HQ Player

[Miska]

4 hours ago, Hensema said:

I like to oversample 1x together with LNS15 noise shaping. However, when Nx is selected without filtering (NOS), is the LNS15 in that case also automatically disabled? 

 

No, dither/noise-shaper is completely separate and always active as selected. Why are you not using oversampling for higher PCM rates? Any content below 352.8 kHz badly needs oversampling...

 

[bipet]

Hi @Miska

 

I wanted to pick your brains! I have had some issues with my Allo USBridge Signature player as I have been hearing pops/crackles. I have been in discussion with Allo for a while now and have posted on the Signature thread. They have made lots of changes and as a result have released a new version of Dietpi  (6.26.3) recently. Lots of pops recently, and less now after their changes.

 

On their request (they wanted to see if I get pops across all sample rates, PCM, DSD etc) I did some testing this week, I was still using Roon (on an Intel i7 NUC) and HQ Player (on my i7 9700 desktop PC) connected over ethernet to the USBridge, itself feeding a Topping D90 DAC over USB.

 

I found that pops are present in different scenarios, with the first thing noted being they do not appear to be present when upscaling to PCM only to DSD. However the remaining pops, it used to be worse, seem to mainly occur at DSD512, no matter what modulator. When I play at DSD 256 (I have tried DSD7 256+fs, DSDv2, ASDM7EC and ASDM7) I don’t seem to hear them.

 

Any thoughts on why these pops are more prevalent on DSD 512 but not on DSD 256? Any thoughts how I might eleminate them?

 

My current settings are

 

[antonellocaroli]

12 minutes ago, rossco said:

Hi @Miska

 

I wanted to pick your brains! I have had some issues with my Allo USBridge Signature player as I have been hearing pops/crackles. I have been in discussion with Allo for a while now and have posted on the Signature thread. They have made lots of changes and as a result have released a new version of Dietpi  (6.26.3) recently. Lots of pops recently, and less now after their changes.

 

On their request (they wanted to see if I get pops across all sample rates, PCM, DSD etc) I did some testing this week, I was still using Roon (on an Intel i7 NUC) and HQ Player (on my i7 9700 desktop PC) connected over ethernet to the USBridge, itself feeding a Topping D90 DAC over USB.

 

I found that pops are present in different scenarios, with the first thing noted being they do not appear to be present when upscaling to PCM only to DSD. However the remaining pops, it used to be worse, seem to mainly occur at DSD512, no matter what modulator. When I play at DSD 256 (I have tried DSD7 256+fs, DSDv2, ASDM7EC and ASDM7) I don’t seem to hear them.

 

Any thoughts on why these pops are more prevalent on DSD 512 but not on DSD 256? Any thoughts how I might eleminate them?

 

My current settings are

 

 

are more with dsd512 because the traffic on the network is higher...
try to put a download a file (on the system where you have pops) from the internet while doing upsample dsd256...you see that you have equal pops....
t

[ambre]

58 minutes ago, antonellocaroli said:

 

are more with dsd512 because the traffic on the network is higher...
try to put a download a file (on the system where you have pops) from the internet while doing upsample dsd256...you see that you have equal pops....
t

Did you checked with another DAC due to the fact that the Topping 90 is brandnew and might have teething problems!

[bipet]

1 hour ago, antonellocaroli said:

are more with dsd512 because the traffic on the network is higher...
try to put a download a file (on the system where you have pops) from the internet while doing upsample dsd256...you see that you have equal pops....
t

I am not sure its network traffic. I just did a quick test. Using the same settings as above. I set a game to download via steam on the same machine as HQ Player is installed on, then I also set a torrent to download at the same time. I then went and listened to my test tracks and no popping. I am not sure what this signify's, maybe capacity in the network??

[Yviena]

14 minutes ago, rossco said:

I am not sure its network traffic. I just did a quick test. Using the same settings as above. I set a game to download via steam on the same machine as HQ Player is installed on, then I also set a torrent to download at the same time. I then went and listened to my test tracks and no popping. I am not sure what this signify's, maybe capacity in the network??

Try the Allo usbridge sig image with the older kernel, that fixed the soft popping issues for me, seems that the newer kernel.has pops on all of my dacs.

[bipet]

11 minutes ago, Yviena said:

Try the Allo usbridge sig image with the older kernel, that fixed the soft popping issues for me, seems that the newer kernel.has pops on all of my dacs.

Thanks. I have tried the older kernel I think and this version does seem a little better. I was using the DietPi_v6.18_41492_RPi-ARMv7-Stretch_AlloGUI.img image previously.

[Hensema]

4 hours ago, Miska said:

 

No, dither/noise-shaper is completely separate and always active as selected. Why are you not using oversampling for higher PCM rates? Any content below 352.8 kHz badly needs oversampling...

 

High res recordings sounds very good on my pcm1704 r2r dac in NOS. However, most of the time I use mqa-mp.

[Miska]

3 hours ago, Hensema said:

High res recordings sounds very good on my pcm1704 r2r dac in NOS. However, most of the time I use mqa-mp.

 

It still doesn't remove the fact that reconstruction quality is not so great at 352.8/384k rates, and even worse at lower rates. If there is sufficient analog filter, it causes problems in audio band, with less than sufficient analog filter is causes strong correlated images in ultrasonic band.

 

[Solstice380]

7 hours ago, rossco said:

I am not sure its network traffic. I just did a quick test. Using the same settings as above. I set a game to download via steam on the same machine as HQ Player is installed on, then I also set a torrent to download at the same time. I then went and listened to my test tracks and no popping. I am not sure what this signify's, maybe capacity in the network??

 

I don’t have the same network and other hardware as you, but I did experience the “pops” once when I updated the firmware in the NAA.   It was eliminated by increasing the Buffer so you might try that.  It has been recommended to go with Default first and start increasing it if you have issues.  Yours is only at 10 mSec.

[Hensema]

8 hours ago, Miska said:

 

It still doesn't remove the fact that reconstruction quality is not so great at 352.8/384k rates, and even worse at lower rates. If there is sufficient analog filter, it causes problems in audio band, with less than sufficient analog filter is causes strong correlated images in ultrasonic band.

 

I thought that 8x oversampling is sufficient in general. How much improvement can be expected with oversampling to 1536khz? Denafrips and Holo Audio make that possible with their new usb interfaces. The new Holo May might take it even further , but it is untested. I think HQplayer is a this moment limited to 1536khz? 

[zouyu513]

Dear miska
I have a 9900ks, but I run it in linear power supply and fanless way. I found that when the Turbo frequency is 4.2g, the PCM of any sampling rate can be upsampled to DSD256 with poly sinc xtr + asdm7EC, but dsd128 cannot be smoothly upsampled to DSD256. Does the DSD128 upsampling need more calculation?
In addition, if direct sdm is selected, will the convolution file be applied?

[Miska]

3 minutes ago, Hensema said:

I thought that 8x oversampling is sufficient in general. How much improvement can be expected with oversampling to 1536khz? Denafrips and Holo Audio make that possible with their new usb interfaces. The new Holo May might take it even further , but it is untested. I think HQplayer is a this moment limited to 1536khz? 

 

It usually still leaves plenty of images. But for example at 1.5 MHz Holo starts to become clean. But I think better way is to use DSD instead.

 

6 minutes ago, Hensema said:

I think HQplayer is a this moment limited to 1536khz?

Depends on the OS and backend, there's practically no technical limit as such. On Linux at the moment PCM limit is 3.072 MHz. On macOS there's no limit (to be exact the limit is a bit over 4 GHz). On Windows the limit is at the moment 1.536 MHz. If needed, it is easy to extend the rates on Linux and Windows.

 

[Yviena]

I see that the new AKM 4499 has DSD filters at Filter1 (fc = 37 kHz, DSD64 mode) Filter2 (fc = 65 kHz, DSD64 mode)

do you still recommend using 37khz, or 65khz, i remember that you yourself use 150khz on the RME dac.

[Account Closed]

18 hours ago, Miska said:

 

It usually still leaves plenty of images. But for example at 1.5 MHz Holo starts to become clean. But I think better way is to use DSD instead.

 

Depends on the OS and backend, there's practically no technical limit as such. On Linux at the moment PCM limit is 3.072 MHz. On macOS there's no limit (to be exact the limit is a bit over 4 GHz). On Windows the limit is at the moment 1.536 MHz. If needed, it is easy to extend the rates on Linux and Windows.

 

How do you extend the limit on Windows?

[Miska]

8 hours ago, bobflood said:

How do you extend the limit on Windows?

 

I just add more rates to the rate tables. ASIO doesn't have practically any rate limit since the rate is expressed in 64-bit floating point. WASAPI has the 4 GHz limit, same for ALSA on Linux.

 

[Miska]

10 hours ago, Yviena said:

I see that the new AKM 4499 has DSD filters at Filter1 (fc = 37 kHz, DSD64 mode) Filter2 (fc = 65 kHz, DSD64 mode)

do you still recommend using 37khz, or 65khz, i remember that you yourself use 150khz on the RME dac.

 

I would listen between the two and use the lower setting for less ultrasonic noise output if you cannot notice the difference. If you can notice, use the one that sounds better. At DSD64 rates it would be likely quite audible. But at DSD256 or higher when the corners become 148 and 260 kHz the difference is likely very small. With the ADI-2's I have, I still notice slight difference and have been thus using 150 kHz setting which sounds a very slightly more open to me.

 

[Yviena]

15 minutes ago, Miska said:

 

I would listen between the two and use the lower setting for less ultrasonic noise output if you cannot notice the difference. If you can notice, use the one that sounds better. At DSD64 rates it would be likely quite audible. But at DSD256 or higher when the corners become 148 and 260 kHz the difference is likely very small. With the ADI-2's I have, I still notice slight difference and have been thus using 150 kHz setting which sounds a very slightly more open to me.

 

I see this filters are scaled up right, so it becomes either 37khz x 3, or 65khz x 3 at DSD256 right?

[Miska]

45 minutes ago, Yviena said:

I see this filters are scaled up right, so it becomes either 37khz x 3, or 65khz x 3 at DSD256 right?

 

Of course x4 since 256 / 64 = 4. Not 100% sure they scale up, but IIRC they do. Depends how they are clocked.

 

[Yviena]

1 hour ago, Miska said:

 

Of course x4 since 256 / 64 = 4. Not 100% sure they scale up, but IIRC they do. Depends how they are clocked.

 

Hmm so 37khz would be 148khz at DSD256 while 65khz would be 260khz at DSD256, does this mean that the RME at 150khz is at 600khz at DSD256?

[Account Closed]

3 hours ago, Miska said:

 

I just add more rates to the rate tables. ASIO doesn't have practically any rate limit since the rate is expressed in 64-bit floating point. WASAPI has the 4 GHz limit, same for ALSA on Linux.

 

Thanks for the reply. I thought that maybe it was a setting in Windows. I ask because I have not been able to make 1.536 PCM work with my Spring 2. I have Roon/ HQP 431 on Win 10 and NAA 4 on an ultraRendu. When I select for 1.536 PCM, I get indications on Roon and HQP and the Spring 2 that it is processing at 1.536 but the network output showing in the Task Manager is only showing as 65Mbps when for 1.536 PCM (using a 32 bit word) it should be 98 Mbps. At 768 PCM it correctly outputs at 49Mbps and I get perfect sound at the DAC. When I select for 1.536, I get garbled sound at the DAC as it is not getting the full data rate called for. I don't know if this is a Windows problem or something else in the chain. If you have any ideas, please let me know.  Thanks 

[Miska]

1 hour ago, bobflood said:

Thanks for the reply. I thought that maybe it was a setting in Windows. I ask because I have not been able to make 1.536 PCM work with my Spring 2. I have Roon/ HQP 431 on Win 10 and NAA 4 on an ultraRendu. When I select for 1.536 PCM, I get indications on Roon and HQP and the Spring 2 that it is processing at 1.536 but the network output showing in the Task Manager is only showing as 65Mbps when for 1.536 PCM (using a 32 bit word) it should be 98 Mbps. At 768 PCM it correctly outputs at 49Mbps and I get perfect sound at the DAC. When I select for 1.536, I get garbled sound at the DAC as it is not getting the full data rate called for. I don't know if this is a Windows problem or something else in the chain. If you have any ideas, please let me know.  Thanks 

 

Yes, likely it won't work with the official Rendu's OS. You could try with my NAA image if it works better (I have not tested myself).

 

It also depends on the particular USB hardware. It should work with Intel USB controllers though, with my HQPlayer OS and NAA images.

 

When you are using NAA, the OS where HQPlayer is running doesn't  matter at all from that perspective since output is to a network endpoint bypassing the OS audio stack entirely.

 

[Miska]

2 hours ago, Yviena said:

Hmm so 37khz would be 148khz at DSD256 while 65khz would be 260khz at DSD256, does this mean that the RME at 150khz is at 600khz at DSD256?

 

No, I think the filters in the previous chipset work quite differently. I need to get the new one to measure how it actually behaves.

 

[Yviena]

23 minutes ago, Miska said:

 

No, I think the filters in the previous chipset work quite differently. I need to get the new one to measure how it actually behaves.

 

Hmm maybe 37khz at DSD256 is now 150khz from old AKM 4490?

[Miska]

1 minute ago, Yviena said:

Hmm maybe 37khz at DSD256 is now 150khz from old AKM 4490?

 

It is hard to say, but I think they have made some drastic changes to the implementation. I'll try to get some new DACs that have filter switching over the spring. Both Topping and ADI-2 still retain this feature?

 

I have just the EVGA NuAudio card, but it doesn't have option for switching the filter. But since volume control doesn't work in DSD mode, I assume it is in DSD Direct mode.