A novel way to massively improve the SQ of computer audio streaming

[sig8]

26 minutes ago, Exocer said:

Thanks for sharing your expertience. Very curious about going this route myself... Hopefully you don't mind the barrage of questions 😃.

 

How would you compare the sound of the 10th gen i9 to the 9th gen i9?

 

Do you have any experience with 8th gen CPUs to compare?

 

Some have reported that the K version CPUs can sound better due to better binned parts and/or enhancements with cooling capabilities due to a difference in the manufacturing process between the k and non k versions. It would be great if you could compare non-k and k on the same motherboard...

 

End rant 👍

 

Thanks

Comparison between 9th and 10th gen is pending, I have access to i9-9900k/Gigabyte Z390, and i9-10900k/Gigabyte Z490. This comparison will happen in next few weeks, it is first on my list of things to do. That will be a very close comparison, Z390/i9-9900k vs Z490/i9-10900k, both Gigabyte AORUS Master boards, same RAM, PSU, etc.

 

I did have 8th generation CPU's also, don't remember exactly which ones, but at that time I think lot of people including me thought 9th gen sounded better, and unlocked K's sounded better than locked CPU's. Before that conventional thinking was that low powered CPUs sounded better. 

[emailtim]

2 hours ago, sig8 said:

I will give a brief update for now, because some more A/B is required. After i9-10900/ASUS, I got to try a i9-10900k with a Gigabyte Z490 AORUS Master. This certainly sounds better than i9-10900/ASUS, and I think it is the Gigabyte motherboard, not the differences in the processors. 

 

All in all, in the past year I have tried two different ASUS boards with i9-9900K, one ASUS with i9-10900, one Gigabyte AORUS Master Z390 with i9-9900K, and the current i9-10900k with Gigabyte Z490. Both Gigabyte boards sound better than any ASUS combo. I have to compare current i9-10900k/Gigabyte with my i9-9900k/Gigabyte. That will be in next 2-3 weeks. Current system is still in burn-in process, it needs to settle down in 2-3 weeks.

 

For now, I have concluded that Gigabyte boards sound better in my system. YMMV.

 

FWIW, this review of the Z490 mini-ITX boards indicate the MSI and Gigabyte boards have the strongest VRM implementations for OCing, but indicates it is not necessary for normal stock speeds.  I wonder if the stronger VRM power designs equates to better sound when underutilized. 

 

 

[kelvinwsy]

I hv the 10700k on the Gigabyte Aorus Ultra - 1 step down from the Master. The 14 phase VRM capable of 60amp peak current vs the 90amp for the Master and Ultimate ( the relative peak amperage is the important stat here not tge actual number)

SQ is rich, deep and effortless!

Kelvin

[flkin]

On 8/5/2020 at 3:31 AM, Exocer said:

 

Would the buffer queuing available in Stylus render all network tweaks moot if one only streams from services like Tidal/Quboz?

 

On 8/5/2020 at 4:20 AM, austinpop said:

 

Sadly, no. I wish that were the case.

 

As many have observed, and certainly in my own experience, there appears to be some SQ cost to merely the presence of a network connection, even when playing local content. So the upstream quality and mods still matter. However, that said, the ability to buffer network streams does raise their SQ.

 

I'm a little more fortunate, network tweaks in my system doesn't seem to make a difference.

 

Both Tidal and Quobuz buffers 2 songs entirely in Stylus when playing (without HQP that is) and so you can see if the physical LAN connection is passing noise into the streamer simply by disconnecting the cable whilst playing - the track will continue to play for the current and next song both streaming or local. I am unable to hear any difference with and without the LAN cable connected.

 

I think that this is a mandatory test before entering the deep network enhancement rabbit hole - at the very least, if you hear a difference, you know what sound correction to aim for.

[Nenon]

51 minutes ago, flkin said:

 

 

I'm a little more fortunate, network tweaks in my system doesn't seem to make a difference.

 

Both Tidal and Quobuz buffers 2 songs entirely in Stylus when playing (without HQP that is) and so you can see if the physical LAN connection is passing noise into the streamer simply by disconnecting the cable whilst playing - the track will continue to play for the current and next song both streaming or local. I am unable to hear any difference with and without the LAN cable connected.

 

I think that this is a mandatory test before entering the deep network enhancement rabbit hole - at the very least, if you hear a difference, you know what sound correction to aim for.

 

Unfortunately it's not that simple. There is more than the noise coming from the network cable to be considered, especially when you stream Tidal / Qobuz. I have been comparing two identical Buffalo switches in the past couple of days - one with the stock clock and one with upgraded clock (PinkFaun ultraOCXO).

You can do that same test in my system, and although you disconnect the cable while playing, the buffered tracks using the switch with the upgraded clock sound much better. In other words, even if you disconnect the network cable, whatever good or bad things your network does is already embedded in the buffered track somehow. Don't ask me how or why... I wish I could explain all that, but I can't. 

Sorry to tell you that, but I don't agree with your simple theory. I wish it was that simple, but it's a lot more complex than that.

BTW, replacing the clock on this switch made almost as much difference with streaming services as replacing a clock on a USB card... go figure

[jacky5555]

On 7/25/2020 at 11:38 AM, ray-dude said:

with a vlan, I’d have to tag traffic and the vlans would be fully isolated without routing/forwarding rules. Since I want my audio server to be able to talk to Roon remote on my laptop and phone (and my file server), that would have been a pain. I had it set up this way initially, but the way I had firewall rules basically required session initiation (I used ping) before the connection would be open. Probably better ways to do it, but not very civilian or audio guest friendly.

 

I'm interested in an Edgerouter X for making a vlan for audio server. My question is: if we isolate the audio server from others, how the Roon remote (on my iphone) connected via wifi access point can talk to the server? Can you pls share you solution!

[lyndonlim]

On 5/30/2018 at 12:53 PM, lmitche said:

Nope, it's this one:

 

http://gothamcables.com/en/gothamcables/starquad/11301gac41ultrapro

 

 

 

Two layers of shielding separated by a thin insulation layer.

 Hi guys,

 

I just received my Gotham cable so I was wondering if you could clarify some detail around achieving the JSSG360 effect. In my GAC-4/1 cable I can see Shield 1, 2 but no shield 3. I can see shield 5 and a bunch of copper hair strands after shield 5 is peeled away. 

 

Can I confirm that shield 1, 2 and 5 are to be bunched up together (terminated) ?  What do I do with the copper hair strands inside shield 5 ?  Just to clarify that I am not referring to the OFC Central conductor when I refer to the copper hair strands. 

 

Thanks for you input.

[lyndonlim]

Hi guys, 

 

The attached picture showing the inner copper strands is what i was referring to in my previous email. 

 

Thanks 

 

 

[zerung]

6 hours ago, Nenon said:

BTW, replacing the clock on this switch made almost as much difference with streaming services as replacing a clock on a USB card... go figure

Emil in his SGM Extreme did not use clock and opined that it was not important for his design. Of course he was right and I fully agree you are right too. That is what is wonderful! CAS is being attacked from all front and making heavy inroads.....

 

IMHO the Denafrips Gaia and Audio GD DI20HE are the hen's teeth for CAS....

 

 

[StreamFidelity]

16 minutes ago, zerung said:

Emil in his SGM Extreme did not use clock and opined that it was not important for his design.

 

It's not that simple with Taiko Audio. As far as I know, the ASUS WS C621E SAGE Board is installed in the Extreme. And if it is true what @OAudio writes, then the clock problem is solved excellent.

 

On 5/29/2020 at 7:48 PM, OAudio said:

1. The clock subsystem. Wow this is where Intel have really helped out in the design. 

• A single oscillator on the PCH generates the CPU, RAM, DMI, PCIe, SATA, USB and Network (for on-chip N/W) clock domains signals,
• All these domains and subsystems now locked to each via the single clock - the PPLs generating the clocks for the domains all reference the same master,
• The master clock is even an integer multiple of the USB clock frequency ! so exceptional for good timing of on board USB interfaces for music if you want,
• Finally the power system for the master clock is very different to consumer motherboards, the design effectively carves out almost all of the clock subsystem, (its PLLs and transmission buffers) away from most of the motherboards power rails and other systems.

[ray-dude]

6 hours ago, jacky5555 said:

I'm interested in an Edgerouter X for making a vlan for audio server. My question is: if we isolate the audio server from others, how the Roon remote (on my iphone) connected via wifi access point can talk to the server? Can you pls share you solution!

 

I think the term VLAN can be confusing...it is a specific way to configure a network, while most folks are speaking to the general concept of isolating (fully or partially) a network, not setting up an actual VLAN.

 

With the goal of minimizing network traffic seen by my music server (mainly background broadcast traffic), I looked at the following approaches:

 

1 (LAN) - Isolate my audio server on a dedicated audio network by configuring a dedicated network for a specific ethernet port on my ERX

2 (VLAN) - Isolate my audio server on a dedicated audio network by configuring a VLAN

3 (Subnet) - Isolate my audio server on a dedicated audio subnetwork

 

I started with 1.  In this situation, there was a firewall between my home network and my audio network.  To allow my laptop (running Roon remote) to connect to my audio server, I set up a firewall rule that said that if I initiate a network session from my laptop to the music server, the firewall will allow traffic between the computers.  Worked well, but I had to ping the music server to initiate a session before being able to use Roon remote.  OK for me, but definitely not civilian friendly.   I did however hear a marked improvement in SQ with the music server isolated from all the background network traffic, so I was encouraged to keep moving forward.

 

With 2, in a VLAN you configure the network interface on devices to assign them a VLAN ID (1, 2, 3, etc).  The router can then use the VLAN ID to configure virtual networks where only devices with the same VLAN ID can see each other.  Same idea as 1, but you don't need a shared physical ethernet connection to manage isolation, so easier to move wired devices between VLANs without having to rewire the network.  From a routing perspective, you need to play the same firewall rule tricks to allow traffic from one VLAN to the other VLAN.  No net benefit to me over 1

 

I ended up with 3.  I configured my ERX to have the ethernet port for my audio network on its own subnetwork (I happen to use 192.168.9.x), with a DHCP server for that subnet.  In this configuration, routed network traffic can freely go between the subnet with my laptop/phone etc and the subnet with my music server (no firewall hacks needed).  However, broadcast network traffic is not routed between subnets.  Works like a charm.

 

I should note that in all these scenarios, Roon relies on broadcast packets to discover roon devices/remotes/endpoints on the network.  When you fire up Roon on your laptop, you will not see the Roon server on its own LAN/VLAN/subnet.  On the Choose Your Roon Core it will sit there looking and nothing will show up.  The trick is to hit the help link, which will allow you to enter the IP address of your Roon Core (on a different network).  Once you do that, Roon will use routed traffic to make the connection with the Roon Core, and it will show up and work as normal.

 

Net net, at this point if I disconnect the network connection to my music server, I honestly can not tell the difference (maybe the barest hint of a difference, but so vanishingly small that I have zero confidence in calling it a difference).  

[OAudio]

Hi,

 

My first post in this epic thread so hello to all ! 

 

1 hour ago, StreamFidelity said:

And if it is true what @OAudio writes, then the clock problem is solved excellent.

 

 

@StreamFidelity hi, perhaps I would not go as far as saying this is "solved". It is an excellent starting point, the C620 architecture is the best I have come across so far and in large part I think because of its clock subsystem design. This comment applies to Intel systems, I haven't used or looked deeply at AMD. The C620 is still a starting point for me however. A few posts further down in the "Building a DIY Music Server with custom made parts" thread I also say. 

  

"I used to use expensive commercially available clocks but found that they do not meet the specifications I want so I build my own these days. There is simply a huge amount or work needed to work out how to make a clock really work well within a specific audio server environment."

 

The latter comment was in part aimed at my C621 server design. C620 servers can give absolutely great sound out of the box, but things can still be done to improve the system's timing, both with and without use of a clock. I would not be surprised to find Taiko has done the first steps I have in mind but its possible to go past this but IMO getting to the very highest performance is not a bolt in operation. The issue I found was that C621 boards can be ultra transparent and applying the wrong clock design or put the right clock in but applying it in the wrong way can give beautiful music but unfortunately with standout problems from the clock implementation. The musical equivalent of owning the Mona Lisa painting but with a boil painted on her nose . It took more than 3 years to develop the clock I use now in my C621 server, but the musical results are nice.

 

So perhaps timing is not completely solved but there is now an excellent starting point with the C620 boards. 

 

O Audio.

 

 

 

 

 

[flkin]

8 hours ago, Nenon said:

 

.. whatever good or bad things your network does is already embedded in the buffered track somehow. Don't ask me how or why... I wish I could explain all that, but I can't. 

Sorry to tell you that, but I don't agree with your simple theory. I wish it was that simple, but it's a lot more complex than that...

 

Actually this is not my theory, the concept is in John Swenson’s White Paper on how the EtherRegen works. A complete buffering of the music track should solve the accumulated noise problem but doesn’t solve leakage current issues from a connected lan cable. 
 

If your switch is somehow correcting the ‘dirty bits’ that would be surprising! But stranger things have happened in audio 😄

[Exocer]

1 hour ago, OAudio said:

Hi,

 

My first post in this epic thread so hello to all ! 

 

 

 

@StreamFidelity hi, perhaps I would not go as far as saying this is "solved". It is an excellent starting point, the C620 architecture is the best I have come across so far and in large part I think because of its clock subsystem design. This comment applies to Intel systems, I haven't used or looked deeply at AMD. The C620 is still a starting point for me however. A few posts further down in the "Building a DIY Music Server with custom made parts" thread I also say. 

  

"I used to use expensive commercially available clocks but found that they do not meet the specifications I want so I build my own these days. There is simply a huge amount or work needed to work out how to make a clock really work well within a specific audio server environment."

 

The latter comment was in part aimed at my C621 server design. C620 servers can give absolutely great sound out of the box, but things can still be done to improve the system's timing, both with and without use of a clock. I would not be surprised to find Taiko has done the first steps I have in mind but its possible to go past this but IMO getting to the very highest performance is not a bolt in operation. The issue I found was that C621 boards can be ultra transparent and applying the wrong clock design or put the right clock in but applying it in the wrong way can give beautiful music but unfortunately with standout problems from the clock implementation. The musical equivalent of owning the Mona Lisa painting but with a boil painted on her nose . It took more than 3 years to develop the clock I use now in my C621 server, but the musical results are nice.

 

So perhaps timing is not completely solved but there is now an excellent starting point with the C620 boards. 

 

O Audio.

 

 

 

 

 

 

Thank you.

 

Edit: I see in the datasheet for the C620 Chipset.

 

PCH PCIe, CPU Bclk, CPU PCIe, DBxxxzl all share the same source clock

 

Interesting.

 

 

[OAudio]

17 minutes ago, Exocer said:

 

Thank you.

 

Edit: I see in the datasheet for the C620 Chipset.

 

PCH PCIe, CPU Bclk, CPU PCIe, DBxxxzl all share the same source clock

 

Interesting.

 

 

Indeed,

 

Its 3000 odd pages long but there are a few gems in there. It was bedtime reading for me for some time when I was working out what platform to move to after developing the X99 server.

 

Sign of an obsessed audiophiles I think 😀

 

 

[jean-michel6]

13 hours ago, Nenon said:

 

Unfortunately it's not that simple. There is more than the noise coming from the network cable to be considered, especially when you stream Tidal / Qobuz. I have been comparing two identical Buffalo switches in the past couple of days - one with the stock clock and one with upgraded clock (PinkFaun ultraOCXO).

You can do that same test in my system, and although you disconnect the cable while playing, the buffered tracks using the switch with the upgraded clock sound much better. In other words, even if you disconnect the network cable, whatever good or bad things your network does is already embedded in the buffered track somehow. Don't ask me how or why... I wish I could explain all that, but I can't. 

Sorry to tell you that, but I don't agree with your simple theory. I wish it was that simple, but it's a lot more complex than that.

BTW, replacing the clock on this switch made almost as much difference with streaming services as replacing a clock on a USB card... go figure

I made similar experience in a different context . I used to have a dual pc set up and the music files were stored on an ssd on the server pc , the music files are buffered in ram before being sent through the network to the pc player . As a trial I did stored also mt music files on a small hdd connected through usb to the pc server .

I did buffer in ram the exactly same music file either coming from the ssd or the hdd .

Every time the file coming from the hdd was sounding better . ...

It is obvious that I had in ram exactly the same info coming either from the hdd or ssd , it means that something (some noise ... ) is embedded in ram withe the digital info .

Mysterious but real ....

[sandyk]

27 minutes ago, jean-michel6 said:

Every time the file coming from the hdd was sounding better .

 

 

SSDs need to have their cases securely connected to the 0 volt ("Earth") from the PSU, with their power leads screened too if possible as they emit far more RF/EMI back into the PSU, as well as radiate HF rubbish even into the Wi Fi region. (see attachment)

 They also sound better if their supply is pre -regulated down to a very low noise +5V from the the main supply rail to help further isolate them from the rest of the PSU area. Their PSU needs to have as low and flat an output impedance from almost DC to 1MHZ,.otherwise HF noise and detail will be accentuated.(Low ESR Electrolytic capacitors, by design, have a much lower Impedance at 100kHz than the VF range )

14A2-B4.pdf

[Superdad]

FYI for those owning M12 Switch Gold and JS-2 (or thinking about a JS-2):

These are made for us by a good friend of @Nenon. UpTone makes zero $ on the sale. Just something I wanted to offer to JS-2 owners. 

[HeeBroG]

12 hours ago, ray-dude said:

I ended up with 3.  I configured my ERX to have the ethernet port for my audio network on its own subnetwork (I happen to use 192.168.9.x), with a DHCP server for that subnet.  In this configuration, routed network traffic can freely go between the subnet with my laptop/phone etc and the subnet with my music server (no firewall hacks needed).  However, broadcast network traffic is not routed between subnets.  Works like a charm.

 

I should note that in all these scenarios, Roon relies on broadcast packets to discover roon devices/remotes/endpoints on the network.  When you fire up Roon on your laptop, you will not see the Roon server on its own LAN/VLAN/subnet.  On the Choose Your Roon Core it will sit there looking and nothing will show up.  The trick is to hit the help link, which will allow you to enter the IP address of your Roon Core (on a different network).  Once you do that, Roon will use routed traffic to make the connection with the Roon Core, and it will show up and work as normal.

 

Net net, at this point if I disconnect the network connection to my music server, I honestly can not tell the difference (maybe the barest hint of a difference, but so vanishingly small that I have zero confidence in calling it a difference).  

 

Hi Ray,

 

Did you use the setup "wizard" WAN+2LAN or WAN+2LAN2? 

 

Geoff

 

 

[jean-michel6]

6 hours ago, sandyk said:

 

 

SSDs need to have their cases securely connected to the 0 volt ("Earth") from the PSU, with their power leads screened too if possible as they emit far more RF/EMI back into the PSU, as well as radiate HF rubbish even into the Wi Fi region. (see attachment)

 They also sound better if their supply is pre -regulated down to a very low noise +5V from the the main supply rail to help further isolate them from the rest of the PSU area. Their PSU needs to have as low and flat an output impedance from almost DC to 1MHZ,.otherwise HF noise and detail will be accentuated.(Low ESR Electrolytic capacitors, by design, have a much lower Impedance at 100kHz than the VF range )

14A2-B4.pdf 384.34 kB · 3 downloads

I really did take good care of my ssd : 

 - they were the best Samsung pro 

 - power was with a dedicated 5v psu mpaudio lt 3045 with good shielded diy psu cable 

- sata cable was the excellent Pachanko pure reference which is heavily shielded 

- the case of the ssd was attached to a heavy aluminium bloc as I did found it sounded better like that , however the case was not grounded as you recommend .

 

Even with all that hdd was better and no specific care for the hdd ( usb powered and ordinary usb cable ) 

 

Thank’s for the interesting attached document .

Altogether sata on mobo is also an EMI nest and noisy. I have totally disable sata on my demat pc’s with great sq results .

[flkin]

7 hours ago, jean-michel6 said:

I used to have a dual pc set up and the music files were stored on an ssd on the server pc , the music files are buffered in ram before being sent through the network to the pc player .

 

Do you have a way to buffer the data on the pc player side? Otherwise the data has to pass through the network again defeating buffering. 

[sandyk]

58 minutes ago, jean-michel6 said:

I really did take good care of my ssd : 

 - they were the best Samsung pro 

 - power was with a dedicated 5v psu mpaudio lt 3045 with good shielded diy psu cable 

- sata cable was the excellent Pachanko pure reference which is heavily shielded 

- the case of the ssd was attached to a heavy aluminium bloc as I did found it sounded better like that , however the case was not grounded as you recommend .

 

Even with all that hdd was better and no specific care for the hdd ( usb powered and ordinary usb cable ) 

 

Thank’s for the interesting attached document .

Altogether sata on mobo is also an EMI nest and noisy. I have totally disable sata on my demat pc’s with great sq results .

 

At the suggestion of one and a half (Gary) recently, I made sure that the cases of my internal Samsung 860 Evo ( OS SSD) and 850 Music SSD had a VERY low measured voltage drop between their cases and 0 volts at the SMPS PSU itself when switched on.

As Gary found, doing this did result in an appreciable improvement in  apparent noise and Soundstage. even though both were powered by in my case , dual <4uV noise +5V supplies regulated down from the internal +12V supply

 BTW, the typical LT3045 implementation does have a decreasing output impedance at >100kHz and may result in a little harshness /added HF detail due to this. Ideally it would also have a parallel considerably higher value NON Low ESR capacitor at it's output for this reason.

[lyndonlim]

19 hours ago, lyndonlim said:

Hi guys, 

 

The attached picture showing the inner copper strands is what i was referring to in my previous email. 

 

Thanks 

 

 

Please ignore my question. I went back to the diagrams and now understand that all the copper wire shields should be joined at the ends.

[elan120]

20 hours ago, lyndonlim said:

 Hi guys,

 

I just received my Gotham cable so I was wondering if you could clarify some detail around achieving the JSSG360 effect. In my GAC-4/1 cable I can see Shield 1, 2 but no shield 3. I can see shield 5 and a bunch of copper hair strands after shield 5 is peeled away. 

 

Can I confirm that shield 1, 2 and 5 are to be bunched up together (terminated) ?  What do I do with the copper hair strands inside shield 5 ?  Just to clarify that I am not referring to the OFC Central conductor when I refer to the copper hair strands. 

 

Thanks for you input.

I post this some time ago that could be an interest to you building the JSSG360 DC cables.

[lyndonlim]

2 minutes ago, elan120 said:

I post this some time ago that could be an interest to you building the JSSG360 DC cables.

Thanks Elan. That's a great illustration. It's clear now but when I saw the sheath foil material it confused me a little. I just cut away the sheath material and joined all the copper wire shielding. 

 

Thanks again for responding to my question Elan.