The optical/coax duel

[Norm Strong]

<font color='#000000'>Back in the early days of A/V The Toshiba company made a mistake; they invented TOSlink, an optical digital method of interconnect that seems to have no perceptible advantages in consumer equipment--just costs more. &nbsp;

Since it's here, I suggest that all digital audio inputs be of the type that can handle either an 1/8&quot; TS connector for coax or a 3.5mm optical connection. &nbsp;Buyers will then be able to use all of the digital inputs. &nbsp;It will no longer be necessary to make sure your receiver has, not only enough digital inputs, but the right kind as well. &nbsp;

You think this doesn't happen? &nbsp;I've had to pass on 2 excellent receivers because they had 3 optical and 1 coax input. &nbsp;Unfortunately, I have 2 devices that require coax inputs.</font>

 

[dis]

<font color='#000000'>Your solution is to go with seperates...The preamp I'm looking at right now has 3x coax in, 2x optical, 1+1 coax&amp;optical outs.</font>

 

[Yamahaluver]

<font color='#0000FF'>Even the now ancient Yamaha RX-V 800 had three co-axial inputs as well as a phono input. Most Yamaha receivers, especially the high end ones have more than one co-ax inputs.</font>

 

[Audioholics]

<font color='#000080'>This is a personal pet peeve of mine, one that I added to our pre-pro/receiver wish list. My Yamaha RX-V3000 only has 2 coax inputs (its international counterpart has 3). Not enough for me. Here's where coax is essential:<ul><li>CD players<li>digital cable boxes<li>HTPC/computer digital audio cards<li>DVD players (most have optical, however)</ul>In fact, I simply do not understand the proliferation of optical inputs at the expense of coax. I simply do not see this in the real world, where I find more coax digital SPDIF than TOSlink.

Great topic!</font>

 

[Guest]

<font color='#000000'>It is a simple task to get coax-Toslink converters, even bi-directional and multiple ports are available, so you never have to pass on any piece of equipment you really want.

Here are some examples;

One
Two
Three
Four</font>

 

[Audioholics]

<font color='#000080'>I guess if your in a pinch and want to spend the bucks. I'm glad, however, to see some companies opting for including more coax on their newer receivers and pre-pros.</font>

 

[Guest]

<font color='#000000'>IMHO, $30-$60 is not much to spend for flexibility with anything, but agree it would be nice to have it all in the original unit.  This method (Toslink&lt;-&gt;coax converters) does allow one to choose equipment preferences without worrying about the number of Toslink/coax inputs though (when selecting equipment based on how it sounds).</font>

 

[Guest]

<font color='#000000'>hi all,
&nbsp; &nbsp; &nbsp; &nbsp; &nbsp;my question. why can't I buy toslink cables longer than 3 meters? I need 50 feet. &nbsp;Is there a problem using connecters between 3 or 4 cables? &nbsp;Companies that deal with communications use much longer runs than consumers would ever consider. Whats up with that?
I know that I could and should use a converter, but I WAS JUST WONDERIN................... thanks in advance</font>

 

[Khellandros66]

<font color='#000000'>I don't care much for the sound that Optical cables produce. &nbsp;Its sounds compressed and the soundstage isn't as open.

:0~

Bob</font>

 

[PaulF]

<font color='#000000'>JOE H50

Looks like this Canadian site sells Toslink cables up to 50ft.

The price for 50ft is $192 + tax ( I assume Canadian).

http://www.sysconcept.ca/product....s_id=68

Khellandros66

Are you saying that you can hear a difference between Toslink and Coax regardless of product as a general rule? I find that difficult to believe. I have seen the reports that say that the E/O/E conversion adds jitter and therefore affects the high frequency response. My question is then, what is the good of all that reclocking they put into receivers if it doesn't remove the jitter? Other than that what could make them sound different?

Paul</font>

 

[gene]

<font color='#000000'>Sounds like I need to write an article about this

&nbsp; In short, I recommend coax for short runs and toslink for longer runs.

Cable Budget Guidelines</font>

 

[JOEH50]

<font color='#000000'>Thanks PAULF for that site. I'm still wonderin if there would be any big difference in sound due to any signal ( if any ) loss from the coupling of several toslink cables. any ideas out there?</font>

 

[PaulF]

<font color='#000000'>JOEH50,

I can guarantee there will be signal loss if you couple multiple cables together. There is a thing called insertion loss each time the signal reaches a break in the cable.

However I don't have any way of telling what level of loss will create audible problems with your setup. It might be worth finding out the optical output level of your receiver's Toslink transceiver to get an idea of what distance it can drive.

You will also need to buy couplers to patch each of the cables together, I'm not sure it's worth the cost.

Paul</font>

 

[JOEH50]

<font color='#000000'>A toslink coupler I beleive is just a hollow connecter which would allow light to pass through following its path to the next cable unimpeded. &nbsp;You would assume since there really isn't a true interrupt in the flow, no loss would ocurr. I thought a digital signal didn't suffer the same degradation as analog. Like satillite tv for example, you either got a signal or you didn't. No snow, ghosts, etc. I guess it comes down to a few factors like how many strands per cable ( nothing on cable packaging indicates that ), and as you mentioned, output level at the receiver (YAMAHA 3300 ) , and distance. What numbers should I be looking for?
thanks again for your input PAUL</font>

 

[PaulF]

<font color='#000000'>JOEH50,

I guess the short answer would be to call or email Yamaha for a recommendation. I'm not fobbing off your question but even with all the variables taken into consideration, we would still not know at what point any of the theory makes a practical impact on your system's performance. Or you could simply try it and return the cables if it doesn't work.

If you're interested in reading further here are a few notes on the theory behind the optics.

When two cables are joined together via a connector there will be an insertion loss. This loss is caused by the glass to air to glass transition, in technical terms a change in refractive index. The only way to avoid this is to splice the cables together effectively fusing the glass. In some cases gel is used between the connectors. Even then there will still be a loss. The effect of this transition is to reflect some of the light back down the fiber.

In communications systems connectors have ultra polished ends that minimize the reflection but these connectors often cost more than an entire audio cable. Typically these connectors are angled so that the reflected light is sent into the cladding of the fiber where it is absorbed rather than reflected into the core.

Another thing is that for low cost consumer electronics we are probably dealing with plastic or synthetic fiber. This fiber is likely to be multimode meaning signals can propagate down the fiber along various paths. Mutiple modes cause dispersion along multimode fiber (the longer it is, the worst it gets). The dispersion will alter the shape of the pulses making it difficult for the optical receiver to properly determine the beginning and end of each pulse. This will cause bit errors.

By adding connectors, the reflected light will add to the dispersion characteristics of the cable further reducing the length of the cable that can be used before bit errors occur.

Paul</font>

 

[JOEH50]

<font color='#000000'>thanks for the enlightenment PAUL.</font>

 

[GermanMan]

<font color='#000000'>I am not convinced that optical cable is worse than coax.
Its the same digital data traveling down the medium.
The problem is that neither of the methods carry a clock signal, as I understand it, so that clock drift can become an issue. &nbsp;

The failure of some optical connect systems is, for reasons beyond me or you, they use cheap optical connector ends on the hardware and possibly cable, which aggrivates this situation.

However, technologically, there is no reason optical would be worse than coax, if the proper components are used in the connection from equipment a to equipment b.</font>

 

[_audiouser_]

<font color='#000000'>Paul &nbsp;(or anyone)

I'm wondering if Tsolink does or does not have built in error correction and buffering to compensate for bit errors?

I can't recall the source but I do remember someone stating that it did.</font>

 

[Eric]

<font color='#000000'>What you're calling &quot;Toslink&quot; is really either Sony/Phillips Digital Interface or AES/EBU. Toslink is simply the type of optical connector used to hook up gear.

GermanMan:

SPDIF and AES/EBU use biphase mark encoding. In this encoding scheme the bit clock is recovered from the data stream. So, the clock is carried with the data.

_Audio_user_:

Consumer gear (SPDIF) has no CRC (or other bit error recovery scheme).

AES/EBU (some call it SPDIF Pro but it's not really SPDIF) does have a CRC. On pro gear, I'm not sure if the CRC check applies only to the channel status data or the audio data as well. It probibly applies to the audio data as well (since that is a sub-frame within the status block).

A good start on how SPDIF works:

Digital Audio Data Structures</font>

 

[PaulF]

<font color='#000000'>I'm not an expert in this area but from what I've seen this seems correct. The Toslink component simply does electrical to optical conversion or visa versa. These devices come rated from 10m (30ft) to over 1000m (3000ft), so the length of cable that can be used will depend largely on the Tolink component employed in the CD player and receiver.

As with most transmission systems the clock is recovered from the data singnal as there is generally only one fiber or wire (plus ground) to carry the signal. The SPDIF or AES/EBU seems no different. From the attached link it apears that consumer mode transmission does not include the CRC, but even with, CRC is capable of detecting errors. not fixing them. There was also single bit parity bits for error detection of an odd number of bit errors.

There is plenty in the DVD standard to account for bit error correction but this would take place during the decoding process.

Paul</font>