StackAudio SmoothLAN Ethernet Filter

[theveterans]

If I’m not mistaken, the Volumio Primo does have a WiFi receiver. So it was exposed to the ambient WiFi “noise”.

Hopefully they revisit this with testing through WiFi instead. My point though in the first place is that there are some conditions where Ethernet filtering and this thread

StackAudio SmoothLAN Ethernet Filter​

can definitely have audible effect in sonic perception and one potential/not definite is the effect of jitter and CM EMI going through Ethernet on some streamers where it can benefit from this Ethernet filter

 

[XTF1]

Hopefully they revisit this with testing through WiFi instead.

I don’t have data or measurements disproving this.

can definitely

That’s the whole issue with this type of measurements: “can” does not equal to “will” in every possible scenario. In addition, even if it causes measurable jitter, I didn’t see them assessing the audible impact.

If every streamer requires a filter to work properly… it is just poor design or flat out robbery!

 

[theveterans]

As I’m more of a subjectivist but with the curiosity for speculation through science. I have my opinion /belief that there are scientific tests that can potentially infer audibility but will never agree in claiming absolutes that this specific test method (e.g. FFT 1KHz SINAD) is the only absolute method that everything that meets a certain threshold is audibly transparent and that I’m deluding myself for biases from hearing audibly different

So with regards to Ethernet filtering, it’s not really about working properly or we won’t have internet at all but more of whether there’s an audible difference with the filtering and Ethernet cables on some streamers that aren’t designed to eradicate as much EMI or have their crystal oscillators be immune to jitter and noise as much as possible. We are talking very small amounts of jitter differences as a whole, but I do not infer that jitter is the only potential factor if there’s a subtle audible difference. I myself pretty much is posting my opinion and of course YMMV in everything

 

[bfreedma]

It has NOTHING to do with sound perception other than sounds cutting in and out during buffer overrun.

You don’t seem to understand buffering and why the numerous times data is buffered would eliminate any relationship between electrical noise that in theory would somehow avoid the galvanic isolation existent in every spec wired ethernet connection.

And where did “buffer overrun” come from? Buffer overflow is a condition that used to occur when more data was written to a buffer than the memory allocated to that buffer allowed. That has zero relationship to this discussion whatsoever.

 

[theveterans]

You don’t seem to understand buffering and why the numerous times data is buffered would eliminate any relationship between electrical noise that in theory would somehow avoid the galvanic isolation existent in every spec wired ethernet connection.

You also don't seem to understand that buffering has poor (I'm not claiming zero) correlation to audibility differences perceived with Ethernet filtering

 

[gregorio]

It has NOTHING to do with sound perception other than sounds cutting in and out during buffer overrun.

Correct, it has nothing to do with sound, let alone sound perception and yet you keep arguing that it does!

Above are objective measurements that can really affect the sound perception through different cables and different switches and Ethernet filtering.

So let’s be clear, the threshold of audibility of jitter with music is typically around 200,000 - 500,000ps (pico seconds), the lowest ever detected was 27,000ps and you’re claiming jitter of about 8ps “can really affect the sound perception”? Regardless, there is no jitter at all when the data is buffered, there obviously cannot be! There is no time between samples when the data is just sitting there in memory, so obviously there cannot be any timing inaccuracy (jitter). Obviously the data has to leave the Ethernet buffer at some point to be fed into the DAC chip and as it’s just sitting there without any time, that’s why you need a DAC clock.

You also don't seem to understand that buffering has poor (I'm not claiming zero) correlation to audibility differences perceived with Ethernet filtering

Buffering has no correlation at all to any audibility differences and it cannot have. You cannot hear data just stilling in a memory buffer.

G

 

[theveterans]

So let’s be clear, the threshold of audibility of jitter with music is typically around 200,000 - 500,000ps (pico seconds), the lowest ever detected was 27,000ps and you’re claiming jitter of about 8ps “can really affect the sound perception”? Regardless, there is no jitter at all when the data is buffered, there obviously cannot be! There is no time between samples when the data is just sitting there in memory, so obviously there cannot be any timing inaccuracy (jitter). Obviously the data has to leave the Ethernet buffer at some point to be fed into the DAC chip and as it’s just sitting there without any time, that’s why you need a DAC clock.

The measurements indicated that jitter at the DAC clock is being influenced by different Ethernet cables and filtering. Obviously buffering does not fix it since if it is, you can literally put a 10 ns jitter through the Ethernet and yet the DAC clock should be immune to that cause muh Ethernet buffer... (yet it's not objectively!)

 

[gregorio]

The measurements indicated that jitter at the DAC clock is being influenced by different Ethernet cables and filtering.

The measurement demonstrated a difference between the unshielded twisted pair (UTP), extremely cheap CAT5 cable and CAT7 cables (items 2-6) of around half a pico-second. That’s a surprise to me, I thought it would have been more, half a trillionth of a second is excellent. It’s also, from what I can tell, completely irrelevant compared to the Allan Variance of the clock itself. And, where did all the jitter from the switch go, it’s just disappeared, it must be magic?

Lastly, you did not answer the question, the audible threshold of jitter with music is typically >200,000ps, so how does 8ps, or the variation of just half a ps “really affect the sound perception”?

G

 

[phonehom]

I was curious about the company, took a look at their website, and AFAICT, it's mostly isolation/signal preservation products. They've been around since 2013, so I guess they're doing ok. I'm skeptical that the Smoothlan would make that big a difference.

 

[theveterans]

I was curious about the company, took a look at their website, and AFAICT, it's mostly isolation/signal preservation products. They've been around since 2013, so I guess they're doing ok. I'm skeptical that the Smoothlan would make that big a difference.

The Smoothlan regenerator would make more of a difference than the filter. However, your streamer must be very transparent to notice it. Even a regular PC will benefit from Ethernet filtering but only if using a galvanic isolated USB receiver built in the DAC

 

[gregorio]

Even a regular PC will benefit from Ethernet filtering but only if using a galvanic isolated USB receiver built in the DAC

Because galvanically isolating something twice must be better than just once!

G

 

[theveterans]

@alexaldente



Looks like the SmoothLAN Ethernet filter is a dud in his system since it makes the sound less focused and articulate. However, the reviewer recommends the SmoothLAN regenrator instead which reclocks (like all switches do BTW) the signal and while at the same provides cleaner signal to your streamer

 

[gregorio]

Looks like the SmoothLAN Ethernet filter is a dud in his system since it makes the sound less focused and articulate.

Someone on YouTube said so, why didn’t you say, it must be true then! lol

G