Hello,

I'm trying to split the outgoing coaxial signal from a Kemper profiling amp in two.

The reason for this is that one of the coaxial signals is going to return to my interface, while the other will be used as a clock sync input for an external SRC converter I plan connect to the Kemper.

So I'm thinking of implementing the following:

Kemper (as master clock) --> SRC of my Audio Interface (44.1 to 96)
_____________________--> Clock sync input of an External SRC

...on the SPDIF signal out of the Kemper, and the following:

Audio Interface --> External SRC clocked to Kemper's clock (96 to 44.1) --> Kemper.

...on the SPDIF signal into the Kemper.

However, I need to find a way of splitting the Kemper's outgoing SPDIF signal (without introducing latency, artifacts, etc).

So I'm looking for a high-end SPDIF splitter.

Any recommendations?

Note: I initially found the following splitters:

https://www.monoprice.com/Product?c_...seq=1&format=2

...or:

https://www.monoprice.com/Product?c_...seq=1&format=2

...but given that they are passive (and look pretty cheap), I think I'm asking for trouble going with either one.

Thank you,
Nelson

I got the cheap monoprice ones for an idea I didn't end up implementing, but I tested and it seemed to split the spdif signal with no issues

Hello BGood,

Thank you for the heads up.

Can I assume that you took a look at the SPDIF signals to see if they nulled or at least looked similar enough?

It's just odd that something so cheap can work so well.

Thank you,
Nelson

Upon further investigation, it looks like there is a loss in quality.

When using a passive splitter, it seems that there's a loss in gain (ex. 3db) and a loss in amperage (but not voltage).

So it looks like I need to look for a powered amplifier / splitter of some sort.

Nelson

Yah... I ran signal through and it worked and recorded a few seconds of audio and that worked fine too... Nothing too scientific, as it was a quick check... I recall doing the test because I didn't think it'd work either ha

Just read what you wrote about level loss... If it's predictable (always 3db) maybe just boost the input by that much on the receiving device?

Hello BGood,

Out of curiosity, could you try the test again and see if both signals null (if not, no worries)?

As for the signal loss, it's only something I've heard from a few people, I haven't yet tried it myself (but it seems to make sense).

Thank you,
Nelson

Hi
There are many aspects to the distribution of digital signals and the tolerance of various pieces of gear's inputs is usually pretty wide.
The biggest 'killer' is usually cable length (capacitance) which can either distort the signal (round off what should be square) or attenuate it.
For short runs almost anything can work reliable. Losing 3dB in a 'splitter' is also no big deal as the signal level tolerance is usually around 10dB or more.
Thus for a simple setup with short runs and a splitter you should be fine.
Matt S

Hello Matt,

Thank you for the information.

So just to clarify, if I use short 75 ohm SPDIF cables (with the proper shielding / gold tips, ex. Mogami) and a short gold platted splitter:

https://www.svsound.com/products/gold-rca-y-adapter

...then it's highly likely that the only degradation to my SPDIF signal will be a drop in gain, correct (rather than a distorted / artifact-filled signal)?

If so, recording a bit hotter (ex. plus 3 db) to compensate for the db loss shouldn't be a problem.

Thank you,
Nelson

Hi
The level of a SPDIF AES or any other DIGITAL signal stays CONSTANT whatever audio levels might be. Thus a 'loss' in a splitter or long cable is only important if the digital receiver cannot recover the signal (it is just too small).
A SPDIF signal (I can't remember details) might be 0.5 Volt peak to peak at the TRANSMIT end but the receiver can correctly work with 0.1 Volts peak to peak so can tolerate a significant loss.
Matt S

Hello Matt,

Thank you very much for clarifying that.

So basically a digital audio signal is no different than most other digital signals in that, the data integrity is maintained, but the longer the distance of the cable, the greater the likelihood that the digital data will never reach its intended target.

Sort of like how HDMI or USB cables (I think) can only be of a certain length.

If you exceed the length (without using an amplifier), the signal doesn't degrade, but rather simply never reaches its target (whether it be a monitor or USB port).

Thank you for all your help,
Nelson

Hi
Not quite.
There are so many variables involved it is difficult to predict the max length of cable or the point of data corruption so often it is 'try it and check'.
Matt S

Hello Matt,

Understood.

But is the corruption absolute or partial?

In other words, if the signal corruption is partial, then the receiving device will receive a SPDIF signal, but you'll hear random pops, clicks, distortion etc.

Whereas if the corruption is absolute, then the receiving device won't receive anything.

I'm not that worried about an absolute corruption, it's the partial corruption which worries me.

Thank you once again,
Nelson

Hi
Try it. If it sounds OK then use it.
Unless you are going to test it properly with a full digital signal analyser you will never know for sure.
There is more to life than worrying if a cheap digital signal is 'perfect', move on.
Matt S

Hello Matt,

Understood.

Thanks,
Nelson

Sorry to necromance this thread, but I figured someone might appreciate this info (I found this post while searching for this a couple of weeks ago):

There is no signal loss when splitting Coax SPDIF with a 1 in, 2 out connector. I bought a gold-plated Monster branded connector from eBay for $6 bucks and it did the trick. I needed to split Coax SPDIF out of my HEDD to another converter and to a Clarity M meter I recently bought.

Before splitting the signal, I ran a 1k tone on my VU meters (set on the analog outs of my monitoring DAC, which takes the SPDIF signal from the HEDD). After splitting the signal, meters still registered at 0VU. SPDIF cable lenght is 3' between converters and the breakout cable of the Clarity M is about 4'.

I didn't do any other measurements, but can say I know my rig pretty well and would have registered a quality drop. Properly resumed with life after that. I hope this helps anyone who might stumble on this post during a search like I did!

Thx... I think the OP was hung up on thinking somehow that expensive “high end” spdif cables were inherently better than cheapies... but, it’s digital, so, it either fires or it doesn’t is how I’ve always treated digital cabling

Hi
There will be a DIGITAL loss of level by using a 1:2 splitter but this will only affect the audio if the loss is sufficient to cause the receiving 'amplifiers' to miss any digits. It MIGHT make the link more susceptible to 'noise', meaning that for example a very strong AC hum field close to the cable might increase the digital jitter or other artifacts.
The SPDIF (and all digital inputs) have a defined tolerance that they have to meet to be 'approved' however many bits of gear are considerably better than minimum specifications.
Matt S

Thanks for this info. I haven’t noticed anything strange since doing this, probably because of the short cable lengths and because I have everything grounded properly and all cabling is “neat”.

I was asking my recording engineer friend. He said

No you can't split s/pdif.
S/pdif is a digital 2 channel communication between units.
S/pdif is a 75 ohm unbalanced connection, and AES is a 110 ohm balanced version.
There may be converters to convert between the two but you certainly can't split the signal because the digital connection also sync's the 2 machines for digital communication...

A digital signal like spdif or AES absolutely may be split passively as Matt said previously. Spdif is a one way signal sent from one device to another. There is no two way device communication. If you need bidirectional signal flow you have to use two cables.

The audio is represented with ones and zeros transmitted via a nominal peak to peak voltage. When this voltage is split, no perceived audio signal attenuation occurs and as long as the cables aren’t so long that the voltage shrinks (or capacitance grows) to the point that the ones and zeros cannot be accurately read by the receiving device. When this happens, a valid signal will not be received and you will either hear complete silence or the signal may be intermittent.

Most digital audio equipment manufacturers publish specs on their digital inputs and outputs (input should be a nominal X volts peak to peak and a minimum of Y volts peak to peak). Likewise, a digital output peak to peak voltage is usually specified (We will call this Z volts peak to peak). Feel free to research the nominal output voltage of your digital output versus the nominal and minimum input voltages expected by the digital input. You will probably find that voltage Y is far smaller than voltage Z, even if you split voltage Z several times...

All of this assumes that cable lengths are reasonable. If I’m gonna try and go 30’ or more, I’d rather not split even though it should still work with decent cables... but at 6’, I wouldn’t hesitate to split a spdif or aes signal passively.

Now if anyone is concerned about impedance, don’t be.... as long as cable lengths are reasonably short. The specified impedance is all about the characteristic impedance of the cable. A 75 ohm or 110 ohm cable does not introduce a load in the circuit, it is a cable which has been designed to see a termination impedance. This termination impedance allows the cable to have a flatter frequency response (we are talking about megahertz or gigahertz, way out of audio band) and maintain the integrity of the signal over a longer distance.

Bottom line: the shorter length and higher quality the cabling is, the less important these theories and principals become.

All of this said... if you want an active splitter, just search for a distribution amplifier. This will actively distribute a single signal to multiple destinations while keeping proper termination load and nominal signal voltage. I for example, would use a passive splitter in the control room, but a distribution amplifier at a large venue with long runs to multiple mission critical destinations.

Hope this helps!

-B

Roughly speaking most 'transmitting' devices (Sp/Dif outputs) put out around 1 Volt P-P (the manuals would say) and most Receiving ends are happy with anything over half a Volt so as they may have a 'termination' inside the unit it would suggest that one output can usually feed 2 inputs.
The issue is that whatever is received must have a good 'clean' waveshape for the data to be decoded precisely. The classification of 'high quality' cable is a moving feast but the overriding 'factor is that it should be 75 Ohm, although the 'tolerance' can be quite wide which again gets back to length issues. A 6 foot (or less) length of any old crap will likely work as well as any 'high spec' cable.
The effects of poor cable and whether it is terminated correctly are not directly equivalent to an analogue audio signal. Gold plating is largely a marketing ploy unless your gear is used in a damp or corrosive atmosphere in which case you will have other problems anyway.

Totally right although I thought it was ‘S/Pdif’ (as in ‘Sony/Phillips Digital InterFace) but I could easily be wrong and this is minutiae anyway!!!

As long as the cable can deliver a clean square wave above the ‘minimum acceptable voltage’, everything else is irrelevant. Delivering a clean square wave means good bandwidth (as bandwidth decreases, a square wave will eventually become a sine wave).

To understand this, just use any virtual instrument in your DAW of choice that can produce a square wave. Then add a filter plug-in after the instrument and start lowering the filter frequency. As you lower the filter frequency, hear how the square wave gets rounded off and starts to sound more and more like a sine wave! This is exactly what a poor quality cable does to a digital signal (though at a much higher frequency).

At some point, the square wave gets rounded off to the extent that the receiver cannot recognize the beginning and end of a ‘bit’. In order for digital audio to work, the shape of the wave must remain relatively square to a point that the receiver can accurately determine the beginning and end of each bit. The more rounded (more sinusoidal and less square the waveform is), the more jitter will be introduced since the receiver cannot determine the exact ‘start and end’ of the waveform.

If you are running inches or a few feet, any scrap of wire will work, but if you need distance, use proper cables terminated with proper connectors with proper termination load.

-B

Yes, Sony/Philips Digital interface. But it was early morning for me!
Strictly speaking phono (Cinch) connectors are a travesty and 75 Ohm BNC (or other proper 75 Ohm connectors) should have been specified, but I suppose the intended 'few feet' for 'domestic' equipment got hijacked.
Matt S