USB Cable type AB recommendation

[sidvee]

Guys wrt what is inside a USB cable here is link to Gordon Rankin's (Wavelength audio) description. He does indicate transmission of noise is possible as well as current running in the cable (not sure if it is Ac/DC).


Guys remember what the cable does and why they make a difference - Gordon Rankin - Computer Audio Asylum

BTW - great deal of information being shared on the thread. I am learning a lot about the humble usb cable! Thanks and keep it coming guys.

Cheers,
Sid

 

[manoj.p]

its all about the interface. The power cannot be bleed into data stream inside the cable. But it can bleed into DAC only IF the USB pin interface is poorly designed. The dac should be designed not to interface the power pin at all. The cable alone can't bleed the power in the data, and the digital data would remain intact.

 

[Audiodoc]

I too believe there would be not much difference if any in case of a digital USB cable though I believe good quality analog interconnects as well as proper gauge speaker cable do impact sound quality.
A standard USB cable with a ferrite core should do fine. Do invest the money in a good DAC instead.

 

[Thad E Ginathom]

Prankey, we all seem to forget that all forms of digital transmissions is error prone. Cat 5, USB, SATA were all designed for speed or convenience and not for accuracy. Why? Simply because, digital transmission is a two way process, and both sides can talk to each other. I am sure you have heard of data loss? Accuracy is achieved though checksum and re-transmission. The receiving size can calculate and make an exact assessment whether the received packet is bit perfect. If not, it simply asks for a re-transmission.

And how often does that happen?

 

[thatguy]

"Cat 5, USB, SATA were all designed for speed or convenience and not for accuracy. "

This one needs citation. As someone who works in the computer/networking industry, I am very surprised by this statement.

Even in Ethernet, collisions/packet drops are a thing of the past.

 

[venkatcr]

The cable alone can't bleed the power in the data, and the digital data would remain intact.

If the cable is not properly shielded, the power being carried so near will affect the data movement. Remember data is also being carried as mild electrical current. If there is interference, the data will also get affected. When I said bleeding, I did not mean that the power will move into the cables carrying data. If that happens, it is short-circuiting. What I meant was interference and corruption of data.

Thad, if you do an analysis of data transmission between any two computers connected on a LAN, you will see data losses. In the days of Novell, we used to actually analyse this on-line, regularly. I am not sure anyone bothers any more. If you look at torrents, for example, they constantly display 'wasted' data through 'hashfails'. As one member said, it may be less, but it cannot be denied that there is data loss in digital transmission.

Cheers

 

[ajinkya]

If the cable is not properly shielded, the power being carried so near will affect the data movement. Remember data is also being carried as mild electrical current. If there is interference, the data will also get affected. When I said bleeding, I did not mean that the power will move into the cables carrying data. If that happens, it is short-circuiting. What I meant was interference and corruption of data.

Thad, if you do an analysis of data transmission between any two computers connected on a LAN, you will see data losses. In the days of Novell, we used to actually analyse this on-line, regularly. I am not sure anyone bothers any more. If you look at torrents, for example, they constantly display 'wasted' data through 'hashfails'. As one member said, it may be less, but it cannot be denied that there is data loss in digital transmission.

Cheers

Venkat,
Sorry to chime in at such a late stage but I think two different issues are being confused here. The data losses that you are talking about are because of packet collisions created due to the same network/cable/data path being shared by multiple hosts. In today's switched Ethernet LANs, this is eliminated. The reason you still get data loss on the Internet is because of the fact that everything is not "circuit" switched in the path from the other computer to yours. This has NOTHING to do with how a particular cable (Ethernet or USB) carries its individual data bits.

Regarding the discussion about interference, the whole point of "Certifying" something to a standard means that all these factors have been measured, analysed, and adhered to. If a cable advertises itself as USB 2.0, it automatically implies that it is capable of supporting the required bandwidth for 2.0 data rates, it has power and shielding protection to do so (which is also specified in the hardware standard for USB 2.0) and that it adheres to the clock specifications of the standard. Of course, if the manufacturer has lied about what the cable does, there is no way of knowing. And that is why all of us prefer to go for a branded cable ( as you and Thad and others have pointed out), because we are paying for the reassurance of quality, NOT because the branded cable has any better data transmission (or audio) qualities, compared to a well-implemented, non-branded cable.

I would also like to point out that unless one has a solid understanding of electromagnetism, which is a subtle and complex phenomenon, one should not try to imagine only current as being the carrier of information, nor should one talk about interference. Unless one understands how EMI occurs, and how much of it is really measurable compared to other sources of noise in the audio chain, all such discussions are superfluous and actually very misleading. I have seen many discussions like this on other forums, where most people are trying to describe electrical properties without having any idea about Maxwell's Laws, which is what a cable acts according to, as a short transmission line. And at the distances we are talking about (Laptop to DAC, DAC to Amp etc.), these effects have to be measured before saying they are causing noise.

Sid, just to add a bit of information to the cable discussion:

We "audiophiles" deride any other standard except pure analog for passing "exact" signals between components. In particular, because most "audiophiles" (and here I mean the Stereophile generation) come from an analog circuits background, they have a difficult time understanding computer-based audio, which is why they convert that misgiving into misleading skepticism about anything digital/computer-based. This is based on my reading many of their articles, and seeing the subtle msgs interspersed.

So, to help solidify their digital nightmares (along with anybody else who still thinks digital is inferior/dangerous/lossy/poorly understood), remember that the next time you fly the Airbus A380, one of the most advanced, safe, and beautiful machines to fly through our skies thus far, you are trusting your life to the Ethernet standard, because the A380 avionics bus is Ethernet- and UDP/IP- based! (Search for "Ethernet" in the link below)

Airbus A380

My point is, if Airbus engineers who deal with EMI, losses, and noise on MILES of cable, which are designed to carry millions of passengers safely,
use and trust the standard Ethernet (along with UDP/IP!) through years of experience, I think as "audiophiles", we should be a little humble and recognise that a few nanoseconds of jitter on a USB cable (if it even exists) will not spoil our lives or our music enjoyment. Our ears are not that sensitive compared to the sensitivity needed by stability control algorithms on a massive airplane, which require much more precision, and much less jitter than the average audio replication chain.

-Jinx.

 

[sidvee]

While we are on the topic of power flow disturbing the data in a USB - please see attached an interesting review of the acoustic revive USB cable that has an unique design that separates data and power
6moons audio reviews: Acoustic Revive USB cables
BTW the cost of the cable approx US$630. Is it a bargain for the unique design and offered benefits or is it Insanity?! Let the jousting (I mean debate) begin:lol:
Cheers
Sid

 

[reignofchaos]

An amazing idea! One can feed the power bit to a clean 5V power generator and the signal bit to the actual device.

However is it worth 600$? Probably not! Something like this can be built for not so much.

 

[venkatcr]

My point is, if Airbus engineers who deal with EMI, losses, and noise on MILES of cable, which are designed to carry millions of passengers safely, use and trust the standard Ethernet (along with UDP/IP!) through years of experience, I think as "audiophiles", we should be a little humble and recognise that a few nanoseconds of jitter on a USB cable (if it even exists) will not spoil our lives or our music enjoyment.

Ajinkya, thank you for the clarifications.

What I was trying to say is that assuming there is no data loss in digital transmission is not correct. That is all.

When it comes to Airbus, please remember, the amount of testing they do for systems TO be safe is enormous. They assume that systems WILL fail and test for recovery from a failed situation. Everything they use is of the highest quality and specially made for them. The cables would be fire proof, heat proof, and who knows what else proof. And everything is backed up 6 times of more. Yes, every cable has 5 backups or more. I am sure all cables are specially manufactured for Airbus and are tested under extreme conditions before being used.

Cheers

 

[thatguy]

While we are on the topic of cables, here is a 1.5m Ethernet cable for $499 that promises to bring out all the nuances from your recordings

DENON US | AK-DL1

And this is Denon, no less.

The following short article talks about some research done on why people like paying more for things:

How being swindled can make you feel better

I work for the computer industry and used to scoff at the techniques used by our salesmen. Until I started exploring audio. The amount of snake oil for sale by AV companies is mind boggling.

 

[manoj.p]

this thread is the epitome of what's wrong at this forum. No offence, OP. I am glad its brought up.

Why do I get a feeling that I stand vindicated.

 

[Asit]

Ajinkya,

I do not have a direct interest in USB cables for my audio needs. I cannot contribute directly to the healthy discussion here. However, I have a few clarifications as follows.

... packet collisions created due to the same network/cable/data path being shared by multiple hosts. In today's switched Ethernet LANs, this is eliminated.

I do not think collisions are eliminated in switched networks. Ethernet/FastEthernet/GigabitEthernet is a broadcast-based network. Whenever there is broadcast, there is bound to be collisions. In a switched network, assuming the switches have level-3 functionality, there is an assured path for a packet from an IP-address to another IP-address in the LAN. The chance of collision is reduced with a smaller LAN. This has given rise to the concept of the so-called VLAN.

Unless one understands how EMI occurs, and how much of it is really measurable compared to other sources of noise in the audio chain, all such discussions are superfluous and actually very misleading. I have seen many discussions like this on other forums, where most people are trying to describe electrical properties without having any idea about Maxwell's Laws, which is what a cable acts according to, as a short transmission line. And at the distances we are talking about (Laptop to DAC, DAC to Amp etc.), these effects have to be measured before saying they are causing noise.

Electromagnetic interference is a common phenomenon. In a lab situation, where there may be strong magnets around, there are pockets where you would lose signal of your mobile phone. When you are addressing an audience on a microphone, try to bring your mobile phone too close to the microphone and you will hear effects of the em interference. At times, during riyaz at home, I put my mobile too close to my electronic tanpura, and immediately hear strange noises. Ambient strong em fields seriously interfere with wifi signals, I see these things on a daily basis (exactly opposite to my office, there is a lab with strong magnets).

Regarding audio signals, try connecting your turntable with an unshielded cable to the phono-stage, there will be tremendous hum due to em interference. Luckily, unlike gravitational fileds, em fields can be shielded.

Agreed, beyond the phonostage, the signal is line-level and a lot bigger as with all other line-level signals (e.g. from CDP to amp), and the effect is not as drastic. But of course there is a finite possibility of undesirable effects, because shieldings are not always ideal.

Now, regarding the em interferences in an USB cable, I do not have a proper idea how serious this effect is. There is no reason to think that just because the wave-form is square-wave pulses or whatever in the digital transmission, there will be no interference, or the transmission will never get affected by the domain-walls (created by imperfections in the crystal structure of the conductor).

I am sorry, I do not have enough first-hand knowledge in computer-based music to contribute to the main discussion here. However, I found your above comments a bit misleading. May be I misunderstood in my late night reading of your post.

Regards.

 

[thatguy]

The reason such threads become long and inconclusive is because people don't have to put their money where there mouth is. E-warriors armed with Google will throw anything at you to justify their (possibly flawed) point of view.

Here is a proposal. Why not arrange a DBT with real money at stakes. If someone claims that USB cables make a difference and can tell a Rs 500 USB cable from one made of unobtanium just by hearing, he gets real money. If they can't they pay up and have to shut up. Yes, you shouldn't make claims you can't substantiate.

I am willing to bet Rs 1000 they can't tell the difference. But I have very limited experience with HiFi and would not mind if proven wrong.

It has been a while since I studied statistics, but we can sure find someone to design a good experiment.

Anyone else interested?

 

[theironhorse]

we all know it.. a good built digital cable will work well.. rest all is marketing gimmick to do some research make a better cable that wont do any good that any other digital cable but sell it at insane high prices....

and these companies always have the theoretical aspect of it.. but none claims that human ear heard the difference.. probably the one staying the village can.. metro cities are full of pollution of all kind.. i doubt our ears will be that good to identify the difference made by the 500$ cable..

 

[thatguy]

I do not think collisions are eliminated in switched networks. Ethernet/FastEthernet/GigabitEthernet is a broadcast-based network. Whenever there is broadcast, there is bound to be collisions. In a switched network, assuming the switches have level-3 functionality, there is an assured path for a packet from an IP-address to another IP-address in the LAN. The chance of collision is reduced with a smaller LAN. This has given rise to the concept of the so-called VLAN.

I (and my colleagues at work) make switches for a living. Ajinkya knows what he is talking about and is right. You might want to do some more research on how modern switches work.

 

[ajinkya]

I do not think collisions are eliminated in switched networks. Ethernet/FastEthernet/GigabitEthernet is a broadcast-based network. Whenever there is broadcast, there is bound to be collisions. In a switched network, assuming the switches have level-3 functionality, there is an assured path for a packet from an IP-address to another IP-address in the LAN. The chance of collision is reduced with a smaller LAN. This has given rise to the concept of the so-called VLAN.

Electromagnetic interference is a common phenomenon. In a lab situation, where there may be strong magnets around, there are pockets where you would lose signal of your mobile phone. When you are addressing an audience on a microphone, try to bring your mobile phone too close to the microphone and you will hear effects of the em interference. At times, during riyaz at home, I put my mobile too close to my electronic tanpura, and immediately hear strange noises. Ambient strong em fields seriously interfere with wifi signals, I see these things on a daily basis (exactly opposite to my office, there is a lab with strong magnets).

Regarding audio signals, try connecting your turntable with an unshielded cable to the phono-stage, there will be tremendous hum due to em interference. Luckily, unlike gravitational fileds, em fields can be shielded.

Agreed, beyond the phonostage, the signal is line-level and a lot bigger as with all other line-level signals (e.g. from CDP to amp), and the effect is not as drastic. But of course there is a finite possibility of undesirable effects, because shieldings are not always ideal.

Now, regarding the em interferences in an USB cable, I do not have a proper idea how serious this effect is. There is no reason to think that just because the wave-form is square-wave pulses or whatever in the digital transmission, there will be no interference, or the transmission will never get affected by the domain-walls (created by imperfections in the crystal structure of the conductor).

I am sorry, I do not have enough first-hand knowledge in computer-based music to contribute to the main discussion here. However, I found your above comments a bit misleading. May be I misunderstood in my late night reading of your post.

Regards.

Asit,

Perhaps I should have been more explicit about the kind of Ethernet I was referring to. I meant full-duplex, switched Ethernet which does eliminate all collisions since the links now become point-to-point. If it had been a bus/broadcast medium, with no time-sharing, then you would be right in saying that collisions must occur with finite probability. Here is a link explaining the military-spec Ethernet:
Ethernet Suits Up for Military Avionics Apps - COTS Journal

Your examples regarding EM are illustrative and easy to follow. Another one which happens to me whenever I listen with headphones with a mobile nearby is that I can make out whenever an incoming sms/call is about to happen before the phone rings, since I can hear a distinct set of clicks, courtesy the EM waves inducing voltage in my headphones.

However, I was not suggesting that EM is some mysterious phenomenon that cannot be measured or observed. What I was saying was that the reasons for EMI and the noise created by it are subtle and the effect has to be measured before jumping the gun and using EMI as an excuse to charge exorbitant amount for cable shielding. In car engine controllers, where the spark plug and engine environment produce horrible EMI, the CAN cables are highly shielded. In other areas of a car network, the shielding is much less. The shielding is done proportional to the noise rejection needed, which is the right engineering approach. And having worked with people inside some well-known car companies, I know that they don't spend $500 on cable shielding, because the automotive market is cut-throat in terms of price. Still, we do not hear of car engines dying out or the car having "jitter" in its performance because of "delay or lost packets" in the engine control cabling.

Many audio companies use the dreaded "jitter" to induce fear in customers about digital audio, and make them pay for something that may not be needed for their work. Of course, if you put an electric motor next to your DAC, then yes, you may need heavy shielding around the setup to make sure your USB cable can transmit anything at all. But I assume that most of us do not have a DC motor or a car engine sitting on their desk while working or listening to music

Putting $500 or more as a price to justify USB/SPDIF/Firewire shielding and conductor purity is a crime, in my engineer's perspective, since it is a wasteful utilisation of resources and method, and false advertising. That is what I was trying to convey in that paragraph, which may not have been clear to all readers.

I agree with your example of turntable hum, because there small signals are used, which are more susceptible to the detrimental noise picked up. And I am not suggesting that the shape of the wave (square, sine, or sawtooth) has an effect on how much noise is picked up. What I am reminding all of us is that the whole point of digital transmission is that the noise rejection is much higher than that of analog, because the signal waveform is interpreted at the receiver again. So, a much larger proportion of noise is needed to invert a bit from 0 to 1 or vice-versa. The analogy would be similar (though not exact) to why FM transmission is more noise-robust than AM, although at the cost of higher bandwidth in terms of sidebands.

That does not mean digital is immune from noise. As a physicist, I know you are aware that satellite processors used in space undergo radiation hardening procedures precisely because bit-flips can occur because of strong radiation exposure. However, in our home environment, I think a reasonably shielded digital cable (one costing much much less that $500) is more than adequate for all our listening pleasure, at least in my practical engineer's perspective.

-Ajinkya.

 

[sidvee]

Guys at this stage I would like to interject something:
1. The world is full of people willing to pay for something other people would consider excessive - Does not mean either one is wrong, just different perspective, value system, affordability et al. And this is just not in audio - it is present in almost all fields.
2. A debate by its very nature does not have to be conclusive. People can present their points of view and one can take away what he likes and leave what he dislikes.
3. Audio is a subjective field - people hear differently, have different systems, room environments etc. a lot of variables. So in the end a person has to hear for himself to arrive at the conclusions.
4. In audio there is a tenous link between science, creativity, genius and insanity. It is almost like religion - so explanations will not and does not suffice.
5. DBT's have been always bandied about in Audio maybe from past 50 years, but yet there are disbelievers who continue to buy and believers who refrain. Who is right?
With that out of the way - what can I say "Great discussion"! I learnt more about USB in a few days than past couple of weeks of research. These debates have a way of filtering the topics that matter the most rather than noise (pun intented:lol of which there is a lot of out there in the www.
Cheers,
Sid

 

[marsilians]

Great and very pertinent points. Will be interesting to see what you end up with.

Guys at this stage I would like to interject something:
1. The world is full of people willing to pay for something other people would consider excessive - Does not mean either one is wrong, just different perspective, value system, affordability et al. And this is just not in audio - it is present in almost all fields.
2. A debate by its very nature does not have to be conclusive. People can present their points of view and one can take away what he likes and leave what he dislikes.
3. Audio is a subjective field - people hear differently, have different systems, room environments etc. a lot of variables. So in the end a person has to hear for himself to arrive at the conclusions.
4. In audio there is a tenous link between science, creativity, genius and insanity. It is almost like religion - so explanations will not and does not suffice.
5. DBT's have been always bandied about in Audio maybe from past 50 years, but yet there are disbelievers who continue to buy and believers who refrain. Who is right?
With that out of the way - what can I say "Great discussion"! I learnt more about USB in a few days than past couple of weeks of research. These debates have a way of filtering the topics that matter the most rather than noise (pun intented:lol of which there is a lot of out there in the www.
Cheers,
Sid

 

[sidvee]

Great and very pertinent points. Will be interesting to see what you end up with.

Thanks Marsilians. Thanks to this thread and inputs from your good self and other esteemed members, one take away that I have concluded on, is not to jump and buy a relatively expensive cable right away. Appears that the use of this cable specifically for audio applications is in its infancy and there is no clinching evidence for the use of an expensive cable, though there seems to be a logical position that a well built cable from a trusted brand is desirable. That being said, I will only commit to a purchase after auditioning and in this case I think an audition will play an even more significant role. Though I wonder in this Great Land of our's (the Land of Non-Instant Gratification), how I can get my hand on different cables to audition. So like everything else, another paradox to be tackled, but that is what makes life fun isn't it (the thrill of the chase).
Cheers,
Sid