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

[Theobetley]

Anybody successfully bridge a wifi and ethernet nic on windows 10 after the 1709 update? So far it has eluded me and I don't like the sq loss.

[mozes]

Has anyone compared the sMS-200 ultra and the Auralic  Aries G2 streamer?

[seeteeyou]

I wish that's just nothing more than April's Fool but apparently it's Con Te Partirò for real this time

 

https://www.oppodigital.com/farewell.aspx

 

Oppo Shock: Popular AV Brand Announces That 'It's Time To Say Goodbye'

https://www.forbes.com/sites/johnarcher/2018/04/02/oppo-shock-popular-av-brand-announces-that-its-time-to-say-goodbye/

 

Oppo Digital Pulls The Plug, Interdyn Already Planning Ahead

https://www.channelnews.com.au/oppo-digital-pulls-the-plug-interdyn-already-planning-ahead/

[auricgoldfinger]

1 hour ago, seeteeyou said:

I wish that's just nothing more than April's Fool but apparently it's Con Te Partirò for real this time

 

https://www.oppodigital.com/farewell.aspx

 

Oppo Shock: Popular AV Brand Announces That 'It's Time To Say Goodbye'

https://www.forbes.com/sites/johnarcher/2018/04/02/oppo-shock-popular-av-brand-announces-that-its-time-to-say-goodbye/

 

Oppo Digital Pulls The Plug, Interdyn Already Planning Ahead

https://www.channelnews.com.au/oppo-digital-pulls-the-plug-interdyn-already-planning-ahead/

 

You better hurry up and buy yours. 

[seeteeyou]

1 hour ago, auricgoldfinger said:

You better hurry up and buy yours. 

 

Let's look at the bright side, we still have Cambridge Audio CXUHD if Roon weren't a necessity.

[rickca]

1 minute ago, seeteeyou said:

 

Let's look at the bright side, we still have Cambridge Audio CXUHD if Roon weren't a necessity.

Maybe.  Quote from John Archer's article in Forbes

I've also asked Cambridge Audio if Oppo's announcement will impact its 4K Blu-ray player business, given that the British brand's CXUHD players have Oppo electronics at their heart.

[Johnseye]

1 hour ago, auricgoldfinger said:

 

You better hurry up and buy yours. 

 

Yea this makes me question whether to buy one now.  I love Oppo and have many generations of them, but if the firmware updates start slowing down, and support will end at some point, it's a risk.  A real shame because they're the best made Bluray players.  Kind of like when Pioneer stopped making TVs after the Elite line of plasmas.  Those were the best TVs made at the time, they just cost too much and there wasn't enough demand for high quality, expensive plasma once cheaper LCD came out.

[jay]

Sorry to jump back probably 9 months or longer in terms of topic newness for this thread, but I've got what I hope is a quick question regarding running the Mac Mini off of the SD card.  I've successfully (?) done that this morning, but my question is where I should be running Roon from, as this is also my Roon Core server.  I assume the "applications" folder is by default with the OS, so it appears it's on the SD card with the OS where I booted from, and consequently I'm assuming when Roon installs in that folder, then Roon runs from the SD card too in that state?  Is this where I want to run it, or do I want it running from the internal hard drive?  I assume the latter?  I am not an OSX expert by any means...I made a folder on the main hard drive and tried both to move it there, which didn't work, and then to install it there, which also didn't work (app wouldn't load in either case).  Can anyone smarter than I give me some quick answers?  Thank you!

[mozes]

Just to have a fresh perspective in SQ improvements on this thread but of course  not to derail it.

We have been focusing on optimizing our servers whether they are a custom built PC, streamer, or an end point. There is no doubt that we can always harvest some improvements in that space (power,  clocking, DC cables, isolation, etc.). However, I recently chnaged my approach and decided to add a pair of Rel Tzero subs. I have a near field two channel setup where Dave directly drives Omega Super 8XRS speakers. These subs are tiny in all aspects with a 6.5 inch drivers and can be hidden behind the Omegas. The improvement I got from these subs is really jaw dropping! It took some careful tweaking to blend them in, but once finished the results are astonishing.

The sound stage took new dimensions in all directions, vocals became more natural with space and air to die for. The highs are sweeter and you feel like they just reach out to fill all the space in the room.

Bottom line, this is by far more effective than anything else I have done to tweak my system. Don’t listen to the old view of audiophiles that subs are bad. Keep an open mind and try a quality sub, I am glad I  went with Rel. These are very fast with tight bass.  I hooked them up to the Omega binding posts directly through their high level connection, so no long messy wires.

[Ralf11]

Yes, attention to acoustics will usually be more rewarding than attention to electronics (at least, initially).  But be carefull with the rel's -- they may burst into flames, as happened to @wgscott

[jay]

@mozes +1 on the sub recommendation.  I added a Zu Undertone to my Soul Supremes last month, and had the same experience.  Ironically, I had 2 Paradigm $1700 subs that are supposed to be musical in the mix (1, then 2, then back to 1), and I never got them dialed in, so I eventually just took them out.  I bought the Zu on the reviews, and it was so easy to integrate with the Zu speakers.  So the matching and tweaking, to your point, is totally necessary.

 

And apparently not having them catch fire.  Yowser.

[Johnseye]

1 hour ago, mozes said:

Just to have a fresh perspective in SQ improvements on this thread but of course  not to derail it.

We have been focusing on optimizing our servers whether they are a custom built PC, streamer, or an end point. There is no doubt that we can always harvest some improvements in that space (power,  clocking, DC cables, isolation, etc.). However, I recently chnaged my approach and decided to add a pair of Rel Tzero subs. I have a near field two channel setup where Dave directly drives Omega Super 8XRS speakers. These subs are tiny in all aspects with a 6.5 inch drivers and can be hidden behind the Omegas. The improvement I got from these subs is really jaw dropping! It took some careful tweaking to blend them in, but once finished the results are astonishing.

The sound stage took new dimensions in all directions, vocals became more natural with space and air to die for. The highs are sweeter and you feel like they just reach out to fill all the space in the room.

Bottom line, this is by far more effective than anything else I have done to tweak my system. Don’t listen to the old view of audiophiles that subs are bad. Keep an open mind and try a quality sub, I am glad I  went with Rel. These are very fast with tight bass.  I hooked them up to the Omega binding posts directly through their high level connection, so no long messy wires.

 

I use a single JL Fathom f112v2.  It took a bit to dial in, and I still keep it on the low side as it's too easy to overpower certain music.  It adds a lot more.  Specifically from 21-47Hz more.

 

I believe the magic of the Rel is in its ability to simply crossover.  Is that the case with your Tzeros?

[jay]

In case useful to the thread at this point (additional relative data points I think), here is what I've done this week, and the results...focused thus far in following the recommendations on improving the Mac Mini as a Roon server.

 

In prep for my incoming SOTM Trifecta, MMK linear power for the Mac Mini, etc., I did the following:

 

1.  Loaded OSX High Sierra to a 128G high-speed SD card rather than running it from the internal hard drive, and then tried to optimize the Mini as much as possible by turning most everything off, as recommended here, and in the myriad of sites on what to do.  I run it headless so it doesn't drive a display unless I remote into it.  I was still a little surprised at the difference doing all this made (even though this thread suggested it would)...mostly because of how bad my network is between it and the network player (I thought the bad, complicated network would mask any improvements).  I would say it definitely lowered the noise floor, but in doing so, it also gave the appearance of a little less weight/meat on the bone, so to speak?  I think I lost a little bass energy.  I would still say it improved overall--definitely could hear parts in well-known songs that had some nuanced layers I hadn't noticed before.  I think a couple folks here correctly termed is as "relaxed" and in my case it pulled out a little more detail as well.  Cool.  I did NOT overwrite the hard drive, so at some point I can boot as normal and test again without the SD mods...I don't think that's hurting anything on the SD-boot side, since none of it is running.  The server is still running with wifi on, over wifi in a remote room, then through 3 (count them, 3) additional Netgear switches and probably 100 feet of 10-year-old CAT5 to get to the UR.  Hardly an ideal network path to the UltraRendu/ISO Regen/Schiit Yggy.  That's obviously optimization to come.

 

NOTE:  The results appeared to be more pronounced late night and early morning, where I might assume either the power coming to the house is cleaner, the power circulating in my house is cleaner, and/or the house was quieter and therefore I could hear better.  Probably some combo or all of the above.  I do have dedicated lines and balanced Equi=Core power, but I'm not convinced that time of day doesn't still matter.

 

2.  Added an Apple Airport Extreme in the LAN path between the old CAT5 that was going to the UltraRendu.  Wireless is off, so I'm only using it as now a 4th switch.  It is NOT plugged into the balanced power as it's on a wall-wart.  Shockingly, I couldn't hear a difference...I expected it to be more noisy and congested (worse), but it actually sounded just as clean, maybe even the slightest bit more relaxed?  Perhaps the clock in it is not as bad as the clocks in the netgear switches before it?  Let's conservatively say "no additional harm."  This was to prep for the next 4 tests, which I will try to get done tonight:

 

3.  Move the Mac Mini to the listening room, hanging it off of the Airport Extreme switch.  This will put only this switch between the server and the Ultra Rendu, rather than a wireless switch and 3 hardwired switches.  Tidal will still be flowing through the 3 netgear switches, so 3a will be local files, and 3b will be remote Tidal.

4.  Putting the switch on the balanced Equi=Core/dedicated line vs. plugging it into the standard house lines.  I can repeat 3a and 3b as 4a and 4b if the result of 4 is interesting.

5.  Turning off Wifi on the Mac Mini.

6.  Using a firewire/Ethernet adapter to bridge the UR to the Mac Mini, thus eliminating the Airport Extreme switch between them.  5a will be local files, and 5b will be remote Tidal.

 

Jay

[mozes]

1 hour ago, Ralf11 said:

Yes, attention to acoustics will usually be more rewarding than attention to electronics (at least, initially).  But be carefull with the rel's -- they may burst into flames, as happened to @wgscott

Thanks for the caution

[mozes]

1 hour ago, Johnseye said:

 

I use a single JL Fathom f112v2.  It took a bit to dial in, and I still keep it on the low side as it's too easy to overpower certain music.  It adds a lot more.  Specifically from 21-47Hz more.

 

I believe the magic of the Rel is in its ability to simply crossover.  Is that the case with your Tzeros?

They have a nice built-in cross over that is easy to work with together with the high/low volume. The Tzero doesn’t go deep, maybe nothing less than 30Hz I guess, but this is enough for music. 

[jay]

Some findings from last night.

 

a) As noted in my previous post, even though I have dedicated lines (3) and Equi=Core balanced power (1800), I still hear worse sound during the day than I do at night and early morning.  I'm sure I do not have optimized power in this house, even with what I've done.  This affected my second round of testing, and I had to re-do the tests last night for apples to apples.  This was true in previous testing I'd done a year ago with the Audience power conditioner I've been using for the last few years.  Equi=Core was better, but didn't totally isolate bad power in the house.

b) plugging my Audion Golden Dream 300B monos into a dedicated line was NOT as smooth and balanced as running them through the Equi=Core balanced power, on a dedicated line.  I had not tried this, and I know many report hearing improved dynamics and clarity with amps going direct to the dedicated lines.  That's one everyone should play with (if you've made the investment in dedicated lines), as I'm sure it's very system- (and amp-) dependent.

c)  Out of necessity (it wouldn't power up after unplugging it), I took the IsoRegen out of the mix to first test the introduction of the Apple Airport switch into the mix (#2 in my previous post).  It was a LOT of SQ degradation vs. having it in (I couldn't get it to power up once I re-plugged it in, so for a couple songs I thought I had killed it, but it revived itself thankfully and I put it back in).  I hadn't pulled that out of the system since putting it in to replace the USBRegen--so it's been years without having one of those in my chain.  At least with an UltraRendu/MicroRendu, the IR is a must in my experience if you're going USB into your DAC...in my system with 3 different DACs I've tried, anyway.

 

With the rest of the house asleep, I redid the tests...interesting results?  Maybe, maybe not.

 

3.  Placing the Mac Mini near my Equi=Core and power cord loom didn't seem to be good...it was still on wireless, but it sounded less clean on the top-end.  I'll assume it was either picking up RFI being that close, or it was introducing it's own noise back in the Equi=Core.  Maybe some combo or something else entirely.  I tried to connect it via ethernet, turn off wireless (I jumped to 5), and it sounded identically bad.  So I moved it over to the other end of the rack away from all the power cords (but next to the Apple Airport switch)--perhaps a little less "bad," but still not as good as when it was running in a remote room.  Then I unplugged it altogether from the Equi=Core and put it into it's own dedicated line, off of balanced power.  Some improvement.  In the end, I believe that it was roughly the same as having the server in a completely other room, going through all those switches, even for local files.  I could not discern any notable differences between local files (3a) and Tidal files (3b), btw.

4.  Putting the Apple Airport switch on balanced power...hard to tell if this made any difference, so I pulled it back off of the balanced power conditioner, especially since this switch was unnecessary once I get to #6.

5.  Again, switching network connections between ethernet and wireless (and then turning off the service), I didn't hear a lot of improvements--there was one flip from wireless to ethernet where I heard an audibly discernible "calming" but I wasn't really able to recreate it, so I'm going to conservatively call this a draw in my current environment too.  It's definitely not worse through a crummy Apple Airport switch (wifi off on the switch) and crummy CAT5.

6.  Bridging the MacMini and the Ultra Rendu using the Thunderbolt connector.  Perhaps just a little improvement.  Pulling the Apple Airport switch out of the mix again, maybe a little more improvement.  No discernible differences in local files (6a) vs. Tidal remote files (6b).

 

Net-Net Punchline:  The SD card OS and the Thunderbolt/ethernet bridging, on their own, have not made "OMG" improvements in my system, but I would argue for the $100 investment plus a couple hours of time, those sticking with Mac Mini as a Roon server should do them.  I got a little cleaner if not a little tiny bit leaner sound (after some settling in, I'm not longer convinced of this slight loss of "weight"), but definitely a little more detail, and I wouldn't say any less musical (I won't sacrifice musicality and PRAT for detail, ever).  This seems to be at least in the direction of most of the previous findings.  I don't think you will go backwards doing these two changes.  These may pay off bigger dividends with some of my incoming SOTM, Uptone and Sonore purchases that have not yet arrived.

 

Of related note, for me, IsoRegen still has a massive effect when going USB.  Sablon USB on top of that was also a giant USB improvement in terms of SQ.  Those two investments, while sizable, have had the most demonstrated improvements in this current chain of late.  I'm feeding a Schiit Yggy with the USB5 upgrade--different DACs may have different results.

 

Next up:  My Uptone MMK power supply upgrade for the Mini, plus a JS-2 linear power supply should arrive later today.  I'll see how quickly I can get the MMK installed, but before I do that, I will move the JS-2 around the system (recognizing it won't be at all burned in) and see if I can hear any differences.

[str-1]

Suspecting that the answer is still ‘no’, I’m wondering if anyone has had chance to compare the Mutec Ref 10 with the sCLK-OCX10?  I need to take a decision on ordering an sCLK-OCX10 very quickly if I’m to benefit from a free sPS-500.

 

I would also be interested in any comparison of the sCLK-OCX10 with standard caps vs Evox caps.  May has told me that the difference that can be heard from the Evox caps is more about a change in character (e.g. more energy) than any technical improvement (though to some, more energy would be just that).  I’m not sure that I need more energy with Chord Blu 2/DAVE at the heart of my system.

[auricgoldfinger]

4 hours ago, str-1 said:

Suspecting that the answer is still ‘no’, I’m wondering if anyone has had chance to compare the Mutec Ref 10 with the sCLK-OCX10?  I need to take a decision on ordering an sCLK-OCX10 very quickly if I’m to benefit from a free sPS-500.

 

I would also be interested in any comparison of the sCLK-OCX10 with standard caps vs Evox caps.  May has told me that the difference that can be heard from the Evox caps is more about a change in character (e.g. more energy) than any technical improvement (though to some, more energy would be just that).  I’m not sure that I need more energy with Chord Blu 2/DAVE at the heart of my system.

 

I have two sPS-500's, one with the Evox caps and the other without.  When powering a tX-USBultra, I prefer the supply with the Evox caps.  The sound is more open as if a veil were lifted.  I didn't spend time doing a more detailed analysis because the difference was quite obvious when I first made the substitution.

 

I'm not really sure what May means by more energy, but it's hard for me to imagine why better caps would be detrimental to an audio device, and particularly, a reference clock.  I have never heard anyone complain after upgrading caps.

[rafa]

Is anyone tried to build own 10MHz WCLK? 

 

There are some not so expensive Abracon OCXO 10MHz 2-5PPB oscillators on the market. With good quality PSU it could be something maybe worth to test?

[str-1]

2 hours ago, auricgoldfinger said:

 

I have two sPS-500's, one with the Evox caps and the other without.  When powering a tX-USBultra, I prefer the supply with the Evox caps.  The sound is more open as if a veil were lifted.  I didn't spend time doing a more detailed analysis because the difference was quite obvious when I first made the substitution.

 

I'm not really sure what May means by more energy, but it's hard for me to imagine why better caps would be detrimental to an audio device, and particularly, a reference clock.  I have never heard anyone complain after upgrading caps.

Thanks.  I have little technical knowledge but can more easily imagine upgraded caps bringing clear improvements to a power supply.  The comments May offered me were in respect of the sCLK-OCX10.  Would the impact you noticed be similar to what could be expected by upgrading the caps in a master clock?  I really don’t know, so just asking the question.

[Cornan]

6 minutes ago, str-1 said:

Thanks.  I have little technical knowledge but can more easily imagine upgraded caps bringing clear improvements to a power supply.  The comments May offered me were in respect of the sCLK-OCX10.  Would the impact you noticed be similar to what could be expected by upgrading the caps in a master clock?  I really don’t know, so just asking the question.

 

If you want to try a really great capacitor, try the KEMET A750MS567M1EAAE015 aluminium polymer caps. The best caps that I have tried with digital audio for sure.

[auricgoldfinger]

I also am not technical, but as I mentioned earlier...

 

2 hours ago, auricgoldfinger said:

it's hard for me to imagine why better caps would be detrimental to an audio device, and particularly, a reference clock.  I have never heard anyone complain after upgrading caps.

 

Sorry I can't be of more help.  My sCLK-OCX10 with Evox caps and eABS-200 is scheduled to arrive on Monday.  This will be my 3rd unit because SOtM has screwed up my order twice!  Hopefully, this one will be configured to my specifications and will also work properly. 

[ALRAINBOW]

Has anyone’ here used a CUST built server for 

nas , network roughter and roon server running anology hack 

I use this and a second pc for hq to dac via usb ?

[seeteeyou]

On 4/7/2018 at 12:01 AM, rafa said:

Is anyone tried to build own 10MHz WCLK? 

 

There are some not so expensive Abracon OCXO 10MHz 2-5PPB oscillators on the market. With good quality PSU it could be something maybe worth to test?

 

Well, I learned something new from reading this post by @romaz and maybe the question is more like this - why don't we go for an OCXO system clock directly (provided that we're able to source an OCXO with the right frequency) instead of resorting to 10MHz reference clock with a significant degradation?

 

https://www.head-fi.org/threads/review-comparison-of-5-high-end-digital-music-servers-aurender-n10-cad-cat-server-totaldac-d1-server-auralic-aries-audiophile-vortex-box.787020/page-79#post-14120317

Quote

One of the things I learned about the REF10 is that Mutec's published phase noise measurements were taken from the BNC output and not from the clock itself and so comparing the phase noise plots of REF10 to another external master clock may not be a true apples to apples comparison and this highlights my point. I predict that by the time the clock signal goes through whatever length clock cable you use to connect the REF10 to something like the sCLK-EX board in my tX-USBultra and by the time that clock signal reaches the USB hub in my tX-USBultra, that signal may have become significantly degraded and so this is the potential benefit of physically placing a good clock in the same location as the old clock.

 

Before sCLK-EX was introduced, we could only find a very limited number of components with 10MHz clock inputs by default. Now we do have an option to modify anything with sCLK-EX ourselves or pay SOtM to do that for non-audio devices. However, the clocks inside sCLK-EX itself will be disabled once we sync them with 10MHz reference clock and therefore we might wanna ask ourselves how could we pay so much for those relatively expensive clocks in the first place when we aren't really using them? On top of that, we're also paying for a REALLY expensive 10MHz reference clock while it's inevitable that signal is gonna be degraded significantly as mentioned above.

 

Anyways, the improvement is still there and it ain't even a subtle difference for sure. Though part of it might simply owe to the fact that most system clocks should be pretty lousy to begin with.

[Hauser]

This is what I have been thinking for some time now too.  Should it not be asked of SOTM, why they don't offer a TX USB with a direct connected 10MHz clock to the card rather than selling expensive upgrade clocks, making the previous ones redundant and adding more and more boxes and cables?

 

Martin.