spdif cable length

[hendrixstyle]

Hi

I have a short RCA 75ohm S/PDIF going from my computer's soundcard to my Avocet monitor controller. All is well.

I can't seem to find a definitive answer to whether or not there is a max length, and what that would be.

I just plugged a 35 ft. cable, and everything sounds fine. I am a "if it works it works" kind of guy, but am still wondering if there should be a concern. I am thinking about getting a db meter, and testing high mid and low freqs with both cables, to see if there is a difference. Any thoughts on this?

Thanks!

 

[voodooless]

I am thinking about getting a db meter, and testing high mid and low freqs with both cables, to see if there is a difference. Any thoughts on this?

That’s a futile idea. The cable does not affect the frequency response. If the data integrity fails, you’ll clearly know: you get dropouts, clearly audible artifacts or nothing at all.

As long as you have a proper coaxial cable with the correct impedance, you can bridge significant distances without issue.

 

[staticV3]

Are you transmitting SPDIF? Then measuring frequency response, distortion etc. is pointless, as digital doesn't work that way. It's data packets devoid of the concept of audible frequencies.

There is no mechanism for losing just bass in SPDIF transmission.

There is only a mechanism for packet loss (signal dropouts), as well as jitter.

 

[DVDdoug]

I am thinking about getting a db meter, and testing high mid and low freqs with both cables,

Digital audio doesn't work like that. If the data gets corrupted you might get nasty clicks or pops, or dropouts, or it won't work at all.

 

[chelgrian]

Hi

I have a short RCA 75ohm S/PDIF going from my computer's soundcard to my Avocet monitor controller. All is well.

I can't seem to find a definitive answer to whether or not there is a max length, and what that would be.

I just plugged a 35 ft. cable, and everything sounds fine. I am a "if it works it works" kind of guy, but am still wondering if there should be a concern. I am thinking about getting a db meter, and testing high mid and low freqs with both cables, to see if there is a difference. Any thoughts on this?

Thanks!

S/PDIF over coax is specified to 10m however in reality it will work with far longer cables than that.
If it works it works, if it doesn’t work you’ll either get nothing at all or random dropouts if it’s just on the margins of the tolerance of the transceivers.

 

[hendrixstyle]

Thank you everyone.

Digital audio doesn't work like that.

It's not digital when it comes out of the monitors.

That’s a futile idea. The cable does not affect the frequency response.

This wouldn't be a frequency response check, but a LEVEL check.

 

[Devnull]

Thank you everyone.



It's not digital when it comes out of the monitors.



This wouldn't be a frequency response check, but a LEVEL check.

Level doesn't change, it's just part of the bitstream.

 

[hendrixstyle]

Level doesn't change, it's just part of the bitstream.

Oh, I see.

 

[Beave]

To be clear, the level of the digital bitstream sent across the cable could change, but that doesn't affect the level of the resulting audio once that bitstream is converted to analog.

As stated above, if the cable is too long, you will get pops, clicks, or dropouts, or it won't work at all.

 

[LTig]

That’s a futile idea. The cable does not affect the frequency response. If the data integrity fails, you’ll clearly know: you get dropouts, clearly audible artifacts or nothing at all.

Yep.

As long as you have a proper coaxial cable with the correct impedance, you can bridge significant distances without issue.

At 44 kHz samplerate I had no problems with 16 m of RG59.

 

[IsaacOscar]

As others have said, if you don't hear the sound dropping out or popping (which is just a really short dropout) don't worry at all.

I've had analog cables cause buzzing/static/noise, but digital cables don't have such problems (which is often why they're better: if you get a working connection, the signal is going to be 99.9% accurate; bits can still flip occasionally, and I don't believe SPDIF has error detection)

It's not digital when it comes out of the monitors.

Right, which is why if you really wanted to waste time checking the accuracy of your SPDIF cable, what you need is some kind of recording device with an SPDIF input. Then record some music with it, and check that it's the same music as you sent it (in theory, if your recording it in the same PCM format as you're playing it, the bits should be identical; but some minor variation is unlikely to be audible).

Trying to measure the output of your monitors would just be making your measurements unnecessarily inaccurate.

This wouldn't be a frequency response check, but a LEVEL check.

If your playing pure tones (i.e. sine waves) to check the levels, then this is exactly a frequency response check.

 

[hendrixstyle]

So, I have been listening to an album for about 45 min, and no dropouts, clicks, pops, nothing. I would like to say I am good to go, but is it possible that there is some loss that I am not perceiving, because it is slight - a difference that I might hear if the ability to do a proper A/B test existed?

 

[IsaacOscar]

So, I have been listening to an album for about 45 min, and no dropouts, clicks, pops, nothing. I would like to say I am good to go

Yes, but I have had optical cables dropout occasionally. Don't worry though until a problem actually occurs.

but is it possible that there is some loss that I am not perceiving, because it is slight - a difference that I might hear if the ability to do a proper A/B test existed?

The ability does exist, you just need to get a really short cable and have someone change cables on you behind your back.

None of this is worth worrying about or checking.
But there could theoretically be some slight jitter, some bit flips, or just a delay (totally inaudible as electrical signals travel at roughly the speed of light)

 

[DVDdoug]

It's not digital when it comes out of the monitors.

The S/PDIF connection which you asked about is digital.

but is it possible that there is some loss that I am not perceiving, because it is slight - a difference that I might hear if the ability to do a proper A/B test existed?

No. Did you read what everybody said? Another analogy I like to use is that one "flipped bit" in your bank account is equally likely to cause a 1-cent error or a billion dollar error! It might be a 1-cent error that you don't notice but it's more-likely to be a very noticeable amount.

To be fair, audio is a bit different because "CD quality" audio has 44,100 samples per second and one very-badly corrupted sample is not likely to be noticed. On the other hand, if the audio stream is interrupted or delayed too long, you'll get clicks, pops, or dropouts. When you check your bank balance it's not time-critical and if there is an error it gets re-transmitted.

Sometimes there are glitches with streamed audio or video going all around the world through all kinds of multiple connections (and again they are usually very obvious.). As a rule, digital is very reliable and robust and when it goes bad it usually goes very bad.

to do a proper A/B

You'd need to do a blind, statistically valid ABX test (We often imagine that we hear differences and proper repeatable blind listen tests take out those biases and errors.)
Also see Controlled Audio Blind Listening Tests (video)

 

[hendrixstyle]

The S/PDIF connection which you asked about is digital.

Fully aware of that, thank you.

Did you read what everybody said?

No dude - I hear only the voice in my own head. The responses that were given were AI.

 

[sam_adams]

I can't seem to find a definitive answer to whether or not there is a max length, and what that would be.

From (PDF).

And:

From (PDF).

S/PDIF is standardised in IEC 60958 as IEC 60958 type II (IEC 958 before 1998).

 

[IsaacOscar]

View attachment 509400

From (PDF).

Yes but how do they calculate that maximum length? Oftentimes it's just a specification, i.e. any longer is "illegal".

It's important for long cables apparently that they are 75Ω. So if you find such a cable longer than 10m, it should be ok. E.g. this 30m cable.

You can also get TOSLINK cables up to 30m, despite the spec limiting them to 10m.

 

[skankhobag]

This is one of those Your Mileage May Vary kind of deals. If the S/PDIF transmitter and receiver are sufficiently robust and the cable has the proper impedance, you can run much longer cables than the industry's spec. I found that to be the case with AES/EBU.

Quite often, those specs are very conservative to account for worst case scenarios.

As stated several times earlier, if you don’t hear any noises or drop outs, you’re good.

The only thing to measure is re-sent data packets that were caused by data corruption. I’m sure that can be done, but who’s going to own equipment like that?

 

[IsaacOscar]

The only thing to measure is re-sent data packets that were caused by data corruption. I’m sure that can be done, but who’s going to own equipment like that?

Although most digital protocols do do that, I don't think S/PDIF does? I.e. if there was corrupted dsta, thats what gets played.

 

[voodooless]

Although most digital protocols do do that, I don't think S/PDIF does? I.e. if there was corrupted dsta, thats what gets played.

It does not. It’s a one way street. If something is detected as corrupt,, usually the sample is dropped and interpolated. If there are many, you get what was already described.

The responses that were given were AI.

What? None of the responses were AI generated.