Digital Audio Cables: Wire or Optical

Sean is right. Output from transport is electrical digital and stays that way until it reaches the DAC chip then converted to analog inside the chip.

If you use toslink, optical signal when input into the receiver on the DAC has to be converted to electric digital before it goes into the DAC chip hence the extra step.

On the transport side, the laser pickup DOES NOT output the optical signal. Signal from laser pickup is converted to electrical then sent to the optical LED to light up the TOSLINK cable.

Yes, coax has it's own problems with the RFI and grounding which the TOSLINK does not have.

Ultimately, it is up to you the listener to use whatever cable you think sounds better to your ears.

With no real attempt to clarify all of the confusion, please keep in mind that the bandwidth limitations stated are those of the "typical" transmitters/receivers usually found in consumer equipment at the time the information source was written. The optical cables are capable of far greater bandwidths with commercial grade lasers or even higher quality, current technology opto-converters. Single-mode fiber has transferred near terra-bit data rates in specialized experiments. The typical negative associated with Toslink is the poor connector pair mating. Any misalignment will result in signal loss, but also signal reflection. Toslink's plastic shells allow too much play between the connector mating. Also, multi-mode fiber is used in most Toslink cables. (Actually, I'm not aware of any Toslink using single-mode fiber, but I haven't taken a survey of all manufacturers. The Toslink shells that I'm familiar with are designed for the larger diameter multi-mode fiber.) Multi-mode uses a "reflective boundary" characteristic to propogate the signal down the fiber. This will induce additional jitter. Of course, until your equipment is completely thermally stabilized, the timing source will produce far greater levels of jitter.

Conversely, ST-type connectors typically use ceramic that is precision ground to achieve much better tip alignment. Also, single-mode fiber inherently should minimize any timing error due to the cable. However, the quality of the glass thread used in single-mode can vary. The specifications would suggest that there should be no practicle difference, but how often do we hear differences in copper cables. Just something I thought I'd toss out to further muddy the waters. In both cases, the quality of the tip polishing is very important to maximize signal transfer.

As for opto-electrical conversion time, this should be irrelevant. Jitter is the relative timing error between the leading and falling edges of the serial digital data or clock, either of a single bit or from one bit to the next. As long as the conversion time is constant, there will be no increase in jitter. Reclocking of the received data from the FIFOs is very common in digital circuitry to reduce timing errors. Extreme accuracy crystals are available, though expensive. Phase lock loops (PLL) are also routinely used to minimize clocking variations.

I guess what I'm trying to get across is that you should be careful not to be too quick to blame a particular type of cable system as much as the quality of the components used within the equipment to implement that cabling system.
Enjoy the music.

Response to Genesis168 and Sean:
Is it possible you are confusing two different types of electrical signals and lumping them into "electrical digital" ? My impression is that signal from pick-up/laser is some form of "pulse-wave electrical" [is this electrical digital?] and then gets converted to a different kind[voltage?]. Anyway - I do not recollect much about this now- too many years have passed. In lieu of that I will make this my last post on this topic. The final point I wish to make is that one should refrain from making blanket statements like "coax is better" OR "coax is superior to optical" since there are a whole lot of other parameters involved (see post by Jcbtubes). My fear/feeling is that a majority of users jump on the coax bandwagon just because it tends to be driven by the reviewers/manufacturers who both have their own axes to grind in this matter. Cheers.

I think that digital signals are PCM ( pulse code modulated ) but i'm not sure. Never really paid a lot of attention to how they did this, only the results. Either way, they are "modulating" or "super-imposing" the digital data stream onto an rf based signal. If i remember correctly, i think that this signal is somewhere around 3 to 4 MHz or so. As such, coax is still a pretty efficient conductor at that range and should work quite well.

You do bring up a good point though that i had forgotten about. There actually has to be another stage that we've overlooked. Obviously, laser is a light based energy. As such, the signal has to start off as optical and is then converted to digital. At this point, the data can be extracted / seperated between the optical laser signal and the digital data that was superimposed on that light stream. From there it can be sent from transport to dac via the rf based coaxial jack or via optical on a toslink jack. Obviously, there must be technical problems ( probably scientific limitations ) with trying to leave the signal optical from the transport and then directly it at the DAC directly. This would drastically simplify things if possible.

After thinking about all of this, it is no wonder that it has taken us 20 years to get good sound out of redbook players. While i do believe we started off with something that was not well thought out to begin with, there is obviously a LOT going on within these players. As such, i see a lot of room for error and the necessary refinement that takes place when you have such errors.

Maybe we should be grateful for what we have and be glad that technology is working to our benefit. Then again, if they would have simply started off with a higher sampling rate to begin with ..... : ) Sean
>

Genesis -

The bandwidth of Toslink is at least 6Mbits/sec, assuming that you have your transport and DAC within shouting distance of each other. Note that as Jcbtubes said, plastic fiber links (including Toslink) can run much faster. The datastream for CD runs under 1.5Mbits/sec.

If you prefer coax to Toslink, no problem. If you're looking for a technical reason that would make coax superior in audio applications, bandwidth isn't it.