Euphony OS w/Stylus player setup and issues thread

[davide256]

17 minutes ago, Mike Rubin said:

Very good question.  I would like to know this, too.  I have a lot more money tied up in my Signature Rendu SE than I would in a second NUC. (I assume in this configuration the NUC's connect via ethernet rather than USB.  If the latter is the required connection, this won't work for me because server and end point are on different floors of my house.)

Its the first case, the physical setup doesn't change.  It was a conflict for me selling the microRendu, always felt that it was a  SW use case limitation where

if it ran Stylus endpoint or Stylus server supported UPNP that the outcome could flip.

[davide256]

4 minutes ago, edwardsean said:

I leave my system running, but I've also rebooted many times. I'm pretty sure the change in sound is due to the update. Euphony, like many audio companies, do not seem to flag all modifications in SQ in their change logs. 

 

How this shift affects you, or to what degree, will obviously vary heavily from system to system .

 

I would venture to say though that if your system already errors on the analytical side I would make sure I can revert to the version before 20210216.

?? No system in profile or signature... kind of a data point with no x-y coordinates or scale  😉

[davide256]

2 hours ago, edwardsean said:

 

Yup. I'm real mystery wrapped in an Enigma. 

 

It's a system built around a DAVE+SJ power supply, if that provides some reference. I haven't really listed it all out because from there I've gone a little bit on my own way. Cheers. 

Thanks, good to know I would probably have to spend a lot more money before the latest version could pall 👍

[davide256]

14 minutes ago, mantis07 said:

Hello, I'm trying to use RAMROOT with Stylus and am having the issue below. I'm using a Ryzen 4800u NUC with 32GB of RAM. Any ideas? 

 

Thanks!

Tony

 

 

 

When fault isolating always best to start with simplest case first, not full bells and whistles. Does RAM root work with copy app config

unchecked? If so, perhaps try clearing the queue before invoking the copy app config option... perhaps Euphony is trying to copy

a queue buffer?

[davide256]

I’m trying Stylus UPNP streaming now with Qobuz, running into issues with some songs truncating early, advancing to next song using an Oppo 103, seems to happen at same place when I repeat play, Oppo song timer shows there was more to play before song quits. Will try a Sony player next, wondering if anyone else has tried this and what their outcome was?

[davide256]

3 hours ago, dminches said:

 

To be fair, Stylus has no issues playing gapless on its own.  Asking each software vendor to keep up with other software packages is asking a lot.  Every time there is change to HQPlayer Željko has to make sure it still works with his software.  I am sure he will get there but it is a burden.

 

 Basically Euphony is like JRiver was like originally.. designed to play to physically connected devices. Solutions like Roon don't have this

issue because they were built for network play (latency).  Proper flow control logic is universal, so whatever is discovered with HQPlayer

should be applicable to whats missing/weak with UPNP. Whats baffling me is why some Qobuz songs in Stylus are 1 second shorter

in song length as buffered using UPNP at an  Oppo 103, causing premature song termination. An example would be the sound track to

Star Trek - Nemesis,  happens on tracks of around 2 minutes.

 

 

[davide256]

Hadn't noticed any treble issues before upgrading DDC's but I'm noticing now a "hot pepper" like effect in the treble. Is this what others are experiencing

as different with the latest Euphony version?

[davide256]

10 hours ago, Lukasluis said:

I have done extensive and extended (trial extension) testing regarding this 2 box Euphony config. The best I have so far is Euphony Server on NUC 8i3 to a Windows UPNP Endpoint on Intel NUC 7i3. The endpoint is runs on Windows Server 2016 core and running Jplay Femto. This 2 box setup is much better than a stand alone Euphony in my system. It may be due to the DAC driver which is developed specifically for Windows. The generic Linux DAC driver on Euphony could not even do Native DSD 256. Both the machines are powered by oversized linear power supplies. 

Curious as to how you are running a JPlay Femto on an i3 NUC without a PCIE slot? I’d really like to improve the USB out on my i7 NUC

[davide256]

3 hours ago, Holzohr said:

 

It is possible with a m.2 to PCIe adapter. The drawback is you can not use a m.2 ssd in this case. I am just playing with this solution: m.2 to PCIe adapter plus JCAT USB XE. The NUC (NUC7i7DNKE) is booting either from a Diretta stick or from a NAA stick. I think about purchasing a HDPLEX H1 V3 case plus NUC Kit plus m.2 to PCIe riser.

 

 

 

Thanks!  Would lose Optane boot using the M.2  riser for PCIE but that sounds like a minor sacrifice to get clean 5v USB out +better USB out hardware.

[davide256]

4 hours ago, al2813 said:

 

Hi, I assembled a NUC with an HDPLEX H1 V3 and the NUC kit + riser cable. I ended up selling the combo becuase i decided to keep my second hand server in a Streacom F10 Alpha for the time being as a combo Stylus server / end point with my JCAT USB card. While the HDPLEX box is a great product and the only solution in the market allowing you to run a fanless NUC with a PCI card, there one point to take into account. You need an external power supply for the JCAT card, since the the riser cable delivers 3,3V and it seems to not be sufficient to power the JCAT. You will also need to get a SATA splitter to allow the single SATA port in the NUC to feed both the JCAT and the SSD (and indeed M2 is not possible). 

 

I have been in contact with Larry on this one and I believe my info is checked. The person who bought my NUC with the case confirmed the same findings to me. He had no issue as he had DIY LPS foreseen both for the NUC and the JCAT card (in his cae net card)

JCAT site doesn't list specs for current draw on the Femto, any idea how much amperage is needed for external LPS? Any reason you can't do SATA

boot if you have external LPS?

[davide256]

14 hours ago, di-fi said:

When changing to a track with a different bit depth the next track plays with heavy distortion

 

I am hoping that someone in this great and resourceful community (many thanks for sharing!) could give me a clue how/where to troubleshoot. With my setup (details in my profile)  I did not have any problem until end last year.  Then suddenly my (1 box) Euphony Stylus (ram root and buffer) started randomly playing distorted tracks when changing from 16 bit to 24 (and 24 to 16) from a play list or 'radio' function. Basically in my day to day listening (all my music is on a NAS) I can only play 1 album (same bit-depth either 16 or 24) or make playlists with one bit depth. When it happens I can hear the song play, but it is extremely distorted, a lot of high, no bass. Sometimes the song will play distorted and at a faster pace. 

Without blaming Euphony, this distortion began december last year after Euphony/ Željko was troubleshooting so I could have all tracks of 1 album (instead of only 2) loaded in memory. (That has been fixed since for everyone).  Another change I made in that time was the installation of the new UNISON USB card in my Yggdrasil dac.

In December Željko did some testing for the distortion. I could provoke the distortion while changing tracks, he listened to a distorted song, could not see a bug in Euphony code and concluded this must be some kind of unfortunate incompatibility between OS and my dac. He stated that only if more users have this problem (which I honestly do not wish to anyone ;-) there might be a solution.

 

I did not find any indication Unison is not explicitly supported by Linux and relies on generic USB drivers. I tried changing dac delay in Stylus settings, changing sample rates or not at all. I am  hesitant to uninstall the unison card but it could confirm where (not) to find the error. I could separate/eliminate all 24 bit files (my DAC plays 16bit wonderfully :-)) , but I do not consider that a real solution. 

Does anyone have an idea? Thanks in advance,
Paul



Sound file below (you can hear me change tracks). You do not need to listen for long, it sounds pretty awful:

https://drive.google.com/drive/folders/1aMrj1bH0Naeh5AHoFlj1_nYGuyuOCD4u?usp=sharing

 

Please note I can flawlessly play up to 196kHz - PCM. But no DSD due to Yggdrasill dac. 

Do you have the same issue when you are using the USB ports native to the SuperMicro MOBO vs those on the Startech USB card? Can't say that I encountered

this when I had a Gumby with the Unison upgrade connected to NUC. I seem to recall that the Gungnir always audibly clicked when speed rate changed.

[davide256]

2 hours ago, di-fi said:

SJJ,

You mentioned SPDIF and AES. For 3-4 months I played SPDIF - Yggy with a SingXer (LPS 1.2) with and wo IsoRegen and it was great, no Issues. I changed back to UNISON USB and LUSH2 usb cable with and wo IsoRegen and that seemed at least equal in quality. It was only then that the distortion occurred. 

The SingXer has been sold, so I can not verify if it is an USB vs. SPDIF issue. Your setup must sound great, I was always curious how Yggy/AES would sound.  I would like to try Engineered eRED dock, but ideally with Euphony. 

It looks like you are using a splitter cable from the LPS 1.2, or at least your profile shows MOBO + Startech card powered by LPS 1.2. If you are

using a splitter cable you may want to sub in a different supply to the MOBO, see if that resolves the issue. Could be crosstalk from the MOBO

connection causing issues with the startech USB out.

[davide256]

12 minutes ago, NanoSword said:

a- with this feature the problem is You have to buffer the queue before disconnecting network . (( this will add network activity  in buffer)) .

b- but I think is better to disconnecting  network first then  buffer .

with option (b) you are buffering  without network activity.

 

I test this with same song two copy I find (b) less harsh more natural more clear . can you make same test from your side .

 A queue is just a playlist until its locally cached or buffering is complete. So perhaps you are caching first?

[davide256]

12 hours ago, NanoSword said:

 

 

I am enabling cache in my server I use my internal optane drive.

 

But I am trying to explain that we need enhance the method of disabling network with euphony .

I think for better sq we need to disable network first .

second buffering the queue .

this should apply in sequence .

 

So basically you want a macro command that

a) caches the queue

b) takes network connection down

c) buffers queue from cache

d) plays entire queue after buffer completes

e) brings network connection up after play completes or upon any error

 

I will admit I haven't been tempted to use the feature in place for network-less play, prefer something more automated/ idiot proof.

[davide256]

12 hours ago, di-fi said:

Hi Davide256,

 

I put up my audio profile recently, sorry if it was not clear. I have no splitter cable. Thank you for mentioning the crosstalk, I'll keep in mind. 

 

My SuperMicro X10SBA MOBO has 2 separate 12v power feeds:

• Uptone LPS 1.2 (w. custom DC cable) through PICO psu 12V ATX dc-dc feeding the MOBO
• FaradSuper3 (w. custom DC and AC cable) through 4pin PSU connector. 
• SSD separately powered through Uptone LPS 1

 

The Realtek USB (PEXUSB311AC3) card was a recent addition for testing. I installed in PCIe slot and later added 12v through SATA connector on the Realteck card (w. Uptone LPS 1.2). Second option did not change much (I think it is also powered through PCIe slot?), although it is a general improvement (3% like) I decided to return the card (and save for JCAT USB although that doesn't make sense $ wise in my setup, but who didn't say that before?

most troubleshooting is a process of elimination. What I'm wondering is if the LPS 1.2 is too low in current for error free operation powering MOBO and its devices.

Are you able to do a split cable from the Superfarad3 to  both 12V inputs?  The manufacturers site says the CPU consumes 10W but I can't find specs for the MOBO.

However the Superfarad3 should have far more wattage available than the LPS 1.2, even after subtracting the CPU demand. If the problems stays, that would rule

out a PS transient current draw issue.

[davide256]

2 hours ago, Johnseye said:

 

The data is the data.  Nothing gets added, although some people will tell you ripping on a quiet PC sounds better than on a noisy one. YMMV.

 

There are a lot of variables that can impact SQ.  Noise is one.  Jitter is another.  Latency is another.  Reducing all of those to as little as possible has been a goal of digital music reproduction.

 

I don't know how the quality of RAM impacts SQ but probably by one of the things I mentioned.  I don't know enough about it personally.

 

If the network is on it can create unwanted noise.

I tend to think of this as system electrical noise that causes the USB out electronics to behave less well and which can be transmitted to the downstream USB receiver causing similar “stressed” performance. Or does your experience point in a different direction?

[davide256]

22 minutes ago, di-fi said:

Nenon, I was referring to your example that NanoSword quoted (below). In that context I was talking about good or bad things added, I should have said what good or bad the network does. I am as fascinated by this as you are. I am just trying to breakdown in laymen's terms what exactly is happening in my Euphony Stylus in the cache and in the buffer and why and when - as in NanoSwords example - there is a difference. I do not expect that all things (can) to be explained, but I know that you can hear them. Thanks again for your patience.

 

quote: I have been comparing two identical Buffalo switches in the past couple of days - one with the stock clock and one with upgraded clock (PinkFaun ultraOCXO). You can do that same test in my system, and although you disconnect the cable while playing, the buffered tracks using the switch with the upgraded clock sound much better. In other words, even if you disconnect the network cable, whatever good or bad things your network does is already embedded in the buffered track somehow. 

An annoying phenomena to explain. One would think that the data buffered is just a file with CRC error check, that it should be identical. Need to understand better what’s actually buffered... is the file actually processed, no longer in its original format

[davide256]

Deleted

 

[davide256]

10 hours ago, Exocer said:

Here is the explanation:

 

"The fix tries to prolong the time during which player switches from one song to the next as I expect that DAC does not handle that fast change very well. This is done by preventing automatic loading of the next track into the player during play of the first one. When you switch between songs back and forth it defeats the purpose because then both tracks are downloaded and buffered and change happens much more quickly."

by definition there is no rate change in gapless. This sounds like bad approach. The normal way to handle this is a DAC delay parameter. 50ms delay is common value

and not that noticeable.

[davide256]

55 minutes ago, di-fi said:

Both Exocer and me changed dac delay through the Euphony menu until 1000 ms. That did not change. 

I don't believe DAC delay is working properly with Euphony. I just did this experiment, put 4 tracks of different rate/bit depth in queue,

set DAC delay to 5000 msec and observed. What I saw

 

1. Song lengths were increased by 5 seconds, with 5 seconds silence at the end of each song

2. rate change occurred after the 5 seconds, exact start of next song

 

my understanding of DAC delay is that the delay/silence is required at beginning of song to avoid disrupting synchronization, doesn't help if

done before the rate change

[davide256]

3 minutes ago, di-fi said:

I just put 6 tracks of 2 different bit depths (16 and 24 bits alternating ) in queue. DAC delay 5000. Euphony Stylus plays them without distortion (problem solved?). My DAC delay set to 5000ms now, is not working (I get 0ms), maybe due to the fix?

 

This is the fix applied:

quote- Since I think the cause is a very fast change from song to song due to gapless playback I turned it off so the song is buffered only when the previous song finishes. This is the only thing I can do. If it helps, then this is it - there is no way to have both - this fix and gapless playback.- end quote

 

My understanding with this fix as described, besides eliminating distortion,  I also loose buffering of whole queue , ''buffered only when previous song finishes''. But in my experiment Euphony Stylus will still buffer to queue (blue icons besides tracks). Which is great of course. I have to test further to confirm the gapless disappeared. 

If a DAC is slow to resynch at rate changes it can go into "mumble mode" if it gets data before resynch has completed. Euphony's implementation of DAC delay

with buffered tracks is broken and should be fixed before anything else is tried... you should see the delay at the start of a song when resynch is occurring.

[davide256]

56 minutes ago, di-fi said:

Hi, I reported this message last Friday (see post in this thread) it popped up once (with v. 20210108) in my quest to have a distortion free version of Euphony. I’ve seen this message in the past when it had to do with Euphony having problems to read from SSD due to a SATA cable connection. Maybe a contact/hardware issue? You know you can send logs to Euphony so they can troubleshoot?

@Anwar  The Op is using an iFi micro USB hub for external drive connectivity + DAC connection and a fast NVMe drive for OS. Probably should go direct

to NUC with external drive to see if problem goes away, could be hardware timing issue between NUC and external drive with the hub in the drive access path

[davide256]

albums per row -Just learned something thats helping me, adding since I don't find in the thread. Using a widescreen with Windows browser I was getting

a 3x2 album cover display from Euphony GUI which wasn't useful  for album browsing.  if you  click repeatedly at the bottom of the screen just below

the last album row you can change albums per row, I was able to bump up to 7

[davide256]

5 hours ago, KATARRAZZI said:

Can anyone here using a USB to AES / SPDIF adapter behind the Euphony server?  I am considering buying a Berkeley Alpha USB to further enhance the input quality of my DAC

 

I'm using a Denafrips Iris for USB to SPDIF, worlds better than the Schiit Unison/Eitr and Gustard U12 DDC's I've had in the past.  It has AES and pretty much any other alternative

digital out you could need, uses FEMTO clocks for reclocking integrity. If you have the cash, they offer 2 better models, Hermes and Gaia

 

https://www.denafrips.com/ddc

[davide256]

14 hours ago, di-fi said:

This used to work for me but since a few months ago when I reduce to 100 Mbps Euphony Stylus (ram root) becomes too slow and randomly unresponsive. Building thumbnails and buffering takes forever. So I have to leave at 1000 Mbps and unfortunately skip the benefit of less noise.
I’m playing local files from a NAS (up to 700-800 Mb size per album max.) and my setup is very low power compared to most users here, so YMMV.

The latest version does appear to have issues, slower, unresponsive at times to button clicks and gives a sqllite error if you click buffer when Stylus is already buffering. I may go

back to prior version again.