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

Hi guys

 

sorry but is there a sticky post which specifies what parts to buy to make an AL end point, which nuc components are recommended, as that would be handy.

 

be handy to see 3 types, end point only, server only, combined server and end point.

 

thanks in advance for endpoint nuc component recommendations, in terms of budget, I’m happy to spend £700.

 

also, if running roon server in ram, you can’t make changes to the library? Like adding playlists? How would they be persisted ?

Link to comment
23 minutes ago, Aberrant-Decoder said:

Hi guys

 

sorry but is there a sticky post which specifies what parts to buy to make an AL end point, which nuc components are recommended, as that would be handy.

 

be handy to see 3 types, end point only, server only, combined server and end point.

 

thanks in advance for endpoint nuc component recommendations, in terms of budget, I’m happy to spend £700.

 

 

 

Hi @Aberrant-Decoder and welcome to AS!

 

Dunno about a sticky but that would be a good idea.

 

For endpoint basics the following are popular:

 

Boards

NUC7CJPH or NUC7PJYH - with Akasa Newton fanless case

 

or

 

NUC7i7DNBE/KE/HE with Akasa Plato X7 case fanless case

 

The first two are celeron based and cheaper. The i7DN models are i7 based, more expensive but potentially better.

 

4 or 8GB of RAM will suffice for an endpoint. 

 

You don't need anything else other than a decent PSU for the endpoint.

 

Probably worth checking out the following thread:

 

 

Any of the above models would also work for a server but you'll want to consider 16 or 32GB RAM.

 

Also if you want to do heavy processing like upsampling you'd be better going with the i7 models.

 

An SSD is optional if you want to store the music in the server (as opposed to on a NAS or suchlike).

 

Finally, re this question:

 

33 minutes ago, Aberrant-Decoder said:

also, if running roon server in ram, you can’t make changes to the library? Like adding playlists? How would they be persisted ?

 

If you're using AudioLinux then you can save the system and settings to your USB memory stick that you boot off. So it's persisted in terms of how often you back it up.

 

Hope this helps.

 

Cheers,

Alan

 

Synergistic Research Powercell UEF SE > Sonore OpticalModule (LPS-1.2 & DXP-1A5DSC) > EtherRegen (SR4T & DXP-1A5DSC) > (Sablon 2020 LAN) Innuos PhoenixNet > Muon Streaming System > Grimm Audio MU1 server > (Sablon AES) Mola Mola Tambaqui DAC > PS Audio M1200 monoblocks > Salk Sound Supercharged Songtowers

Link to comment

Has anyone found a device with a better  hardware implementation than NUC which can be used with AL as Roon endpoint?  I do find it freaky/worrisome that just removing the USB  drive after boot to RAM makes such a difference. And intriguing that an RPI can come so close to  Pentium NUC as endpoint. Still hoping that Sonore makes AL an option for xRendu...

Regards,

Dave

 

Audio system

Link to comment
3 hours ago, jbparrish said:

 

Hi Em2016

 

I forgot to ask how is your DNHE performing? I mean how does it sound with HQPE? Is it better than other options in its price range and/or performance range?

 

Thanks

 

Hi jbparrish

 

Working really well, alongside Roon Server on the same machine for me.

 

 I haven’t compared with other options because I initially built it inside a fanless case to run RoonOS (ROCK). This model NUC is officially supported by Roon to run ROCK, so that’s why I selected this model.

 

But I was later surprised I could install Debian Stretch Server and run both Roon Server and HQP Embedded on the same machine and have no drop in performance. It can’t do poly-sinc-xtr-2s at DSD512 though... not enough grunt.

 

Link to comment
14 minutes ago, Johnseye said:

Has anyone compared using the USB 3 ports direct to the NUC board vs the optional USB 2 ports?  Or has anyone compared USB v2 to v3 in general?

I compared the integrated USB2 ,3.0 and 3.1 on a desktop Mobo and later 3.0 and 3.1 ports in form of PCIe cards with low noise power supply.  Even for USB2 devices the 3.0 and 3.1 do , to my ears at least, provide better sound quality in all cases, 3.1 being the best. 

 

Link to comment
20 minutes ago, Johnseye said:

Has anyone compared using the USB 3 ports direct to the NUC board vs the optional USB 2 ports?  Or has anyone compared USB v2 to v3 in general?

 

Signal integrity (seen on eye-pattern) is generally a bit better from USB 3.1 chips--even when just doing USB 2.0 480Mbps high-speed--as the internals of those more modern chips have to be very carefully designed so they can achieve USB3 SuperSpeed.  But they do vary, and the SuperSpeed power circuitry can generate a bit of excess noise and current bounce on the ground-plane.  Do NUC BIOS adjustments offer the ability to turn off USB3 SuperSpeed?  That would be a good thing.

Link to comment

Thanks Alex for your feedback @Superdad and @Johnseye for bringing that subject forward, I was about to ask that question partly.

I am going from a Celeron HQP NAA via USB to a Denafrips Terminator DAC, before I was using a Project S2 with your LPS1.2 Alex, superb performance, the actual use of your LPS1.2 made me a more objective person in these subjects.

Now I am using the JS2 to power my 2 NUCs (1 NUC running Linux own install with HQPE+Roon and the other NUC running NAA HQP)

anyways, I think that possibly I could enhance the SQ of my NAA --> DAC connection going the spaghetti way (which I have been trying to stay away from, didn't wanted to open that Pandora's box due to my obsessive behavior)

There are many different alternatives here and combinations. 

1. Replace the NAA by a different NAA which could allow for enhance USB circuitry for example a mini ITX computer board with tweakable BIOS settings to minimize EMI and maybe an addon USB card like these femto usb cards, Kitsune audio have one https://kitsunehifi.com/product/femto-jcat-usb-card/ or Sotm tx-USBexp https://www.sotm-audio.com/sotmwp/english/portfolio-item/tx-usbexp/. Also the NAA could be replaced by an Ultrarendu etc. but I like to tweak my Linux installs and have my own packages installed etc. Sometimes I use the NAA as Roon Bridge sometimes as HQP NAA etc.

2. Keep the NAA and add a USB decrappifier (ISORegen, tcUSBultra) with Lush^2 included and JSSG360 cable etc. full spaghetti

3. Keep the NAA but avoid USB like the plague and change the approach and go I2S (with a Singxer or Matrix https://www.audiophonics.fr/en/digital-interfaces/matrix-x-spdif-2-usb-interface-xmos-u208-32bit-768khz-coaxial-aesebu-i2s-hdmi-lvds-p-11975.html) units. 

4. Also I can use clocks feeding these devices in the chain etc.

5. I even consider looking into custom replacing the Crystek clocks that comes stock with the Terminator by a Neutron Star clock, some guy in Russia did it (I hope Alvin @alvin1118 not looking in this forum :) but that will most likely void the warranty on the DAC and I would be on my own which I don't want.

 

Summarizing pretty much everything discussed in the 500 something pages of this thread.

 

Of course money is an issue so my budget is very limited and whatever I choose to do I would like to have something which I don't need to discard down the road if I get something else.

 

Any comments based on experience are appreciated, I'm dealing with SQ issues now, I moved from a quiet neighborhood house fed by a hospital substation less than 300 mts away to a townhome community with dirty mains, I hear buzzing all over the place, I purchased a Furman cheap line conditioner but it has good reviews and without wanting to trash their name made things worse than no conditioning at all. My system sounds like crap now and I have to go back to what it was and establish a baseline.

 

Thank you all

 

 

 

 

Link to comment

Reading my post again I realize that of course I can try everything I have mentioned before and still get cumulative improvements (or detriments) it is hard to determine what will provide the more impact, so any path is valid. 

 

What I will be most likely doing is something new, experimenting with a new board capable of handling USB spec 3.1 (thank you Alex) with bios options tweakable for this effect. In this way at least I can contribute with something new to the forum.

 

I will not do I2S yet as that is very specific to a few DACs Since going I2S will be more specific to a few and the majority is using USB. Additionally to this I will be tieing myself to a specific vendor combo which I am trying to avoid 

 

Software wise will be most likely running @Miskas bootable NAA realtime image and also AL ramroot for comparison purposes. 

 

 

Link to comment

https://audiophilestyle.com/forums/topic/55159-best-alternatives-lpsu-to-paul-hynes-sr7/?page=2&tab=comments#comment-919513

2 hours ago, wanta911 said:

Thanks! I have a fully built one now that is powering my Chord Qutest....I'm not sure if I've heard the Qutest sound this good but I'll give it a bit more time before I make any definitive statements. For the record - I own a JS-2, an Ultracap 1.2, an SBooster and an HDPlex 200W.

 

https://www.diyaudio.com/forums/power-supplies/297610-low-noise-pre-amp-dac-power-supply-mje15034-tl072-regulator-based-studer-900-a.html#post5253455

Quote

I recently bought one of these "Studer 900" PSUs, and surprisingly, it made a very big difference in the sound quality in my system. I used it to power the Sotm sMS-200ultra endpoint renderer at 9V DC. It was surprising because I was comparing it to my Uptone Audio Ultracap LPS-1 that is "energized" by a 12V linear PSU, and the Studer 900 PSU improved the clarity and detail of the SQ of everything by a big margin, similar to how the LPS-1 improved the SQ when I moved from a regular LPS (based on the sigma 22 design) originally powering the sMS-200ultra.

 

I was stunned by this, and I can't explain it. I also changed the energizing PSU of the LPS-1 to the original Meanwell SMPS that came with it, and it made no difference. The Studer 900 was still clearly better.

 

https://www.diyaudio.com/forums/power-supplies/297610-low-noise-pre-amp-dac-power-supply-mje15034-tl072-regulator-based-studer-900-a-2.html#post5302522

Quote

Reading further on the Web, I believe phantom stabilizers are traditionally used to power electret condenser microphones, and since mics are very sensitive to electrical noise , the PSUs powering them have to be very low noise PSUs.

 

Another thing I am guessing that could be responsible for the high SQ is a very low output impedance.

 

Dude, could that be a winner or what?

 

https://www.aliexpress.com/item/STUDER900-Regulator-linear-power-supply-DC12V-2-5A-30W-DAC-Audio-Decoder-Professional-Power-Adapter/32821841363.html

https://www.aliexpress.com/item/STUDER900-Regulator-linear-power-supply-DC12V-2-5A-30W-tube-Preamplifier-Audio-Decoder-Professional-Power-Adapter/32822732608.html

 

DIY ones

 

https://www.audiophonics.fr/en/linear-regulated-psu/module-alimentation-lineaire-regulee-mje15034g-faible-bruit-5v-a-24v-2a-p-11131.html

https://www.aliexpress.com/item/reference-STUDER900-power-line-LM317-regulator-5A-rectifier-ON-15034-output-dual-op-amp-TL072-Regulator/32824440580.html

 

We could also do better with some Audio Note Kaisei or Mundorf Mlytic AG caps etc.

Link to comment
15 minutes ago, austinpop said:

it doesn't remove the additional benefit that bridging still seems to provide

How are you bridging on the i7 NUC, are you using a USB ethernet adapter?

Pareto Audio AMD 7700 Server --> Berkeley Alpha USB --> Jeff Rowland Aeris --> Jeff Rowland 625 S2 --> Focal Utopia 3 Diablos with 2 x Focal Electra SW 1000 BE subs

 

i7-6700K/Windows 10  --> EVGA Nu Audio Card --> Focal CMS50's 

Link to comment
15 minutes ago, austinpop said:

Bridging still helps!

 

More than 2 years into its tenure, I yet again found that the technique that launched this thread - Ethernet bridging - makes a big difference in my system.

 

Background

 

Of late, I had not been using bridging in my system, for 2 reasons:

  1. from just over a year ago and until recently, I had moved from my SOtM trifecta to a Zenith SE-based all-in-one approach, so there was no network path between my Roon server and my Roon endpoint, since they were both running on the same machine.
  2. I moved my rig to another room when I put in a dedicated circuit, so my faithful Dell server was not conveniently located to run an Ethernet cable directly from its location to the listening room. All the rooms in my house have Ethernet home runs (in-wall) back to a wiring closet, so this is where my modem, router, and NAS live - far away from my listening room.

The NUC era

 

With my recent return to the distributed model (I've come full circle), I resurrected my Dell server (now running AL in RAM, with Roon DB in Optane SSD, and all other SATA devices disconnected) to be my Roon Core, but due to its location, could not accommodate a bridged setup.

 

Here is the baseline layout. A couple of notes:

  • while I have it, the TLS DS-1 has usurped my i7 NUC as the endpoint
  • While not shown, the tX-UBSultra is reference clocked by the Ref 10
    • Side note - for me, the tX-U reference clocked by the Ref-10 is still a HUGE improvement vs. going direct to the DAC from the NUC/AL/RAM endpoint.

Since my evaluation DS-1 has freed up the i7 NUC to try as a server, I recently reported in Larry's Audiolinux Server thread on comparing this i7 NUC as a Roon Server with the Dell. The latest post in that sequence is here: https://audiophilestyle.com/forums/topic/55247-audiolinux-server-configurations-software-hardware-and-listening-impressions/?do=findComment&comment=917305

 

Which leads me to bridging. Since my system has been in so much flux, my most recent configuration looks like this:

 

 

Baseline topology

                                           
                                          Router (HDPlex 100W) ---------- NAS (SMPS)
------                                           |
|      | ______________________ |

|      |
|      |------------------------------- i7 NUC (sPS-500 @ 19V)
|      |
|      |-------------------------------- DS-1 (SR-4) > tX-U (LPS-1.2) > QX-5 > amp
|      |
|      |_____ JSGT ground

------

TLS

OCXO

Switch (LPS-1.2)

 

Note that with this configuration, my i7 NUC is in my listening room, configuring a bridge on it was trivial.

 

Bridged topology

 

As you would expect, with bridging, the topology now looks like:

                                           
                                          Router (HDPlex 100W) ---------- NAS (SMPS)
------                                                         |
|      |                                                        |

|      |                                                        |
|      |------------------------------- i7 NUC (sPS-500 @ 19V)
|      |
|      |-------------------------------- DS-1 (SR-4) > tX-U (LPS-1.2) > QX-5 > amp
|      |
|      |_____ JSGT ground

------

TLS

OCXO

Switch (LPS-1.2)

 

Listening Impressions

 

I won't belabor the point. The bridged configuration sounds noticeably better. There is more air around instruments, a larger image, and a small reduction in "harshness," which at this point in my system is strictly a relative term, as my baseline system does not sound harsh by any means!

 

For me, the lesson was that while all the other optimizations that this thread has covered after bridging - power, clocking, OS, isolation, etc - all make a positive difference, it doesn't remove the additional benefit that bridging still seems to provide.

 

Explanations - none. But do try for yourselves and see what you think.

 

what is your definition of bridging?

 

i am thinking of something like this CISCO Linksys WES610N

 

cheers!

Link to comment
21 minutes ago, austinpop said:

Bridged topology

 

As you would expect, with bridging, the topology now looks like:

 

Thank you for the report Rajiv!  May I ask, what are you using for the second ethernet port on your NUC server?

 

I'm currently in the middle of linux-fu to get a WiFi/ethernet bridge working on my server NUC, just to keep another device off the USB controller on the server, but keep hitting roadblocks.

 

(oops, I see the response above now...thank you!)

 

 

 

ATT Fiber -> EdgeRouter X SFP -> Taiko Audio Extreme -> Vinnie Rossi L2i-SE w/ Level 2 DAC -> Voxativ 9.87 speakers w/ 4D drivers

Link to comment
7 minutes ago, ray-dude said:

 

Thank you for the report Rajiv!  May I ask, what are you using for the second ethernet port on your NUC server?

 

I'm currently in the middle of linux-fu to get a WiFi/ethernet bridge working on my server NUC, just to keep another device off the USB controller on the server, but keep hitting roadblocks.

 

(oops, I see the response above now...thank you!)

 

 

 

 

Hi Ray,


Yes, I am using the same USB 3 Ethernet dongle I bought 2 years ago to bridge my Dell machine at the time. The chipset is:

Device-2: Realtek RTL8153 Gigabit Ethernet Adapter type: USB driver: r8152 

           IF: enp0s20f0u2 state: up speed: 1000 Mbps duplex: full 

 

 

EDIT: It's still on Amazon: https://www.amazon.com/gp/product/B00NOP70EC

 

Link to comment
2 minutes ago, elan120 said:

I bridged my setup in the following way.  Not sure how close this is to your bridged topology, but the impression is similar where the bridged configuration sounds better.

 

1364770111_MusicPCBuildBOMList(1).thumb.jpg.d8bee041ad29f9e2b0a866837374c128.jpg

 

Hi Kevin,

 

Yup that's very similar. Only difference in my case is that my NAS (with 4 noisy HDDs and still on its original SMPS) is far away and attached to the router.

 

Nice setup, btw! I have been hearing excellent things about that JCAT Net Femto card.

Link to comment
14 minutes ago, austinpop said:

 

Hi Kevin,

 

Yup that's very similar. Only difference in my case is that my NAS (with 4 noisy HDDs and still on its original SMPS) is far away and attached to the router.

 

Nice setup, btw! I have been hearing excellent things about that JCAT Net Femto card.

I also like the JCAT Net Femto card, and having it clocked with sCLK-EX and sCLK-OCX10 now is really elevated SQ higher.

Link to comment
7 minutes ago, elan120 said:

I also like the JCAT Net Femto card, and having it clocked with sCLK-EX and sCLK-OCX10 now is really elevated SQ higher.

Beautiful setup.  My question is whether the EtherREGEN will reduce the need for such a sophisticated server.  I'm hoping.

Pareto Audio AMD 7700 Server --> Berkeley Alpha USB --> Jeff Rowland Aeris --> Jeff Rowland 625 S2 --> Focal Utopia 3 Diablos with 2 x Focal Electra SW 1000 BE subs

 

i7-6700K/Windows 10  --> EVGA Nu Audio Card --> Focal CMS50's 

Link to comment

@austinpop, @elan120, both of you are using an optimized switch in the path music server - endpoint according to your diagrams. Did you try it without (not asking you to, just curious) or do you have the experience from a previous setup that the switch still helps things forward. I did see Romaz writing the same so it would not surprise me but I’d rather not make assumptions. :)

 

Thanks for sharing your experiences BTW, very helpful and interesting!

Link to comment
14 minutes ago, austinpop said:

 

Hi Ray,


Yes, I am using the same USB 3 Ethernet dongle I bought 2 years ago to bridge my Dell machine at the time. The chipset is:

Device-2: Realtek RTL8153 Gigabit Ethernet Adapter type: USB driver: r8152 

           IF: enp0s20f0u2 state: up speed: 1000 Mbps duplex: full 

 

 

EDIT: It's still on Amazon: https://www.amazon.com/gp/product/B00NOP70EC

 

 

Thank you Rajiv!  I'll snag one to play with as I work on WiFi/Ethernet bridging (frustratingly close)

 

To your question about why it makes a difference, does anyone have a sense of what happens on bridged segments at the physical layer?  That is, is the only electrical signal that gets sent on the bridged port packets intended for the the device on the other side of the bridge?

 

Pure speculation, but our experiments with large buffers with SqueezeLite on the end point NUC clearly points to ethernet traffic having a very noticeable impact on SQ on the end point.  My hypothesis (untested) is that the end point responding to interrupts from the ethernet controller is somehow impacting the connection to the DAC over USB (data and audio have different messaging interfaces in USB, so there may be some contention in the USB controller?)

 

There is some network-related mechanism underlying the SQ difference we hear with large buffers.  To narrow it down, on my near infinite "to test" list is see if copying a large file to the large buffer SL end point during playback causes SQ to fall back to what we heard with small buffers.  If it does. next test would be to see if pushing heaving traffic across a shared switch (non-bridge config) impacts SQ on a large buffer SL end point connected to that switch.  

 

Alas, I'm a couple weeks away from having time for the next series of tests, but it is great to read that bridging still has an impact with the NUCs.

 

ATT Fiber -> EdgeRouter X SFP -> Taiko Audio Extreme -> Vinnie Rossi L2i-SE w/ Level 2 DAC -> Voxativ 9.87 speakers w/ 4D drivers

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...