HQ Player

[manishex]

19 minutes ago, tboooe said:

Have you reset all AO optimizations and restarted to see if qwave runs correctly?  Is it only after to run AO that qwave stops working?

I haven't had a problem with playing the newest update of HQ player. Before the AO optimisation, QWave was running active. 

[tboooe]

2 minutes ago, manishex said:

I haven't had a problem with playing the newest update of HQ player. Before the AO optimisation, QWave was running active. 

So when you run AO, qWave service stops?  In AO did you choose to disable services?

[manishex]

20 minutes ago, tboooe said:

So when you run AO, qWave service stops?  In AO did you choose to disable services?

So I have ws2016 core mode, prior to activating AO, QWave can be enabled. Correct, I chose the disable all drivers and services options and enable anything recommended. 

Signature X, X and not have anything running, including process lasso and fidelizer. 

Overall you can see I only have 9 services enabled. 

[tboooe]

Just now, manishex said:

So I have ws2016 core mode, prior to activating AO, QWave can be enabled. Correct, I chose the disable all drivers and services options and enable anything recommended. 

OK so i seems in AO you will need to not select disable drivers and services.

 

BTW, I am using Server 2016 in Core with AO. I  havent enabled qwave service yet though so your experience is very helpful.

[HardrockInMiniMac]

On 6/3/2018 at 8:35 PM, Miska said:

 

The difference is mostly that HQPlayer alone you don't need to route all audio through Roon.

 

 

On 6/3/2018 at 2:42 PM, Sevenfeet said:

 

I don't get it.  Why would HQP alone be "ahead" of the same software with a Roon front end?

 

And the closest thing to a remote control for OSX is the iTunes-HQ Player frontend by @Geoffrey Armstrong combined with Apple's Remote app.  Its not perfect, but it works.

 

Well, I do not know why, but I can switch on the same song in just seconds and it is absolutely significant. The combination of Roon+HQP ist better than Roon alone but worse than HQP alone by far. However, to me it only matters for serious listening, for all other requirements I run Roon + HQP. 

[luisma]

17 minutes ago, HardrockInMiniMac said:

 

 

Well, I do not know why, but I can switch on the same song in just seconds and it is absolutely significant. The combination of Roon+HQP ist better than Roon alone but worse than HQP alone by far. However, to me it only matters for serious listening, for all other requirements I run Roon + HQP. 

How dare you? we only do serious listening around here .... all other requirements are a distraction ....   

[tboooe]

7 hours ago, manishex said:

I haven't had a problem with playing the newest update of HQ player. Before the AO optimisation, QWave was running active. 

May I ask how you enabled qwave in Server 2016?  I keep getting an error.

[manishex]

3 minutes ago, tboooe said:

May I ask how you enabled qwave in Server 2016?  I keep getting an error.

I can only enable it while Hqplayer is open

[tboooe]

3 minutes ago, manishex said:

I can only enable it while Hqplayer is open

May I ask what command you used to enable qwave?  Ive tried dism and install-windowsfeature but keep getting errors.  Like you I am using Server 2016 in core mode.

[manishex]

12 minutes ago, tboooe said:

May I ask what command you used to enable qwave?  Ive tried dism and install-windowsfeature but keep getting errors.  Like you I am using Server 2016 in core mode.

In powershell I used the dism install feature method. Have you done get features? You need it to have the exact name. 

[ChrisVH]

On 6/3/2018 at 1:35 PM, Miska said:

 

The difference is mostly that HQPlayer alone you don't need to route all audio through Roon.

 

 

On 6/3/2018 at 3:56 PM, Sevenfeet said:

 

Well with heavier filters and non 2s variants, you'd be hard pressed to hear good sound on most setups period due to CPU/system bus bandwidth issues.

 

I haven't tried Roon yet...but I think I'll be pulling the trigger in the next few days along with a HQPlayer license now that I've been playing with it for months now.  This combo is a bit expensive for me but I'm hoping it's all worth it.

 

Why is it that all the 'processing' can't be completed by HQ Player first, then sent out after being buffered/completed? I've heard the arguments about the simultaneous (background) processing causing potential problems, but don't understand why we can't have all processing completed, and then played from that 100% buffered file... Seems to me this would solve all the 'computer load' sonic arguments... 

 

 

 

 

 

 

 

 

 

 

 

 

 

[tboooe]

20 minutes ago, manishex said:

In powershell I used the dism install feature method. Have you done get features? You need it to have the exact name. 

yes I have used get features and the name is qwave.  Not sure why I keep getting errors.  I have tried to enable it with HQP open and closed.

 

EDIT - restarted my machine 3 times a row and I was able to enable the feature.  Thank you!

 

Also, I applied AO, restarted and qWAVE service is still running.

[Miska]

20 minutes ago, ChrisVH said:

Why is it that all the 'processing' can't be completed by HQ Player first, then sent out after being buffered/completed? I've heard the arguments about the simultaneous (background) processing causing potential problems, but don't understand why we can't have all processing completed, and then played from that 100% buffered file... Seems to me this would solve all the 'computer load' sonic arguments... 

 

 

OK, let's take an example. Let's say you want to listen 60 minutes long album and processing it in realtime on the fly at the moment generates 50% CPU load. IOW, the processing runs at 2x faster than playback speed.

 

If we would process it first before playback begins, you'd then have to sit and wait for 30 minutes for the processing the complete before you would hear any sound. If we would do this per track basis and you have two 5 minute tracks on the playlist, you would have 2:30 long gap between two tracks.

 

Needless to say this would consume massive amount of RAM or storage space, especially if you have let's say 6 hours long playlist (113.55 GB at stereo DSD512). And what would we do with streaming things like Roon/Tidal where there is no definite end in sight at all. And even with just HQPlayer playlist, you could be adding more tracks to the playlist while it is already playing. We cannot complete something upfront that didn't exist yet at that time.

 

[Allan F]

On 6/1/2018 at 9:07 AM, jimdukey said:

I've never enjoyed HQ more than with the new 3.21...

 

I wasn't expecting better sound quality from 3.21 but, from the first few albums, it was clear that there was an audible improvement. I don't know how you do it, Miska, but you keeping getting more out of HQP. Much appreciated.

[ChrisVH]

3 hours ago, Miska said:

 

OK, let's take an example. Let's say you want to listen 60 minutes long album and processing it in realtime on the fly at the moment generates 50% CPU load. IOW, the processing runs at 2x faster than playback speed.

 

If we would process it first before playback begins, you'd then have to sit and wait for 30 minutes for the processing the complete before you would hear any sound. If we would do this per track basis and you have two 5 minute tracks on the playlist, you would have 2:30 long gap between two tracks.

 

Needless to say this would consume massive amount of RAM or storage space, especially if you have let's say 6 hours long playlist (113.55 GB at stereo DSD512). And what would we do with streaming things like Roon/Tidal where there is no definite end in sight at all. And even with just HQPlayer playlist, you could be adding more tracks to the playlist while it is already playing. We cannot complete something upfront that didn't exist yet at that time.

 

Thank you for your response, Miska. I have a followup Q, if you don't mind:

 

"If we would do this per track basis and you have two 5 minute tracks on the playlist, you would have 2:30 long gap between two tracks."

 

But... while the first track is playing, can't the processor be working ahead on a PORTION of the track that isn't playing, and then once completed,  it moves into the playable queue. This would only add a few seconds to the beginning of each track, and could be executed on a track by track basis... So essentially this would be a 'read ahead' loop that then feeds chunks of the data at least X secs in advance of when the data is actually needed.

 

Or is the issue that once the bits are 'delivered to be played', the background processing can have an effect on those already processed bits that are being executed during playback?

 

 

[Miska]

6 hours ago, ChrisVH said:

But... while the first track is playing, can't the processor be working ahead on a PORTION of the track that isn't playing, and then once completed,  it moves into the playable queue. This would only add a few seconds to the beginning of each track, and could be executed on a track by track basis... So essentially this would be a 'read ahead' loop that then feeds chunks of the data at least X secs in advance of when the data is actually needed.

 

That is already begin done? Processing goes ahead of the playback, at playback speed (so your CPU doesn't sit at 100% all the time).

 

6 hours ago, ChrisVH said:

Or is the issue that once the bits are 'delivered to be played', the background processing can have an effect on those already processed bits that are being executed during playback?

 

Not an effect as such, but keeping processing running full steam ahead would keep the CPU at 100% load all the time (or until end of playlist) and it would also keep the RAM usage growing so at some point you could run out of RAM. And this wouldn't be so great with streaming like Roon/Tidal either.

 

I rather keep CPU load steady below 100% while playback is ongoing.

[Geoffrey Armstrong]

13 hours ago, Allan F said:

 

I wasn't expecting better sound quality from 3.21 but, from the first few albums, it was clear that there was an audible improvement. I don't know how you do it, Miska, but you keeping getting more out of HQP. Much appreciated.

+ 1

Immediately obvious this is another step up on the previous version. More natural with greater sense of space and openness. Bravo Miska!

[manishex]

14 hours ago, tboooe said:

yes I have used get features and the name is qwave.  Not sure why I keep getting errors.  I have tried to enable it with HQP open and closed.

 

EDIT - restarted my machine 3 times a row and I was able to enable the feature.  Thank you!

 

Also, I applied AO, restarted and qWAVE service is still running.

Do you hear any difference with enabling and disabling it? 

[ChrisVH]

13 hours ago, Miska said:

I rather keep CPU load steady below 100% while playback is ongoing.

 

OK. Thank you. Do you have a target CPU load reference to stay under, for best SQ? IN other words, if I use task manager on my main workstation (in another part of the house) when running Roon Core, pushing the bits to my UltraRendu in NAA mode (located in listening room, via ethernet) at what CPU % load should I become concerned about too much CPU overhead?

 

My main machine has a dual quad core Xeon processor as well as 128 GB of ram, so can't imagine I'll hit a wall. but would like your take, to be certain...

 

Thank you.

[Outlaw]

If HQPlayer runs fine on CPU alone should I enable Cuda

[freesteve]

Only the very best from HQ Player...   wonderful investment.

[Miska]

On 6/7/2018 at 11:07 PM, Outlaw said:

If HQPlayer runs fine on CPU alone should I enable Cuda

 

Not unless you have some need to free some CPU time and have enough GPU time to spare instead...

 

[Miska]

On 6/7/2018 at 12:07 AM, ChrisVH said:

OK. Thank you. Do you have a target CPU load reference to stay under, for best SQ? IN other words, if I use task manager on my main workstation (in another part of the house) when running Roon Core, pushing the bits to my UltraRendu in NAA mode (located in listening room, via ethernet) at what CPU % load should I become concerned about too much CPU overhead?

 

My main machine has a dual quad core Xeon processor as well as 128 GB of ram, so can't imagine I'll hit a wall. but would like your take, to be certain...

 

Nothing specific, risk of drop-outs/stuttering usually become relevant when CPU cores get loaded higher than 75% and really a risk at 90% core load or more. This while playing. Total load is usually less important, while load of the highest loaded cores is more important.

 

[ChrisVH]

52 minutes ago, Miska said:

 

Nothing specific, risk of drop-outs/stuttering usually become relevant when CPU cores get loaded higher than 75% and really a risk at 90% core load or more. This while playing. Total load is usually less important, while load of the highest loaded cores is more important.

 

 

Excellent. So... if a person was to make a 'stress test' to ensure they remain within these parameters, what circumstance would you recommend to construct such a stress test? Example: upsample to DSDXXX with XYZ filter, etc... 

 

Very helpful information. Thank you.

[Le Concombre Masqué]

Miska,

 

I'm very happy sound wise with newly designed filters 

but I have some concerns linked to convolution :

 

Though IR is only 4.9+ *, I have to go down 13 dB (26/0) not to get limited (on a nasty but favorite track by Jose James on Blue Note; I probably have nastier...) : are my speakers in danger ? 

 

In the process of A/Bing (or so) I had another corrupted file (no big deal i had the source) + HQP happened to quit ; Music files are wifi streamed from a GTech HD

 

My gut feeling is that it's all triggered by switching convolutions filter too often too fast and that I'm about to live happy for a while but I prefer to report/ask

 

* but Target level is 10 dB up and there's now a 5 dB @ 600 Hz Low Shelf vs 3@110 previously ; however, predicted bass SPL below 30 dB is less than before