The DAC is a just run-in NuForce uDAC2 using the standard USB cable that came with it.

The sound is confused, splashy, poorly timed, quite harsh - rough as a badger's arse really. Every time I turn it up, I want to turn it straight back down again. I'm on the point of returning the NuForce frankly as I don't see how this will improve the already excellent headphone output of my laptop :(

Enter the Kimber USB...

:scratch:

:doh:

:eek:

Well smack my gob with a wet kipper. The sound now has timing. My toes are tapping. I can hear individual cymbols again. The soundstage is deeper. The roughness is mostly gone (well what can you expect for a £123 DAC). I can turn it up without wincing now http://www.technobear.btinternet.co.uk/emoticons/techno.gif

To keep things in proportion, this is not even close to Transporter+Caiman. I think it would certainly give Squeezebox+Caiman a run for its money though.

Now what was I saying...

Ah yes, USB cables. :lolsign:

This Kimber USB ain't goin' nowhere :ner:

[Ammonite Audio]

Funnily enough, after reading a couple of positive comments here, I have a Kimber USB cable coming too! I did e-mail Item Audio about availability of the Cryoparts USB cable, but they did not respond, so sod 'em. The Kimber cable is much cheaper anyway.

[gramofone]

Here is a recent article, which discusses asynchronous USB and other issues in relation to computer audio.

http://www.audialonline.com/html/articles/usb2010/

It is written by someone who is capable of both measuring objective performance of circuits AND discerning subjective differences at a very fine level.

Can't remember if he discusses the impact of USB cables on audio performance in this article.

If anyone is interested in knowing his view on the topic I could ask him.

Certainly shielding is a consideration even with digital cables ... a 7m 75 Ohm mil.spec digital cable of mine - with a braided shield - still managed to pick up radio stations until I shielded it with aluminum foil tape.

Braids are useless against RFI, you need foil shielding on your USB cable.

[Rare Bird]

Sorry chaps but my Kimber USB is better sounding than the cheap arse USB! I work by my ears not speks!

[technobear]

In all honesty, this Kimber USB cable has saved the day here.

I thought I was going to be in low-volume audio pergatory until my new DAC arrives. Thanks to the Kimber USB, it's now quite tolerable - really quite decent in fact.

This NuForce uDAC2 is not a bad piece of kit but you're gonna have to budget for a decent USB cable if you wanna hear what it can do.

[Mark Grant]

Interesting :)

Anyone have any links to places that sell USB type A and B cable mount connectors ( not PCB mount) to make a cable ? :)

[Welder]

Isatso Andre.
I’ll see yer bloody Kimber, with or without the pretty ponsey ferrites and raise yer my Belkin Gold at any length; even a ten metres! :ner::lolsign:

[technobear]

Interesting :)

Anyone have any links to places that sell USB type A and B cable mount connectors ( not PCB mount) to make a cable ? :)

You wont find anything worth buying from the usual suppliers.

These need to be of decent construction (no steel) and have precious metal plated contacts, preferrably made of pure copper or silver. And that's before we get to the cable itself. How many shields? Read that article linked above. It's good. This one won't be as simple as sticking a Neutrik RCA on the end of a length of Belden I'm afraid :lol:

[Vinyleyes]

Well I was just asking myself this same question and I found this thread ... and what a thread indeed :) ,,, ..... But to a non techy like me who is thinking to purchase the HRT Music Streamer II USB DAC and who needs a 1mt and 5mt run .. from DAC to preamp and DAC to laptop respectively it sounds like i should go for the Kimber.
Does anybody know a good supplier ?
Oh, and while I'm here can anyone help me with .. how do I stream movies and movie sound from my computer to my AV system HD TV ...
Thanks

[Rare Bird]

Russ Andrews
available in 0.5-5.0 mtr...'B' or mini 'B' USB,

[Vinyleyes]

Ouch !! ,,, I won't be getting the silver ... Thanks Andre :lol:

[Mark Grant]

This one won't be as simple as sticking a Neutrik RCA on the end of a length of Belden I'm afraid :lol:

:) I know :)

I have a large quantity of a very nice cable that might be suitable, it is a US Mil spec silver plated copper teflon insulated twisted pair in silver plated copper screening and a teflon outer jacket and is small size, I just need to find a use for it as I have lots of it :) just no connectors....

Two lengths might be ideal I think for the 4 cores.
Or 1 length for the data pair and another two core cable for the power.

Would only be an experiment.

[Welder]

But Mark, doesn’t it take years of development and expensive test equipment to develop a cable? :eek:
Surely you don’t think you can have a successful product you just knocked up out of a few bits of scrap cable you had lying around?

You’ll need to test for jitter and noise rejection and see if the electrons flow better with silver plating and check the oxegen content of the copper if you use it and, and :eyebrows:

Oh, hang on, all this has been measured already by people with the right equipment and specialist knowledge in the field. Apparently all you have to do is make sure the wires are correctly soldered to the correct terminals wrap a bit of mylar foil around the outside and your good to go with under 5m lengths :rolleyes:

F**k it! I’ve got loads of bits of wire knocking around. See you down Maplins Mark. They sell bare A and B plugs :)


(All I need now is some mug to say how much more depth there is to the sound stage and the highs got all fairy tinkle and I’ll be in the money) :lolsign:

[The Grand Wazoo]

:) I know :)

I have a large quantity of a very nice cable that might be suitable, it is a US Mil spec silver plated copper teflon insulated twisted pair in silver plated copper screening and a teflon outer jacket and is small size, I just need to find a use for it as I have lots of it :) just no connectors....

Two lengths might be ideal I think for the 4 cores.
Or 1 length for the data pair and another two core cable for the power.

Would only be an experiment.

Mark,
Is this cable marked in black print with either:

M27500 -14xxxxxx 12703
or
SS7615-2212S9

??

If so, I've used it for (non-USB) purposes & I have some thoughts on it that may interest you.

[Rare Bird]

I bought a load of wires which were mil spec Teflon coated Ag plated from a guy in the USA, i made a phono to 5 din for the Quad, sounds superb, i'm sure he had multicore in it..

[gramofone]

Let me reiterate regarding digital cables:

I'm afraid mil.spec this, silver-plated that, or teflon-jacketed the next thing doesn't mean a monkey's a... if you pick up police radios or mobile phones in the middle of your Chopin.

Of course they are nice-to-haves, but you need to protect yourself from picking up parasitic elements, particularly if you want place you laptop away from your sound system using longer cable runs of 3 - 5M say.

At these frequencies the wavelengths are too short for braided shields to help you - you will need foil.

[Marco]

Malcolm,

He doesn't listen to Chopin, and prog rock is immune from interference ;)

Marco.

[Mark Grant]

But Mark, doesn’t it take years of development and expensive test equipment to develop a cable? :eek:
Surely you don’t think you can have a successful product you just knocked up out of a few bits of scrap cable you had lying around?

You’ll need to test for jitter and noise rejection and see if the electrons flow better with silver plating and check the oxegen content of the copper if you use it and, and :eyebrows:
:lolsign:

Its just a bit of fun :)
If I make something and it works people can borrow it, would not be worth making them to sell as would take too long to make them affordable.
Cant compete with chinese prices :)

[Bonky]

Stan B recommends a cheap Dell USB lead - with a RFI filter.

See:
http://cgi.ebay.co.uk/SMALL-80CM-A-B-USB-CABLE-FILTERS-DELL-/160509083267?pt=LH_DefaultDomain_3&hash=item255f163e

Worthy of an experiment?

BW

Richard

[StanleyB]

Cant compete with chinese prices :)
You can on quality and technology. But you have to be careful. I designed a few cables years ago under the THOR brand ( the brand is still going strong without me...). Sent the design to a little company called Maxcable. A few months later they started selling similar cable designs to the likes of monster cable. It's a gamble doing business with the chinese.

[Sketchy]

my kimber usb cable also makes me think it sounds better.

the waveforms from the recorded music, and my ears tell me its all in my head (and wallet)

[technobear]

I was a USB cable unbeliever before this thread :mental:

I've swapped a few times between the Kimber and the decent enough looking cable that came with the NuForce uDAC2. It's almost unlistenable with the standard cable but actually sounds pretty decent with the Kimber. I can even turn it up and rock out with the Kimber. If I do that with the standard cable, it's ear shredding.

I bought the Kimber USB after reading the thoughts of others on this thread and knowing that Russ would take it back if I wasn't satisfied. I don't think he gets many back. He isn't getting mine.

[Welder]

I tried the Kimber USB. It didn’t have the ferrites on any more (removed by cable owner)
At the time I compared it to a 0.5m Belkin Pro. I couldn’t hear any difference.

I think if there are any audible sonic differences between USB cables then perhaps such differences could be dependent on the Dac you use.
I think most accept that data integrity is ensured by a correctly constructed cable, of any make. The only possible variation I can believe could possible influence sonic characteristics given the above is the amount of noise an inadequately shielded cable could transfer to the Dac, perhaps.

The two Dacs I’m most familiar (HRT 11+ and HRT Pro) both have galvanic isolation and are asynchronous and I have yet to hear a difference between any USB cable and I’ve tried a few.
So, perhaps if the Dac you use isn’t particularly well designed with regard to noise rejection a cable that may potentially introduce less noise gives the audible benefit some have reported (?)

[lovejoy]

Good to see that people are giving these cables a try and some are finding differences. As with everything else Hi-Fi, YMMV.

I've tried the Kimber with both my rDAC and my modded Caiman and found a significant improvement in both, even over the Belkin Pro cable that lots of people are getting good results with.

I also found that removing the ferrites from the Kimber took some of the thickness in the bass away, so to my ears it's better without them.

[gramofone]

Marco


Malcolm,

He doesn't listen to Chopin, and prog rock is immune from interference ;)

Marco.

Prog. rock - I thought that was interference ...;) ... now a bit of Prog. House that might be nice ...:)

[gramofone]

Has anyone on this forum tried this:

http://locus-design.com/polestar_usb.html

I have read some saying that it is better than the Kimber offering ...

He is making more expensive variants these days, but I don't know of anyone using them.

[Reid Malenfant]

Could anyone let me know where these Kimber USB cables are available from & how much they cost?

A link or two would be great, thanks in advance :)

[keiths]

Russ Andrews

http://www.russandrews.com/category.asp?lookup=0&region=UK&currency=GBP&customer_id=PAA1375024911633WWGFOPBFVFHOPOSH&cat_id=pacccomp

[Reid Malenfant]

Cheers Keith, that's like a load more than i payed for a USB cable & USB to S/PDIF convertor though :eyebrows: Both came to £18.75 delivered from Hong Kong :scratch:

I'll try the thing out feeding a dCS Purcell/Delius combo but it was pretty damn good feeding my old DAC in all honesty :D

Lets just say i'll keep an open mind, but not an open wallet ;)

[John]

Mark it looks like the same cable, definite same blue cover as the Russ Andrews No idea if its the same cable but as you say when the converter works that well who really cares