Even if this is another thread about SQ with different inputs, I will narrow it down quite a bit here and focus on the oscillator in use.
Background: An (XO) Ether Regen is on its way. Now there is the server/player that’s the missing fix in my setup. I will need a new server and am totally open if the server should contain a player as well, or if I should stick to the CI Roon player. I could think of building the server(player) or buy a commercial product. One switch, one multirail LPSU, one server(+player) is the “box count limit” on the streamer side of things.
Back to OCXO oscillators:
S/PDIF out in a Server + Streamer controls the audio clock into the DAC. That will make it possible to let OCXO devices such as Aurender N10 or Pink Faun SPDIF bridge PCIe own the audio clock. But this will not happen with the Expert Pros, right? Is it not so that the Expert Pros inputs, including S/PDIF, will change/process the signal and the OCXO oscillator precision will be lost? Put it in another way. What is the benefit of OCXO oscillator precision in a device into an Expert Pro?
No one knows the oscillators in use in Expert Pros right?
Now ethernet:
If building a server I would like to use the OCXO 5Volt dual port PCIe JCAT NET Card XE. Ether Regen B-side - NET Card XE – Expert Pro. Again, how to make the most out of that high precision JCAT oscillator in a server in a Expert Pro setup?
For example, the RAAT protocol states that the DAC controls the clocking, right?
I hope that OCXO oscillators still are utilized in a good way? If so, what to think of implementing them in our Expert Pro systems?
Best regards
Best regards
Please note that improving the clocking in the EtherRegen (and USB) devices is different to the clocking of the audio signal in the Aurender, Pink Faun S/PDIF bridge etc. You cannot use one type of clocking with the other.
As for the possibility of external clocking for the Expert Pro, Devialet will never, ever divulges plans (as you know!).
With S/PDIF, the clock is with the source signal and needs to be "extracted" by the receiving device. So for S/PDIF alone, there is very good reason why a source with an accurate clock might be beneficial.
With USB or Ethernet, the Devialet is fully in control of the audio clock and the audio data needs to pass through a FIFO type buffer. So in these cases it is very difficult to explain why an OCXO type clock in the "sending device" should make any difference whatsoever over any other reasonably accurate low jitter "to spec" clock.
So considering all of the above, the logical conclusion would be to not worry to much about the source (except in terms of functionality), use USB or Ethernet and enjoy the music.
Moving to the subjective, I personally much prefer the sound from my Devialet using a OCXO clocked device feeding the Devialet via S/PDIF. (AES/EBU) So maybe the method that the Devialet uses for S/PDIF clock extraction does afford some benefit for a low jitter S/PDIF source? That said, I do not think anyone outside of Devialet has any specific details of how the Devialet S/PDIF clock recovery works or if a "clocked source" can provide an objective improvement over USB or Ethernet.
If anyone does have some details re exactly how the Devialet S/PDIF clock recovery works - please let us all know!
I am aware of async. - S/PDIF differences IN GENERAL regarding clocking. Since I am in OCXO-modus here S/PDIF therefore may seem like the best solution in this case. But as said. I got a fairly strong hunch that OCXO-clocking is not kept by Devialets input processing in S/PDIF. But I hope I am wrong. "If anyone does have some details re exactly how the Devialet S/PDIF clock recovery works - please let us all know!" AMEN to that Confused!
On the other hand I think of it not being so unlikely that OCXO oscillators might reduce jitter despite being in devices that later on are "overrun" by the Expert Pro. Meaning that there could be some benefit to OCXO oscillators anyway. Input on when/where they might have made a difference would therefore be interesting to hear about. Personally I am interested in experiences with JCAT NET Card XE and the Pink Faun SPDIF bridge into the Expert Pros.
I don't know any technical details of how the Devialet processes the input signal but I know that Devialet's chief engineer confirmed that the best input would be AES/EBU, followed by spdif, and lastly USB. He said that Devialet doesn't have such a high-quality clock as some of those streamers today (OCXO). This is not something I read on the internet but I was there at the "manufacturer meets users" event in Munich during the high-end show but at a different location, when we asked this question. About 6-7 of us were there from this forum. (CI board was not discussed then). They said that when Devialet was in the design phase OCXO clocks would have been too expensive to implement.
I can also confirm that my SOtM SMS-200 ultra streamer sounds better if reclocked by a Mutec MC3+USB which also converts the SOtM's USB output to AES/EBU. Both are clocked by a Mutec Ref 10 OCXO 10Mhz reference clock. I'm just replacing the MC3+USB with a Denafrips GAIA which has two OCXO clocks for 44.1 and 48kHz and multiples of those and is considerably better than the MC3+USB.
So all I can say is: get the best possible OCXO clocked source or DDC and use the AES/EBU input.
The CI board has better clocking than earlier models. That is one of the improvements that comes with the CI upgrade. According to Mathieu Pernot there is also a reclocker in use on the CI. This will make all input sound better than pre CI amps. That said, a good external DDC with reclocking, like a Mutec MC-3 +USB/Ref10 combo, can still improve sound quality on SPDIF (AES/EBU) inputs of a Devialet amplifier.
(18-Apr-2022, 14:50)zdenes Wrote: [ -> ]I don't know any technical details of how the Devialet processes the input signal but I know that Devialet's chief engineer confirmed that the best input would be AES/EBU, followed by spdif, and lastly USB. He said that Devialet doesn't have such a high-quality clock as some of those streamers today (OCXO). This is not something I read on the internet but I was there at the "manufacturer meets users" event in Munich during the high-end show but at a different location, when we asked this question. About 6-7 of us were there from this forum. (CI board was not discussed then). They said that when Devialet was in the design phase OCXO clocks would have been too expensive to implement.
I can also confirm that my SOtM SMS-200 ultra streamer sounds better if reclocked by a Mutec MC3+USB which also converts the SOtM's USB output to AES/EBU. Both are clocked by a Mutec Ref 10 OCXO 10Mhz reference clock. I'm just replacing the MC3+USB with a Denafrips GAIA which has two OCXO clocks for 44.1 and 48kHz and multiples of those and is considerably better than the MC3+USB.
So all I can say is: get the best possible OCXO clocked source or DDC and use the AES/EBU input.
Hello
I am very interested to know your opinion about the Denafrips GAIA vs Mutec MC3+USB.
Please share when you can.
Thanks
(22-Apr-2022, 20:45)jbp Wrote: [ -> ] (18-Apr-2022, 14:50)zdenes Wrote: [ -> ]I don't know any technical details of how the Devialet processes the input signal but I know that Devialet's chief engineer confirmed that the best input would be AES/EBU, followed by spdif, and lastly USB. He said that Devialet doesn't have such a high-quality clock as some of those streamers today (OCXO). This is not something I read on the internet but I was there at the "manufacturer meets users" event in Munich during the high-end show but at a different location, when we asked this question. About 6-7 of us were there from this forum. (CI board was not discussed then). They said that when Devialet was in the design phase OCXO clocks would have been too expensive to implement.
I can also confirm that my SOtM SMS-200 ultra streamer sounds better if reclocked by a Mutec MC3+USB which also converts the SOtM's USB output to AES/EBU. Both are clocked by a Mutec Ref 10 OCXO 10Mhz reference clock. I'm just replacing the MC3+USB with a Denafrips GAIA which has two OCXO clocks for 44.1 and 48kHz and multiples of those and is considerably better than the MC3+USB.
So all I can say is: get the best possible OCXO clocked source or DDC and use the AES/EBU input.
Hello
I am very interested to know your opinion about the Denafrips GAIA vs Mutec MC3+USB.
Please share when you can.
Thanks
Well, the Denafrips GAIA is so much better than the Mutec MC3+USB with linear power supply (Uptone Audio JS-2) and even clocked with the Mutec Ref10 that I already bought one and removed the Mutec from my system. I left the Ref10 still in as it is also clocking the SOtM SMS-200 ultra and the Uptone Etherregen switch.
With the GAIA, everything sounds more life-like. Timbre and tone improved the most and made the system sound a bit warmer which is a really good thing with the Devialet and the YG Acoustics speakers. Over 30 years into this hobby, I have gone through hundreds of A/B testing of all sorts of things. Sometimes I had to go back and forth a few times to be able to fully evaluate differences and decide which one I like better. Not this time. 30 secs into one of my test tracks, I knew that it was so much better, that I didn't even need to test it longer. Well, the Mutec gave me a similar 'wow' factor when I first put in between the streamer and the Devialet. The GAIA is probably better because it has two OCXO clocks inside that provide all audio frequencies (x1024, so it can do DSD1024 theoretically, practically DSD512) and there is no complicated PLL needed as in the Mutec. MSB explains it very well:
https://www.msbtechnology.com/dacs/clock-options/
(24-Apr-2022, 04:48)zdenes Wrote: [ -> ] (22-Apr-2022, 20:45)jbp Wrote: [ -> ] (18-Apr-2022, 14:50)zdenes Wrote: [ -> ]I don't know any technical details of how the Devialet processes the input signal but I know that Devialet's chief engineer confirmed that the best input would be AES/EBU, followed by spdif, and lastly USB. He said that Devialet doesn't have such a high-quality clock as some of those streamers today (OCXO). This is not something I read on the internet but I was there at the "manufacturer meets users" event in Munich during the high-end show but at a different location, when we asked this question. About 6-7 of us were there from this forum. (CI board was not discussed then). They said that when Devialet was in the design phase OCXO clocks would have been too expensive to implement.
I can also confirm that my SOtM SMS-200 ultra streamer sounds better if reclocked by a Mutec MC3+USB which also converts the SOtM's USB output to AES/EBU. Both are clocked by a Mutec Ref 10 OCXO 10Mhz reference clock. I'm just replacing the MC3+USB with a Denafrips GAIA which has two OCXO clocks for 44.1 and 48kHz and multiples of those and is considerably better than the MC3+USB.
So all I can say is: get the best possible OCXO clocked source or DDC and use the AES/EBU input.
Hello
I am very interested to know your opinion about the Denafrips GAIA vs Mutec MC3+USB.
Please share when you can.
Thanks
Well, the Denafrips GAIA is so much better than the Mutec MC3+USB with linear power supply (Uptone Audio JS-2) and even clocked with the Mutec Ref10 that I already bought one and removed the Mutec from my system. I left the Ref10 still in as it is also clocking the SOtM SMS-200 ultra and the Uptone Etherregen switch.
With the GAIA, everything sounds more life-like. Timbre and tone improved the most and made the system sound a bit warmer which is a really good thing with the Devialet and the YG Acoustics speakers. Over 30 years into this hobby, I have gone through hundreds of A/B testing of all sorts of things. Sometimes I had to go back and forth a few times to be able to fully evaluate differences and decide which one I like better. Not this time. 30 secs into one of my test tracks, I knew that it was so much better, that I didn't even need to test it longer. Well, the Mutec gave me a similar 'wow' factor when I first put in between the streamer and the Devialet. The GAIA is probably better because it has two OCXO clocks inside that provide all audio frequencies (x1024, so it can do DSD1024 theoretically, practically DSD512) and there is no complicated PLL needed as in the Mutec. MSB explains it very well:
https://www.msbtechnology.com/dacs/clock-options/
Hello
I have just received the denafrips gaia. Yes, WOW. Really a great upgrade. Also using a transparent reference aes/abu.
When I switch from pcm to dsd and vice-versa I can hear a loud pop. Any ideas how to solve this?
Using daphile with dop.
(24-Apr-2022, 04:48)zdenes Wrote: [ -> ]Well, the Denafrips GAIA is so much better than the Mutec MC3+USB with linear power supply (Uptone Audio JS-2) and even clocked with the Mutec Ref10 that I already bought one and removed the Mutec from my system. I left the Ref10 still in as it is also clocking the SOtM SMS-200 ultra and the Uptone Etherregen switch.
Thank you.
The Denafrips Gaia DDC is not necessarily better than the Mutec MC3+ with Ref10 combination for everyone. This is largely dependent on what DAC it is feeding into. Your findings do suggest that, given a high quality signal, the AES/ EBU input is better than the USB input on the Devialet.
Furthermore, the Gaia has a big advantage of being a simple one box solution while the Mutec combination with a LPS is a mass (& mess!) of boxes and cables. And the Gaia cost significantly less (but I do recognise the Ref10 can be used on other devices, and JS2 can power 2 devices).
For owners of the Expert Pro, the Gaia is almost a must have...