Which is better, coax or optical cale connection from DVD to reciver? Why?

Jim Middleton
middlton@starlink.com

The guy at my hi-fi shop said that optical isn't necessary good. Among them, the signal has to be converted from electrical to light, and converted back from light to electrical. Also, there are good optical cables and bad optical cables. Good optical cables are very expensive. Coax cable lets the signals through via one standard interface, although you may get all the 'problems' that cables get. But there are also good coax cables and bad coax cables. So he made me one using the best audio (Nakamichi) cable and plugs he's got, and I'm very happy with the results.

I believe the english Audiophiles think Coax is the only way to go; but I like anything digital and cannot detect a variance between the two..?

Theoretically you cannot win..consider the paths the signal travels; both DIG(light) & COAX experience switching and converting...

Even though an electrical signal is being converted to light surely the fact that it is a digital signal means it arrives at the detination in exactly the same state it left no matter which method you use to get it there..? a 01010101 data segment is still the same if it equals 01010101 when it gets to the AMP.

JH

John,

True...but keep in mind, if you bent the optical cable, no signal goes through...plus...For me, the coax cable cost me $50 (I got it almost at cost), while $50 gets me the cheapest, thinnest optical. The thicker ones are about $150. Since I didn't know if the thicker ones will make that much a difference, I went for the coax.

Just had a thought on what you said: Both outputs are digital anyway...which means the datastream gets to your receiver/amp, and the amp/receiver translates it into sound. Then there shouldn't be any difference since digital is digital ('1's and '0's), I mean, computer networks uses coax cables too.

I think it is generally accepted that coax is better than optical. Optical cables are more fragile, cannot be used for very long runs, etc. However, there are concerns that coax is susceptible to magnetic interference. In any event, when it comes down to it, I don't think it's going to make that much of a difference. I guess what you really have to be concerned about is the types of inputs in your receiver.

The DAC in most DVD players are either 20-bit or 24-bit type, which poses no threat to either coax or optical transmission. Both are shielded sufficiently and are short enough that there must be major interference in order for the transmission to have a bit-error.

The battle comes down to bandwidth, in which coax holds the advantage. Optical cannot runs for extended length without repeater assistance.

In case you're interested, here are some cable description:

Optical cable (also known as Toslink cable), contains fiber optic "wire" which transfer data as light. For those who are unfamiliar with fiber optic, light "bounces" inside a fiber optic "wire", due to the its special angle of refraction, just like a diamond, it "traps" light within. The light pulses got translated into electrical impulses at the receiver end. Plastic fiber optic has some jitter, the glass ATT type are much cleaner.

Coax cable are usually built using 75Ohm cable with RCA connectors at each end. Sometimes 300Ohm RCA are used, but rare. Because of the high shielding of the coax, signal jitter in coax are very low, making ideal for data transfer.

Some other cable consideration are: (although may not be applicable due to equipment consideration)

AES/EBU. This is the digital standard use in most professional equipment. The cable is a 110Ohm cable, with connectors similar like that of an XLR connector (balanced microphone).

Sony TCD-D3/7/8. This is for Sony professional application only.

RANKING:
1st: AES/EBU
2nd: Coax
3rd: Optical

Coax is a superior transmission line compared to coax digital. You can get a reasonabel optical cable for 40 bucks, same with a coax, If you dont want to buy a coax digital cable, a good quality video cable has the same requirement as a digital cable. A simple test is to try using the optical output of your dvd player and the coax.Use two separate inputs on the receiver or DAC. There is a noticable loss in resolution as well as signal level. It is very rare to be able to kink a fiber optic cable and more likely to damage the termination point on either end. I have persormed personal tests of XLR connectors versus high quality single ended coax and found no difference. Of course I may be deaf, but thats a different issue.
The issue of ones and zeros is not as simple as it sounds because you have frequency to voltage convertors on each end of each optical transmission. ( Converts light to an electrical impulse ). The data rate is assumed to be constant but is not. Just as a CD player can be sloed down and sped up according to the data rate. Assuming the transmission of the CD player and the conversion of the receiver are the same, you may have a reasonable transmission. It is my experience that direct electrical(coaxial) transmission and conversion has the least amount of variance in data rate and generally sounds fuller.

I'm glad I ended up with the 'better' of the two...But now I'm curious:

Optical cables/interface came out later than coax. If coax is superior, why bother with optical?

I'm also curious about this, there must be some advantage to optical that we are missing. All of the DVD players i see and most AV receivers have the optical and not the coax. Maybe its because optical is less expensive to impliment, or can optical carry more channels without being decoded.

Well, I don't think optical is less expensive to implement, because for one, it has to put in a converter to convert electrical signals to light. And as Daniel said, coax is superior in bandwidth, so why optical?

As I mentioned in my earlier post, coax is superior than optical. Let me explain in more detail why this is so. At the end of this post I will explain the significant of optical. This is an extremely long post (Sorry Daniel Tonks), so bear with me.

A coax cable comprises of a solid conductor (wire) in the middle, surrounded by an insulator (non-conductive material), then shielded with a braid-like mesh of conductor. Now the material may vary, either copper, silver, alloys, or gold can be used as the conductive material. The basic principal is this: There is a basic impedance to a coax cable. For example: Let's take a 75Ohm coax cable. That would means if we were to take an INFINITELY LONG 75Ohm cable, and measure the impedance between the core conductor and the shielding conductor, it would be 75Ohm.

That's pretty darn good!. And that's the whole purpose of coaxial cable. The shielding is complete (totally surrounding the cable), virtually eliminate any signal leakage (from the inside out), and any interference (from the outside in).

That's why for TV Coaxial Cable (those 75Ohm coaxial in your home), you can run for very extended length without signal loss. Why do we need signal booster?. In theory, we don't. But we live in the real world. And believe it or not, most of the signal degradation came from the connections (splitters, joints, couplers...), the cable are fine. This is due to impedance mismatch and inadequate shielding.

Impedance mismatch came from several sources. Let's say you're joining 2 coaxial cable with a coupler. It would be ideal if the material of the core of coaxial 1 be the same as that of coaxial 2, also the shielding material be the same. That's because each material (silver, aluminum, gold, copper) has a unique impedance. You would think that buying that gold-plated coaxial to extend your existing coaxial cable would be excellent right? IT'S GOLD! (THE BEST CONDUCTOR IN THE WORLD!). WRONG!!!. Not if the coaxial you're connecting to is of different material. Ideally you wouldn't do this, just get rid of the old coaxial cable and buy yourself a long coaxial.

The next source for signal degradation came from inadequate shielding. The coupler, the splitter. Think of these as "stations". The less "stop" the signals must travel through the better. Note that at these "stations", the signals are not completely shielded. Most, but not completely, so watch out, don't use too many of those.

Optical works on the concept of light pulses. First, a little theory about light. I hope that all of you know that light does not travel in a straight line. It bends. Yes!. By the time you can see the sun set, it has already set. The light bends as it entered our atmosphere. Light bends more or less depending on the material it travels through. More on that later. Second property of light is that it reflect. In optical material (fiber optic), there is an angle of refraction, which when increased to a certain amount, the light instead of going through the material, got reflected back. Some light still escaped, but most got reflected internally.

The diamond. The most exquisite material in the world. Because of its angle of refraction, it "traps" light internally. Light entering the diamond cannot escape (some of it). The diamond will shine by itself in a totally pitch dark room. Try it. You will then really appreciated it.

So optical fiber (fiber optic) was created as a data transfer medium. It works on the concept of a transmitter and a receiver, one converting electrical impulses to light pulses, the other converting it back.

Optical fiber has the following advantages over coaxial:

1. Size. Coaxial cable is too big. You can fit hundreds of fiber optic in a place of a coaxial cable.

2. Reliability. It has to happen. Coaxial or any other kind of "metallic" wire will oxidize (react with oxygen) to form rust, to breakdown its "metallic" property, and degrade performance. Thus, coaxial needs to be replaced from time to time.

3. Distribution. Optical has the advantage here. Let's say you're distributing a signal (splitting a signal into 30 signals), then optical is much better, because for coaxial, doing so simply destroy the signal.

4. Price. Here's the kicker. Yes, it was expensive at first, but optical costs less now. Why?. Optical does not have to deal with the shielding aspect, plus the impedance aspect. Coaxial driver meanwhile must be very good in order to deal with the various coaxial cable which will be connected to equipment. For example, Bevan may be using 75Ohm coaxial, while Daniel is using 77Ohm coaxial, both must work. So the cost factor is really why coaxial is superior.

The disadvantages of optical are:

1. Repeater. As mentioned earlier, light reflected internally inside of the fiber optic, but some will leak out. Eventually, repeaters are needed to bring the "signal strength" back to normal. The longer the run, the more repeaters.

2. Bending. A coaxial will probably still work after being bending, but an optical cable will have a limit, and we all know where that is (no, it's not when it's snapped, it's when you bend enough that the angle of refraction is no longer sufficient for the light to bounce internally).

3. Jitter. This has to do with optical advantage #3. Coaxial has this complete braided shielding from one end to the other, while optical has none. Then how does the signal ground of the equipment 1 got connected to the signal ground of equipment 2 (that's what shielding does)?. It doesn't. So therefore there always will be a little bit of signal offset between the equipment, simply because the ground planes of each equipment is at a different potential (voltage). Trust me, the ground plane (unlike its name) does not stay at 0.000Volts.

==============================================
Now I will answer some of the questions from all of the earlier posts:

Jim: For DVD, no difference. But from a comparison point of view, coaxial is better.

Bevan: Glass optical cable are much better than plastic optical cable. Much less jitter. But again, for DVD, no difference. Also thicker optical cable will only get you thicker insulation, not necessarily better performance. And you're right, computer networks (at the backbone) uses coaxial. But between the computer and the switching hubs, twisted-pair cables are used. Optical has the price advantage. Not performance.

John: Yes, no difference performance wise for DVD, between coaxial and optical.

JEF: Optical is more fragile, but it does hold an advantage in long run. It's not that coaxial is bad (IN THEORY, INFINITE LENGTH WITHOUT SIGNAL DEGRADE), but it's more susceptible to interference over the long run. Optical (beside the repeater issues) are very good over long run.

HiFiGuy: Sorry, but no RCA-type video cable, no matter how good, can hold a candle to coaxial. You can definitely tell in DVD application. Regarding XLR connection, it is byfar the most quality connection you can use. Because it is "balanced", unlike any type of connection which we currently use, which are all "unbalanced". Also, the CD data rate does stay constant. 16-bit data at 44.1KHz. Always. I think what you meant by CD data rate variant is relating to Computer CD Data rate, not Audio.

JL: Yup, it all comes down to this. Price. I'm not sure what you meant by more channels though. Are you referring to more optical outputs?.

==============================================
CONCLUSION

So there you have it, the difference between coaxial and optical. Coaxial still holds the performance advantage, but optical does held its ground. It boils down to the application at hand, which I believe is DVD.

For DVD, the plain truth is that there is simply not enough bandwidth, bit-error rate, signal to noise ratio, to test the limit of both.

I would be really surprise if for DVD application that anyone can tell the difference...

SO WHAT'S DANIEL USING:
CD-Player & Receiver: Optical
Tape & Receiver: Standard Audio (RCA connector)
VCR-Audio & Receiver: High-Quality Audio (RCA)
VCR-Video & Receiver: S-VHS
DVD & Receiver: Coaxial
LD & Receiver: Optical
Receiver & Speaker: 16-gauge oxygen-free premium quality cable. The best money can buy. Mmmmmm. Enough said...

The characteristic of a good video cable is good sheilding with a 75 Ohm termination and the same holds true. When I refered to RCA, I meant a composite coaxial cable.Refer to the Sterophile archives in terms of the best value High quality digital cable. Its a monster M1000V VIDEO cable. Considered a best buy for both applications.
Secondly, XLR connections can be very deceptive. While it is true that the XLR by design is capable of lowering the noise floor by 12db, it does not mean that just because you have those type of outputs, that you will achieve this result. The reason why is because many manufacturers put XLR connections on a unit without the internal balanced design. In other words, its not just the cable, but the total design of the electronics. I you own Adcom, Lower end Mcintosh, etc, they are not truly balanced designs. So buyer beware, when you buy a unit with an XLR, many times its only the jack on the chassis that you are buying.
Coax cable will do just fine in most household over distance. 100 Feet is not a problem. Usually, distances over 20 feet, XLR is suggested, which is typical in high end audio systems where the amps are as close to the speakers as possible minmizing the speaker cable length.
Optical fiber can be run for 60 miles in a single run unitl a repeater is needed. Coax is every mile .
Anything transmitted digital to your receiver or DACs whether its DVD or CD can suffer the degradation of optical. I have done the tests side by side.
I have converted two of my own devices from optical to coax digital by tapping the coax run before it hits the optical voltage to frequency convertor. On my Sat dish and my CD player. There are kits out ther for those of you that wish to do the same. I use an Analogue way switcher that maintains a 75 ohm load between all sources as my external D to A has only one coax input. All of my sources are COAX, except for my VCR and television.
Lastly, as for Gold as a conductor, there is one problem. It is the issues regarding plating and the reaction of disimilar metals. While gold is great, typicaly the connector is merely flash plated to prevent oxidation of the actual contact surface. Also in the customers eye, it represents value. In fact, Gold plating can be a far poorer connection then Tin on certain types of connectors. Audioquest uses silver for this reason. If youve ever seen gold plating come off a watch or jewelry, youll know what Im talking about. The gold connectors you see on the back of most receivers are steel plated with gold. Some use a nickel underplate. Sony ES knew that years ago and provided tinned jacks instead because they saw the inherent problem. Copper to gold direct is fine, as they are related, steel is not. Bottom line is that good uniform purity connectors with a copper or brass bass metal and possibly gold plating are some of your best consumer grade connectors.
If you are very intersted in wire design and signal transmission, try reading some of the studies done by Bell labs in the 30s. Much of the technology today is derived from their findings.

The last thing which I ever want to do on any forum is to give out false or misleading information on any subject, let alone making any kind of recommendation without thorough knowledge of the subject at hand. Remember, always consider the application at hand, and that is DVD.

I cannot agree with HiFiGuy on his idea of converting/tapping the audio signals from the coaxial sources, bypassing the optical converter because optical is "inadequate".

Both SP-DIF IEC958 Coaxial & Toslink can handle data rate of upto 48KHz at 24-bit resolution, more than "adequate" to handle CD-quality, 44.1KHz, 16-bit resolution, 2.8224MBit/sec audio.

Can you honestly say that you can hear the difference between audio data transmitted by coaxial and that transmitted using optical???.

I would recommend you seriously look at your equipments, your knowledge on this subject, your research sources, before even suggesting that optical is "inferior" such that one should convert/tap the signal from coaxial sources...

As far as studies go, ahem... you should consider a course in Transmission Line Theory, where you will find that audio is no where near that rate. Also regarding your comment on "gold can be a far poorer connection than tin...", you might want to retract that.

Before you answer, you should consider looking at my post at "An interesting article DVD vs DIVX", and other forum here to find out about my background, and remember, I don't kid around about what I do.

A course in Transmission line theory is just that. Theory...Been there by the way.
The consumer end product versus design engineering is a topic unto itself. You may design VCRs for panasonic, but it doesnt mean you havve an ear for sound. As for the tin comment, contact Audioquest or Sony Engineering and ask them directly. Being that we use copper PC boards and use a combination of lead and tin as solder, I cant imagine it being all that bad. Some use silver which in itself is not pure because it to is alloyed. You can check, but my best friend is one of the top Metallurgists for Jorgensen Forge. He might know more then you about the bonding of metals.
As for Coax versus optical, in the final evaluation, opticalis relatively lack luster in its overall imaging capabilites. Its not debatable at the high end audio level, its common knowledge and I have experienced it first hand. If it wasnt, Id be designing fiber optic distribution systems instead of coax. Its an easy comparison with a piece of music you are familiar with.

As for my system, maybe we should compare notes: My strict two channel system is:

Classe CA100 Amplifier
Classe CP35 Preamp
Meridian 508 20bit
Rel Strata Sub
Audioquest Lapis cable.
Audioquest Lapis XLR from CD to AMP.

Its not the best system in the world and I dont claim it is, but its no slouch. I am happy with it. Go to a better Stereo store and have yourself a listen, you might be pleasantly suprised.

Daniel,

Two questions:

1) From what you've said about coax, I presume that unless you can have a cut in the cable somewhere along the length of it, the oxidization you were talking about occurs at the ends, where it's more suseptible to air. So if I use good gold plated connectors at the ends, does it stop the oxidization?

2) Not all computer networks uses hubs...a small office can do very well with a thin Ethernet system, which just uses t-connectors and terminators and coax cables. So while they maybe subject to interference, I haven't come across the case where in a normal office environment, external interference has caused data corruption (even with a CheaperNet system). And so when it comes down to digital output, we're still talking about '1's and '0's. And it is still up to the DAC on the amp/receiver to interpret those '1's and '0's to convert it to sound that we hear. What I'm trying to say is that while I agree that using coax to transmit analog signals (like TV) is suseptible to interference and therefore influence the output, surely, in the case of digital output, external interference (under normal situations) shouldn't be that much of an influence?

I understand and agree with your pros and cons with optical...and from your explanation, it seems that you're saying that even with the optical/electrical converters built in, optical is still a cheaper solution?