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

[LTG2010]

1 minute ago, elan120 said:

Question1:  Would the latest RUFUS version 3.3 be an issue?  Or should I download version 2.18 instead?

 

Question2:  Should I download Audiolinux lxqt unless I want to go headless, otherwise no need to go with Audiolinux headless, correct?

1. Best Use Rufus 2.18 as this has the DD option even though Piero has fixed some issues in his latest version.

2. If you buy the headless version you can download both versions (but check with Piero) If your comfortable with comand lines then go straight for headless otherwise start with the lxqt to familiarise yourself with the system.

The good thing about the lxqt version is that its a desktop and you can plug your headless USB key in and edit its files in the desktops file editors.

Also you will need to mount your Hard disk/ file storage etc.  instructions are on Audiolinux website, its a simple edit of a file with your hard disks details.

[LTG2010]

1 minute ago, elan120 said:

Once I finished mounting hard drive / file storage etc., will switching back to Windows be the same as earlier instructions?  Remove USB stick and then boot from hdd where Windows resides?

Yes when you reboot without the USB you'll be back into Windows. Look for the NTFS script. You can basically mount any drive in the linux desktop, but Roon will not see it unless you edit the: etc/fstab file.

[LTG2010]

5 minutes ago, seeteeyou said:

Let's see who's gonna be the first to try AudioLinux in ramboot mode with an Innuos

Yep, technically you only need plug in your monitor to the rear of the SE and USB keyboard. Boot into bios and change the boot order to boot from USB then plug in your Audiolinux USB. Pull out USB and then your back into the Innuos OS. No harm done....... ?

Anyone here brave enough ???

[LTG2010]

37 minutes ago, stinart said:

Are you sure you are not having any drawbacks from moving to battery power?

Could be very poor mains quality, hence the improvement with battery.

[LTG2010]

1 minute ago, Confused said:

I noticed similar using the sPS-500 with my microRendu, and switching to the LPS-1 removed the issue.  Although I guess the LPS-1 also serves to isolate from the mains, as should any decently designed LPSU.  Mains may be a factor I agree, but I think it is pretty good where I live, but it’s impossible to be sure.  

An isolation transformer might do wonders for your system, if you can borrow one well worth a try.

Neotech OCC copper is very detailed but smooth sounding, if you can get a DC lead made up with say AWG 16, should not matter if its 500mm.

[LTG2010]

34 minutes ago, Confused said:

I gather that the DC cables offered by Paul Hynes are 22 SWG.  Surely adequate for 0.5m length?  Maybe not?

The resistance of the AWG16 wire will be approximately 1/4 of a 22AWG wire so it should in theory sound better, but I also don't know much about Paul's cables, I expect they will sound great.

[LTG2010]

9 hours ago, rickca said:

The USB reclocker board is already done as it's the same as used on the Statement - we designed it from the beginning to be used both internally and externally. 

I thought it was a PCIE USB card if its a reclocking board, most of you have the tx-USB Ultra.(Maybe Nuno means one and the same) The most benefit to the Zenith SE would be to fit a quality PCIE USB output, with OCXO clock. There's no space unless the top is taken off to mount it vertically, or some sort of riser cable could be used.

[LTG2010]

37 minutes ago, auricgoldfinger said:

Thanks, Alex.  The wire diameter is definitely a concern.  Hopefully, we don't have to go smaller than 18awg, but even 20awg would be an improvement over what I have now.

I would solder a small piece of 20AWG to the terminal /connector then wrap the 20AWG round the end of your 16AWG and solder that as close as possible to the terminal - rather than drop down to 20AWG for the entire cable.

A better option is to solder the 16AWG directly under the terminal socket and bypass the connectors, I dont know how feasable that is for you.

[LTG2010]

1 hour ago, Mark62 said:

As far as what it is, until someone has a more clear idea on "why" what some folks heard helps - then we're just chasing after ghosts in hope of finding something better.  Net, I'll wait and see instead of trying more w/bridging or swapping out other gear.

In audio, the shortest route offering the least resistance / lowest latency results in the least distortion. The direct route from PC to renderer will give the best performance / degrade the signal the least. A PC cannot be connected via an ethernet connection direct to a renderer unless its part of that network. Therefore the network bridge is just the link to the network, the signal follows the direct path from PC to renderer. I found the difference in sound quality  not subtle when I first tried this with a microrendu, especially with content stored directly on the PC. Even if you place an audiophile switch in the pathway that acts as a reclocker it still follows the direct path.

Regarding your original post, Roon is quite buggy in terms of locating storage other than the conventional path. I would look at the usb connection from NAS to router that might be causing a problem. I dont know if it's possible to connect up via ethernet.

[LTG2010]

59 minutes ago, Mark62 said:

Out of curiosity, prior to your bridging, what was you network path from router to PC and ultimately your microrendu?

Router/modem via long ethernet cable to netgear gs105 switch. Pc and microrendu  connected to switch via 1m cables. Then connected microrendu direct to pc ( removing connection to switch) and bridged pc's ethernet ports. I don't have a NAS.

[LTG2010]

11 hours ago, Soma said:

So I am wondering what's the difference between a 10MHz and 100MHz clocks. Is this a thing? 

Hi, the 100MHz is likely to be the frequency of your DAC's clock, of which the OCXO will replace.

10MHz is the Master / World clock frequency of the the Mutec Ref10.

The Mutec Ref10 does not replace any clock it just acts to regulate any clock attached to it in order to improve its performance, closer to that of the OCXO in the Ref10.

Most clocks have a master clock pin so there needs to be a Master clock input on any device in order for it to connect to the Ref10 or similar, such as that on the Mutec MC3+ or SOTM's sCLK EX board. The good thing about the sCLK EX is that you can 'upgrade' up to 4 devices, so when attached to a REF10 it becomes cost effective, 1 OCXO regulating multiple devices with lower spec clocks.

Regarding the OCXO for your DAC it is the most important clock in the chain, however its all in the implementation, circuit design, components and power supply and also the quality of the DAC itself. For example some $5K DACS use a $25 Crystek Femto clock with very good results.

Placing the Mutec ref10 or similar prior to an OCXO in a DAC in theory should have minimal effect due to the asynchronous data flow, but in practise we know it can have dramatic effects depending on the equipment in the chain. For example it seems to have less effect on FGPA DACS like Chord DAVE but larger sound quality increases on other DACS and its also dependent on the quality of the equipment preceding, computers, NAS, Switches etc.

[LTG2010]

11 hours ago, Bricki said:

I assume it's going to require more than the 12v 1.1a of the lps1.2? Have you tried to boot it off the lps1.2?

If you limit the RAM to say 4GB then the LPS 1.2 might work well.

[LTG2010]

6 minutes ago, rickca said:

I'm surprised that Roon Server can run completely diskless.  Usually a database needs to do some I/O operations to maintain integrity (checkpoints, logging, etc) but I guess if you aren't adding any music or updating metadata then it doesn't have any transactions to commit.  

 

Doesn't Roon Server ever try to do some automatic maintenance?

When running from RAM, Roon can update etc, but its all saved to RAM. If you have your USB stick installed there is a ramsave command which will save the changes to your USB drive. Regarding databases when using ramboot I just load a few files in Roons storage options, I don't mind not having the full database the sound quality is worth it. Audiolinux/ Roon needs to be configured to save the database to non volatile storage and not be part of the boot files.

Roon is not entirely necessary, you can use HQplayer desktop on a high powered pc and HQplayer's network audio adapter on your NAA/ NUC in RAM alternatively HQ Player embeded controled remotely. Jriver also sounds fantastic. There are many possibilities.

Also great work Larry.

[LTG2010]

1 minute ago, austinpop said:

What this means is that if there is a more creative option, there may be further SQ to be gained. I like @Dev's idea of running ramsave on a timer. Kind of like a syncd, except not every 30 seconds! But this still puts responsibility on the (super)user to ramsave before a reboot. And of course, you're always vulnerable to a data loss due to a crash.

Its also benefitial to clear the RAM once in a while and for huge databases it takes forever to boot. If there was an option in Roon to choose a location for the database files the could be kept anywhere on the network, it can be done.

[LTG2010]

2 minutes ago, austinpop said:

Perhaps a network attached SSD?

Thanks for the links, yes something like that, its simply library management.

[LTG2010]

3 hours ago, rickca said:

I wonder why people only recently jumped on this AudioLinux/NUC solution.  Hasn't AudioLinux been around for quite some time?

RAMROOT funtion is more recent ( 6 months) that tipped it for me, plus reading about it here.

As for the NUC I haven't gone there yet simply because the Supermicro X10 SBA motherboard sounds so good with Audiolinux, so no rush.

[LTG2010]

44 minutes ago, greenleo said:

Say the headless is 10, the SQ of the GUI would be ?

8.5

[LTG2010]

7 hours ago, simonklp said:

 

I see. For your information, I use the 3.0 version on Windows 10 Pro without problem. Cheers.

Rufus V3 works fine but it does not have the option to create a DD image, on some machines ( a minority)it may cause an issue so a safer bet is V 2.18 if you run into problems.

[LTG2010]

6 hours ago, One and a half said:

Can Roon core run on a USB stick of AudioLinux? 

Both versions of Audiolinux have Roonserver and Roonbridge installed, just need starting or enabling via a command (or a prepared script in GUI version).

I found it simply more transparent coming from windows with all the optimizations. It works pretty good with my Celeron J1900 based system.

Experimenting with various machines, high powered ones increase noise, and tend to give a more dynamic but much harder sound, therefore Pink Fauns approach of high (ish) power with extremely good power supply and improved clocking yields better results.

However when you use the higher power machine as a server and the lower powered machine as the endpoint, the isolation provided over ethernet, mitigates some of that noise pollution and things really get interesting. Still a lot of experimentation is needed.

[LTG2010]

1 hour ago, romaz said:

Is higher latency leading to increased jitter, increased noise or dropped bits?

Absolutely software as well as components are subject to jitter lowering latency and as a result jitter would account for the increased clarity.

The second point regarding the RAM OS is simply the removal of the very noisy OS SSD/HDD which introduces a large amount of jitter in the playback chain.

As for the NUC there is no magic in 'NUC' as such it's the simplicity of it's SOC processor, avoidance of PCH etc. RAM as well as low latency is also direct to CPU on most motherboards hence the good overall results with audiolinux.

[LTG2010]

12 hours ago, ElviaCaprice said:

1.  Format limitations:  JRiver video/hires playback is a need of mine. 

You can use Jriver in Audiolinux lxqt it sounds fantastic and outperforms Roon with PCM at least.

The jump up in quality from Windows to Archlinux is equivalent to going from a standard setup to a sclk-ex one. The better the components the more the improvements become apparent such as the standard motherboard USB output to that of a tx USB- exp with sclk - ex.

Get off that fence

[LTG2010]

31 minutes ago, ElviaCaprice said:

Your claims are a bit over the top which makes me hesitant about this OS.

Lol.

You are hesitant to pay $29 and burn the image to a USB stick and try it for yourself.

Once you pull the USB stick out you are back to your old OS.

[LTG2010]

18 minutes ago, romaz said:

This isn't just about the noise that a drive makes.

I didn't say it was- but its part of it as well as the latency which is well documented by Piero on his website. Most users have SSD / HDD OS systems and are hearing the benefits when removing them. I believe Innuos play back from RAM to cut out some of the SSD noise.

[LTG2010]

3 minutes ago, sandyk said:

You can also cut out most of the SSD noise using the +12V supply and regulating down  to +5V with a low noise voltage regulator.

 This prevents most of it getting back into other areas via the PSU.

 The attached JLH PSU add-on separately powers my OS SSD and Music SSD (12V IN, and 2 x separate +5V Out )

Yes, Ive seen this and it looks very well thought out. I also use dedicated 5V rails for SSD and HDD from a linear power supply.

Just that the sound is so much clearer when the SSD has been removed with playback from RAM, I have no affiliations with Audiolinux I just spotted one of Larry's posts a while back mentioning it and tried it out myself.

[LTG2010]

7 minutes ago, lmitche said:

I have machines here that cover all these bases and still don't sound as good as the NUC.

I have tried out the little NUC and it does sound great, so I don't intend to knock it, Ive also seen your website and good luck with that.