HQ Player

[bibo01]

@MiskaSome years ago for Metrum Musette you were suggesting upsampling to 352.8/384k with closed-form and DAC bits set to 16.

For the present Metrum Onyx, which it also has 2 DAC chips (although new model), would you revise your recommendations, perhaps newer filters? Thanks   

[luisma]

On 6/26/2021 at 7:07 AM, Hensema said:

Just upsample pcm to 24 bit with dithering with your filter of choice. If you want to use a noise shaper instead of dithering, I would  set it to about 19 bit and use the highest sampling rate possible.

@Miska the Project S2 DAC based on the ESS9038Q2M chip, best results should be achieved with DSD256 and ASDM5EC or DSD512 with DSD5, in the case of using PCM and noise shapers (NS / LNS) bits should be set to 19 and in case of dithering to 24 bits as well?

 

Just checking with you if that would be accurate.

 

I'm using the fast rolloff out of these filters:

• Optimal transient (Pro-Ject preffered
• Fast Rolloff (Linear phase)
• Slow Rolloff (Linear phase)
• Minimum Phase Fast
• Minimum Phase Slow
• Linear Apodizing
• Hybrid Filter
• Brickwall Filter

 

And no onchip distortion compensation.

 

[luisma]

EDIT: duplicated post removed

[Miska]

17 hours ago, luisma said:

@Miska the Project S2 DAC based on the ESS9038Q2M chip, best results should be achieved with DSD256 and ASDM5EC or DSD512 with DSD5, in the case of using PCM and noise shapers (NS / LNS) bits should be set to 19 and in case of dithering to 24 bits as well?

 

Just checking with you if that would be accurate.

 

I'm using the fast rolloff out of these filters:

• Optimal transient (Pro-Ject preffered
• Fast Rolloff (Linear phase)
• Slow Rolloff (Linear phase)
• Minimum Phase Fast
• Minimum Phase Slow
• Linear Apodizing
• Hybrid Filter
• Brickwall Filter

 

And no onchip distortion compensation.

 

 

ASDM5EC at DSD256 or ASDM5 at DSD512.

 

With DSD inputs filter selection doesn't matter, since Pro-Ject doesn't expose setting for the DSD filter. But if you use PCM inputs, use 705.6/768k input rates which bypasses the filter. And for lower rates use Fast Rolloff.

 

Distortion compensation should be enabled, otherwise you get significantly increased distortion.

 

[pier]

@Miska I'm sure they already asked 10000 times but.. what do you suggest for Ares II ? DAC bits? PCM for sure (DSD is 6bit on Ares) ? 

 

thanks!

[Miska]

16 minutes ago, pier said:

@Miska I'm sure they already asked 10000 times but.. what do you suggest for Ares II ? DAC bits? PCM for sure (DSD is 6bit on Ares) ? 

 

Why PCM? I'm running it with DSD256 from ASDM7EC modulator. On my math it would be 5.0444 bits though.

 

[pier]

1 hour ago, Miska said:

 

Why PCM? I'm running it with DSD256 from ASDM7EC modulator. On my math it would be 5.0444 bits though.

 

 

I was checking this thread and a friend stated that since it has no 1 bit hardware,  DSD is converted to 6 bit internally.  So I was thinking that 16bit to 1 bit to 6 bit are too many conversions. 

 

Guess I'll go back to DSD testing... Are you using 16bit too?

 

[Miska]

3 hours ago, pier said:

 

I was checking this thread and a friend stated that since it has no 1 bit hardware,  DSD is converted to 6 bit internally.  So I was thinking that 16bit to 1 bit to 6 bit are too many conversions. 

 

Guess I'll go back to DSD testing... Are you using 16bit too?

 

 

Since I use DSD output, "DAC Bits" has no effect. But I've done testing with DAC Bits set to 20, but you could use 16 bit as well.

 

[Miska]

On 6/27/2021 at 6:07 AM, bibo01 said:

@MiskaSome years ago for Metrum Musette you were suggesting upsampling to 352.8/384k with closed-form and DAC bits set to 16.

For the present Metrum Onyx, which it also has 2 DAC chips (although new model), would you revise your recommendations, perhaps newer filters? Thanks   

 

I would keep output bits to 16 and use suitable noise shaping. But choice of filter shouldn't really matter from technical perspective.

 

[luisma]

9 hours ago, Miska said:

use 705.6/768k input rates which bypasses the filter.

Yeah unfortunately there are issues playing at that rate with recent HQP versions, it is something I haven't addressed before but yeah, stable I can do 384 but nothing further than that, I get some ringing noise IIRC or just white type noise.

If you are interested in troubleshooting I can capture a log and send it over.

9 hours ago, Miska said:

ASDM5EC at DSD256 or ASDM5 at DSD512

Understood

9 hours ago, Miska said:

Distortion compensation should be enabled, otherwise you get significantly increased distortion

This is onchip and I read somewhere it was worst using the setting but obviously not with HQPlayer so I will follow your advice, logically I figured that leaving distortion management to HQP and not the DAC could improve things but my logic was flawed :)

 

Thanks Jussi

 

[Miska]

8 minutes ago, luisma said:

This is onchip and I read somewhere it was worst using the setting but obviously not with HQPlayer so I will follow your advice, logically I figured that leaving distortion management to HQP and not the DAC could improve things but my logic was flawed :)

 

I have measured this twice and the feature works as intended.

 

Here's the DAC with distortion compensation disabled:

 

And here's the DAC with distortion compensation enabled:

 

[luisma]

Thank you for posting your measurements, definitely slight improvement but improvement regardless

[Miska]

11 hours ago, luisma said:

Thank you for posting your measurements, definitely slight improvement but improvement regardless

 

About 10x improvement (~20 dB).

 

[luisma]

12 hours ago, Miska said:

About 10x improvement (~20 dB).

True, only 2 and 3 Ghz? or the peaks are just samples performed at 1 khz interval and not continuously throughout the band, since I see roughly peaks at integer khz intervals possible these are the samples meaning that most likely between 1 and 3 is the section that affects the most

EDIT: I'm just not familiar with the testing methodology ...

[DancingSea]

Aloha @Miska & Everyone,

 

I recently bought a Marantz SACD 30n.  If I use a direct USB connection between the M1 Mac Mini to the Marantz and play a DSD file using Roon with hqplayer, the DSD makes it the Marantz as DSD. 
 

If I connect the Mac to the network and use a microrendu 1.4 to connect to the Marantz, the DSD becomes pcm at whatever the PCM rate is set to in hqplayer.  This happens if hqplayer is set to source, pcm, or DSD. 
 

Remove the mR 1.4 and go back to straight USB, with the Roon/hqplayer combo, and DSD again reaches the Marantz normally.  
 

All settings in Roon & Hqplayer remain the same throughout. 
 

I can’t figure out why going through the mR is not allowing DSD to pass through. 
 

Any ideas?

 

Thanks

 

[Zauurx]

🤪

Help!!! CPU load under 20bits!!!
I'm testing my new dac that I've just put in run-in.
I chose a Musician Pegasus R2R which replaces my RME ADI-2.
For the configuration, it's plug and play with my Allo USBridge under Gentooplayer (usb audio).

So I just changed the 32bits output to 20bits with DSD256, gaussian-xla and ADSM5EC.
I let it run and I come back to see just for fun and there... horror !

CPU at 100% (core i5-8400 which was running at 32 bit at 65% max).

Is this the wrong setting? or is it related to the switch from Roon 64bits to 20? easier to go to 64 > 32.

 

EDIT :
 

After investigation... it seems that the problem is related to the radio streams passing through Roon.
Source :192kps stream (AAC or MP3) >> 100% CPU.
If I take HiRes flac, it works without load change (70%).
Same thing for DSD and no problem with Tidal.

I only use radio streams from Roon on my laptop and I haven't tried it for a long time on my living room system with HQP

 

[Zauurx]

Ex.  

http://icecast.radiofrance.fr/fipjazz-hifi.aac

So, a Roon problem... or between Roon and HQP?

 

 

[Miska]

1 hour ago, Zauurx said:

🤪

Help!!! CPU load under 20bits!!!
I'm testing my new dac that I've just put in run-in.
I chose a Musician Pegasus R2R which replaces my RME ADI-2.
For the configuration, it's plug and play with my Allo USBridge under Gentooplayer (usb audio).

So I just changed the 32bits output to 20bits with DSD256, gaussian-xla and ADSM5EC.
I let it run and I come back to see just for fun and there... horror !

CPU at 100% (core i5-8400 which was running at 32 bit at 65% max).

Is this the wrong setting? or is it related to the switch from Roon 64bits to 20? easier to go to 64 > 32.

 

EDIT :
 

After investigation... it seems that the problem is related to the radio streams passing through Roon.
Source :192kps stream (AAC or MP3) >> 100% CPU.
If I take HiRes flac, it works without load change (70%).
Same thing for DSD and no problem with Tidal.

I only use radio streams from Roon on my laptop and I haven't tried it for a long time on my living room system with HQP

 

 

If your output is in SDM mode, "DAC Bits" setting has no effect. Only for PCM outputs.

 

From HQPlayer point of view it doesn't make difference where Roon is getting the content, it looks the same to HQPlayer.

 

[Miska]

8 hours ago, DancingSea said:

I recently bought a Marantz SACD 30n.  If I use a direct USB connection between the M1 Mac Mini to the Marantz and play a DSD file using Roon with hqplayer, the DSD makes it the Marantz as DSD. 
 

If I connect the Mac to the network and use a microrendu 1.4 to connect to the Marantz, the DSD becomes pcm at whatever the PCM rate is set to in hqplayer.  This happens if hqplayer is set to source, pcm, or DSD. 
 

Remove the mR 1.4 and go back to straight USB, with the Roon/hqplayer combo, and DSD again reaches the Marantz normally.  
 

All settings in Roon & Hqplayer remain the same throughout. 
 

I can’t figure out why going through the mR is not allowing DSD to pass through. 
 

Any ideas?

 

Is DoP enabled for microRendu case too?

 

[Miska]

9 hours ago, luisma said:

True, only 2 and 3 Ghz? or the peaks are just samples performed at 1 khz interval and not continuously throughout the band, since I see roughly peaks at integer khz intervals possible these are the samples meaning that most likely between 1 and 3 is the section that affects the most

EDIT: I'm just not familiar with the testing methodology ...

 

Test tone is 1 kHz sinewave. Harmonics at multiples of this frequency (2 kHz, 3 kHz, etc) are called harmonic distortion. In perfect case you would only see 1 kHz test tone and very low noise floor and nothing else.

[luisma]

21 minutes ago, Miska said:

Test tone is 1 kHz sinewave. Harmonics at multiples of this frequency (2 kHz, 3 kHz, etc) are called harmonic distortion. In perfect case you would only see 1 kHz test tone and very low noise floor and nothing else.

Thank you

[DancingSea]

5 hours ago, Miska said:

Is DoP enabled for microRendu case too?

Where does one enable DoP for the mR?

[Miska]

31 minutes ago, DancingSea said:

Where does one enable DoP for the mR?

 

In HQPlayer settings, same way for all backends.

[GMG]

 

How to prevent HQPlayer OS going to sleep?

 

Hope someone can guide me what needs to be done on a laptop running HQPlayer that will keep going to sleep whenever the lid is closed.

The HQPlayer is run from a USB drive on an old lenovo T420

Can't find anything to change in the BIOS - all power settings are configured for max performance

I disabled the sleep function from the Windows installation on the main HDD - no effect. not related, but I tried :-)

On HQPlayer OS I edited /etc/systemd/logind.conf and enabled lid ignore parameter - did not help

 

Anyone?

Many thanks

[GMG]

1 hour ago, GMG said:

 

How to prevent HQPlayer OS going to sleep?

 

Hope someone can guide me what needs to be done on a laptop running HQPlayer that will keep going to sleep whenever the lid is closed.

The HQPlayer is run from a USB drive on an old lenovo T420

Can't find anything to change in the BIOS - all power settings are configured for max performance

I disabled the sleep function from the Windows installation on the main HDD - no effect. not related, but I tried :-)

On HQPlayer OS I edited /etc/systemd/logind.conf and enabled lid ignore parameter - did not help

 

Anyone?

Many thanks

Well, I guess I will answer myself :-)

I got it.

Had to add the following param to /etc/systemd/logind.conf

HandleLidSwitch=ignore

 

This param was not in the original file and previously I changed a different param that was similar: LidSwitchIgnoreInhibitor

 

Seems to work now