HQ Player

[dericchan1]

2 hours ago, dericchan1 said:

I actually never tested dsd512 (used to work with ASDM512 +fs but honestly I hardly ever use it anyway) on either beta versions. I will give it a try and report back

 

thanks

 

Deric

Sorry Miska, my bad, didn’t realize you were referring to the EC modulators, which never worked for me at dsd512. I gave it a try anyway, @dsd512 asdm5ec, ext2, the stuttering actually improved from the past, it only stuttered every 3-4 sec as opposed to every 1-2 sec with 4.15 😂 

[Fleecyears]

Thanks Miska.

 

Once I removed the UpTone EtherRegen device, I'm now able to play DSD256 with poly-sinc-mp & ASDM7EC without any dropouts.

 

It's actually quite disappointing, as the EtherRegen added a little bit of gestalt, which is now no longer present. The Soundstage has somewhat collapsed and the music has taken a step backwards in terms of the organic presentation there previously.

 

Anyone know how to get round this limitation, as EtherRegen may just end up being an expensive paperweight now? :-(

 

Thanks. 

 

[jamesg11]

20 minutes ago, Fleecyears said:

Anyone know how to get round this limitation, as EtherRegen may just end up being an expensive paperweight now?

See what the Uptone people have to say ...

 

Not a problem for me, thus far.

[Fleecyears]

The drop outs only occur when using DSD256, ADDM7EC and poly-sinc-mp. As Miska stated, the UpTone device is likely maxing out it's available bandwidth ~ 400m/bits

[Fleecyears]

Still there has to be some other settings I can check, like the Asus Ryzen CPU BIOS, NVIDIA Graphics Driver, Ubuntu OS, etc…..

Just not sure what I need to be looking for 

[LoryWiv]

8 hours ago, bogi said:

Would it be hard to to switch between beta1 and beta2 behavior within one executable?
Beta1 for lower performance HW, beta2 for higher?

Is it clear that beta 1 is better for lower performing HW and beta 2 for better HW? I ask because I am on an older i7 now but expecting my i9 12900 rig next week. I'll tru beta 1 now and see where things stand next week...

[ted_b]

Well, I found out that, through my mobile VNC program, I can GUI through to uninstall HQP and then reinstall it in a matter of seconds (like 30)...all from my listening position in my lower level audio room (my HQP server is upstairs, as is my Windows office desktop machine where I do any maintenance)  So it has made the task of comparing beta1 to beta2 fairly easy and straightforward..  And yet this comparison stopped dead in its tracks when I was in the midst of playing the DSD version of "Workin On A Building" from Cowboy Junkies Trinity Sessions.  I was acutely aware of the warm and enveloping lines from Anton's heavy bass work (having just completed the act of folding in a pair of Rythmik subs to my 2 channel system).  This was with beta2.  When I quickly uninstalled it and instead installed beta1 the playback was nothing even close.  The bass became one-note and not at all warm, with overall less impact/gain throughout.  30 seconds later I was back to beta2 and all was well.  My comparison unscientifically stopped.  I'm a beta2 believer.  😎  No need to go further.  This was DSD64 to DSD256 with the ASDM7ECV2 modulator.

[LoryWiv]

Like @ted_bI just tried both beta versions and overall prefer sound of beta 2 and to my pleasant surprise both versions play ASDM5EC 256 without stuttering. I could not achieve this with 4.15.3. However, cpu utilization is  almost double on beta 2 versus beta 1, 40% range...so I'm amazed there are no stutters.  Will stick with beta 1 for now to keep resource use / temps in lower range, but new i9 12900 en route and I look forward to revisiting beta 2 (or whatever else @Miskahas in store for us) soon!

[MasterWarzombie]

it starts to get complicated with different versions. we have beta with different versions now.
people will spend more time testing than actually listening to the music .....

[dericchan1]

15 minutes ago, MasterWarzombie said:

it starts to get complicated with different versions. we have beta with different versions now.
people will spend more time testing than actually listening to the music .....

I have to disagree on this. The beta testing is a great way to make sure customer satisfaction in upcoming releases. These betas provide night and day improvements over the previous full version for a lot of us which would not be possible if we never communicated our concerns and providing proper feedbacks to Miska

 

Deric

[Superdad]

3 hours ago, Fleecyears said:

Anyone know how to get round this limitation, as EtherRegen may just end up being an expensive paperweight now?

As mentioned, you can try the built-in Ethernet port of your NAA machine rather than the JCAT card. We had one other client with a new JCAT network card who had similar trouble yet the EtherREGEN feeding his machine’s built-in port worked just fine.

 

And I’m curious if you experience the issue with HQ Player only—skipping Roon.

Also, have you experimented with buffer size settings in Roon?

 

I run DSD256 all the time and a good number of our users do DSD512, so something else is going on.

I know this may seem a strange question, but are you using shield-tied Ethernet cables? (those where the shield of the wire is tied to metal shells at both ends) Such have been proven to sometimes cause issues—that is why most of the high-end streamer manufacturers (Lumin, Aurender, dCS, Auralic, etc.) as well as Miska advise against use of shielded cables.

[dericchan1]

1 minute ago, Superdad said:

As mentioned, you can try the built-in Ethernet port of your NAA machine rather than the JCAT card. We had one other client with a new JCAT network card who had similar trouble yet the EtherREGEN feeding his machine’s built-in port worked just fine.

 

And I’m curious if you experience the issue with HQ Player only—skipping Roon.

Also, have you experimented with buffer size settings in Roon?

 

I run DSD256 all the time and a good number of our users do DSD512, so something else is going on.

I know this may seem a strange question, but are you using shield-tied Ethernet cables? (those where the shield of the wire is tied to metal shells at both ends) Such have been proven to sometimes cause issues—that is why most of the high-end streamer manufacturers (Lumin, Aurender, dCS, Auralic, etc.) as well as Miska advise against use of shielded cables.

Not a etherregen user but just like to point out that my Cisco 2960 switch only supports upto 100M and runs fine with dsd512 asdm 512 +fs and dsd256 ASDM7ECv2 connecting to a rpi4 NAA Ethernet port

 

i did take Miska’s advice to use cat6 utp cables (ones with plastic connector)

[Yviena]

@Miska Hmmm i still have the issue with either right, or left channel being silent when matrix pipeline is turned on, i think it happens when it's switching between 44.1 and 48khz rates as i can reliable trigger it if i playback a 44.1 file switch to 48khz, and back to 41 khz again, you may need to do it a couple of times.

 

edit: also seems that moving the volume knob a couple clicks up or down also fixes the lost sound in channel when it happens.

[audiodan]

Hi all. In my digital system (double PC with both using win server 2019 core mode) I've compared my resident server machine  to the Win server 2019 GUI mode and windows 10 21H1.

The final result was that win server 19 in core mode in the server is largely better sounding than others O.S.

Now, from 4.14 version of HQ player is not more possible to update newer version after 4.13.1 on a win server core mode PC.

Thinking about this, I've compared  core mode + 4.13.1 with Win server 19 GUI mode + 4.15.3 but core mode plus 4.13.1 remains largely better sounding, so it's really a pity to kill the best sounding O.S. for lack of support.

It will to hope a future support of future versions of HQP Server Core mode or is it become impossible?

It will be possible to add the suitable DLL or It exists a not resolvable structural problems ?

 

 

[Fredc]

I tried both beta versions as well on windows 11 desktop. Beta 1 uses 10-15% less watts than 4.15.3 and beta 2 only slightly less. For me, neither beta can run 512x48 7ECv2 without dropout. I previously can with 15.3 (on 11900) but suffer from temperature getting too high too quickly. Betas though allow me to use ext3 on Nx 256x48 (I use sinc-L on 1x) without dropouts which I can't do on 14.15.3.  I  feel beta1 is more to my taste may be because the power usage and hence cpu temperature is lower. The sound seems more natural.

[Von Brechts]

Hello Miska!

 

I love HQPlayer, and I really think aubout spending almost 240 € for it but...

...the last working (trial) version for me (without any dropouts) was HQPlayer Desktop 4.15.0-2 with ASDM7ECv2 44,1x256 and poly-sinc-gauss-long.

All other versions after (event the latest betas) won't be able to do more than 44x128 with the same modulator/filter settings no matter if I checked, unchecked, greyed out any combination of CUDA-Support and MultiCore DSP, if I use it with roon or standalone. (Alienware Laptop, Intel Core i7-10750H, 6 Cores/12 Thread, 32 GByte RAM, nVidia GTX 2070 8 GByte, RME ADI2-DAC FS with AKM 4493). Besides this I'm not able to combine any EC-Modulator with any Sinc-x Filter since 4.13.1...music simply won't start.

I also tried the latest HQPlayer on different OS ( Server 2019, Server 2022, Win10, Win 11, Win 11 Insider - always clean installed out of the box, no optimizations, just installed drivers and latest OS updates) with no changes.

Could it be your AlderLake optimizations made HQPlayer giving us dropouts?

 

I really hope you'll give us a new version like 4.15.0-2...(and maybe you could make EC/ECv2 work with Sinc-Mx again)

 

Regards...

[cesio]

7 hours ago, Yviena said:

@Miska Hmmm i still have the issue with either right, or left channel being silent when matrix pipeline is turned on, i think it happens when it's switching between 44.1 and 48khz rates as i can reliable trigger it if i playback a 44.1 file switch to 48khz, and back to 41 khz again, you may need to do it a couple of times.

 

edit: also seems that moving the volume knob a couple clicks up or down also fixes the lost sound in channel when it happens.

I have the same problem since versions 4.15... But 4.14.0 doesn't cause this issue fo me. So it would be nice if you can try it out.

[Miska]

3 hours ago, audiodan said:

It will to hope a future support of future versions of HQP Server Core mode or is it become impossible?

 

There has never been support for Server editions of Windows. And it is not recommended to use I/O bandwidth optimized OS for latency sensitive applications like HQPlayer. Windows desktop editions are optimized for such workloads.

 

If you want slimmed down OS, then HQPlayer OS is best option. You will never get any Windows version to such level.

 

[Miska]

Any AMD CPU users testing build 1 vs build 2?

 

[Luca72c]

Hello and sorry for being a bit late, i wanted to be sure reporting about beta 1 and beta 2 on my i9-10850K Win10 server (no GPU, ac oc to 5,2 ghz).

About performance, Beta 1 is someway similar to 4.15.1 in load, but it seems that equal load leads to minor dropouts, so i can push load further than 4.15.1 still avoiding dropouts. Beta 2 has much better spread load on cores: playing a test redbook PCM file (converting to DSD256 using ASDM7ECv2 and gauss xla), i have max. load 59% on 2 cores with beta 1 and max. load 39% on 2 cores with beta 2. Other cores load is very similar with both versions. So performance with beta 2 is much better than 4.15.1 and than beta 1 too, in my system. Beta 2 clear winner here.

About sound (sorry, i know this is strange, but i have to report what i hear), beta 1 sounds a bit "damped" when compared to beta 2 and to 4.15.1. Soundstage is less clear and more confused, dynamics seem to suffer too. With beta 2 sound is at its best like in 4.15.1, with very good soundstage and instruments body/placement and great dynamics.

So for me beta 2 is the way to go.

Thanks Miska, you did an incredible work in a flash! Chapeau... This is GREAT customer support!

[dericchan1]

Sounds like a tough call with supporters for both betas!!! 

[Zauurx]

With my i5-8400 and the 4.153, I can still play DSD256x48 with the ASDM5EC.
With version 4.154, beta 1 or 2... this is no longer possible 😔

 

Miska, you work for the founders ! 😛

[cesio]

11 hours ago, Yviena said:

@Miska Hmmm i still have the issue with either right, or left channel being silent when matrix pipeline is turned on, i think it happens when it's switching between 44.1 and 48khz rates as i can reliable trigger it if i playback a 44.1 file switch to 48khz, and back to 41 khz again, you may need to do it a couple of times.

 

edit: also seems that moving the volume knob a couple clicks up or down also fixes the lost sound in channel when it happens.

Looks like problem with Matrix pipeline has been solved with new version 4.1.5.4. b1 or b2. Haven't noticed anything unusual. My sincere thanks to Miska.

[semente]

Hi @Miska, I'm getting drop-outs with 4.15.3 doing ext3 on my Mac mini.

It works fine on 4.15.0 with the exact same settings.

What could be causing this?

[bogi]

2 hours ago, Zauurx said:

With my i5-8400 and the 4.153, I can still play DSD256x48 with the ASDM5EC.
With version 4.154, beta 1 or 2... this is no longer possible 😔

Did you switch to Multicore greyed?