At what point does a "Master Clock" make a difference....

An external clock just to improve the sound of a single interface? IME, no. It was a topic of much online debate years ago. 'Feels nostalgic posting about it now...

Atomic clocks bringing on the nostalgia like a warm glowy tube.....now ironic...

Unless you like a bit of jitter no, also the Carbon has AVB clocking so between multiple units so there's no reason to ever have an external clock. If you've got multiple disparate units that need sync'ing together that's when you'd buy a clock so they're not out of time with each other. Or I guess you could try MAC's aggregation but I'm not sure about it TBH..

It won't. External clocks were used to sync a bunch of different devices. At some point people seemed to like what they did to the audio so they started using them to "fix" the sound.

But actual measurements show that external clocking, with the cable/connector/termination impairments and parasitics make EVERY external clock technically worse than internal clocking.

What I've found over the years is that deterministic phase noise is not necessarily a problem in any modern system in the sense that the primary tone of interest is quite stable with sideband power being very low and easily filtered out. Almost all sigma-delta converter ICs need 128-512fs clocking, and any residual phase noise would also be divided by those amounts before the sampling hardware sees the clock signal. Again, very, very low chance that jitter is the "cause" of a difference in sound.

However, in external WC cabling, there is a possibility of random external noise being introduced.

There's also the question of *what exactly is happening to the WC signal* once it's brought into the device.

1. Used directly.. Which would be more rare today since sigma-delta converters simply can't use the low frequencies directly.
2. Upconverted by PLL/DPLL to 128-512fs for use by the sigma-delta converters.. Which is more likely but also a worst-case scenario.
3. #2 but following the PLL is a "jitter cleaner" which is another PLL with a specific design to decouple the dirty reference clock from the output.

#2 is a bad situation alone. Taking a noisy WC signal and using it as a direct reference for a PLL would MULTIPLY the jitter by the amount that you'd need to multiply the signal to get 128-512fs for the converter clocking. Imagine multiplying your jitter 512x..

So #3 would be the only viable situation, by multiplying the WC signal up to say 100MHz with an aggressive loop filter(low pass feedback) then using that as a reference for a jitter cleaner IC that divides back down to something like 22.579Mhz for the converter ICs to use.

You can see that the converter IC clocks are now extremely decoupled from the WC input and other than being in sync, have very little to nothing in common with other devices also being clocked in a system. It also has very little to nothing to do with the quality of the WC signal.

So modern converters don't need external clocking at all unless you need some semblance of sync between devices, which I don't think is necessary in any normal use case because DAWs allow you to nudge tracks to line up in phase anyway.

yeah and is svart can explain, it doesn’t do anything at all in converters that utilize asynchronous sample rate conversion to a rate that’s not a multiple of 44.1 or 48 khz like benchcmark, apogee, crane song, and weiss. All that matters is that they get the samples in order. They ignore the clocking altogether. If they have word clock outs or options, it’s to spoof another device into thinking it’s the master clock. The asrc chips are not as good as software now but are still cleaner than the multibit delta sigma converter chips, which are still linear to their noise floor. These converters are among the best products you can buy.

I don't think anybody is saying it's not different.. But I think a lot of people actually like MORE jitter than less jitter.  More jitter would cause more harmonic excitement and would sound "wider" and more "full" and probably have more "detail" in the form of small amounts of distortion.

I watched this interview with Tchad Blake and the interviewer kept asking technical questions and searching for answers on quite scientific levels ….

Tchad kept shrugging his shoulders and saying, honestly I don’t know - I don’t analyse this stuff - I just go with what sounds best.

The interviewer seemed quite disappointed there wasn’t a formal method to Tchad’s success.

I think I fall into that camp - if it sound right - it’s right.

I’m not knocking teachnicel knowledge, obviously not, I wouldn’t have great gear if it wasn’t for the likes of Dave Hill and Rupert Neve (to name but two) of this world.

But what I love about creativity is, it’s very pure, you can just do it without having to “read the manual” so to speak

No it doesn't, science doesn't often change because you want it to unfortunatley.

Satire my friend.    I've never use an external clock.