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

[Johnseye]

7 hours ago, tims said:

Hi @jean-michel6

 

I also want to feed my tx-USBexp card with an external psu but I will have to open up my SOtM server to disconnect the tx-USBexp cards internal  supply as you have done.  This may invalidate my warranty doing this so do you recall if the SQ was markedly better when you disconnected the internal supply of the tx-USBexp card?

 

Thanks

 

This shouldn't invalidate your warranty.  Check with May at SOtM.  Let me know if you need any help.

[Johnseye]

5 hours ago, tims said:

Great, thanks for the information. 

I have a PH SR7 on order that will eventually supply the mobo and the txUSB with separate feeds.   I wonder if I disconnect the internal supply from the txUSB (as you have done) and power up/power down the SR7 to turn the server on and off that should be OK?

I will talk to May from sotm about it.

 

 

 

I use and SR7 to power my motherboard and tX-USB-exp.  Leaving the SR7 on and powering from the PC's power will keep continuous power to the tX-USBexp.  Even if you shut off the SR7 it will power both down simultaneously just as it would if the tX-USBexp were internally connected.  I've had no issues whatsoever.

[Johnseye]

12 hours ago, vortecjr said:

I spoke with Andrew at Small Green Computer and Roon today about server performance. I knew there were specific reasons why Andrew makes high performing music servers when there are less expensive and lower performing hardware options available. Understand that Andrew's music servers are designed primarily for streaming to an endpoint and dedicated to this task so he is not worried about server noise with network streaming. These music servers run a host of server software and each has it's specific minimum system requirements. However, the comments below are primarily for Roon server.   

 

This is a summary of the discussion:

When it comes to Roon Core performance, there are two things that are important to remember:

- The most demanding workloads aren't happening constantly, so subjective experiences once the system is "up and running" isn't the best indicator that things are "all good".

- Roon grows in its capabilities over time.

 

The heaviest stuff that the Roon Core does is mostly non-interactive and is not happening constantly:

- Importing

- Performing audio analysis on your library

- Database migrations during software updates

- Altering library settings or batch-editing lots of content

- Re-extracting tags, either manually, or because lots of files were changed at once

- Updating library metadata in the background

- Syncing with your TIDAL library in the background

- Recursively scanning folders to pick up changes to your library

 

Sometimes core performance simply determines how long these things take, but it might also determine how reliable things are while this stuff is happening, or how responsive the system is during those times. Not everyone minds sluggish performance, but Roon doesn't make recommendations based on getting people to a sub-par experience or one where you have to think about too much stuff instead of just using the system and trusting that the resources are there to make it work.

 

Weaker machines will get bogged down during some of these times. Importing content can be slow. Database manipulation can overload the machine to the point where playback is failing or remote connectivity is unreliable. Screens may take a while to load because of library database manipulations happening at the same time as a result of some of that stuff. Weaker core hardware is repeatedly implicated in support situations-

-usually atoms, celerons, and when people are using NAS's that don't meet the minimum specs.

 

Library size also matters--a lot. There is more leeway to use somewhat underperforming hardware with a small (say 10-20k tracks) library, than a large one (say, 100k+ tracks). Storing the library on a local drive is generally less demanding than the same library on a NAS--since the NAS creates a bunch of extra network traffic, and generally takes longer to do filesystem operations than a local drive. 

 

Roon's DSP features are an additional factor. These can produce pretty substantial CPU load even during single zone playback when configured in certain ways.

 

Finally, when you're making a decision about core hardware, it isn't just about right now. It's also about how the system will feel down the road. Roon is constantly growing/evolving/expanding its product offering. It's important to be prepared so you are not hurting next time they do something big that disturbs the performance equation--like the release of DSP features last year.

 

5 hours ago, Dev said:

 

I am not so sure if this is totally true. Over at Jplay, most has been using a dual jplay setup for a while (long before microRendu or the likes came out) - the control PC, which is equivalent to the music server, is also very important in contributing to the SQ as is the audio PC, which is equivalent of a streamer, like mR. If you have to optimize one, the one closer to the DAC must be done first but a noisy music server can also pollute the SQ. Thoughts ?

 

I would say that all of us who have made improvements upstream from the endpiint have found they have an impact. 

 

Jesus if you haven't tested and begun a path to develop a solution which follows a path similar to what's being discussed here you're missing out. There is real opportunity here, but maybe Sonore isn't interested in that path. 

 

Our most recent discoveries have shown us that SQ can be improved without the endpoint when specific components are used. 

[Johnseye]

6 hours ago, Dev said:

 

By any chance, have you experimented with two identical server with those recipe (which eliminates the need of endpoint) and connected them through network - one acting as server and another acting as endpoint and found the SQ did not improve ?  

 

Two identical servers where one was the core and the other the endpoint?  If that's what you're asking, no.  I wouldn't invest in duplicate hardware to test that.  I did test a server (without modifications, tX-USBexp, sCLK-EX, tX-USBultra) and endpoint (microRendu & sMS-200) against the same exact one (with modifications, tX-USBexp, sCLK-EX, tX-USBultra) but without endpoint and it was without a doubt, no question a significant improvement in SQ.  I still have the sMS-200 and adding it instead of the USB output, but still gaining the sCLK-EX utilization on the motherboard system and NIC clocks does not sound as good as with the tX-USBexp and tX-USBultra.

 

With so many variables and possible combinations it tough to test them all.

[Johnseye]

19 minutes ago, austinpop said:

@mozes Moussa,

 

Can you describe where you’ve tried the Bybee IQSE in your system, and what you heard?

 

In the same vein, any experience with Synergistic Blue or Beeswax fuses? 

 

Nothing like some Romaz discoveries to get the mind thinking.

[Johnseye]

6 minutes ago, mozes said:

No need for $$$. A $20 power bank really transformed my switch. The switch is 2 feet away from my server, but the LPS-1 bettered the battery. Does the PSU matter for a switch from a SQ perspective? Yes for sure. My ears don’t lie.

 

What power bank are you, or were you using for your switch?

[Johnseye]

3 minutes ago, mozes said:

This one

 TP-Link 6700mAh Portable USB Battery Charger, Power Bank for Smartphones and Tablets (TL-PBG6700) https://www.amazon.ca/dp/B01GJ782FC/ref=cm_sw_r_cp_api_f7MEAb89CC8PD

 

this is special as it doesn’t disconnect power when you charge it, so I leave it plugged in to the charger 24/7

 

What cable did you use to connect that to your switch?

[Johnseye]

25 minutes ago, jean-michel6 said:

Which modifications is it ? 

The server clocks driven by the sCLK ex ?

 

Yes.  System and LAN clocks driven by the sCLK-EX.  Also, tX-USBexp clock driven by the sCLK-EX.

[Johnseye]

1 hour ago, Dev said:

 

are you using the network during playback, like streaming or using NAS ? if not, how does the LAN clock mod help ?

 

The music sources come from both my NAS and Tidal, through the NICs.  It's SOtM's opinion that all the clocks have an influence.  I don't know the logic behind this.  I didn't bother with the USB clocks because I'm not using them.  Since the music passes through the NICs I thought that worth while.  To do it for the motherboard USB clocks I'd need to buy another sCLK-EX and I don't think that's worth it, not to mention I'd need a place to house another one.

[Johnseye]

3 minutes ago, lateboomer said:

Hi Johnseye,

Could you clarify further as my English is not that good? You said you tested server with endpoint and without one, then which setup you think improved significantly? From your no doubt statement I can't figure out which one you are referring to. 

 

Without endpoint.  Using the sCLK-EX taps for my motherboard, tX-USBexp and tx-USBultra for a direct USB connection was better.

[Johnseye]

1 hour ago, hurka said:

what do you think about iso R and singxer F1  clock with sclk ex board? I cant clock my sparky sbc  this time. 

 

I don't know as I've never tried them with the sCLK-EX.  I did have an iso Regen but it didn't play well with my other equipment.  For the times I got it to work I liked what I heard, but that was before the sCLK.

[Johnseye]

3 hours ago, LTG2010 said:

Have you guys seen this- SOTM have released their new Master clock. It will be available for shipping from 19th February just after Lunar holiday and pre order 12th February.

There's a promotional price of  $3500 with sPS500 (a free sPS500) and $500 for the cable ($200 discount), for the first 100 units.

Press Release_SOtM_sCLK-OCX10.pdf

 

For $4k I question whether the gain in SQ is equal investing that amount into better speakers, a DAC, preamp or amp.  My guess is I'm better off putting it towards any one of those components.  Whether that's a Chord Dave, Blu2, or a myriad of other components, $4k is a tough nut to swallow.  That's just me though.  YMMV. 

[Johnseye]

2 minutes ago, auricgoldfinger said:

 

The introductory price is $3500 which is comparable to the Mutec REF10, which @romaz has gushed over.  Maybe he will chime in with an answer to your question.

 

It's not really a question.  I'm fairly certain I'd rather have a new Chord DAVE than a master clock.  The costs compared are not equal, but if I were offered a DAVE at a $4k discount I'd probably bite.  We all have our priorities.

[Johnseye]

22 minutes ago, austinpop said:

 

I'll chime in and say that - for me, my system, and my ears - my Cybershaft OP-14 made a bigger sonic difference than what I heard replacing my Codex with a loaner DAVE.

 

I know Roy thinks I'm nuts for not hearing more from the DAVE, but as I've repeated many times, the reason I am so active in this pre-DAC part of the digital chain is that the improvements I'm hearing are a bigger bang for the buck than I hear with DAC upgrades - and I've tried the Hugo 2, DAVE and the Ayre QX-5.

 

Does that mean I won't ever upgrade my DAC? Of course not! What I'm finding is that the old adage of "improve your source" applies here in full force. With my latest round of improvements upstream of the DAC in place, I felt like I hear bigger differences between DACs now.

 

Everything matters.

 

YMMV.

 

I think we both agree YMMV.  Who knows, SOtM's master clock could be the bees knees and we'll all want one.  I just look at what I paid for other components.  For example, Zenith SE or master clock?  Same price point, which is better?  It probably depends on what your needs are.

[Johnseye]

Just now, austinpop said:

 

Having done that one, I know my verdict. With my previous baseline of the SOtM trifecta, if I were given a choice of only one of the Zenith SE OR the Cybershaft, it would be the Cybershaft without a doubt.

 

Interesting.  I wonder if we could compile a list of better to best in terms of SQ improvements as rated by people who've tested.  I'm not asking you to do this of course.  At some point I may consider doing it as it will take some time.  Maybe this summer.

[Johnseye]

On 2/7/2018 at 3:28 PM, Cornan said:

 

Fair enough Alex! I use AHP gold fuses on several spots with brand new devices and they have improved everywhere that I have tried them. However, the best spot I have ever experienced a fuse replacement was with my "vintage" AVM Evolution A1 integrated amp. The sonic improvement there just blew me away. I'll guess the fuse holder as old as well, but I always use DeoxIt on all my fuses and fuse holder contact surfaces.

It reminds me of my tube days though. Replacing the valve sockets to new ceramic and golplated ones was a surpricing improvement with my Siemens E88CC gold-pin NOS tubes (with Duende Criatura tube rings) on my Musical Fidelity X-Can V2 with Rock-Grotto mod. Everything matters! ?

 

What version of DeoxIt do you use?  I was just talking with a tech over at Audio Research (ARC) as I have their LS-28.  He actually recommended DeoxIt as well.  In fact, when I mentioned I was going to try a Synergistic Blue Quantum fuse he said you can absolutely hear a difference in fuses.  Not only that but the direction of the fuse as well.  So this is from a well respected company who has to install fuses in all of its products.

[Johnseye]

1 minute ago, auricgoldfinger said:

 

I use G5.  Does a nice job on my tube pins.

 

So formulated for gold or not, it doesn't really matter?

[Johnseye]

1 hour ago, lmitche said:

 

Rick, the microAtx 6700k dsd512 solution running here with the dual PSUs is sounding spectacular. I would argue it is the scaled up configuration you are looking for. It doesn't have a Jcat or Sotm USB card, but those were tried and eliminated. I haven't tried a Jcat Ethernet nic, but the USB powered nic used here gets it's lt3045 power through the 4 port USB card, and sounds great.  Tidal SQ = HDD SQ, so no need to do more. I'll add a 3 amp lt3045 and a couple 13.5 volt sigma 11 lpsus in the next week and I'm expecting to be done.

 

Scratch that original question, Rick asked and I missed that post.

 

What made you pick that board?

[Johnseye]

13 minutes ago, austinpop said:

Speaking of Axpona, I know @lmitche and I are going. Anyone else on this thread going to be there? I'm going to propose another CA meetup. The one at RMAF was a lot of fun, but too short!

 

As for manufacturers, I believe Innuos will be there, and SOtM was a maybe, last I checked.

 

Ummm....yea.

[Johnseye]

1 hour ago, lmitche said:

The Asus motherboard was the only matx board without radios and other gizmos that would not be used in a music server. I prefer Asus branded products over the competition.

 

I've taken a similar approach and have taken into consideration the size of the board to limit the length of the lanes.  I wonder if the inclusion of components like wifi have any impact when disabled in the BIOS.  I certainly take it into consideration, but wonder if I should look at those features as limiting for an audiophile purpose.

[Johnseye]

19 minutes ago, d_elm said:

You could put a JSSG braided shield on what you have to lower RF interference, if any.

 

How do you connect the shield at the ends?

[Johnseye]

1 hour ago, lmitche said:

You don't. You connect a new drain wire between the endpoints of the braid. Nothing is connected to whatever is in the shield. The impact is surprising. Just Google jssg for details.

 

Works on DC, Ethernet USB and Spdif cables.

 

I thought I read somewhere the braid you bought.  On Amazon?  Would you mind linking it or sharing the name?

Googling jssg doesn't hit much.  I'll search through this forum.

 

thanks

 

Edit, this was a good explanation.

 

On 12/20/2017 at 12:55 AM, JohnSwenson said:

The whole concept of the JSSG is a separate wire outside the shield that connects one end of the shield to the other. The shield should NOT be connected to anything else. SO in an existing shielded starquad cable connect the four wires in the starquad geometry to the power plug, do NOT connect the shield to the plugs at all, and connect a separate wire to one end of the shield and the other end of the wire to the other end of the shield. It does NOT have to be a thick super duper wire. I use a 24AWG wire in mine.

 

You can either just lay the wire in parallel to the main cable, or loosely wrap it around. Loosely wrapping around main cable is fine, I use 3 turns per foot or so.

 

John S.

[Johnseye]

3 minutes ago, d_elm said:

The shield goes on top of the existing cable, is not connected to anything, and solder an insulated wire, around 20 AWG or thicker, at each end one or two spirals down the outside of the shield.  Then add a sleeve on top of the shield.  Ends of shield and slieve are secured with heat shrink.  1/2" braid and shield can be installed over an ethernet connector.  If one connector is not installed 1/4" braid and sleeve can probably be used.

 

That is worth 1000 words.  Thank you!

 

Where did you get your braid?

 

[Johnseye]

2 hours ago, sandyk said:

 John

 Recently I purchased a 2M - Canare L-4E6S, Premium Balanced Quad XLR Microphone Cable .

 I cut a 2 feet length from it , with the shield left open at both ends, and compared it against a similar length of 2 core mains cable at the output of my JLH PSU add-on before taking it further with the extra wire to both ends of the shield.

 I lost most of the 3D effect on better than average Quad Reel to DVD-A stereo downmixed material, so I didn't bother taking  the final step.

I was using a 12V 15,000mAH Li Ion battery into a JLH PSU add-on to power a USB memory stick via an Uptone USB Regen.

 I believe that the gauge of wire used in the microphone lead was way too thin to compensate for the much lower resistance of the same length of power cord.

What I am suggesting here is that for best results you may need to use heavier gauge wire Quad leads.

 

Regards

Alex

 

I'll be trying it around my USB cable to start. It's interesting that it had such a profoundly negative effect on your sound.

[Johnseye]

4 minutes ago, mikicasellas said:

Does anybody have experienced the Roon Rock solution? i am confused NOW regarding having a Roon Rock which i have been reading that it is a huge step over a Regular PC/MAC, so how much of a difference would someone experience having a Zenith MKII instead of a Roon Rock Nuc? price is way cheaper with A Ronn Rock NUC.

 

i know the PS on the Zenith is very good and so, but as a person who is going to extract the best possible SQ from Roon, a NUC Roon Rock seems to be a great alternative too and a lot cheaper is not it? also you can put more RAM, processor and bigger SSD for less and keep it flexible for future updates...

 

What am i missing here?

 

ROCK is the OS unless you're referring to a NUC sold by Roon, which I'm not aware if they are selling those or not.

 

There are many ways to skin this cat.  If you want simple and great sound, get a Zenith.  If you want to build your own for great sound you can use ROCK.  If you just want simple, buy a standard NUC with ROCK on it.