HQ Player

[Miska]

4 minutes ago, jhwalker said:

 

As I said, that's what I hear in my system - the -xtr filters sound like someone has thrown a velvet curtain over my speakers (headphones, in this case) - which is lovely and warm and all that but feels like missing an awful lot of air and energy in the upper frequencies.

 

Or it could be removing some artifacts that I tend to like hard to tell without measurements to go by.

 

Well, it could be the age old time vs frequency domain thing. poly-sinc-short and poly-sinc-xtr are on the different ends (while being within "reasonable" bounds) and poly-sinc is somewhere in the middle. I could at some point add poly-sinc-long which would be somewhere between poly-sinc and poly-sinc-xtr, but I'm not sure it would make enough difference and at least I haven't yet figured out what would be the specs to work against.

 

[Dr Tone]

1 hour ago, Miska said:

 

Compared to what filter in HQPlayer? Of course I could make the CPU load much lower, if that would be the measurement stick...

 

 

I don't get any either. But I don't use Roon, I use HQPlayer standalone.

 

Ok changing tracks works fine in HQPlayer alone.  It seems to be more of an issue from Roon. I've started a thread on the roon forum.  Roon must be bring HQPlayer to a full stop between track changes or something?

 

[Miska]

4 minutes ago, Dr Tone said:

Ok changing tracks works fine in HQPlayer alone.  It seems to be more of an issue from Roon. I've started a thread on the roon forum.  Roon must be bring HQPlayer to a full stop between track changes or something?

 

IIRC, yes it does that. HQPlayer Embedded has a kind of workaround for that, but Desktop cannot easily do the same.

 

Generally for controlling applications it is better to add a new item to the playlist (all with queue-flag) and then instruct HQPlayer to proceed to the next track. Or alternatively just cut the current stream and HQPlayer will automatically proceed to the next playlist item. Instead of stop-clearplaylist-additem-play cycle. Queued items are automatically removed from the playlist once played.

[57gold]

Figured out "problem" with poly-sinc-xtr and DSD128 on my system...

 

Clue was the red ring round the volume control...have had to lower output by 8db.  Using -2 version as straight version uses a bunch of CPU on 2.6 quad core Mini.

[Dr Tone]

1 hour ago, Miska said:

 

IIRC, yes it does that. HQPlayer Embedded has a kind of workaround for that, but Desktop cannot easily do the same.

 

Generally for controlling applications it is better to add a new item to the playlist (all with queue-flag) and then instruct HQPlayer to proceed to the next track. Or alternatively just cut the current stream and HQPlayer will automatically proceed to the next playlist item. Instead of stop-clearplaylist-additem-play cycle. Queued items are automatically removed from the playlist once played.

 

I can get around it by hitting pause first in Roon and then selecting the next track track to play or moving the playback location in a track.  So I have to think that that Roon could possibly do the same by default?

 

[zoltan]

8 hours ago, jhwalker said:

 

I did some careful listening this morning and discovered (in my system, at least) poly-sinc-xtr drastically cuts treble content - which does sound more "analog" and "warm / smooth"  but not sure why I'd want to cut off all the high frequencies in my listening.

3

I totally agree with this observation. I've done tests now p/s/xtr vs. p/s/shrt vs. p/s and p/s shrt is the most detailed and realistic at hight frequencies. At the same time also more tiring as some commented. As it should be. I never understood when people say "you can listen to it all day". Listening to live music all day (I'm not talking about rock festivals but actually sitting down in a concert hall or a jazz club), you'll get very very tired, probably don't want to listen to anything after 3 hours. It's a lot of information to process, just like DSD upsampling in HQPlayer. Heavy on the CPU (our brains). I can listen to my car stereo all day. But it's crap. I can't listen to more than two albums on my main rig. Too much to process...

[AudioAnt]

Hi, need some troubleshooting help.

 

Recently I upgraded from 3.16 to 3.16.1 and NAA from 3.5.0 to 3.5.1. Now I have strange behavior since then with one of my dacs.

With NAA 3.5.1 I cannot have sound at all. Strange message "push to queue timed out!". With 3.5.0 all fine!

Tried to downgrade to 3.16, but the NAA 3.5.1. behavior maintains itself and same message within NAA. It's a windows NAA with ASIO driver.

 

I've tried direct also from PC (windows also) to dac and got some strange behavior also. If I have the default to SDM first time I click a music to play, it doesn't! Second time i click on another music it starts to play, if on the same music it doesn't. After that if I click again on the first clicked song it plays!

 

I think this message "clHQPlayerEngine::Execute(): push to FIFO failed" from the log is the one when I click the first time in the first chosen track.

 

Settings on PC:

 

Any help appreciated.

 

Thanks

 

 

 

 

 

 

 

 

 

 

 

 

 

[Miska]

7 hours ago, AudioAnt said:

Recently I upgraded from 3.16 to 3.16.1 and NAA from 3.5.0 to 3.5.1. Now I have strange behavior since then with one of my dacs.

With NAA 3.5.1 I cannot have sound at all. Strange message "push to queue timed out!". With 3.5.0 all fine!

Tried to downgrade to 3.16, but the NAA 3.5.1. behavior maintains itself and same message within NAA. It's a windows NAA with ASIO driver.

 

Can you email me full log from the NAA side? There are very little changes between 3.5.0 and 3.5.1 NAA, mostly behavior in error cases is changed.

[Hazard]

19 hours ago, lmitche said:

Many thanks to Ted, Zorntel, Jussi and everyone else for contributing answers here on the limits of polysinc-xtr and DSD512 hardware.  I'll going to play around for with the new graphics card for another day or two and then likely return it.

 

Ted, your clarification on PCM is helpful.

 

Zorntel, thanks for testing with your son's graphics card.  May he never know you stole it for a few hours!

 

Jussi, thanks for the explanation of polysic-xtr limits and your monster machine.

 

If anyone has a i7-7700k and a Z270 motherboard, it would be useful to know if polysic-xtr pcm upsampling to dsd512 is possible with a GPU.

 

Thanks again. Larry

Hi Larry, I have O7-7700K and Z270 mobo.  I upsample to DSD512 to my diy DAC (based on DIYinHK usb boaed).

 

With -xrr filter I can upsample DSD64 to DSD512 with no problem.  CPU utilisation is around 25 to 30%.

 

However I get stuttering when I play PCM with this filter.  CPU utilisation is only 50% so I dont understand why it is stuttering.  My video card is an ancient GT9600 so CUDA offload is probably doing nothing.

 

PCM to DSD512 works fine with -xtr-2s filter.  CPU utilisation is around 30%.  I just spent $1200 to build this machine and honestly, everything sound great using -xtr or -xtr-2s filter.  I am not going to spend abymore money chasing the last bit of performance.

 

Oh and as a general comment about Daphile (not in response to your post).  I tried Daphile.  For about 5 min.  Was clearly inferior to HQP in my opinion, in my system.

[Miska]

2 minutes ago, Hazard said:

However I get stuttering when I play PCM with this filter.  CPU utilisation is only 50% so I dont understand why it is stuttering.

 

If you look at  core loads, it may explain. 7700K has HyperThreading, so on total load figures 50% means all real cores fully loaded (8 virtual cores of 4 fully loaded). In per-core load you may see that every second of these 8 cores is close to 100%.

[blue2]

2 minutes ago, Hazard said:

My video card is an ancient GT9600 so CUDA offload is probably doing nothing.

Might be doing less than nothing  I believe someone else had this problem where CUDA offload to low power GPU was worse than no offload.

[blue2]

10 hours ago, zoltan said:

I totally agree with this observation. I've done tests now p/s/xtr vs. p/s/shrt vs. p/s and p/s shrt is the most detailed and realistic at hight frequencies. At the same time also more tiring as some commented...

 

13 hours ago, jhwalker said:

 

As I said, that's what I hear in my system - the -xtr filters sound like someone has thrown a velvet curtain over my speakers (headphones, in this case) - which is lovely and warm and all that but feels like missing an awful lot of air and energy in the upper frequencies...

Can you guys give us any listening notes i.e. point to any particular tracks where you experience this? Or is it just a general impression?

[Miska]

8 minutes ago, blue2 said:

Might be doing less than nothing  I believe someone else had this problem where CUDA offload to low power GPU was worse than no offload.

 

Yes, if the GPU is slower than CPU, then CPU just ends up sitting and waiting for GPU to complete it's work. If it takes longer to do the work than playing the results out, it means that it's constantly falling behind delivery deadlines -> stutter...

[semente]

Hi Miska,

 

Do you think this Arcam network player would work with HQPlayer?

 

Playing from the network

(...)

Network music playback may also be controlled from

your computer, phone or mobile device using your

favourite UPnP control point application. 

 

http://www.arcam.co.uk/ugc/tor/cds27/User manual/DISPLAY_CDS27_MANUAL_SH269E-F-D-NL-ES-R-IT-SC_Issue2.pdf

 

R

[lmitche]

57 minutes ago, Hazard said:

Hi Larry, I have O7-7700K and Z270 mobo.  I upsample to DSD512 to my diy DAC (based on DIYinHK usb boaed).

 

With -xrr filter I can upsample DSD64 to DSD512 with no problem.  CPU utilisation is around 25 to 30%.

 

However I get stuttering when I play PCM with this filter.  CPU utilisation is only 50% so I dont understand why it is stuttering.  My video card is an ancient GT9600 so CUDA offload is probably doing nothing.

 

PCM to DSD512 works fine with -xtr-2s filter.  CPU utilisation is around 30%.  I just spent $1200 to build this machine and honestly, everything sound great using -xtr or -xtr-2s filter.  I am not going to spend abymore money chasing the last bit of performance.

 

Oh and as a general comment about Daphile (not in response to your post).  I tried Daphile.  For about 5 min.  Was clearly inferior to HQP in my opinion, in my system.

Hazard, thanks for the information on the 7700k and upsampling.  Since posting my question, another 7700k owner confirmed that his rig will also not upsample pcm to dsd512 with the polysinc-xtr filter.  We have two data points which is useful. Thanks again!

[Hazard]

3 hours ago, Miska said:

 

If you look at  core loads, it may explain. 7700K has HyperThreading, so on total load figures 50% means all real cores fully loaded (8 virtual cores of 4 fully loaded). In per-core load you may see that every second of these 8 cores is close to 100%.

Thanks Miska.  Sorry to ask dumb question, but is it possible to see load on individual cores in Ubuntu OS?  Task Manager shows total CPU load, but I dont know how to display load on inividual cores.

 

 

[ted_b]

Happy Easter to everyone who celebrates this Holiday, and Happy Spring to all else.

 

Jussi, a question that hasn't been raised in awhile, and I;m asking cuz I'm not sure of your answer:

Is there any reason to have HyperThreading turned off, if it's already on..or turned on, if it's already off?  I realize it can confuse the reading of actual cores in Resource monitoring, but I also know that you once said you leave it on.  AO's Phil recommends turning it off (for sq) but there we are.  Thx.

[miguelito]

17 hours ago, 57gold said:

Figured out "problem" with poly-sinc-xtr and DSD128 on my system...

 

Clue was the red ring round the volume control...have had to lower output by 8db.  Using -2 version as straight version uses a bunch of CPU on 2.6 quad core Mini.

I don't understand. Are you saying CPU load is lower with a volume reduction?

[jhwalker]

5 hours ago, blue2 said:

 

Can you guys give us any listening notes i.e. point to any particular tracks where you experience this? Or is it just a general impression?

 

For me, it was all tracks.  

 

When -xtr came out (a month or two ago), I switched to it, thinking, "it's new, it must be the best"   and really hadn't listened critically / comparatively until a few days ago, when Roon released their build 216 that includes DSD > DSD resampling, ability to select modulators, etc.  To my surprise, I found the -xtr filter substantially more "muffled" than any of the Roon filters, or any of the other poly-sinc filters variants I tried in HQPlayer, with pretty much any music I played.

 

As others have pointed out, perhaps the others filters are allowing some digital splash / hash that the -xtr filter is not, and maybe I like that sound!  There's no accounting for taste . . . 

 

In any case, I'm going to give it another try with both PCM (to 384k) and DSD (to DSD256, my main DAC doesn't support 512).  It certainly sounds lovely (warm, syrupy) if you're not attuned to what you (may be) missing.

[57gold]

1 hour ago, miguelito said:

I don't understand. Are you saying CPU load is lower with a volume reduction?

 

No, I would not know if that is the case.

 

But, on earlier post, I tried the poly-sinc-xtr at CD rip PCM to DSD128 and after a few seconds of play got hash and the red ring on the V control panel.  I scanned through this thread and found Miska's advice to lower the output to fix overload...and that worked.  

 

The -2 version uses significantly less CPU...so I have been using/liking its sound.

[Miska]

5 hours ago, ted_b said:

Jussi, a question that hasn't been raised in awhile, and I;m asking cuz I'm not sure of your answer:

Is there any reason to have HyperThreading turned off, if it's already on..or turned on, if it's already off?  I realize it can confuse the reading of actual cores in Resource monitoring, but I also know that you once said you leave it on.  AO's Phil recommends turning it off (for sq) but there we are.  Thx.

 

At least I'm leaving it on. IMO, it doesn't hurt anything and helps a little in cases when there are lot of processes/threads running on the OS.  It makes context switching between those processes/threads faster.

 

It just needs to be taken into account when looking at load figures, to realize that half of the cores are virtual and not real ones.

[Miska]

6 hours ago, Hazard said:

Thanks Miska.  Sorry to ask dumb question, but is it possible to see load on individual cores in Ubuntu OS?  Task Manager shows total CPU load, but I dont know how to display load on inividual cores.

 

"gkrellm" is utility that can be always kept visible on the desktop and shows per core load figures and can show lot of other info too without taking much space on the display.

 

Then different desktop environments have some additional utilities to show loads in various ways.

[brother love]

So I gave up trying to achieve DSD256 with Intel i5-4670k PC Ubuntu 16.04 server with NAA Windows 10 mini PC connected to Gustard X20u w/ XMOS driver. Could only get max PCM 384k w/ SDM None & DSD128 w/ SDM DoP (documented in my post this past Friday).

 

Dual booted Ubuntu 16.04 on NAA,fought with Intel mini PC Bios to get Grub screen for Ubuntu priority, installed networkaudiod 3.5.1-35 amd64 & couldn't launch so I installed GDebi package installer 0.9.5.7 all.deb. Opened networkaudiod w/ GDebi and voila ... DSD256 !

 

After cycling thru things to make sure all was well (it was) & enjoying some music for a bit, I broke away for other matters. Came back, fired everything up & I get ...

 

"Software Index is broken ..please check for broken packages with synaptic.". I followed up with 'sudo apt-get update' and 'sudo apt-get installed 'f' & I get:

 

"The package networkaudiod needs to be reinstalled, but I can't find an archive for it."

 

I moved networkaudiod to trash, removed from Apps, installed again, no luck.

 

Help !!!

 

[Miska]

5 hours ago, jhwalker said:

As others have pointed out, perhaps the others filters are allowing some digital splash / hash that the -xtr filter is not, and maybe I like that sound!  There's no accounting for taste . . . 

 

All the poly-sinc filters are technically good, just having weightings on different aspects and it is up to personal preferences and sensitivity to different aspects of sound.

 

I myself currently use poly-sinc-shrt-mp for pop/rock/etc and poly-sinc-xtr for classical music. But it's just my personal subjective preference, not any absolute truth...

 

[zorntel]

27 minutes ago, Miska said:

I myself currently use poly-sinc-shrt-mp for pop/rock/etc and poly-sinc-xtr for classical music. But it's just my personal subjective preference, not any absolute truth...

 

I completely agree at DSD512...with acoustic jazz also going to poly-sinc-xtr.  I'm using the 2s xtr however because of processing demands.

 

Robert