HQPlayer Linux Desktop and HQplayer embedded

[guiltyboxswapper]

Thanks @Miska for the upnp Idle fix, very appreciated. 

 

For those who have used HQPlayer with upnp in the past, and found that changing tracks manually incurred a large (9s+) penalty especially for DSD512 modulators, its gone 🙂

 

Now down to 2s even with poly-sinc-xtr (non-2s) selected as the filter at DSD 512 rates, with the penalty only paid if switching between frequency rates (not so bad with XTR).  Ofcourse you can force a fixed output at all times to get around that switching penalty too....

 

 

[craighartley]

23 hours ago, Miska said:

There is now updated 4.15.1 build of x64 HQPlayer OS image. Let me know if it still has problems...

 

4.15.1 works perfectly for me.  Thank you. 

[craighartley]

On 3/3/2020 at 12:40 AM, Miska said:

There is now updated 4.15.1 build of x64 HQPlayer OS image. Let me know if it still has problems...

 

I spoke too soon. I no longer have stuttering that I had with 4.15 (and not on previous versions), but I still have a large click from the speakers when I press play or change tracks on the same album (manually). I've never had anything like this with HQP before 4.15.

 

(Roon streaming to HQP OS on bootable image streaming DSD via NAA to T+A DAC8 DSD).

[Miska]

5 hours ago, craighartley said:

I spoke too soon. I no longer have stuttering that I had with 4.15 (and not on previous versions), but I still have a large click from the speakers when I press play or change tracks on the same album (manually). I've never had anything like this with HQP before 4.15.

 

(Roon streaming to HQP OS on bootable image streaming DSD via NAA to T+A DAC8 DSD).

 

What kind of NAA do you have? I've been trying to reproduce this with T+A DAC8 DSD:

 

I also tried with Holo Spring and RME ADI-2 Pro, but I'm not getting such behavior.

 

[craighartley]

38 minutes ago, Miska said:

 

What kind of NAA do you have? I've been trying to reproduce this with T+A DAC8 DSD:

 

I also tried with Holo Spring and RME ADI-2 Pro, but I'm not getting such behavior.

 

NAA is MicroRendu running latest Sonicorbiter. But I've just tried it using your naa-401-cuboxi-urendu image instead, with same result. Please note that My T+A has the updated Amanero driver to run DSD512:

 

[craighartley]

Although I'm doing these tests with DSD 256 and 7EC modulator.

[Miska]

3 minutes ago, craighartley said:

Please note that My T+A has the updated Amanero driver to run DSD512:

 

That is something I cannot test, as mine is running the original firmware...

 

4 minutes ago, craighartley said:

NAA is MicroRendu running latest Sonicorbiter.

 

Is it feasible to try connecting DAC directly to the HQPlayer machine?

 

[craighartley]

1 minute ago, Miska said:

 

That is something I cannot test, as mine is running the original firmware...

 

 

Is it feasible to try connecting DAC directly to the HQPlayer machine?

 

No I'm afraid it's not. HQPlayer machine is in cupboard under the stairs in another room and I had an operation on my spine a few weeks ago. I can lift a USB stick, but I have to ask my wife to plug it in!

[craighartley]

6 hours ago, Miska said:

 

That is something I cannot test, as mine is running the original firmware...

 

 

Is it feasible to try connecting DAC directly to the HQPlayer machine?

 

 

Going back to 4.14.1: there is actually a click when starting Play, but there is then a delay before the music starts, so its not so intrusive as on 4.15 (and .1), where the click is at the point that the music starts playing.

 

So I don't know whether it's possible that the change is actually a faster start of the music (an improvement!) rather than the introduction of an unwanted artefact. In any case from the point of view of listening to music 4.14.1 is preferable in my system, so I'll stick with that for now.

 

When my back is better and I can lug the server around, if there is still a click in the latest version, then I'll connect the HQPlayer machine direct to the DAC and get back to you.

 

Is there anyone else out there running Embedded to the T+A DAC8 DSD with the updated Amanero firmware?

[Miska]

11 minutes ago, craighartley said:

Going back to 4.14.1: there is actually a click when starting Play, but there is then a delay before the music starts, so its not so intrusive as on 4.15 (and .1), where the click is at the point that the music starts playing.

 

So I don't know whether it's possible that the change is actually a faster start of the music (an improvement!) rather than the introduction of an unwanted artefact. In any case from the point of view of listening to music 4.14.1 is preferable in my system, so I'll stick with that for now.

 

When my back is better and I can lug the server around, if there is still a click in the latest version, then I'll connect the HQPlayer machine direct to the DAC and get back to you.

 

Is there anyone else out there running Embedded to the T+A DAC8 DSD with the updated Amanero firmware?

 

Are you on the latest firmware version? Output relays are supposed to mute the output during start/stop transitions, happens also when you switch track in Roon because it (unnecessarily) asks HQPlayer to stop first. This mute control comes from the Amanero firmware.

 

[craighartley]

57 minutes ago, Miska said:

 

Are you on the latest firmware version? Output relays are supposed to mute the output during start/stop transitions, happens also when you switch track in Roon because it (unnecessarily) asks HQPlayer to stop first. This mute control comes from the Amanero firmware.

 

I have the latest T+A firmware, but not the very latest (late 2019?) version of the Amanero USB receiver firmware. When I installed the Amanero firmware any clicks were negligible. I didn't have problems with clicks on the music starting until HQP 4.15(.1).

 

I could try the latest version of the Amanero firmware, but would want to read/hear more reports of its stability first. I'll do some research.

[Miska]

Just now, craighartley said:

I have the latest T+A firmware, but not the very latest (late 2019?) version of the Amanero USB receiver firmware. When I installed the Amanero firmware any clicks were negligible. I didn't have problems with clicks on the music starting until HQP 4.15(.1).

 

I could try the latest version of the Amanero firmware, but would want to read/hear more reports of its stability first. I'll do some research.

 

It is just that when you use NAA, HQPlayer is not directly controlling the audio device, it is instead communicating with NAA that deals with the actual audio device. So such behavior is more related to NAA than HQPlayer itself. There may be some slight timing differences though as you noted, but these have indirect effects on the output.

 

[craighartley]

7 hours ago, Miska said:

 

It is just that when you use NAA, HQPlayer is not directly controlling the audio device, it is instead communicating with NAA that deals with the actual audio device. So such behavior is more related to NAA than HQPlayer itself. There may be some slight timing differences though as you noted, but these have indirect effects on the output.

 

From a listening point of view, the timing difference of the click is not small. With 4.14.1 there is at least a second pause after the click before the music starts. With 4.15.1 it interrupts the leading edge of the first note. It’s the difference between tolerable and intolerable. 
 

[Le Concombre Masqué]

4.15.1 on my MBPr crashes after a few minutes 

 

tens of lines appear, first time in a narrow column, second as long  lines; stories told seem completely different 

 

1st crash mentioned preemption disabled at stp, second time it mentions things about cpuidle... 

 

btw, would be nice to be able to back-up matrix ; 8 configurations, 4  being Mch to St, to redo twice in the same hour is boring

[Miska]

1 hour ago, Le Concombre Masqué said:

btw, would be nice to be able to back-up matrix ; 8 configurations, 4  being Mch to St, to redo twice in the same hour is boring

 

It is otherwise backed up, but convolution filters are not backed up now. I need to think about making a ZIP or something out of the configuration + filters... Something for the next version...

[craighartley]

11 hours ago, Miska said:

 

Are you on the latest firmware version? Output relays are supposed to mute the output during start/stop transitions, happens also when you switch track in Roon because it (unnecessarily) asks HQPlayer to stop first. This mute control comes from the Amanero firmware.

 

Further to this. I tried taking Roon out of the equation by playing simply from HQPlayer (controlling via Client) to the NAA. As you would have predicted, there is no click switching tracks on an album. But there is the same click on Play.

 

For comparison I tried using the MicroRendu in its RoonReady mode (rather than as HQP NAA) and playing direct from Roon. So taking HQP out of the equation. All clicks disappear streaming same albums upsampled to DSD256.

[Miska]

1 hour ago, craighartley said:

Further to this. I tried taking Roon out of the equation by playing simply from HQPlayer (controlling via Client) to the NAA. As you would have predicted, there is no click switching tracks on an album. But there is the same click on Play.

 

For comparison I tried using the MicroRendu in its RoonReady mode (rather than as HQP NAA) and playing direct from Roon. So taking HQP out of the equation. All clicks disappear streaming same albums upsampled to DSD256.

 

This is a mystery, I tried the same yesterday on three DACs and two NAA's (both Windows and Ubuntu based) and I cannot reproduce it. Both NAA 3.6.0 and 4.1.0 versions.

 

Taking NAA out of equation would tell which side it likely is.

 

[Miska]

4 hours ago, Le Concombre Masqué said:

4.15.1 on my MBPr crashes after a few minutes 

 

I've been now playing one and half hour of Tidal with HQPlayer OS 4.15.1, running on UpSquared, connected to Holo Cyan DSD DAC. Tidal playing with mConnect Player and upsampled to DSD256.

 

No pops or crashes so far.

[Le Concombre Masqué]

1 hour ago, Miska said:

 

I've been now playing one and half hour of Tidal with HQPlayer OS 4.15.1, running on UpSquared, connected to Holo Cyan DSD DAC. Tidal playing with mConnect Player and upsampled to DSD256.

 

No pops or crashes so far.

back to 4.14.1... will try to pm u the pictures I took after the crashes (btw, I have assumed no screen capture was possible and took photos but maybe I ignore some trick?)

[Miska]

1 hour ago, Le Concombre Masqué said:

back to 4.14.1... will try to pm u the pictures I took after the crashes (btw, I have assumed no screen capture was possible and took photos but maybe I ignore some trick?)

 

There's no screen capture, but if the error doesn't result in complete OS crash or takes long enough, the logs can be recovered from the system journal (you can view with "journalctl -a"). This can be then copied over a network share for example.

 

[callany]

Hi，Miska

How to use USB Audio Class 2 inputs in HQPlayer 4 Embedded 4.15.0

[Miska]

38 minutes ago, callany said:

How to use USB Audio Class 2 inputs in HQPlayer 4 Embedded 4.15.0

 

Most of the important details are described here:

 

Also other hardware that has USB device controller likely works. These are usually Atom, Celeron or Pentium branded SoC's in Intel's product portfolio. With NAA 4.1.0 it is also possible to have such input through a NAA instead of running locally on a HQPlayer machine.

 

[Le Concombre Masqué]

On 3/5/2020 at 3:12 PM, Miska said:

 

I've been now playing one and half hour of Tidal with HQPlayer OS 4.15.1, running on UpSquared, connected to Holo Cyan DSD DAC. Tidal playing with mConnect Player and upsampled to DSD256.

 

No pops or crashes so far.

On UpSqared I suppose you don't run convolution filters ; did you? If not could you please confirm no pops no crashes even when convolving?

[Miska]

1 hour ago, Le Concombre Masqué said:

On UpSqared I suppose you don't run convolution filters ; did you? If not could you please confirm no pops no crashes even when convolving?

 

I didn't try, and it has hard to generalize about convolution filters because there are so many different sizes. For RedBook sources reasonably short ones are not so bad. But DSD256 is pushing the UpSquared pretty much on the edge, so I wouldn't try to put convolution there.

 

For PCM outputs it could likely would though at least to some extent.

 

[Le Concombre Masqué]

43 minutes ago, Miska said:

 

I didn't try, and it has hard to generalize about convolution filters because there are so many different sizes. For RedBook sources reasonably short ones are not so bad. But DSD256 is pushing the UpSquared pretty much on the edge, so I wouldn't try to put convolution there.

 

For PCM outputs it could likely would though at least to some extent.

 

thank you Jussi, my apologies : I was not enquiring about the Upsquared capabilities but suggesting that maybe the crashes I experimented with Embedded 4.15.1 could be revealed only with convolution engaged