lipfix faq

Q: I've always been told optical audio connections are better than using coax connections so both my sources have optical fiber cables. I will need a converter to adapt one of them to the DD540's coax input and am concerned that using the coax input will compromise my sound quality. Is this a valid concern and if so how can I prevent it?

A: The only real advantage optical audio has over coax is its immunity to "electrical" interference and ground loop problems. Counter to what is generally thought, digital audio "coax" signals on short cables are normally cleaner and more stable than their corresponding "optical" signals.

That comment will probably draw some flack on the various Internet forums so I will be expecting it but the reason is obvious if you think about it: All digital audio signals are initially voltage pulses essentially identical to the signals placed on the coax connections. To get the optical audio signal those voltage pulses are applied to LED (light emitting diode) devices (called Toslink transmitters) to convert them to pulses of light for transmission through optical fiber cable to a Toslink receiver at the other end where they are converted back to voltage pulses. That conversion to light pulses and back to voltage introduces slight offsets in time as the light emitting diode transmitters and photo detector receivers can not respond instantaneously. Those slight offsets make no difference as long as the packets get through the specs (and they are if you hear the audio) but my point is that a coax signal of that same digital audio won't have those distortions or offsets (which some call jitter if it varies over time). So, technically, coax digital audio signals should be "better" than optical - not the other way around.

So why is optical considered better? The problem with coax signals arises when there is electrical interference of sufficient magnitude to keep the digital packets of audio information from being recognized by the receiving decoder. Coax cable run near motors or fluorescent lights whose rising and falling electrical fields can induce voltages in the copper wire may experience interference corrupting their digital audio signals whereas optical fiber cables carrying pulses of light would not. Optical cables are immune to electrical interference.

For short runs of cable without a source of electrical interference using coax provides a superior signal with less jitter.

A converter from optical to coax at the DD540's input would only have a very short coax signal path and should be immune to any electrical interference. That might not be true if you converted it and then ran the coax cable 30 feet.

Note that the new four input DD740 has two optical inputs and two coax inputs which in most cases eliminates the need for any converters. Optical/coax coax/optical conversion is always active placing the selected input on both optical and coax outputs simultaneously so you can use either (or both if you choose) for input to your av receiver.

Q: What type of optical cable should I buy to connect the digital audio delay to my equipment to insure I get the best sound quality? Prices range all over the place.

A: I'd buy the cheapest one I could find and try to keep them as short as possible. I have heard salesmen pitch the "sound quality" advantages of $60 optical fiber cables over their cheaper offering and had to restrain myself from interfering with their presentation. I don't think they are usually dishonest people but bet they actually believe what they are saying but in almost every case the cheapest will work as well as the most expensive.

The thing to keep in mind is that digital audio (whether coax or optical) is a very sophisticated digital packet stream similar in many respects to computer networks. These packets have a complex structure including many status bits that "frame" the actual digital audio information. If a packet is "upset" in any way - that is, if any of its "bits" are corrupted by interference - the effect will not be some "marginal" loss of sound quality as one might experience with analog signals but will be a "catastrophic" loss - in most cases causing the packet to not be decoded "at all" with very noticeable audio consequences (popping, loss of sound, etc.). If it happens you will know it. It won't be "marginal".

It is possible, however, that excessive "jitter" in the digital audio packet can cause the digital audio data bits to be presented to the digital to analog converters in your receiver at a slightly varying bit-rate. That can impact sound quality to the extent that a converter reproduces it in the audible analog output. Some D to A converters are much better at handling jitter than others.

Using the optical output will always add jitter but in most cases it won't impact the signals enough to matter. One optical cable that is less transparent than another can reduce the light intensity arriving at an optical receiver causing it to switch later which adds to "jitter" so the clearest optical fiber cable should be the best but that doesn't always correlate to price.

For those interested in the technical details, the digital audio standard specifies no more than 1/4 Unit Interval "jitter". The "Unit Interval" is the 162 ns pulse width of a digital audio bit at 48 KHz sampling rate so total "jitter" should be less than 40 ns. Unfortunately the switching times or propagation delay of most optical receivers and transmitters are specified as 20 ns max each so for optical connections whose components happened to have that maximum they alone would use up the entire 40 ns jitter spec and there could be no other source of jitter at all.

Fortunately in practice most optical receivers and transmitters come in below their stated maximum propagation delay of 20 ns. (Incidentally a "nano second" - abbreviated "ns" - is the same fraction of a second that a second is of 33 years!)

At 96 KHz sampling rate the Unit Interval is 81 ns so the jitter spec is 20 ns max which makes it very difficult to find optical links that work at that speed.

Q: Why won't the automatic input selection feature of the DD540 work with my set-top-box?

A: The automatic input selection feature of any device, including the DD540, depends upon all inputs except the desired one being "off" so the circuitry simply looks for an active input and switches to it. If more than one input is "on" it can't tell which one you want. Unfortunately some set-top-boxes leave their digital audio outputs "on" even when you have turned the set-top-box "off".

If auto-input-selection does not work you can prove this is the cause by unplugging the set-top-box which will stop its digital audio output and the DD540 should then switch to the other input when its source turns on. If you have one of these set-top-boxes you will need to select the DD540's input with its remote or if you have a macro learning remote you may be able to program the discreet IR codes provided by Felston.

Note: Due to so many devices with misbehaving digital audio (always on) Felston dropped the auto input selection feature from the DD740 which now has an input button (ABCD) for each of its inputs.

If you have a Pronto, Harmony, or Universal Remote you can download the Felston discreet codes from your manufacturers website. For the latest in learning remote support from Felston click here for the DD540 and here for the DD740.