Jump to content
IGNORED

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


Message added by The Computer Audiophile

Important and useful information about this thread

Posting guidelines

History and index of useful posts

Most important: please realize this thread is about bleeding edge experimentation and discovery. No one has The Answer™. If you are not into tweaking, just know that you can have a musically satisfying system without doing any of the nutty things we do here.

Recommended Posts

Does any one know the ssh ID and password of SMS-200?

 

Using LAN Scan HD or similar on iPad to scan home network can reveal that SMS-200 has SSH enabled. Through ssh it's possible to assign a static IP to it to eliminate all the Ethernet bridging fuss. Especially now the SD card can be backed up, it's totally safe to mess around.

 

Any idea?

Link to comment

Hi all

Firstly a big thanks to romaz for bringing this tweak to our attention and to everyone else who has helped to build this interesting thread.

I’d had more than a few years of experience with 2 PC Jplay so static address setting is like falling off of a log for me.

But with the bridge set up it does not work so well.

Who knows why?

Seems like we Windows guys may have been working toooo hard to get our bridges up and running. I have Windows 10 pro/process Lasso/Fidelizer pro running on a spare experimental j1900 atom board and a (cheap) 2 way PCI e NIC. I set my bridge up in the usual way but rather than assigning a static IPv4 address I just let Windows handle that. With a USB Wi-Fi dongle inserted but disabled Windows set an address automatically and does most of the work so just grab a coffee and wait a little. Then it’s just a question of switching on network discovery and hey presto. Full internet and network connectivity with all of my network devices showing up including my SMS200.

 

Rock solid after a week or so.

 

Cool.

I’m pretty slow computer wise but managed the following in around 10 mins.

MinimServer running on PC and music stored on NAS. I had tried local disc storage but found that it sounded no different so the NAS stayed.

 

WI-FI disabled. (Weird how the wireless is needed)

On board NIC > router.

PCIe 1 > SMS200

PCIe 2 > Synology DS216J NAS.

 

Running headless.

Compared to connection through my Cisco switch to my ears and in my set up it sounds noticeably clearer and more dynamic.

 

Nice work.

BTW I had previously used TP210 FMC between switch and SMS but decided to take them out as in direct mode they seemed to flatten the dynamics a little to my ears. (this seems to be system dependant as others have had different results)

IMHO this is a great free tweak and trying it is a no brainer as nothing gets broken.

Sounding pretty impressive this end and a definite step forward SQ wise to my ears.

Thanks guys.

As an aside on placing a switch between server and renderer. As I understand it the switch is not just re-clocking the signal but it’s also checking the sent data. Bad packets initiate a re-send from the switch so maybe even closer to the Network ‘’Regen’’ idea.

I’m no expert but happy to learn if anyone has some thoughts on that.

I may put my ‘’ordinary switch’’ in to hear what I does to the sound if anything.

 

 

Link to comment

I suspect the benefits we hear are due to the elimination of unnecessary and redundant hardware of the switch, and it's power supply. A switch is just another computer with multiple NICs a processor, RAM and software. It makes sense that reducing the number of devices touching the music signal leads to improved SQ.

 

This does not mean there may not be specialized switch's optimized for music that will sound great.

Pareto Audio aka nuckleheadaudio

Link to comment
I suspect the benefits we hear are due to the elimination of unnecessary and redundant hardware of the switch, and it's power supply. A switch is just another computer with multiple NICs a processor, RAM and software. It makes sense that reducing the number of devices touching the music signal leads to improved SQ.

 

This does not mean there may not be specialized switch's optimized for music that will sound great.

 

Yes, I always feel a Pang of envy when I hear of @romaz 's fancy switches. :D

Link to comment
Hi,

 

I've successfully made the direct connection between my iMac and microRendu.

 

However, using Roon I can only get it to work in Roon Ready mode but not HQPlay NAA.

 

Any thoughts would be appreciated.

 

Thanks

 

 

Sent from my iPhone using Computer Audiophile

 

Before you set up the bridge, you could play through HQP to the microrendu? When you go into HQP settings after the bridge do you see the microrendu as an option?

12TB NAS >> i7-6700 Server/Control PC >> i3-5015u NAA >> Singxer SU-1 DDC (modded) >> Holo Spring L3 DAC >> Accustic Arts Power 1 int amp >> Sonus Faber Guaneri Evolution speakers + REL T/5i sub (x2)

 

Other components:

UpTone Audio LPS1.2/IsoRegen, Fiber Switch and FMC, Windows Server 2016 OS, Audiophile Optimizer 3.0, Fidelizer Pro 6, HQ Player, Roonserver, PS Audio P3 AC regenerator, HDPlex 400W ATX & 200W Linear PSU, Light Harmonic Lightspeed Split USB cable, Synergistic Research Tungsten AC power cords, Tara Labs The One speaker cables, Tara Labs The Two Extended with HFX Station IC, Oyaide R1 outlets, Stillpoints Ultra Mini footers, Hi-Fi Tuning fuses, Vicoustic/RealTraps/GIK room treatments

Link to comment
I suspect the benefits we hear are due to the elimination of unnecessary and redundant hardware of the switch, and it's power supply. A switch is just another computer with multiple NICs a processor, RAM and software. It makes sense that reducing the number of devices touching the music signal leads to improved SQ.

 

This does not mean there may not be specialized switch's optimized for music that will sound great.

 

You are correct Larry. In particular, it is the dirt-cheap Ethernet PHY chips used in most switches--which often produce marginal signal integrity (thereby causing the receiving endpoints to work harder and thus generate more bursty PS draws and ground-plane noise; same as the issue with USB and the aim of the REGEN and similar devices). The PHYs are better in some of the high-end managed switches (Cisco, etc.), but there they use their own custom, massively integrated chips which each handle a bunch of ports--so it is hard to know if that helps our cause. And of course power supplies, regulators, and clocks in the switches can all be improved.

 

By the way, the Broadcom Ethernet controller chips used in better computers (and in all the Macs) are some of the best available (John knows these from the inside). So going direct does have some basis in principle for why it is better.

 

 

You mentioned a specialized Ethernet switch optimized for audio:

During his stay with me a few weeks ago, John and I selected ideal Ethernet switch chips (they have varying interfaces) and Ethernet PHYs (once that don't have a bunch of analog blocks which are to turn things on and off to save energy), plus his favorite isolator chips to go between the switch chip and the special PHY. It would of course have a great clock, ultra-low noise regulators, and an SFP cage for optical. And for power there would be 2 DC input jacks (optional use, they would be the kind with a third contact switch), so someone could separately power the "clean" side isolated output (with an LPS-1 of course).

 

The above is a broad outline for a truly "audiophile Ethernet switch." Actual engineering of it involves a bunch of other unmentioned complexities (which John explained to me and which mostly went over my head).

But before everyone gets all excited by the prospect of such a product from UpTone, know that we discussed it in relation to another significant Ethernet-related isolation project we have slated for the fall (after his move), and decided that the market for the above-described switch product is much smaller than for the other secret product. So don't expect such from us this year! :)

 

--Alex C.

Link to comment
I suspect the benefits we hear are due to the elimination of unnecessary and redundant hardware of the switch, and it's power supply. A switch is just another computer with multiple NICs a processor, RAM and software. It makes sense that reducing the number of devices touching the music signal leads to improved SQ.

 

This does not mean there may not be specialized switch's optimized for music that will sound great.

 

So this thread has been sooo interesting (even though I have only lived vicariously so far) and makes so much sense. However, it flies in the face of the fiber-ing of my music room (adding a Diablo fiber switch and FMCs, where needed, for my ATT modem/router, Synology NAS, and microRrendu, adding a fiber NIC for my HQP Win10 server). To do what Romaz, and specifically Larry and Shoom, are talking about requires me to "rip it all out", although that process would not be that difficult as the ethernet cables are all still laying there.

 

My project to fiberize my music room seemed to really lower the noise floor of my already high end setup. But yet I acknowledge that going direct does indeed remove hardware and processes, and that maybe copper direct is sonically better than fiber-switched. I wish it were more straight-forward, as I've tried it (only going direct to uRendu) a couple times and gotten nowhere. And afterward justified my failure with "why fix what ain't broken". :)

 

Full non-switch setup success does have a little downside in my setup, in that it dedicates the NAS to my HQP machine (which currently is no big deal). I may PM Shoom to ask what he did in more detail, and will first temporarily turn off AO, as per Larry's recommendation.

Link to comment
Yes, I always feel a Pang of envy when I hear of @romaz 's fancy switches. :D

Pretty funny.

 

The cheap $20 Trend Net switch I sent in to SOtM is becoming a bit more expensive. After opening it up, it turns out this switch has a switching voltage regulator for the DDC and so I had them swap this out for me with a linear regulator (extra $100). They also offered to add a few extra capacitors which makes sense to me as I am a believer that audio electronics work best when they have access to instantaneous current on demand and should have the same impact as lowering impedance. This added another $20 to the price. I will be powering this switch with another LPS-1 which I have on order ($400). Of course, this isn't about the switch, it's about the clock. We'll see where this rabbit hole leads.

 

I'm beginning to formulate my own ideas about where to reclock and where not to reclock. As I stated before, when I placed my reclocking switch before my Mac Mini, the improvement was quite small but I surmise it's because the numerous reclocks that take place within my Thunderbolt hub and my Mac Mini must be of such inferior quality that it amounts to washing your car and then driving it through the mud. All the hard working of cleaning up the signal (reclocking) is lost. So if you're going to reclock, the component that receives that reclocked signal should have its own clock that is at least equal or higher quality. It is also my belief that all clocking points within the audio chain at best adds no jitter and at worst adds tons of jitter. If you have components with bad clocks, it's best to keep these components to a minimum. Unfortunately, from internet modem to DAC, there are many clocking points. Your modem/router has 1-2 clocks. Every LAN connection will have a clock and this includes any intermediary switch or FMC but also the LAN port(s) on your music server. Reclocking will occur at the USB PHY (if you use USB). Same thing goes for the SATA bus and probably the Thunderbolt bus. With SOtM's new superclock, it is actually possible to replace all of these clocks but at considerable expense. It would be more prudent just to avoid adding components to your chain that have bad clocks and if you are going to reclock, the most important reclocking would probably be at the very end of your chain and I believe this is why the microRendu and sMS-200 are such breakthrough devices (they have very good clocks).

Link to comment

Ted, Funny in the past I have often thought about buying a single SFP Intel fiber nic and a second dual SFP Intel fiber NIC and then using them to re-create the bridged copper network topology running here with light. It should work fine, and I expect will sound as good, if not better then a copper solution. At the time I stopped this effort when realizing the dual SFP nic needed a faster PCIE x4 motherboard slot and I only had a x1 slot available in my NAS machine.

 

However in the end I am hoping to eliminate carrying any music connect over Ethernet entirely by moving the music library to my HQplayer upsampling machine. A great sounding single machine with no network, no NAA and no NAS seems best to me, and worth pursuing. The part needed for this test arrived last night, so I'll give this a go in the next few days.

Pareto Audio aka nuckleheadaudio

Link to comment

A few years ago over on the Jplay forum (nice to see some familiar names here) there was a thread running on pretty much the same switch between control and music PC subject.

 

So my ears pricked up when I read romaz posting on the same subject.

 

A few of us noticed a slight improvement in clarity and definition with the switch in place (or we were all suffering from a bad case of expectation biasJ). Some thought that it sounded worse which was fine also.

 

Not huge difference but slight and we only used generic switches (with decent but not extravagant LPS) and not the fancy Dan PP affaires.

 

So the next questions J for those of us who are terminally short of cash and knowledge.

 

Is the clocking arrangement in my Cisco (linear PSU) likely to be better/worse/just the same than my cheapo Amazon NIC?

Does the packet error correction matter more with nothing special NIC’s?

 

A small dedicated 1 in 1 out re-clocker/buffer and error correction device would be soooo nice.

A DIY (sorry it slipped out) solution and a nice super cap supply to power it with even nicer.

 

I’m all ears.

 

Anyhow just a few more things to learn.

 

Link to comment
You are correct Larry. In particular, it is the dirt-cheap Ethernet PHY chips used in most switches--which often produce marginal signal integrity (thereby causing the receiving endpoints to work harder and thus generate more bursty PS draws and ground-plane noise; same as the issue with USB and the aim of the REGEN and similar devices). The PHYs are better in some of the high-end managed switches (Cisco, etc.), but there they use their own custom, massively integrated chips which each handle a bunch of ports--so it is hard to know if that helps our cause. And of course power supplies, regulators, and clocks in the switches can all be improved.

 

By the way, the Broadcom Ethernet controller chips used in better computers (and in all the Macs) are some of the best available (John knows these from the inside). So going direct does have some basis in principle for why it is better.

 

 

You mentioned a specialized Ethernet switch optimized for audio:

During his stay with me a few weeks ago, John and I selected ideal Ethernet switch chips (they have varying interfaces) and Ethernet PHYs (once that don't have a bunch of analog blocks which are to turn things on and off to save energy), plus his favorite isolator chips to go between the switch chip and the special PHY. It would of course have a great clock, ultra-low noise regulators, and an SFP cage for optical. And for power there would be 2 DC input jacks (optional use, they would be the kind with a third contact switch), so someone could separately power the "clean" side isolated output (with an LPS-1 of course).

 

The above is a broad outline for a truly "audiophile Ethernet switch." Actual engineering of it involves a bunch of other unmentioned complexities (which John explained to me and which mostly went over my head).

But before everyone gets all excited by the prospect of such a product from UpTone, know that we discussed it in relation to another significant Ethernet-related isolation project we have slated for the fall (after his move), and decided that the market for the above-described switch product is much smaller than for the other secret product. So don't expect such from us this year! :)

 

--Alex C.

Sounds like you have the recipe for the perfect switch, Alex. I wish I knew some of this before I sent out this cheap Trend Net switch for clock modification. It indeed has this "GREENnet" power saving feature and I was suspicious of it but went with it anyway since I already owned this switch and had it lying around. I'm always looking forward to anything you and John come up with. I hope your audiophile switch becomes a reality sooner than later.

 

Good to know about the Broadcoms. Both my LAN ports (Mac Mini + Thunderbolt LAN) are Broadcoms. Definitely no complaints.

Link to comment
I suspect the benefits we hear are due to the elimination of unnecessary and redundant hardware of the switch, and it's power supply. A switch is just another computer with multiple NICs a processor, RAM and software. It makes sense that reducing the number of devices touching the music signal leads to improved SQ.

 

This does not mean there may not be specialized switch's optimized for music that will sound great.

 

Could it be that eliminating switches or other hardware restores absolute polarity of the audio signal? On high quality systems that could make a significant change in SQ for the better. I for one would rather toss my digital frippery aside than listen to reversed polarity.

 

 

Sent from my iPad using Computer Audiophile

Link to comment
Could it be that eliminating switches or other hardware restores absolute polarity of the audio signal? On high quality systems that could make a significant change in SQ for the better. I for one would rather toss my digital frippery aside than listen to reversed polarity.

 

 

Sent from my iPad using Computer Audiophile

Hmmm, I'd say the SQ impact I hear here is more than a polarity switch.

Pareto Audio aka nuckleheadaudio

Link to comment
Could it be that eliminating switches or other hardware restores absolute polarity of the audio signal?

 

No, that's impossible. All networking equipment transfers the signal unchanged. Nothing would work if it didn't. The worst a switch can do is inject analogue noise riding on the digital signal. Flipping the polarity of an audio stream is as impossible as the switch translating English text to Russian as it passes through.

Link to comment

So, once I got the go ahead from AudioPhil (beta 2.2b4 is fine with bridging) I tried bridging again. I connected my mobo copper ethernet to either the router directly or the rj45 port in the Diablo (didn't matter) and connected the Mellanox fiber NIC directly to the miroRendu's FMC. Did the bridge thing. Sonicorbiter says the microRendu is 168.192.1.193 so I ping'd it from the HQP machine and it sees it. So does the browser, of course (sees my sonicorbiter/Rendu pages). But HQPlayer does not. Under no circumstances (rebooting every known device) could I get HQPlayer to see the NAA (dac connected to the microRendu). But once I swapped everything back to "normal' I was good to go. Ideas?

Link to comment
So, once I got the go ahead from AudioPhil (beta 2.2b4 is fine with bridging) I tried bridging again. I connected my mobo copper ethernet to either the router directly or the rj45 port in the Diablo (didn't matter) and connected the Mellanox fiber NIC directly to the miroRendu's FMC. Did the bridge thing. Sonicorbiter says the microRendu is 168.192.1.193 so I ping'd it from the HQP machine and it sees it. So does the browser, of course (sees my sonicorbiter/Rendu pages). But HQPlayer does not. Under no circumstances (rebooting every known device) could I get HQPlayer to see the NAA (dac connected to the microRendu). But once I swapped everything back to "normal' I was good to go. Ideas?

 

It sounds nutty, but I think (as a few people have figured out), adding a wifi adapter to your network bridge could likely make this work for you.

 

I know of one CA member who was stuck at the exact same place that you are. Adding a tiny, cheap wifi dongle (this one: wifi dongle) made everything work (meaning, HQP was finally able to see his microRendu in NAA mode with a network bridge and direct connection). Even if this doesn't sound so appealing, it's probably worth the $8 expense just to try it.

 

I don't know this for sure, but I think (possibly) that some of the network bridges people are creating are, for some reason, not fully able to handle all the multicast routing that is required by HQPlayer. Whether this is due to their network adapters, their OS, software they have installed, I have no idea...

 

By the way, when I experimented a little (and tried a direct connection without creating a network bridge by 1) putting NIC 2 on a different subnet than my regular network and 2) running Open DHCP Server software to have it provide an IP address to the microRendu that was connected to NIC 2), I got stuck at the same place as you (I could ping the microRendu, could open the microRendu's web interface, could even play music to the microRendu in RoonReady mode, but could not get HQPlayer to see the microRendu in NAA mode).

Link to comment
It sounds nutty, but I think (as a few people have figured out), adding a wifi adapter to your network bridge could likely make this work for you.

 

I know of one CA member who was stuck at the exact same place that you are. Adding a tiny, cheap wifi dongle (this one: wifi dongle) made everything work (meaning, HQP was finally able to see his microRendu in NAA mode with a network bridge and direct connection). Even if this doesn't sound so appealing, it's probably worth the $8 expense just to try it.

 

I don't know this for sure, but I think (possibly) that some of the network bridges people are creating are, for some reason, not fully able to handle all the multicast routing that is required by HQPlayer. Whether this is due to their network adapters, their OS, software they have installed, I have no idea...

 

By the way, when I experimented a little (and tried a direct connection without creating a network bridge by 1) putting NIC 2 on a different subnet than my regular network and 2) running Open DHCP Server software to have it provide an IP address to the microRendu that was connected to NIC 2), I got stuck at the same place as you (I could ping the microRendu, could open the microRendu's web interface, could even play music to the microRendu in RoonReady mode, but could not get HQPlayer to see the microRendu in NAA mode).

 

Yup - that includes me.

 

Lots of issues with 2 x Ethernet bridge. Added a wifi adapter - technically, recreated by selecting 3 adapters (2 Ethernet, 1 wifi) and creating a bridge.

 

Flawless from there.

Link to comment
Also note that users of Paul's clocked switches report a lengthy burn-in time for the clock and so you might want to give it time. Since I bought mine used, this didn't apply to me.

 

As I had mentioned a few nights ago, my first round of comparisons with and without my Paul Pang switch between my Mac Pro and microRendu had me shrugging my shoulders. I thought I might have heard some differences but I could not pick a winner.

 

I left the switch in place since reading the above, as my switch, even though purchased second hand, was said to have never been used.

 

Tonight I thought I'd try another round of comparison. A different reaction this time: my jaw dropped when I heard what resulted from removing the switch. It could best be described as a loss of focus, along with a dulling of transient attacks. Wild.

 

So thanks again to Romaz for another great tip.

 

 

Sent from my iPad using Computer Audiophile

Digital:  Sonore opticalModule > Uptone EtherRegen > Shunyata Sigma Ethernet > Antipodes K30 > Shunyata Omega USB > Gustard X26pro DAC < Mutec REF10 SE120

Amp & Speakers:  Spectral DMA-150mk2 > Aerial 10T

Foundation: Stillpoints Ultra, Shunyata Denali v1 and Typhon x1 power conditioners, Shunyata Delta v2 and QSA Lanedri Gamma Revelation and Infinity power cords, QSA Lanedri Gamma Revelation XLR interconnect, Shunyata Sigma Ethernet, MIT Matrix HD 60 speaker cables, GIK bass traps, ASC Isothermal tube traps, Stillpoints Aperture panels, Quadraspire SVT rack, PGGB 256

Link to comment
Yup - that includes me.

 

Lots of issues with 2 x Ethernet bridge. Added a wifi adapter - technically, recreated by selecting 3 adapters (2 Ethernet, 1 wifi) and creating a bridge.

 

Flawless from there.

In adding the wifi adapter are you enabling it entirely such that it is an active network device with a separate path to a wireless router?

Pareto Audio aka nuckleheadaudio

Link to comment

The really odd thing about the WI-FI dongle in my set up is that I have it disabled so not doing anything much except showing up as a network device.

I was unable to create a working bridge before inserting the USB WI-FI dongle but with it installed (disabled) and bridged, Windows 10 pro set the bridge up automatically.

Weird but works.

 

Link to comment
The really odd thing about the WI-FI dongle in my set up is that I have it disabled so not doing anything much except showing up as a network device.

I was unable to create a working bridge before inserting the USB WI-FI dongle but with it installed (disabled) and bridged, Windows 10 pro set the bridge up automatically.

Weird but works.

 

Do you have any connection to your router / internet?

Link to comment

Yes

 

 

 

Just to re cap

 

I have the following.

 

On board LAN connection to my Modem/Router.

 

A 2 port PCI e NIC connected to SMS200 and NAS.

 

WI-FI dongle disabled but included in the bridge.

 

So full network and internet access.

 

Easy to set up.

 

Just highlight the network devices to include in the bridge including the disabled WI-FI dongle, select create bridge.

 

Let windows do its set up thang.

 

Done.

 

 

 

 

Link to comment
Yes ,Just to re capI have the following.On board LAN connection to my Modem/Router.A 2 port PCI e NIC connected to SMS200 and NAS.WI-FI dongle disabled but included in the bridge.So full network and internet access.Easy to set up.Just highlight the network devices to include in the bridge including the disabled WI-FI dongle, select create bridge.Let windows do its set up thang.Done.

 

Does your setup really works with the Wifi dongle disabled? In my setup the Wifi-USB dongle needed to be enabled but not connected to the Wifi-modem. As soon as I disabled the Wifi dongle, stuttering started again. Although it is disconnected it still keeps on flashing it's led.

Check my profile for my audiosystem.

Link to comment

Anyway : I no longer need the Wifi-dongle : I have installed W2016-core (with AO's optimization) and in core-mode a network bridge is not possible. JMRC DLNA server does work and the SOTM200 is still connected only to NIC2 and NIC1 is connected to NAS/Router. I set the SOTM on a separate network branch as described before. And yes : this sounds again better than the same under W2012R2-core.

Check my profile for my audiosystem.

Link to comment
Anyway : I no longer need the Wifi-dongle : I have installed W2016-core (with AO's optimization) and in core-mode a network bridge is not possible. JMRC DLNA server does work and the SOTM200 is still connected only to NIC2 and NIC1 is connected to NAS/Router. I set the SOTM on a separate network branch as described before. And yes : this sounds again better than the same under W2012R2-core.

Hi Peter,

 

So in the end you are using the DHCP server referenced in the prior post right? If so, have you set it to use the same subnet as your router, or a different subnet altogether? If the latter, the additional SQ improvement may be explained by the the fact that multicast traffic on your router based network will not be seen by the sotm200 anymore, lowering the level of context switching on the renderer.

Pareto Audio aka nuckleheadaudio

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...