|
Post by NigelB on Dec 23, 2023 10:06:11 GMT
Strictly speaking it's a jitter and phase noise reducer. By so doing, it improves the quality of the feed into the DAC through reduced superimposed noise, improving the conversion. It does this by passing the ethernet signal through a highly accurate clock. Even better if a superior external clock is used. That's why I called it a reclocker, but it could easily be called a 'cleaner'. It's very effective, I can hear what it does even though it's upstream of my DDC/DAC combo with its own external clock. There is no money wasted on the superior external clock, the benefits come from both the clock and the moat isolation (and also the additional moat created by the fibre connection from the optical cage). You may have read me saying that in digital, everything matters. It really does. There is no such thing as jitter in ethernet. The reclocking in an ethernet switch does nothing for sound quality.
I'm confused why you mention moats. The "moat" is in the switch, not in an external clock. Moats make a difference by reducing noise.
|
|
|
Post by MartinT on Dec 23, 2023 10:25:41 GMT
Where I disagree is in the ability of ethernet clock accuracy to impact sound quality. I'm not saying external clocks can't make a difference to sound quality, I'm saying clock accuracy can't. The AfterDark clocks you mention: what are they attached to please? We could be arguing semantics here. If by 'more accurate' you mean any kind of long term accuracy or stability, I agree that it's not required for audio. The clock measurements that matter are short term variation (jitter and Allan Variance) and phase noise, group them all as noise if you like. My 1st AfterDark clock is connected to the EtherREGEN to improve its performance. My 2nd AfterDark clock is connected to the Gustard U18 and thence to the Gustard X26 Pro DAC via I2S, and improves the performance of both. By the way, the U18 also has a moat and yet the external clock improves its performance, too.
|
|
|
Post by MartinT on Dec 23, 2023 10:29:55 GMT
I'm confused why you mention moats. The "moat" is in the switch, not in an external clock. Moats make a difference by reducing noise. I'm fully aware that the moats are in the EtherREGEN and the U18, not in the clocks.
|
|
|
Post by NigelB on Dec 23, 2023 14:49:40 GMT
Where I disagree is in the ability of ethernet clock accuracy to impact sound quality. I'm not saying external clocks can't make a difference to sound quality, I'm saying clock accuracy can't. The AfterDark clocks you mention: what are they attached to please? We could be arguing semantics here. If by 'more accurate' you mean any kind of long term accuracy or stability, I agree that it's not required for audio. The clock measurements that matter are short term variation (jitter and Allan Variance) and phase noise, group them all as noise if you like. My 1st AfterDark clock is connected to the EtherREGEN to improve its performance. My 2nd AfterDark clock is connected to the Gustard U18 and thence to the Gustard X26 Pro DAC via I2S, and improves the performance of both. By the way, the U18 also has a moat and yet the external clock improves its performance, too. It's Christmas and I do wish we were merely arguing semantics but I'm afraid we're not. There is no jitter in ethernet; well, I've never heard it called that outside of audiophile circles. In the ethernet domain, clock accuracy and/or stability over the short, medium or long terms do not affect sound quality. The data frames are unpacked into a bitstream by a streamer and at this point clock stability/accuracy becomes a thing of importance (and poor performance is jitter).
I've never used or heard an AfterDark clock but I can absolutely imagine it making a significant difference to the performance of the DDC as this appears to operate in the bitstream world.
Where we differ is on the AD/Etherregen: if the AD clock attached to the Etherregen does anything at all for sound quality, it is due to its being lower noise than the internal clock, not to its accuracy/stability/amount of jitter.
Peace and love!
|
|
|
Post by MartinT on Dec 23, 2023 16:55:33 GMT
Where we differ is on the AD/Etherregen: I can't see how the AD clock attached to the Etherregen can do anything at all for sound quality, as there isn't a mechanism. Hearing is believing. There is a mechanism for a lower noise clock resulting in lower superimposed noise on the ethernet signal.
|
|
|
Post by HD Music & Test on Dec 23, 2023 18:10:35 GMT
Simple lets have a shoot out
|
|
|
Post by stellabagpuss on Dec 23, 2023 19:30:47 GMT
Simple lets have a shoot out Clocks at dawn...
|
|
|
Post by NigelB on Dec 23, 2023 20:46:36 GMT
Where we differ is on the AD/Etherregen: I can't see how the AD clock attached to the Etherregen can do anything at all for sound quality, as there isn't a mechanism. Hearing is believing. There is a mechanism for a lower noise clock resulting in lower superimposed noise on the ethernet signal. I agree, and I hope you will agree that I have either said this before or not ruled it out (I will check what I typed shortly). I thought I had been clear - up to the quoted text at least - that I have no issue at all with the concept of a lower noise ethernet clock but I do take issue with the concept of a higher accuracy/lower jitter ethernet clock. Jitter is not noise.
EDIT: the excerpt above was taken out of context. I've clarified it so it makes sense in isolation. It now reads "Where we differ is on the AD/Etherregen: if the AD clock attached to the Etherregen does anything at all for sound quality, it is due to its being lower noise than the internal clock, not to its accuracy/stability/amount of jitter."
|
|
|
Post by NigelB on Dec 23, 2023 20:48:30 GMT
Simple lets have a shoot out That would be fun and interesting but it might not be relate cause to effect. I have no issue with ethernet devices making a difference to sound quality, I do have an issue with that difference being attributed to clock accuracy, timing, jitter, or other timing-related matters in the ethernet space.
|
|