HQ Player

[LowOrbit]

31 minutes ago, luisma said:

Thank you @LowOrbit for posting your results. There are some problems with Ryzen in regards to full performance on every core

https://www.tomshardware.com/reviews/amd-ryzen-3000-turbo-boost-frequency-analysis,6253-4.html

 

IMO that should not prevent the CPU to deliver some good processing, Turbo for the intel is 4.9 vs 4.4 for the 3700X, Miska have said before and we know the speed per core have more influence that the actual number of cores although for the new modulators it seems the most cores the merrier.

 

I haven't been able to test, I'm in the process of getting a new Ryzen and the practical results of LowOrbit got me a little discouraged.

 

You are using Windows 10 correct @LowOrbit?

 

 

WIndows 10 - yes. 

 

I had an MSI X570 motherboard for the Ryzen cpu - the bios available from MSI did not have the latest AGESA microcode updates from AMD, so it may have been a factor. The 3700X was stable at 4.5ghz, could run any filter and the EC modulators but I could not get past fairly frequent short interruptions at DSD256. No issues at DSD128 with ASDM7EC and it did sound damn fine. With non-EC modulators I could run at DSD512 without issue (using a DSC1/Amanero lashup). 

 

The problem may have been motherboard specific, but I sold the Ryzen and mobo on and the 9700K is performing well without any overclocking or messing. 

[Rune]

1 hour ago, LowOrbit said:

The 3700X was stable at 4.5ghz

You probably only had on core at 4.5Ghz. If you had done an all core overclock to 4GHz it would most likely have worked.

[LowOrbit]

2 hours ago, Rune said:

You probably only had on core at 4.5Ghz. If you had done an all core overclock to 4GHz it would most likely have worked.

That may be the case but was not what Ryzen Master was showing. Manual set all cores. And believe me, I tried many configurations and combinations to try to isolate and offer best performance for the cores which deemed to load heaviest which I assumed to be the ones carrying the Modulator processes. 

 

Anyway, academic now. Not an experiment I will readily repeat.

[Yviena]

16 minutes ago, LowOrbit said:

That may be the case but was not what Ryzen Master was showing. Manual set all cores. And believe me, I tried many configurations and combinations to try to isolate and offer best performance for the cores which deemed to load heaviest which I assumed to be the ones carrying the Modulator processes. 

 

Anyway, academic now. Not an experiment I will readily repeat.

Asdm7EC works fine for me with 3700x no glitches here on 1.0.0.3ABB AGESA single target boost is a little lower 50--60mhz less but Multithead performance went up 100 points, this is with everything on auto meaning ram is at 2133mhz.

[PieterP]

4 hours ago, LowOrbit said:

WIndows 10 - yes. 

 

I had an MSI X570 motherboard for the Ryzen cpu - the bios available from MSI did not have the latest AGESA microcode updates from AMD, so it may have been a factor. The 3700X was stable at 4.5ghz, could run any filter and the EC modulators but I could not get past fairly frequent short interruptions at DSD256. No issues at DSD128 with ASDM7EC and it did sound damn fine. With non-EC modulators I could run at DSD512 without issue (using a DSC1/Amanero lashup). 

 

The problem may have been motherboard specific, but I sold the Ryzen and mobo on and the 9700K is performing well without any overclocking or messing. 

I am not the expert but I wonder whether the 3700x could boost to 4.5 ghz all core? Looking at Silicon Lottery only 21% of their binned 3700x's could reach 4.15 ghz all core ( if i interpret SL right). https://siliconlottery.com/pages/statistics

 

3700x results are at the bottom of their page. 

[k6davis]

2 hours ago, Yviena said:

Asdm7EC works fine for me with 3700x no glitches here on 1.0.0.3ABB AGESA single target boost is a little lower 50--60mhz less but Multithead performance went up 100 points, this is with everything on auto meaning ram is at 2133mhz.

 

This is with PCM 24-192 => ADSM7EC x DSD 256 (x 48) + poly-sinc-ext2   ?

[Yviena]

36 minutes ago, k6davis said:

 

This is with PCM 24-192 => ADSM7EC x DSD 256 (x 48) + poly-sinc-ext2   ?

Yes DSD256 with ASDM7EC works fine for me, i can run all filters excluding xtr as it's too demanding for the gpu.

just tried 96khz track upsampled to ASDM7EC to DSD256 x44.1 so load should be even higher as it's not integer rates/auto family, using ryzen master it shows the cores not even exceeding 3.6ghz while playing.

[ted_b]

Yvienna, what GPU are you using?  And mobo?  Asus ROG STRIX X570-F Agesa 1.0.0.3ABB?  Your posts would be great for the EC tips thread, thanks.

 

[k6davis]

33 minutes ago, Yviena said:

Yes DSD256 with ASDM7EC works fine for me, i can run all filters excluding xtr as it's too demanding for the gpu.

just tried 96khz track upsampled to ASDM7EC to DSD256 x44.1 so load should be even higher as it's not integer rates/auto family, using ryzen master it shows the cores not even exceeding 3.6ghz while playing.

 

Have you tried a 192kHz track upsampled to ASDM7EC DSD256?

[Yviena]

21 minutes ago, ted_b said:

Yvienna, what GPU are you using?  And mobo?  Asus ROG STRIX X570-F Agesa 1.0.0.3ABB?  Your posts would be great for the EC tips thread, thanks.

 

Yeah it's AGESA 1.0.0.3ABB gpu is a 1080ti, and the mobo is actually the Crosshair VI hero, 

4 minutes ago, k6davis said:

 

Have you tried a 192kHz track upsampled to ASDM7EC DSD256?

Just tried a 192khz/24 bit track i downloaded cpu usage from hqplayer is 31-32% so around 4% more usage than the 96khz seems fine so far no drop outs...

[Miska]

Latest Windows 10 update 1903 has specific scheduler optimizations for Ryzen 3. So this will likely make a notable difference when running HQPlayer with Ryzen under Windows...

 

[Rune]

By looking at the R7 3700X screenshot is looks like even a R5 3600X should be sufficient as two cores is not doing anything.

At least with some GPU assistance.

 

[Yviena]

Hmm seems there is still some multithreading overhead cost when multiple cores are used/selected as i can still play 192khz track with ASDM7EC DSD256 EXT2 with only selecting 4 cores on my 3700x which shaves off around 2-5% cpu usage from HQplayer, or it could just be me running the ram at 2133mhz currently which incurs higher thread switching overhead.

[DancingSea]

Aloha,

 

In anyone’s opinion, including Jussi’s, which filter/ modulator produces the widest soundstage?

 

Thanks.

[PieterP]

2 hours ago, Yviena said:

Hmm seems there is still some multithreading overhead cost when multiple cores are used/selected as i can still play 192khz track with ASDM7EC DSD256 EXT2 with only selecting 4 cores on my 3700x which shaves off around 2-5% cpu usage from HQplayer, or it could just be me running the ram at 2133mhz currently which incurs higher thread switching overhead.

@Yviena, you can also upsample from redbook to ASDM7EC DSD256 EXT2 with your 3700X?

[Le Concombre Masqué]

@Miska: in addition to my #15665 Q & observations

 

using D4 on macOS in conjunction with Loopback/Safari:

 

at each restart I have to go back in Preferences and handpick Loopback (the choice does not stick)

 

I then have to fiddle :start without convo, add it later ; go handpick audio:default etc though it appears as selected from the start....

[Miska]

1 hour ago, Le Concombre Masqué said:

using D4 on macOS in conjunction with Loopback/Safari:

 

at each restart I have to go back in Preferences and handpick Loopback (the choice does not stick)

 

I then have to fiddle :start without convo, add it later ; go handpick audio:default etc though it appears as selected from the start....

 

If you change the default while HQPlayer is running, it'll change the ID and thus default is different than earlier default.

 

[Miska]

On 8/15/2019 at 10:30 AM, Le Concombre Masqué said:

* An album that played just fine once D&Ded and then controllable via Client , would not appear in Client after being added to Library. Even stranger : only CD1 and 2 of a 4 box set appeared (all played fine as well once D&Ded). To be thorough, those albums were added to a special folder where I dump recent rips before I dispatch albums where they belong. So that folder has had "Scan" performed on it several times. Maybe I have had missing albums (in Client) before but could not tell after the Scan of my Pink Floyd or Miles Davis folders since they have been madly populated over decades...

 

Have you checked how they show up in the Library folder? I have lot of similar cases for CD rips, because metadata in internet metadata databases differs even for different CDs of same set. Then I've just corrected the metadata in HQPlayer's Library-dialog.

 

On 8/15/2019 at 10:30 AM, Le Concombre Masqué said:

* seems only covers with cover as name make it ; at first thought it looks easier if HQP would take *.jpg, pdf etc in the main folder rather than renaming

 

This is documented, cover image needs to have word "cover" or "folder" in the name and it must be JPG or PNG image.

 

[Le Concombre Masqué]

13 minutes ago, Miska said:

 

Have you checked how they show up in the Library folder? I have lot of similar cases for CD rips, because metadata in internet metadata databases differs even for different CDs of same set. Then I've just corrected the metadata in HQPlayer's Library-dialog.

 

 

This is documented, cover image needs to have word "cover" or "folder" in the name and it must be JPG or PNG image.

 

where is the Library folder?

 

why don't you make HQP accept *.jpg or png in the main folder (excluding subfolders with plentiful of artwork) ?

[Le Concombre Masqué]

23 minutes ago, Miska said:

 

If you change the default while HQPlayer is running, it'll change the ID and thus default is different than earlier default.

 

did not do that. but, being in trial mode, have to relaunch every 30 min. Still, that does not explain why my preference 

for Input device does not stick

[callany]

DSD256 with ASDM7EC,  I can run all filters including xtr 44.1/88/176.excluding xtr 48/96/192.

(CPU:i7 4790 ,GPU:GTX980TI,WIN10)

[Foggie]

Just for some experimenting purposes, I have the following (on fiber network)

- Win10 wkst running ROON server with HQP 3.25 = my current "production" enviro, want to leave as is

- microU endpoint running NAA

- high end wkst XEON W2135, 64GB, P4000, 512GB m.2 etc... = this is the one I want to test out with whatever?

 

I installed Ubuntu Server 18.04.3 LTS on the Xeon = was thinking embedded but I don't have lic for that (yet).  What about AL or Ubuntu desktop.  I could run HQP 3.5 on Ubuntu desktop and send to NAA and keep my Win10 wkst running ROON svr only.  Not sure if this would provide additional "audio" benefit (other than getting away from Win updates). 

 

AL could be an option but I'm confused if I would be able to try this with my current HQP lic/version?

[ted_b]

Over on the EC thread we've been discussing whether the I9-9900k is overkill vs the i7-9700k.  I know Jussi states that although HQP doesn't use the virtual cores created with hyperthreading, the tool does indeed allow OS context switching between processes/threads to be faster.  Is it a plus for the I9-9900k then (as the 9700k does not support hyperthreading)?

[Le Concombre Masqué]

4 hours ago, Le Concombre Masqué said:

did not do that. but, being in trial mode, have to relaunch every 30 min. Still, that does not explain why my preference 

for Input device does not stick

as I have now relaunched several times HQP in order to listen to a lengthy podcast I have to say that it now starts flawlessly as soon as I hit Play

[Miska]

7 hours ago, Le Concombre Masqué said:

where is the Library folder?

 

File menu -> Library? The place where you scan...

 

7 hours ago, Le Concombre Masqué said:

why don't you make HQP accept *.jpg or png in the main folder (excluding subfolders with plentiful of artwork) ?

 

Because there's otherwise no way to tell which of the pictures is supposed to be used. Quite typically there's a front cover picture, back cover picture and so on. You need to indicate the picture you want to use by naming it the way it can be recognized. Not just one of some random pictures that happens to be there.