Why do digital cables sound different?

I have a small confession.. I enjoyed choosing the cables in my system. This enabled me to "customize" an otherwise mediocre system, and approach the high end without spending too much money. I know "too much" is relative. My cabling is valued around 35% of the cost of hardware, but thanks to Audiogon I was able to spend around 20%.

Response to Kthomas about not offering up a hypothosis. I refuse to offer up a purely speculative hypothosis, about an audio issue, that simply has no presently known logical eletrical engineering basis. I don't know how long you have been involved with high end audio, but let me tell you what happens with far reaching hypothosis, regarding electronic issues that don't seem to make any sense :THEY BECOME THE BASIS FOR MANY SNAKE OIL COMPANIES, THAT PURPORT TO BE HIGH END AUDIO MANUFACTURERS, TO INCORPORATE INTO THEIR PRODUCT LINES! I have seen this happen again and again. I remember when Class A amplifiers were considered the "best" sounding amplifiers, so therefore "Class A amplification are the best amplifier designs". The same thing happended with zero negative feedback designs, and more recently, SET amplification. Unfortunately, these so called "facts" are very misunderstood, and EXTREME generalizations at best. Each "fact" has set the high end electroinic industry back more than you can imagine. As audiophiles buy into these facts, they propenciate manufacturers to build these products, that incorporate these topologies, EVEN IF THIS IS NOT THE BEST WAY TO GET GREAT SOUND! The best example that I can think of regards speaker cable design. It has been hypothosized that a great sounding cable needs a ton of current carrying capabilty. So in turn, the majority of the high end cable companies have produced these huge thick cables for us audiophiles. Indeed, many manufacturers don't even consider that thinner guage designs might produce a much better sound. After all, the more current carrying capability, the better bass and dynamics, right? This is a no longer a hypothosis, but a fact, right? Well, this is just not the case . I personaly know three speaker cable manufacturers that are pulling their hair out, trying to design cables that represent what the audiophile community thinks are better (read big and heavy)... BUT ...this is just not true with many better sounding, properly engineered configurations. Other cable companies (that I personaly know...read; big, well known names) have figured out how to make each of their larger cable offerings sound better, as they incorporate more conductors and make each cable more expensive in the process. More importantly, they can then charge an ear, arm and a leg for these huge cables. These guys (that know the real truth) are laughing all the way to the bank! The other cable manufacturers ,that design huge cables. are not necessarily trying to rip us off. They have just not discovered the ultimate truth about cable configurations and conuctor size.(It amazes me how a "fact" can be created overnight in the audio engineering community.) This is why I refuse to offer up speculative, potential non-sense hypothosis that has no engineering basis. When I have a legitimate hypothsis, that has some fundamental electronic explanation, I will always offer it up as a hypothosis (with hope that it will not become a "fact" overnight).

I think I am repeating myself, but continue to get replies which imply that the only reason why digital cables could sound different is bit errors. This is not the reason at all - the reason is jitter - noise-based and time-based distortion. The problem is not about distortion causing a DAC to read a 0 as a 1, or a 1 as a 0. It is about the fact that we are talking about real-time transmission and that a DAC produces harmonic distortion at its output when the arrival times of the 0s and the 1s are not perfectly, regularly spaced. I really am having trouble saying this in as many different ways as I can. It is not about redundancy so that when an error occurs the data can be resent - we are not talking about data packet transmission here. Bandwidth capability is in fact an issue here. Even though the bandwidth for data transmission is low by most standards, if the cable was only just able to transfer the data accurately then the square waves would be very rounded indeed and jitter errors at the DAC would be enormous. Higher bandwidth cables allow sharper corners to the square wave with less undershoot or overshoot. Optical cables are also free from earth noise adding to the signal. It is not about bit errors, it is about timing based distortions. I work with loads of PhD telecommunications engineers but their grasp of these concepts is slight at best, because it is irrelevant for the audio fidelity needs of telephony and irrelevant for data packet transmission. But the best of them acknowledge that their training is insufficient for high quality audio.

Redkiwi: I get it. It is not only that you say yes and no, but "how" you say it as well. As far as I know nothing is perfect and there are always variences to be had. The sub atomic clock in Denver is pretty accurate, but I would assume that DAC's and digital transmission lines are not even in the same ballpark. Perhaps when we get organic based DAC's all cables will sound the same?

Redkiwi, the overshoot and undershoot you speak of is caused by capacitance in the cable. While overshoot and undershoot themselves may not necessarily effect the DAC output signal, the capacitance in the cable may effect the data pulses risetime. This effect, I would assume may be audible. While jitter is a degrading factor in a system of this type the transmission cable is not likely to increase or add jitter to the bitstream. I feel capacitance is the real culprit here. The less capacitance in the cable the better.