Euphony OS w/Stylus player setup and issues thread

[Chopin75]

2 hours ago, ASRMichael said:

Hi, I see there has been 3/4 releases of HQ player since last Euphony update. Do we just need to wait or am I missing something?

Are you referring to when the next Euphony update or HQP update ?

I was told euphony is to update soon to fix some bugs. MY euphony keeps getting stuck when I change track and I need to restart app each time to resolve this. 

[Chopin75]

29 minutes ago, davide256 said:

If anyone else  has the Hans Zimmer sound track "the Last Samurai" and is using RAM mode, you may want to give it a listen in normal mode as well. A lot of sonic issues

with it went completely away when I switched from RAM mode to normal mode, tone color purity hugely improved for me.

If the track is too big and takes up too much RAM with 100% buffer the sound can be worse with Ramroot. BUT I have solved this by using Apacer ECC RAM. Before using it Ramroot was not that good. But perhaps also me using SATA instead of M.2 PCle could affect efficiency of RAMRoot, I don't know. You can try without 100% buffer on RAM root and see. 

[Chopin75]

59 minutes ago, austinpop said:

A couple of excellent features are coming to Euphony/Stylus. One that I have been beta testing is a UI option in the Queue view to buffer all tracks in the queue from the current track on. This makes a noticeable impact on SQ, as you can wait for all tracks to be in RAM before playing the list.

 

The other feature that looks promising is the ability to direct Stylus output to a UPnP renderer. I wish I had had this when I had the dCS Bartók review unit.

 

i gather the Euphony folk are busy working on the launch of a new product - their Essentia server, but no details are known yet.

hmm, I have enough RAM space issue with my current high Res GB tracks running Ramroot, as  a single track can be 1-2 GB. An entire DSD256 album can be 10-16 GB (the classical albums area > 60 min long), there is no way one can run Ramroot with 100% buffer of entire album unless one use like 64 GB RAM (actually I am planning to expand to to 32 GB for DSD512 already but still won't be enough!)  Would a more powerful CPU be needed to run 64 GB RAM ! 

[Chopin75]

3 minutes ago, davide256 said:

CD quality so under 650mb for the entire recording. Buffer/cache absence/presence did not change the symptoms/outcomes.

 

 

Could it be RAM quality, as I am now able to run Ramroot much better with Apacer ECc than with my Gskill, despite using large files

[Chopin75]

2 hours ago, Lukasluis said:

I have the same config for my headphone rig, squeezelite is a little bit better/clearer for music up to 24/96. Roon endpoint is much better for higher res and DSD format. It looks like downsampling is more detrimental to SQ than buffering in memory?

OK, please clarify your question. Why is it surprising downsampling  is worse? Isn't that expected. In general hi Res sounds better. However, upsampling a low res file may not improve things, which can depend on the HW, SW and the quality of the recording. 

I have a bunch of awful DSD256 (like the one I am listening now, which I end up using to test how great a system can be in minimizing the poor quality). A great recording in 16 bit can sound much better than an awful DSD256.  

[Chopin75]

3 hours ago, Middy said:

Gents, I mentioned this on the big thread but got a little lost in the noise.(electrical)

 

Pursuing better power works  but removing/ reducing awitching noise internally really suprised me with adding some IFIi Silencers.  It felt like more of a jump going from windows to Euphony/ fanless. Real detail and depth...quite suprising . Nice re-listening to my music again, 

Mine is a single NUC celeron with now 3 silencers, so I dont know how effective they would be on a well supplied ATX MB/ Paul Hynes / Uptone/ HDplex set up. Or how affective the other USB PCI filters are  but ther inverse filtering part of the i Silencer worked really well on my 5V NUC bus that had a marginal effect on my other transports. 

 

I am only pushing because as they have returns if you try a couple cheap and want others to benefit as much as my set up has.

2 maybe the magic number i bought the third the next day so yet to decide...

 

Good luck

Dave

 

 

 

I have only one Islencer on a USB port that is unused, close to my USB bridge. I have not added more isilencers. The effect is mild only.  I dislike putting the silencer directly in the path of the USB cable (for audio data)  to the DAC. It seems to mask or change the tone somewhat without true benefit.  I suppose adding more silencers to my other unused USB ports may help more. But I don't really want to buy more to try though I know I can return them (but need to pay shipping and possibly restocking fee) 

[Chopin75]

Euphony uses some low tech fanless computer eg the zotac. Which are cheap and only low power CPU chip. Good budget but may.not be adequate for stronger processing power with huge music files. Or do upsampling etc. 

[Chopin75]

3 hours ago, jcn3 said:

 

i'm not clear on your comment.  you don't have to buy hardware from euphony -- you can buy the software and use whatever hardware you want.

Sorry, u are correct. But u can check.out what hardware they use and get an idea what it get yourself. Yes, u can easily get the euphony SW and load onto a hard drive and put into any HW u want. 

[Chopin75]

On 1/15/2020 at 12:25 PM, TheAttorney said:

 

Where is this option located? I can't find it anywhere, but I seem to have it set anyway because there is now no buffering when skipping to next track, and most of my queued items are marked blue. I say "most" because I currently have over 100 tracks in my play queue and only 4GB RAM, which is shown to be over 80% used (and marked red, implying a limit is being reached). Despite this low memory, Stylus keeps playing and adjusting what's marked Blue, so I imagine it's dynamically adjusting what's buffered when it gets to RAM limits.

 

EDIT: I've found it now, at the top of the queue. It displayes that 100 tracks are buffered, out of 117 currently in queue. But there is still a small bit of disc activity when skipping tracks.

 

One small enhancement (that I had requested), that's not listed in the change log, is that the current track playing is now always in display whenever you get back to the play queue. Previously, returning to the play queue always went to the start of the queue, so you then had to scroll down to find the track playing - significant when there's a long queue spread over several pages. This now works well.

 

Click on artist name opens artist's page on LastFM

I'm a bit confused about this one. I had asked for the ability to see all other albums by an artist when you are on the Now Playing screen of that artist (a roon-like feature). This has been implemented when you click on the artist name on the Now Playing page. In practice, this has been done by a filtered search, but you then have to go to library view to see that result. I'd rather you were taken directly to that result if that's possible.

 

Not sure how that relates to the above change log entry, or if that is a different feature that I haven't spotted yet.

 

II don't remember that the current track is now always in display, as like you I can have > 50 tracks on queue. I. just deleted the entire queue and now run only single album as I am testing the buffer queue mode. I am not sure if there is much sonic difference. I can only buffer limited tracks as I play DSD a lot and each album is like 9-11 GB!  Also realize I cannot click to a different track while in Buffer queue, like it would play but no sound is coming out. I then restart app and of course the queue is unbuffered.  Not sure if it is worth the fuss. 

[Chopin75]

4 hours ago, TheAttorney said:

 

When in queue buffer mode, I have no issues with clicking on a different track - it plays immediately. But I do have the much smaller redbook FLAC files.

 

I also don't hear much sonic advantages in queue buffering vs single track buffering. For those that do hear a difference, please elaborate what that is and does it only apply to the first few seconds of each track?

 

There appears to have been an improvement to the single track buffering as well:

Previously, the next track only loaded at the point I skipped to it, and very occasionally I would not hear the first split-second of that track.

Now it looks as if the next rack is always pre-loaded and it's the track-after-next that is buffered when the next track starts. So it's always one step ahead and that seems to work well - so far I've never missed the first split-second of the next track.

I am still unable to smoothly skip to a different track and sometimes it works but other times not. There would just be silence. The sonic difference is there. It seems there maybe be some improvements in entire track after comparing different albums. It brings the sound stage slightly forward, more focus and the distortion or fuzziness is less, a bit smoother. I think it is subtle and maybe only obvious with poorer recordings. I am not sure why it would improve despite using more RAM and also it should only affect the beginning or end of track but the sonic change appears to be constant. Does buffer queue make the system more stable? Less activity from stylus? 

[Chopin75]

Ok the sonic benefit of buffer queue seems to matter more when there is some more poorer quality source, e.g. when the AC mains is noisier. At least it seems so as now with a quieter evening the benefit with Buffer queue is not so clear. Maybe during the day  with the noise from PC server is more beneficial with Buffer queue. Now without buffer queue, it still sounds fine and may have a more naturalistic sound stage. I am confused now...! 

Do people use Buffer queue here ? 

[Chopin75]

1 hour ago, seatrope said:

Has anyone tried using the new StylusEP as a HQplayer "endpoint" for Roon Core yet? Again, no HQPlayer really involved, just a way to preserve Roon's library management and UI while getting the sonic advantage of Stylus.

 

I basically cannot give up Roon's UI mostly so that the wife can stick with what she's used to, but would love to get some of Euphony's sonic advantages. I'm running Audiolinux now and would pay for a license if this StylusEP/HQplayer endpoint feature works (again without HQPlayer, just with Roon). I have a lot of DSD128 and DSD256 albums that I listen to a lot - the StylusEP with Squeezebox emulation doesn't work for me as that's only limited to DSD64.

 

Thanks all!

I could be wrong but given the larger files of DSD128, & 256, I think it is best to use Stylus player itself without Roon or any other SW involved. The timing and large memory of the files that need to be processed  means that staying in a local single box in single SW may work best but I have no experience other than using Stylus player itself. I have files as large as 2GB, so I doubt it can be moved around and travel via ethernet etc...... fast enough without loss of information or timing issue. But perhaps others have tired.

Another thing, is it possible for you to load the file to Ecache directly from Roon? I am not sure if ECache works with Roon  

[Chopin75]

On 1/25/2020 at 4:20 AM, TheAttorney said:

 

When in queue buffer mode, I have no issues with clicking on a different track - it plays immediately. But I do have the much smaller redbook FLAC files.

 

I also don't hear much sonic advantages in queue buffering vs single track buffering. For those that do hear a difference, please elaborate what that is and does it only apply to the first few seconds of each track?

 

There appears to have been an improvement to the single track buffering as well:

Previously, the next track only loaded at the point I skipped to it, and very occasionally I would not hear the first split-second of that track.

Now it looks as if the next rack is always pre-loaded and it's the track-after-next that is buffered when the next track starts. So it's always one step ahead and that seems to work well - so far I've never missed the first split-second of the next track.

If I just load and buffer single track in the queue there is no difference between this and 100% of a single track when there is only 1 track in the queue. perhaps only load 1 track to the queue at a time a and use 100% buffer then? But this won't be practical especially if you have a very short track so you need to load every few min. 

 

Also loading a number of huge tracks into buffer queue has caused my euphony to shut down, it gets stuck into perpetual high RAM use. This is probably because I was trying to load to buffer and play a large track at the same time. I still cannot switch track properly with Buffer Queue. I would have to let each track play in order, like listening to vinyl, is that normal ? 

[Chopin75]

Am I nuts or hallucinating but it appears that buffer queue effects are minimal if not many tracks are in the queue, which makes sense. However, I also tested that smaller number of tracks loaded in the queue (without buffer queue) sound better than if entire album is loaded. (mind you I am testing using huge files DSD256 so an album is like 8-10GB).  A single track only on the queue seems to be best. Now effect is subtle so it may not be reproducible. So maybe keeping the queue as short as possible will help. I have not compared between a buffer queue of a long list of files vs short list yet. 

[Chopin75]

29 minutes ago, mentt said:

I have compared Ephony with Volumio on the same HW and I like sound from Volumio more...

Interesting. Is volumio including it's own OS or u need a separate OS to run on.

[Chopin75]

38 minutes ago, RickyV said:

New Sw update 20200131.

This was released days ago. I still get stuck from time to time with Ramroot + buffer Queue. 

[Chopin75]

4 hours ago, adamaley said:

So, essentially, besides speed and a minimum 8GB space, there are no other factors to consider? 

It also depends if you want to store music in the drive or use Ecache to load the music into the drive. If you want to store music files on Euphony then you need more GB, and there is also space reserved for Ecache, too and the larger the drive the more space there is for it. Using Ecache  means you just load whatever music you want to play at that time into the drive, which can be done remotely via file manager, or via attached external/internal drive, or (I think also via network drive). Since any attached drive can cause noise, remotely loading to it is best method, and for me works best. 

[Chopin75]

31 minutes ago, Anwar said:

With the latest Euphony updates, I am getting better sound quality without ramroot - the music flows more naturally.  With ramroom, music sounds more compressed.

 

Also, with ramroot, the album artwork display is okay sometimes but not okay most of the times.  No such problem when ramroot is disabled.

 

The buffering of an album/folder runs into problem when total file size is greater than 8GB.  I have some DSD256 albums purchased from Blue Coast where Euphony runs out of memory after reading 8 tracks into buffer, so it did not read the remaining tracks.  When I tried to play the 9th track, Euphony shows file not found error.

 

My setup: NUC7i5BNH with 16GB RAM, Euphony O/S in M.2 Samsung EVO 970, music in 2TB external Samsung EVO 860 SSD connected thru iFi Micro iUSB3 which is also used as USB reclocker between NUC and Marantz SA-10 USB audio input.  Music playback with Euphony Stylus only.

I had issue with Ramroot SQwhen I was using G Skill RAM 16 GB, like you. said reg RAM is better,so I updated to Apacer and there is improvement in SQ esp in RAMroot such that it is better than reg RAM. I have also run out of RAM with DSD256. I ended up increasing to 32 GB. I think if you use 16 GB things would improve a lot. To run ramRoot with DSD256, it needs at least 16 GB, depending on how long a track is 

[Chopin75]

1 minute ago, Lukasluis said:

Hi, May I know which Apacer model are you using and links where to puchase please? Looking for my NUC8i7BNH. Thanks.

I got mine that is not available for commercial sale. But in general any Apacer ECC works best. If your NUC does not support Apacer ECC you can get non-ECC. I can't remember off my had what model mine is. Somewhere there is discussion of this ether on this thread or the huge original thread "A novel way to..." 

[Chopin75]

19 hours ago, TheAttorney said:

 

Actually, there are two separate options:

1. Buffer next track (the original option)

2. Buffer the whole queue of tracks to be played (the recently introduced further option).

 

If you disable both options, then there is no track buffering, so 4GB RAM should be enough to play any sized file in theory (I haven't ever needed to test this in practice).

 

However, I find it strange that @organics1 has gone to the trouble of getting huge DSD files (presumably for SQ reasons), yet continues to struggle with a paltry 4GB of RAM. And thereby misses out on a variety of SQ improvement options. IMO, attention to server improvements (not just RAM) is more important to ultimate SQ than the file type of the music tracks. I am frequently astonished by what can be achieved with my redbook FLAC files (a lot does also depend on the DAC design for this PCM/DSD question).

So to run DSD256 or DSD512, both I have succeeded,  yes, you do need large  RAM if you want to buffer entire track or album. Without 100% buffer 4GB is usually enough depending on how long a track is (30 min can go up to 2 GB!) There is obvious sonic benefits with DSD256 compared with DSD64, though it may well be that some DSD256 recordings are just better recorded but I have also heard numerous horrible DSD256. So in the end it really depends on the quality of the recording itself. There are many wonderful 24/96 as well as 24/48 too.  

Also I find buffer queue helps but its benefit diminishes if you have only 1-3 tracks. The best way is to listen to as few tracks as possible that you put on the queue. I suppose if you have many small tracks the buffer queue reduce the need for euphony to continue to work to upload and buffer the next track while you are listening. I am no sure if it is worth the trouble to expand the RAM just to accommodate Buffer queue. Certainly you should at least do 100% buffer of a single track. 

[Chopin75]

49 minutes ago, organics1 said:

It may be getting the cart before the horse as far as some obviously "advanced" super duper golden ear high tech audiophiles like you are concerned. But, as per usual in this hyper consumer market things are often advertised as an improvement over something else. Doesn't necessarily have to be so, but that's the way of the world. Why did the world rush to buy so many SUV's when they first began to flood the market, even though they were the most dangerous vehicles ever created, and obviously couldn't leap tall buildings and climb mountains as advertised? Friends and relatives of the deceased are probably still trying to figure that one out.

In the case of the DSD files, I bought a few and discovered that, indeed, they sounded better than the 196/24 files I had. They didn't always move heaven and earth, but when I could compare them to the 196 files I had of the same recordings I almost always hear an improvement. 

The other thing here is that people and their systems evolve. At least average people seem to. Nice to meet someone who got it right the first time. ALL HAIL!

One should also take into account if the recordings are originally made as DSD with minimal conversion to PCM during editing, and these are usually best. Files converted from PCM may or may not have advantages and it also depends on the DAC too

[Chopin75]

3 hours ago, mnuno10 said:

Hi,

 

I purchase a NUC8i5BEH, 1x16GB DDR4 2400mhz memory and and Optane 16GB. The goal is to use Euphony Player to stream Tidal to my Roksan K3 Dac.

Right now I'm on Euphony Trial and I have some doubts:

- I can't install Euphony in my Optane stick, is this because I'm on Euphony Trial? I have to purchase full version to install it in Optane? Or the problem is another? Note that on BIOS update isn't recognize as a Bootable drive, I don't know if it is normal, because it's the first time that I'm using an Optane stick.

- Euphony only recognizes 8GB of memory, but when I go to BIOS it says 16GB. I only have 1 stick of 16GB, it is some kind of bug in Euphony? Somebody with the same problem?

- I'm using my Mac to access Euphony through Chrome browser and "Euphony web app" is draining my battery like crazy. With a tab with "Euphony web app" open the "Activity Monitor" show Google Chrome process consuming lots of CPU as you can see in the attached picture. Somebody with the same problem?

I'm in the last build of Euphony with every update applied.

 

Thank you in advance.

My Mac also states the Euphony web is draining lots of power but my Macbook PRO is 2018, 16GB RAM. It still functions fine but at times may get hot though that only happens if I am actively loading stuff onto the euphony remotely using File manager (u can't do this with trial version) . You can use Dr. Cleaner to monitor the CPU usage and can optimize as needed when you are running out of RAM. 

[Chopin75]

45 minutes ago, organics1 said:

Bought Euphony 9 days ago, and for a couple of days I was ALL GRINS. But, songs started skipping. It was even skipping from album-to-album. Tech support looked at the logs and couldn't determine much, saying only there appeared to be a few instances of the computer losing touch with the DAC. It was suggested I might have a problem with cables. I didn't find anything. I've run Daphile and Audiophile Linux through the same system for months and never had a similar problem. Back to using Daphile right now, and nary a whimper or hiccup.

Any thoughts?

 

Other than that issue I just want to say that this Euphony OS beats everything else I've used. It does everything better. I'd even go so far as calling the sound SUMPTUOUS.

Up until now Audiophile Linux was king of the hill for me. BUT, not being a LINUXphile I just got tired of chasing its gremlins day after day. By comparison, Daphile has been rock solid and relatively problem free.

 

As I stated in an earlier post, I use a Partaker fanless mini with a Celeron 3150: 1.6Ghz, 4-core. RAM is 4GB. Despite some claims to the contrary, could it be that this little thing just is not up to the task of doing justice to Euphony? It's what I've been thinking, and I'm eyeing another fanless box utilizing the i7-8550U, which is 4-core, 1.8Ghz. It'll take up to 32GB of RAM. How important is base frequency? (BTW, not interested in buying a NUC.)

How big are your music files ? DSD256 can run into issues like you describe if the RAM is not adequate and you have a huge file and running 100% buffer. Try running without buffer and see if that works better. You can also check how much RAM is being occupied when running. - same button u use to check temp/CPU. 

[Chopin75]

Does anyone notice that if you disconnect the ethernet cable form the server/PC where the euphony is, the sound is improved? I am using single box so the entire Euphony/music files are all in the single box and ethernet is only used to control the Euphony remotely.  

Would wireless control be better or worse ? 

[Chopin75]

9 hours ago, Jeff Mann said:

Rajiv (austinpop),

 

I have a number of questions relating to what magnitude of improvement that I can expect if I make certain changes to my Euphony-based online streaming of Qobuz albums.

 

I have been an audiophile (who only listens to classical music and opera) for >50 years, using a "speaker-based + LP-based" system for the first 30 years and a "headphone-based  + CD-based" audio system for the past 20 years. I started to use an Euphony PTS music server to access Qobuz about 8 months ago, but I cannot get a sufficiently good sound quality when using the Euphony PTS unit, rather than a CD-player, as my music source.

 

My audio system is as follows-: CD player (either a Jays Audio CDT2Mk2 or a PS Audio PWT) => Wyred-4-Sound 2v2se 10th Anniversary Limited Edition DAC => Niimbus US4+ headphone amplifier => HiFiMan Susvara headphones. If I assign my CD-player-based audio system as having a sound quality of 10, then I can only obtain a sound quality of 7 when I use my Euphony PTS to stream Qobuz albums - even though I am using a Keces P3 LPS to power both my Euphony PTS unit and my W4S Recovery unit that is placed in the USB line between the music server and my DAC. As an alternative to an USB connection, I also have a Matrix X-SPDIF unit that enables me to connect to my DAC using an i2S connection, but that does not improve the sound quality compared to the USB connection.

 

My first question to you is what level of sound quality do you get using your full-monty computer-based audio system that incorporates an UpTone Audio EtherREGEN unit + Paul Haynes LPS + Mutec Ref-10 clocker - compared to using a state-of-art CD-player (which I presume you own)? If your CD-player-based audio system is rated at a sound quality level of 10, do you achieve a level of 10 (or greater than 10) using your computer-based system that uses Euphony software and an Euphony Stylus endpoint?

 

Going back to my personal Euphony PTS-based audio system that I rate at a sound quality level of 7 (relative to my CD-player as the music source), how much better can I make its sound quality (using a numerical scale) if I make the following changes.

 

The first change that I could make is to replace my Euphony PTS unit with an Euphony Summus unit (https://euphony-audio.com/summus) and according to that webpage Euphony claims the following advantages -" "Activation of Stylus audio player automatically assigns transport of music files to separate CPU cores and turns down high-powered and noisy i7 platform into completely silent audio transport ------ Furthermore, Euphony Summus is booted by default into so called “ramroot” mode (widely embraced by audiophiles worldwide as the best for computer audio), allowing all system files to operate from RAM. Total absence of disk activities combined with near noise free CPU, makes Summus music server a dream come true for computer audiophiles." I don't know if those potential advantages apply equally to Qobuz streaming or only to the playback of local files stored in the Euphony music server's drive. How much positive improvement in sound quality do you anticipate that I will achieve if I replace my Euphony PTS unit with an Euphony Summus unit in terms of numerical magnitude - in terms of moving my Euphony-based from a sound quality level of 7 in the direction of 10?

 

An alternative approach to improving the sound quality of my Euphony-based system is to make the following upstream changes.

 

Step 1 - purchase a EtherREGEN unit.

Step 2 - replace the Ether REGEN unit's stock power supply with a LPS unit.

Step 3 - add a Mutec Ref-1 clock.

 

What degree of sound quality improvement can I anticipate from each step (using the same numerical scale of magnitude that is used to quantify the sound quality improvement that I can anticipate if I replace my Euphony PTS unit with an Euphony Summus unit)?

 

Jeff.

OK, I do have a system of CD player (belt drive state of the art but old CEC TL-1X) so it does play CDs pretty damn good if the CDs are of high quality ($3000 resell price) . I also have  Mitchell Gyro Turntable + SME tone arm + dedicated phono stage (>$5000).  Despite not really having time to tune them up they are are of very high quality and do sound very good if not reference.

Regards to digital audio, I have found it difficult to beat CD player sound but it can be done. I use an Audio-GD R2R DAC for both the CD transport and my Euphony PC so this comparison is fair. 

I use BNC/RCA to connect to the DAC (The CD transport has no BNC out) 

USB to DAC from the PC.

My PC is mainly a AMD based PC with euphony in a single box, using SAMSUNG internal M2 drive (not the best option), I only play music files directly from file. (no streaming/network load or Roon)

Specific features of my PC - HDPlex base, with HDplex LPS, pink faun USB bridge out.

optical USB cable to DAC

Apacer ECC 32 GB RAM

Euphony OS - running RAM ROOT, buffer queue, 

 

There a few issues with your setup

1. Based on your Euphony PTS, - it does not have enough juice, I think last time I checked it is an intel 3 chip, nowadays we use i7, I can't remember the RAM, certainly no high grade RAM. It has no USB audio bridge that I am aware of,

 

2. If you are streaming from Qobuz, you may not get the best sound, this is not my experience so I could be wrong, but from other users who have compared streaming vs local files. Classical music also has huge files so it can make streaming more difficult (but don't quote me,I am not sure. You can play the same downloaded file from Qobuz vs streaming. Now perhaps there is a way to make streaming as good, but note the classical music has very complex music so any real data streaming theoretically is error prone and can affect classical music more than other genre. I am not sure if you can stream and load the entire file onto RAM before playing it, which may help.  

 

3. I am not sure how you set up your Euphony OS but you can explore all sorts of tweaks from expert setting though I have not done that so I am not sure if it would make a big difference. If you are like me who play classical music then the files can be large esp if you use DSD or DXD. U need buffer all tracks, best is buffer queue, and turn off internal volume adjustments, but you need large enough RAM, probably at least 8 GB or more.  High quality RAM makes a big difference. 

 

Sorry there is too much to discuss and not able to put all in detail. But bottom line is you should explore building your own PC or have someone does it for you so you can tweet around

 

For e.g you may want to do I2S --> I2S from PC ---> DAC, using Pinkfaun I2S bridge which can do up to 24/192 (but no DSD) which supposedly much sounds better than USB (even if you use the Matrix-->I2S)

 

Can you eventually beat CD player -- yes and no.   I manage to achieve a point where numerous DSD and PCM recordings clearly beat CD/SACD  but usually they are higher res like 24/96 or higher or DSD256. If I play only 16/44 files then maybe not. (note I do NOS mode) 

But upsampling of low res can help a lot. I would say overall it can beat CD player to 11-12, if CD = 10.   But if CD is of superb/great recordings, then it may be a 8-9 compared to CD = 10. It is difficult to compare, I have done same CD + master files of 24/96 and clearly the high RES files beat the CD but the CD is made in last few yrs and the quality has decreased a lot. (and I don't have the 16/44 files to compare) . Some CD recordings from 1990s-2010 are superb and so they just kick ass and beat the digital files (though again I don't usually have the same music files to compare).

Now if you don't have a great CD transport (<$1000)  then you can easily beat the CD sound. 

I won't go into vinyl, it is just a total different system with different pros/cons.