1. Home
  2. cables
  3. 198107-201402

SPDIF Analyzer : Simulation of AES Coaxial and Optical Cables

Hello,

I have developed a simulation tool that compares the behavior of S/PDIF and AES/EBU cables based on their length, featuring real-time graphical displays (please note a slight calculation delay as the processing is quite intensive).

Link: https://spdif.onrender.com/

This is a precise educational tool. You are free to draw your own conclusions from the results. The suggested findings are purely scientific, mathematically validated, and based on thresholds established by industry experts (not by me).

What the simulator models:

  • Bandwidth: Degradation over length (skin effect + dielectric losses, using the model $BW(L) = BW_{1m} / \sqrt{1+L/8}$).

  • Attenuation: Interpolated between 5 and 10 MHz, calibrated against manufacturer datasheets.

  • ISI Jitter: Inter-Symbol Interference, with a critical threshold at 10 ns (Benjamin & Gannon, 1998).

  • Reflections: Reflection coefficient $\Gamma = (Z_{cable} - Z_{ref}) / (Z_{cable} + Z_{ref})$, calculating up to 5 bounces.

  • EMI Noise: Based on cable shielding and environment (Studio / Hi-Fi / Stage).

Available Cables:

Belden 1694A (broadcast reference), Canare L-5CFB, Mogami 2964, Gotham GAC-1, Van Damme Plasma Grade, low-end RCA cables, plastic fiber TOSLINK, and a dozen others. You also have the option to enter your own custom specifications.

Reference Standard:

IEC 60958-3 (S/PDIF) - receiver threshold 200 mV P-P (§3.2). The practical limit for each cable is calculated based on 4 independent criteria (BW, jitter, amplitude, SNR).

Features:

You can adjust the length from 0 to 100m, select the type of PLL receiver (from the vintage CS8412 to modern ASRC), and watch the eye diagram degrade in real time.

Two levels of documentation are available: a general overview and a full technical version containing all physical formulas.

Documentation: https://spdif.onrender.com/doc

It looks like this:

Dell Latitude spdif onrender com

The tool is free, not monetized, uses no cookies, and only performs anonymous traffic tracking. I am happy to answer any questions you might have.

The database only lists documented professional cables developed and built by reputable manufacturers serving the military, pro A/V, and medical sectors. In comparison, Hi-Fi is a "comfort zone" that is far less demanding than these critical industries.

I pay for the hosting myself and included a small "buy me a coffee" link. So far, I have collected exactly zero, but that isn't why I built this.

You can also input values for cables not listed. Just be sure to check if the data you are entering is per 100ft or per 100m.

djoulfr
ghdprentice's avatar
ghdprentice
9,397 posts
 

That sounds interesting. My scientific side is interested. I like measuring things... I have a seismograph, weather station, 4 air quality sensors. 

Do these parameters have a notable connection with the sonic characteristics of the cables? Or, tend to account for some percentage of the sonic variation. Or is this intended as a starting point for someone to attempt correlation between these parameters and sonic characteristics?

djoulfr's avatar
djoulfr OP
2 posts
 

In reality, the conclusions are clear. But when it comes to sound quality, it’s pretty straightforward: it works, it doesn’t work well, or it doesn’t work at all. The cables sometimes perform very well. At the max...

SPDIF cables are essentially SDI cables (video). The same. 

Professionals use "our" digital cables, and professionals are picky. It’s SDI: coax, BNC, 75 ohms-in short, the ones listed in the tool. We buy those too (I use Belden and Canare).

The PROs: This is for digital video:

- 800mV but with a 10% margin. Basically, they’re happy if they lose 50mV over 50m. For us, it’s 500 mV, but we have a 300 mV margin in SPDIF
- It’s 270 Mbps for 480i and 3000 Mbps for 1080p60... for us, it’s 96 kHz 24-bit: 6.1 Mbps

- It’s 270 MHz for 480i and 3,000 MHz for 1080p60; ours is 12 MHz
- The Pro has an 80mV margin, and we have 300mV
- It’s 20, 30, ... meters and cables sometimes exposed to EMI; for us, it’s 2 meters in a living room.

Just to clarify, when it comes to video, we’re talking about fundamental frequencies. So you have to divide by two, but anyway, we’re still in the same order of magnitude here.

 

What can make a difference is the error rate and jitter. The cable shouldn’t add jitter to the receiver’s natural jitter. Over a few meters, there’s no risk with Belden, Canare, Mogami, or similar cables. But with an older receiver, a good cable matters more than with a modern DAC paired with a latest-generation receiver that compensates for jitter no matter what. 

djoulfr's avatar
djoulfr OP
2 posts
 

I’ve modeled the principle of non-standard connectors—for example, a high-end cable fitted with 50 Ohm RCA/BNC connectors.

As a result, there is a new section in the app:

It covers echo and triple transit effects; since these are explained in detail within the app and the documentation, I’ll skip the technicalities here as it’s quite complex. The idea actually came from someone over at the Mélaudia forum.

I’ve added a 75 Ohm coax with 50 Ohm connectors to the cable list. Otherwise, you can use the "Custom" option. This will help you understand the phenomenon and realize just how much the plugs matter. You can play around with the Custom cable and input your own values. Again, this is designed to help you grasp the impact of specific parameters: take two custom cables and tweak the variables to compare them. You’ll see the effects of shielding, length, attenuation, velocity, and, of course, impedance.

Finally, I’ve included a sensitivity matrix showing which parameters impact which results (Error Rate, jitter, etc.) and to what extent.

zlone
1,317 posts
 

Very nice, thanks for sharing.

More to discover