Digital/optical leads

[johndenim]

Have used them for a while with MD and CD players, but never given it much thought to actually how it works!

 

I'm into a bit of gaming online via PS3 and have recently hooked the thing up properly, and what a difference.

I was wondering if any boffins on here could give a layman's explanation into how it works?

 

5.1 through a little red light?

I know Bobbys a bit of an expert on this subject.

Any thoughts please?

[david.elsbury]

Try these for starters...

http://en.wikipedia.org/wiki/Surround_sound

http://en.wikipedia.org/wiki/Optical_fiber

http://en.wikipedia.org/wiki/Category:Surround_sound

 

I don't know that this is entirely what we deal with here, though, being a little bit more "hi-fi" and less "technical theatre/ backstage".

[paulears]

The thing really is that if you have two items of equipment that carry out their processing digitally, and nowadays apart from the obvious bits of kit, there are now others joining the ranks such as effects processors, mixers, and computers. If you convert digital to analogue, just to stuff it down the cables, then the next bit of kit has to convert it back, process it and then convert it again, then even with the best kit, something gets lost east time. Once the audio is digits, then you can send this electrically down a cable as sp/dif or AES/EBU, but the familiar optical cable is just as capable, using light not electricity.

 

The practical upshot is that audio quality can be as good as the original material, and not degraded by constant conversion. I suppose the real argument is about the 'quality' of the initial digital conversion or generation. 16 bit resolution with 44.1KHz sampling wasn't good enough for the people with golden ears when it first came out, although I subscribed to the view that it was infinitely better than turntable rumble, hum and hiss that we had at the time. My ears have gradually aged and in the studio I rather like Sony Soundforge, and have upgraded many times since I first bought it. I edited something for a client the other day and he called to say that he couldn't play it on his older computer which seemed odd - it was fine here. A bit of investigation showed that the latest version had defaulted to 24 bit 98KHz sampling - and I'd just not noticed. Going back to 16/44.1, even with my old ears, it didn't sound quite the same - so it does make a difference - perhaps a small one, but worth having. With things like video games, it's likely that the entire chain has been digital from conception to your TV - hence why avoiding that last analogue stage make a big difference.

[johndenim]

Thanks for the wiki links David, I'll have a read through.

 

A fascinating insight there Paul, is there a limit as in length?

I know we already have international fibre optics, it this similar or even the same thing?

 

Anyone know why the light is red?

[bruce]

is there a limit as in length?

I know we already have international fibre optics, it this similar or even the same thing?

 

Anyone know why the light is red?

 

Similar, but different wavelength and media. This is a "consumer" grade solution, using red LEDs and plastic 1mm fibre. The fibre has high absorptiojn, so you're theoretically limited to 10m, and 5m is more typical in practice.

 

Why red? - probably because the red LED sources and detectors were more common when the TOSlink standard was defined. It's also a useful visible check that the driver is working. It might be related to the absorption spectrum of the plastic fibre, but I doubt it...

 

Non-domestic fibre applications - such as the long haul comms systems you mention - work on a similar principle, but use much more high precison and expensive components. LED sources are replaced by laser for all but the shortest links, the sources run well into the infrared spectrum rather than visible, the fibres are glass rather than plastic, and the fibre core is much narrower.

[Bobbsy]

With cheap 1mm plastic optical fibres, the maximum length is usually specified as 5 metres though, in practical terms, you can usually get up to 10.

 

You can increase this range with higher quality multistrand fibres or even glass quartz fibres.

 

The undersea fibres you've heard of use slightly different technology--they DO use quartz glass fibres and they're driven by a laser instead of the the LED you'll find in your audio gear. Even so, the undersea fibres have to have repeater stations along the way.

 

I've never really thought about "why red". I'm going to take a guess that it was a combination of red being an easy, cheap LED to get when the technology was being developed and the fact that red is the lowest frequency in the visible spectrum and therefor probably less affected by losses in the inexpensive fibres.

 

Bob

[mackerr]

As others have said, once the signal is digital it can be sent as electrical pulses over various kinds of cable, or as light pulses over fiber optic cable. The lowest bandwidth fiber system is what is commonly known as Toslink, which was developed by Toshiba. It is what is used for consumer 2ch optical SPDIF, and for 8ch ADAT transmission. It uses visible light and a 1mm plastic fiber. It is known to have fairly high jitter characteristics, and should not be used in systems where a "ring" of gear needs to be kept in sync.

 

Once you step up from Toslink you have 2 kinds of professional fiber, Multi mode, and Single mode. The number of modes refers to the internal reflections in the fiber. Multi mode is generally assumed to have a maximum range of 2-3km, Single mode can often go 20km. MM fiber is either 50 or 62 micron in diameter with a cladding that makes the diameter 125 microns. SM fiber is -6- 8 or 9 microns, with the same diameter cladding.

 

Multi mode fiber is becoming more common in pro audio either via systems like Optocore, which is also the system used by Digico in their fiber system, or in systems that use IP technology with switches that use fiber links.

 

If you are an AES member there is a very good tutorial paper in the June 2009 Journal of the AES. Here is the url to the paper.

 

http://www.aes.org/tmpFiles/JAES/20100515/...V57_6_PG461.pdf

 

If you are not an AES member it will cost $20US for the download.

 

Mac

[bruce]

I deliberately avoided going to that level of detail... :D

 

Single mode can often go 20km.

That's without breaking sweat :D links of a hundred km or more without regeneration are practical.

SM fiber is 6 microns, with the same diameter sheathing.

Usually 9, sometimes 8, at least in the UK. Once you get below 9, it makes no difference, only one mode is propagated.