Euphony OS w/Stylus player setup and issues thread

[FelipeRolim]

Good morning everyone. After much delay, I finally decided to try Linux which is the new "darling" around the world: the Euphony Stylus. Having tried the massive majority of free and paid Linux (Daphile, Archphile, Runeaudio, Volumio, Jaguar Audio Design, AudioLinux, among others, including even the most basic ones like KODIbuntu), and failed to try programming since from begining a Linux that was superior to Windows in music playback, I bowed to the testing of the internationally commented Euphony Stylus as the best available. It costs nonsense €249.00 (I find it ridiculous to be a Linux), but for all I heard, who calls it "best of the day" is, in fact, right.

 

After much reading to verify the best method of tuning, I used the setting indicated by most users as being superior in sound quality. To do this, (i) loaded the operating system into RAM, (ii) set the buffering of songs to RAM, and (iii) used Stylus mode, which, among others available, has apparently outperformed Roon in native mode and Roon combined with HQPlayer. The Stylus interface is extremely simple, but, because I already passed Roon's free trial, I opted to try Euphony exclusively in Stylus mode.

 

It's the best "finished and ready operating system" I've ever experienced. It's far superior to AudioLinux and any other because it doesn't have that overly clean, polished "tinny sound" that is typical of Linux, which "stageophiles" and "bassophiles" love. Euphony may be the only Linux I've known since I learned how to properly configure a Windows which I really liked. Still loses, by a good margin, to Windows Server 2019 Datacenter Core Mode (17763.593) with Audiophile Optimizer v. 3.00, JPLAY FEMTO and Minority Clean X installed. In direct comparison, Euphony still denotes some dryness and harmonic poverty, while Windows Server 2019 delivers much richer, fuller sound. It's a much more beautiful and organic sound. The voices are correct, which doesn't exist with Linux. So while it's less exhibitionistic than the others and doesn't appeal to excessive clarity to create a false sense of detail and definition, it still has the same timeless inaccuracy that is common to Linux. It's amazing how explicit it is.

 

To hear “audiophile music” sound doesn't easily denote many deficiencies without instant and direct comparison. However, to hear “ordinary music”, which doesn't always have a great deal of attention in recording or (re)mastering, Euphony makes things a little less comfortable but still easily audible.

 

But what struck me most, and I haven't noticed in any other tests I've done, either with free or paid operating systems, is that Euphony limits the music dynamics. The system loses articulation and cannot reach high levels of sound intensity, as required, for example, in Hugh Masekela's Stimela songs and Harry Belafonte's Mama Look a Boo Boo. It sets a clear limit, as if the amplifier clipped, which doesn't exist with Windows and even with AudioLinux. I was intrigued, because never, in an operating system test, did this feature catch my ears. I can't vouch because I don't know the specifics of the project and intend to use the 30 day free trial to exhaust any possibility of improvement of Euphony, but the feeling I got after hours of listening was that in an attempt to "tame" the typical Linux sound that I critique, reducing glare and highlighting in the early harmonics, the developers eventually inserted some algorithm or hit that, while making the overall sound more palpable, limited the articulation, the dynamics.

 

I believe the transport has favored the performance of this Linux, but it's a fact that, for those who don't have the opportunity to draw a direct comparison between a well-tuned Windows and Euphony Stylus, it performs very well. What's more, for those who don't know how to tweak Windows with a deep degree and/or want something that is simple to install, configure (takes less than 10 minutes), play and forget, it's the best I've ever had opportunity to meet. It plays well and needs very few adjustments. Just not cheap.

 

By the way, any other tips for improving the playback level? Thanks.

[FelipeRolim]

30 minutes ago, Nsxturbo said:

Greetings Felipe,

 

What a crazy coincidence! I just spent all day yesterday getting Euphony set up on my music PC and listening to Stylus. I too was interested in what everyone else was raving about.

 

For years I’ve been using XXHiEnd on a particular build of Windows 10 running out of RAM. It was actually a necessary part of getting 768k upscaled audio sent to my Phasure NOS1 DAC. Simply sublime playback. I now have an additional system that consists of a Chord M-Scaler into a Dave directly driving a high efficiency horn system directly off the Dave’s RCA outputs without an amplifier. I feed the M-Scaler through a highly optimized usb DDC to optical output unit clocked with Mutec Reference10 master clock.(really to many other tweeks to mention here.) The PC is an optimized Xeon Scalable 12 core/24 thread processor with 48G of ECC Memory in a Supermicro server motherboard running off a custom linear power supply. No hard drive is present and the OS is running in RAM. I’ve still been using XXHiEnd software but playing files in their native resolution. Really sounds Jaw-dropping, head shaking incredible! Just about every adjective others have used to describe accurate and realistic playback apply.

 

So I got Euphony up and running out of RAM on the same PC and system. I’ve got to say it really is a pretty slick program and I find Stylus plenty functional and user friendly. (OK, it’s not Roon, which I use for my whole house audio  ) But there is always Stylus Endpoint for Roon integration. And Stylus does allow Qobuz and Tidal integration and playback. 
 

Now for the listening impressions. I used my highly familiar test tracks and albums. Rebecca Pidgeon- The Raven, Cowboy Junkies- Trinity Revisited, Various Norah Jones albums, Amber Rubarth, Rhiannon Giddens, Sarah K, Holly Cole, Dire Straits, Johnny Cash, Plus multiple orchestral selections, pretty much the whole gamut!

I would describe it as highly resolving, nice black background and good separation and good wide soundstage with a decent sense of depth. Really all the things that you would think would draw you into the performance. But it just wasn’t, something was off I couldn’t quite put a finger on it, it just sounded like the life was missing. 
 

So back to XXHighEnd and Windows out of RAM I went. Now what was missing was readily obvious! The textures, the body, the sense of real depth, the naturalness. It really was no comparison. Stylus was just not even in the same league. Could it get there? I don’t know, but it has a long way to go and I’m not sure if Linux is the underlying issue.

 

I think this gives me the impetus to explore Window Server 2019, Audiophile Optimizer and JPLAY out of RAM. I suspect that it would be more in-line with the level of playback XXHiEnd provides me.

 

Hello. Very interesting your impressions. After many years of experimentation, I'm convinced that much of the audio performance comes, in addition to dedicated hardware (no moving parts, good level linear power supplies, among other things), from the driver provided by the manufacturer, and the software. The perfect integration of the software with the hardware, coming from the specific drivers and their integration with the software (in my case, JPLAY FEMTO, with which I use Ayre QB-9 with its own Thesycon driver in Kernel Streaming mode) leaves the result many higher than using native or generic drivers (as with Linux).

 

I believe this is, among others, one of the reasons why a well-matched Windows has superior performance. However, I think this is a subject that would yield many pages and escape the theme of this thread.

 

22 minutes ago, Jeff Mann said:

Felipe,

 

You wrote-: "Still loses, by a good margin, to Windows Server 2019 Datacenter Core Mode (17763.593) with Audiophile Optimizer v. 3.00, JPLAY FEMTO and Minority Clean X installed."

 

I am not computer-savvy, and I would like to know whether I can purchase a computer that is setup as you described, and that is fully "finished and ready to play"?

 

Jeff.

 

Hi Jeff. I don't know where it would be possible to find something like this. My Windows, installed on my two dedicated computers, have been punctually tuned for many years since I was still using Windows 7 to play music and more basic software like foobar2000. The improvement was slow and careful, and I believe that is why I was able to reach the current stage of soundness and accuracy.

 

Anyway, I don't believe that I've developed something universal (which is impossible with a Windows), and I'm convinced that every computer and sound system requires a specific hit. Only with Linux you can achieve something with sufficient degree of abstraction to the point of creating a universal product, but the specific arrangement needed to extract juice from the system isn't possible.

 

I'm not criticizing. I never liked AudioLinux, because I really think it has tin sound, but I really liked Euphony. It just doesn't beat the Windows I set. For a universal product, easy to set up, it's perfect.

[FelipeRolim]

22 minutes ago, Nenon said:

Hi Felipe,

 

If you are the same person who contributes a lot to the jplay forum, I have read many of your posts there. Thank you for sharing so much of your experience. And if I recall correctly, in one of your posts you mentioned that every time you try to share something useful on this forum you get attacked by sharks (well, not exactly what you said) who has a different vision. I will try to be extra careful with my comments. 

Actually, when it comes to open minded people here on this forum, I am as open minded as it gets. I would try everything as long as it does not pass the border with complete stupidity. I have played with Windows and AO but never the exact combination you have mentioned. That's going on my list of things to to try one day. 

 

But I wanted to say what I think about your "dynamics" limit with Euphony:

 

Euphony works best with very powerful CPUs and fast RAM. The faster the CPU, the more dynamic the sound is. You can clearly hear a difference in dynamics by enabling the Intel Turbo Boost in my Core i9 9900K CPU (vs. disabling it), or even when reducing the CPU frequency by a couple hundred Mhz. The same thing applies to my AMD Ryzen 3700X CPU based music server. Reducing the CPU frequencies reduces the dynamics.

Euphony is using a low latency kernel... I don't think they have "inserted some algorithm" to change the sound. I just think they picked a Linux kernel that sounds good and applied the typical low latency kernel optimizations... but that's just a speculation on my side. However, what is not a speculation is that those types of Linux distributions (if done right) can squeeze a lot more of the hardware resources you have available and process data with significantly lower latency (compared to Windows that is... I do that for living and have a first hand experience). Look at the high performance trading industry, and you would never see people using Windows servers for anything critical. Now, that does not say anything about how one OS sounds compared to the other, but it takes me to the next point, which is why I believe you have limited dynamics with Euphony.

Looking at the system listed in your profile - it seems like you are using an Intel Core i7-4750T. That's a 6-year old CPU that only works with DDR3 RAM, and it is not exactly a fast CPU by today's standards. From my experience, Euphony really started to shine when I tried the top of the line consumer grade CPUs like the Intel i9 9900K and AMD Ryzen 7 3700X with no frequency limiting. I did mention that, but as soon as you start limiting the frequencies, you start losing dynamics. 

I have always looked at the digital source as a complex chain starting from the player, OS, hardware, power supplies, cables, interface with the DAC, the DAC architecture, etc. etc. You change one of these elements, and you have to rethink the entire chain.

You have a highly optimized Windows-based system. It's so well optimized over the years that there is probably no significant improvement you can make by changing a single element of the system. You switched the OS to Euphony, and it sounded worse. No surprise there. If I put tractor tires on Ferrari, it would not perform better. Neither would a truck with Lamborghini tires. If you really want to hear the full potential of Euphony, you should try it with a much faster CPU. 

 

Thanks for the contribution, Nenon. I believe my subscription is out of date. I tried Euphony on both computers I have, the Core i3-8100T, with 16GB DDR4 RAM, and the i7-4785T and 16GB DDR3 RAM. In both cases, taking care to have the JCAT USB Card FEMTO as output. Both computers, when running Windows, work much better with TurboBoost turned off (i3 doesn't), just like Hyper-Threading, C-States, and other elements. I did not make any changes to the BIOS, but it was interesting to comment on it. I'm still in the 30 day trial, and I will try to get as much Euphony as possible. It will have to evolve a lot to overcome Windows. Thanks for saying, I will try it on the weekend.

 

14 minutes ago, Chopin75 said:

Interesting findings, as linux based system supposedly to be best for audio (at least for a finished product), but perhaps endless personal tweeking on windows can achieve same or better effects (ironically some audiophiles would not even talk to someone who still uses windows or Mac!). I am not computer savvy and I hate windows (maybe also a bit personal, as I am Mac fan) so there is no way for me to do windows with then endless pain of difficulty in finding/loading drivers. (yah, I am rather computer illiterate) . Regards the euphony OS you have, are you still testing the OS using trial version on a USB thumb drive? If so that is quite inferior to using a SSD. It appears that octane drive works best with Euphony OS. The registered paid version is also a bit more stable and so is better than the trial  version.   100% buffer is also important.  I also find Ecache helps a lot if you don't have internally stored music on the SSD.

 

Thanks for your comments. As I said: (i) loaded the operating system into RAM, (ii) set the buffering of songs to RAM, and (iii) used Stylus mode. I still have some things to try, so tips like Nenon's are very welcome. I'm going to experiment with changing the processor clock this weekend, and reactivate some features of the motherboard and processors.

[FelipeRolim]

I don't know how productive that would be. Many people are only concerned with making their views prevail, not with learning and sharing their experiences. I confess I am very reluctant to try Linux. The only one I liked was Euphony, but I'm willing to learn and listen, since I don't have to spend my money paying for a license beforehand.

 

I'm still reluctant to adopt certain things that are sold in this forum as solutions. But, I can test what is suggested. Any tips for optimizing the operation of Euphony on my hardware will be very welcome. Nenon's are annotated.

[FelipeRolim]

On my computers (using JPLAY FEMTO dual-PC configuration, and also set up Sylus in "Server" and "End Point" modes), I have six power supplies specifically tuned for the best sound: two for processors (sigma11 with special components), two for motherboards, one for SSD (ldovr single rail) and one for JCAT USB Card FEMTO (ldovr single rail). I have two more ldovr for the JCAT Net Card FEMTO, but I haven't bought them yet. So, from what I've already set up and what you're saying, without changing my current hardware to a more powerful, I can only set the processor core assignments/frequencies and promote the BIOS adjustments. That's right?

 

If that's all, I think I should expect a miracle of these changes, because overcoming today's Windows will be very difficult.

 

51 minutes ago, Chopin75 said:

I see , thanks! Now just to confirm, are you using USB Stick to load the OS to RAM ? This reminds me that I am not sure if SSD makes a big difference with RAMroot (I have not tried RAMroot) but per these guys here, optane + RAMroot works best. 

 

Once the operating system is loaded into RAM, there's no further use on the USB stick. However, just like any other USB stick, I realize that external USB connections that are not exclusively for PC to DAC communication detract from the sound quality. So, my computers have only network connections (I tried to disconnect but never heard any difference, I believe by the isolation between the network and PCI-Express controllers of the USB card) and the DAC's USB connection. There's no mouse, keyboard, monitor and any other gadget, as they all, in my view, impair the quality of the audio. I also don't use moving parts, such as hard drives or coolers, as they inject noise into the circuit, either by conduction (cables) or by dispersion (air).

[FelipeRolim]

I downloaded the latest version of Euphony and restarted the experiments. I set it up as I had done before and added the suggestions colleagues left on the topic. Improved, but aspects that don't relate to what I wrote that Windows is superior. That is, it continued to play incorrectly from the timbre point of view, without the same dynamics and articulation of Windows and with a remarkable poverty in textures and harmonics. With the monitor connected to the computers, I tried everything without success. It was the last time I tried a Linux.

[FelipeRolim]

On 10/27/2019 at 9:28 PM, Lukasluis said:

I too find Euphony Stylus' timing and dynamics to be off in my system thus I use Euphony RoonCore (NUC directly connected to DAC). Can you try and let us know if switching to Roon makes a noticeable improvement in timing and dynamics.

 

Hello. Unfortunately my testing period with Roon is already gone. At the time, I had tried Roon against JPLAY FEMTO, and I considered Roon worse in sound quality. Either way, even better, I'd never pay the price for its lifetime license.

 

On 10/27/2019 at 9:36 PM, AnotherSpin said:

 

Would you please mention which components you use to get such wonderful audio from Windows? Thank you.

 

I didn't understand your question exactly. On my computer I use only Windows Server 2019 Datacenter in Core mode, Audiophile Optimizer v. 3.00, JPLAY FEMTO and Minority Clean, all running from RAM, besides making over dozens network and BIOS adjustments. Thus, the SSD works exclusively for transporting music files. But, let's leave that to when they open a specific thread.

 

The big thing is that in my experiments over the years I've never been able to make the network connection between two computers outperform one computer alone while relying on a router or switch to do the IP and network management work. That is, neither bridge connection, router, nor switch played better than a single computer. What the bridge connection does is simplify the path and allow only one cable, but still has absolute dependency on the router for network management. So these "standard" Linux provided to us don't work as well. But, once I learned how to set it up correctly, without the computers relying on the router's DHCP anyway, then both Windows computers started playing substantially better than anything I experienced. I think one of the secrets is on the network, but only one of them, and there's much more behind it, such as the manufacturer-specific Windows drivers.

 

On 10/27/2019 at 10:08 PM, Chopin75 said:

Sorry, remind me again, are you doing euphony RamBoot from USB stick ?  How do you run windows again? Is it inside a  M2 drive or other drive ? 

 

I don't use M.2 because I consider the SSD, with 5v linear power supply and audio grade SATA cable deliver better sound. I tried the internal and external storage. Since Windows is loaded into the RAM from a .vhd file, I can easily swap it through GRUB4DOS.

 

I'm far from saying that I know how to teach someone how to set up a computer. Adjustments are personal and depend on a number of factors. But in my case, I've achieved something very interesting over almost 10 years experimenting. The fact that Linux has few possible tweaks can help, but it is also a limiter when trying to get the most out of a computer. So, or the user is committed to an "easy ride" solution (Euphony is the best alternative I've experienced with this bias), or an effectively audiophile solution. In this case, personally, I prefer Windows, which to me sounds at the top level.

 

To see how there's a difference: I tried several frequencies of the Core i3-8100T processor. Between 3.10GHz and 2.40GHz there's no sonic difference using Windows. But, with Euphony, using 3.10GHz plays better, as Nenon suggested. Therefore each case is very particular. It's not easy to get the most out, so, I like to have a lot of adjustment variables. I don't spend days or waste time testing more than listening. Adjustments have been occurring gradually over many years and many more hours of listening. That's why I won't waste time on a Linux anymore.