optical out ?

[radish]

Actually, in this case it is, right? The light has to turn on and off at nearly 44,100 times every second at a rate that allows for a sensor to accurately read and understand. How do you make that light bulb go from off to full intensity and then back down to completely off? Also remember that the bulb is being turned on/off by an electrical signal. To do this without adding jitter is certainly very complex...

The solution to that, as has been mentioned, is to not use a bulb, you use an LED or laser diode. My mention of a bulb was simply an example of a device for converting electrical energy to optical (well, photonic) energy. In reality, 44.1kHz is such a slow switching speed for any kind of signalling (fibre optic or electrical) as to be truly "simple" with todays technology.

I'm not going to get into the arguments about jitter (I've had enough of those already) but to make a statement that "electrical to optical conversions are harder than electrical to magnetic ones" is vague and meaningless. And to use that as a reason why a HDD is "better" than a CD is equally absurd. First you have to define "better", and depending on that definition you'll get a different answer.

[pfarrell]

On Tue, 2005-12-06 at 14:06 -0800, radish wrote:
> In reality, 44.1kHz is such a slow switching speed
> for any kind of signalling (fibre optic or electrical) as to be truly
> "simple" with todays technology.

A followup nit, 44.1kHz is the sample rate. The signaling rate
is at least 32 times that, two channels for stereo, 16 bits per
sample.

Still, relative nearly anything these days, this is a trivial.
It is a tad less than 1.4 mHz, or 176K bytes/second
which is of course, the where the 1X CD speed of about 150kB/s
comes from.

And a little overhead, some error correcting and you still have
no challenge. ( I can't remember if SPDIF is raw PCM or has
overhead).

Even cheap CD readers are 16x or 32x these days.

--
Pat
http://www.pfarrell.com/music/slimse...msoftware.html

[PhilNYC]

I'm not going to get into the arguments about jitter (I've had enough of those already) but to make a statement that "electrical to optical conversions are harder than electrical to magnetic ones" is vague and meaningless. And to use that as a reason why a HDD is "better" than a CD is equally absurd. First you have to define "better", and depending on that definition you'll get a different answer.

In the context of the conversations I've had, "better" = "less jitter", so if you don't buy the idea that jitter has an impact on sonic performance, then there's nothing more I can say. But if that is true, then do you then believe that it shouldn't matter whether you use optical vs. coax?

[radish]

In the context of the conversations I've had, "better" = "less jitter", so if you don't buy the idea that jitter has an impact on sonic performance, then there's nothing more I can say. But if that is true, then do you then believe that it shouldn't matter whether you use optical vs. coax?

We weren't talking optical vs coax, we were talking CD vs HDD, that's what the original quote was ("HDDs are better than CDs because electrical/magnetic is easier than electrical/optical").

As for the coax vs optical question, I will dump all I know about the topic right here:

1) Back in the day, every audio magazine I read said coax was better, without any science to back that up.
2) All the fastest data networks I work with are optical. That tells me it's easier to get data quickly and accurately down a fiber.
3) No ground loops with optical.
4) My physics background would lead me to believe that optical is a better bet because, as has been mentioned, you don't have nasties like capacitance and induction (or even crosstalk and interference for that matter). The only issues with optical are signal degredation (not a problem for such short runs) and TIR, which only happens with very high frequencies and tight kinks in the cable. Both of those would degrade (or totally block) the signal, but add jitter? Doesn't seem likely to me - what with c being a constant and all.

[bludragon]

I beleive the reason electrical coax spdif is often said to be better than optical toslink spdif is due to the relativly low bandwidth of the toslink transmitter and receiver. They were originally specced to have enough bandwidth to support spdif data rates, however, the relativly low bandwidth headroom can give rise to large amounts of jitter. I havn't checked how valid this is with current devices.

Saying spdif optical is better than spdif coax because high speed telecoms uses optical makes as much sense as saying my 10 year old renault is going to be better than your 10 year old honda because renault won the F1 constructors title this year.

[PhilNYC]

We weren't talking optical vs coax, we were talking CD vs HDD, that's what the original quote was ("HDDs are better than CDs because electrical/magnetic is easier than electrical/optical").

Ah. I thought we were:

The reason I've heard from some engineers is that the conversion process from one medium to another (eg. optical to electrical) is where there is the greatest probability of adding jitter. With coax, everything stays electrical, but with toslink, the signal must be converted from electrical to optical and back.

[bludragon]

optical vs coax:

http://www.audioholics.com/techtips/...ts/toslink.php

http://www.hifichoice.co.uk/review_print.asp?ID=2720

[LavaJoe]

optical vs coax:

http://www.audioholics.com/techtips/...ts/toslink.php

Strange that they spec toslink at such a low bandwidth (5-10 MHz) - I suspect, as I've read elsewhere too, that it would be the emitter/detector pair that imposes such a low bandwidth. The upper limit of bandwith for the optical cable itself is, of course, the frequency of the light (10^14 Hz), but there are other factors, like the way light bounces around in the cable, smearing the pulse edges:

http://www.commspecial.com/fiberguide-print.htm

Still, it's measured in MHz/km! And the lower limit they quote is a few MHz per km. Given that toslink cables are a lot shorter than a km, even the worst cable, I would think, would handle very high frequencies (around 1GHz?).

[radish]

Saying spdif optical is better than spdif coax because high speed telecoms uses optical makes as much sense as saying my 10 year old renault is going to be better than your 10 year old honda because renault won the F1 constructors title this year.

That's a pretty bad analogy (both Honda and Renault use the same basic technology - the internal combustion engine, and what we're discussing are two very different basic technologies), but I don't really want to get into this discussion. I don't profess to be an expert in any of these fields - I only got involved in this thread when I read the statements about one conversion being "easier" than the other, which I still consider to be rubbish. As to the actual merits of coax vs toslink, I'll leave that discussion to the experts and those who actually care

[radish]

Ah. I thought we were:

If you read my first post in the thread, I was quoting and responding to the second part of your post - about HDDs vs CDs.