HQ Player

[Miska]

Just a short reminder, in case you have a problem with audio output while upsampling on Windows, remember to check your selected power profile to make sure you are running high performance profile:

[blue2]

Good point! Makes a difference even though the detailed settings seem to be the same MS 'plan settings' DSD64 clean as a whistle now on one PC and DSD128 good on the other.

 

Windows does not manipulate the CPU or motherboard or multipliers or FSB or anything. Power Options in the Control Panel is simply a factor that is used during task-scheduling. When processes run, the Windows task-scheduler allocates a specific number of CPU cycles to different processes and switches back and forth to enable pre-emptive multitasking. Normally, it assigns as many CPU cycles as are available in order to run everything as fast as possible. What Windows lets you do is to assign a limit to how much CPU capacity is usable. As a result, there will be CPU cycles that remain unused, which allows the CPU to remain at a lower temperature.

- above is my synopsis of a googled web site

 

I think whatever is going on here the FPU load is not considered.

 

PS BTW the 'High Performance' option is normally hidden in the Control Panel, Power Options dialogue.

[Miska]

- above is my synopsis of a googled web site

 

That is not entirely accurate, because all modern CPUs have adjustable operating point; that is both clock frequency and voltage. For example my CPU is at the moment running at 1.2 GHz instead of the 3.5 GHz.

 

It is commonly called dynamic frequency scaling:

https://en.wikipedia.org/wiki/Dynamic_frequency_scaling

 

The implementation in Intel CPU's is called SpeedStep while the AMD variant is called Cool'n'Quiet.

 

One way to deal with these is to use ACPI P-states.

 

Overall, OS should understand to keep the frequency high enough. But seems like in some cases Windows doesn't understand that the application is "in hurry" to meet audio delivery deadlines and that it should increase clock frequency. On Linux this seems to work better (where you can also choose between different frequency governors), but at least for me the "ondemand" governor seems to work just fine.

[giannicimbalo]

Just a question for everyone else: have you managed to control HQPlayer with TeamViewer?

 

No problems at all with VNC, on ipad and an IR mouse is perfect, btw i use drag&drop from Filexplorer

 

Again..i have a nightmare with NAD 3.4 on Winserver 2012R2. Clean installation, updates, library requested installed..but..nothing....i have "transplanted" the dll's from win10, so..the 3.2 works..nope the 3.4...any help will be welcomed

Sorry for my terrific English..

Ciao!

[bodiebill]

Just downloaded the last HQP Windows version 3140. The last version that worked for me was 3133. All versions after that crash when I start to play a song. Alas, same behaviour with 3140. I did it in Windows Server 2012 R2 GUI mode to make sure it is not a core mode problem.

Any suggestions how I could make this work?

 

Just a short update that I recently installed Windows Server 2016 in GUI mode.

 

The bad news: the above problem is unchanged; HQ vs 3140 does not work; it crashes trying to play a track.

 

The good news: HQ vs 3 sounds better than ever and now competes with my favourites HySolid and Bughead. It seems WS2016 removed the last bit of digital harshness.

 

Any others who tried HQ with WS2016? And if yes, which HQ version?

[EdmontonCanuck]

Until such time as Miska supports Windows Server (any version), I consider it unwise to try and run HQP on it. I just don't need the headache. Win10 sounds good enough to my ears.

[satbox]

Hi,

When I upsample to SDM the speaker setup has no effect (-4dB on the left speaker). On the other hand, when I up sample to PCM, it's OK. Is it normal?

[nikander]

Good morning,

 

I have a couple of (stupid) questions:

 

I have started to use the digital volume controll in HQ Player and it sounds fantastic (so I can get rid of an analog preamp).

As far as I have understood it should be advantageous to convert 16 Bit files to 24 bit files for digital volume controll to avoid dynamic range loss while attenuation.

My first question:

I upsample every file up to 192 khz (using "poly-sinc") and I use for dithering "NS9" - are the files also converted from 16 to 24 bit (on the main Signalyst page this is stated:

"...

•Software upsampling / downsampling with selectable algorithm up to 32-bit 1.536 MHz or down to 16-bit 32 kHz PCM (20 options, most "apodizing")...")?

Second question:

I use Acourate DRC - either with online convolution or offline convolution.

If I do offline convolution the flac files are all converted from 16 bit to 24 bit and also dithered and normalized in advance and then stored.

If they are dithered in advance: is it sensible to use another dithering afterwards in HQ Player (like NS9) or do I have to use that dithering in HQ Player anyways as long as I use its digital volume controll?

 

regards,

 

Nikander

[Nikhil]

If they are dithered in advance: is it sensible to use another dithering afterwards in HQ Player (like NS9) or do I have to use that dithering in HQ Player anyways as long as I use its digital volume controll?

 

Noise shaping using dither is an integral part of re-sampling. Anytime you upsample a file dither is used in the noise shaping procees to move the unwanted noise into the (inaudible) ultra high frequency range.

 

On the subject of Volume Control, I would not recommend using the HQP volume control as the primary control. Use it instead as a fine control, Jussi recommends just backing off 100% by a few clicks to avoid clipping. A good analog volume control on your preamp is still the best way to preserve dynamic range. Using only the HQP volume control will make the music flat or thin.

[bodiebill]

Until such time as Miska supports Windows Server (any version), I consider it unwise to try and run HQP on it. I just don't need the headache. Win10 sounds good enough to my ears.

 

Makes sense. However, as I want to run my audio PC diskless (good for SQ), I need a Windows Server OS on my iSCSI server = control PC. And as that runs HQP vs3 without problems, I am happy to use it...

[nikander]

@Nikhil: thanks a lot for your comments. But can you explain then why HQ Player offers "dither - None" separately from upsampling?

 

"...

On the subject of Volume Control, I would not recommend using the HQP volume control as the primary control. Use it instead as a fine control, Jussi recommends just backing off 100% by a few clicks to avoid clipping. A good analog volume control on your preamp is still the best way to preserve dynamic range. Using only the HQP volume control will make the music flat or thin. ..."

 

 

This is a very interesting topic - adding an analog preamp in the chain can be worse due to noise and distortion and worse channel equality/crosstalk ( by the way: Linn has ommitted analog preamp section in their newest products) and I could imagine: if you decide to use a digital preamp solution HQ Player should be one of the world`s best alternatives but I wonder what Miska says himself to that topic.

As far as I remember there are quite a few statements by him concerning this topic and also he has published own measurements (digital vs. analog volume attenuation in a DAC).

 

best wishes,

 

Nikander

 

 

P.S.: I have a Lynx Hilo DAC which has an own analog preamp and 4 outputs (2 with analog volume controll and two line out XLR which only offer digital attenuation but are better equipped (2 DACs instead of 1 per channel). I chose a lowest volume gain as possible (+ 0 dBV) and highest digital output (0 DBFS) so I need relatively little digital attenuation (-30). In this setup the Lynx sounds better than with analog volume controll but I don`t know if this is due to the better equipped line outs).

[nikander]

here are some of his comments:

 

http://www.computeraudiophile.com/f8-general-forum/digital-vs-analogue-volume-control-26184/index2.html

 

P.S.:

I have just noticed that my active monitors have adjustable output gain (just as Miska recommends) - so I could lower that and need even lower digital attenuation.

[Nikhil]

@Nikhil: thanks a lot for your comments. But can you explain then why HQ Player offers "dither - None" separately from upsampling?

I will defer to Jussi to explain how he handles the signal internally in HQP.

 

This is a very interesting topic - adding an analog preamp in the chain can be worse due to noise and distortion and worse channel equality/crosstalk ( by the way: Linn has ommitted analog preamp section in their newest products) and I could imagine: if you decide to use a digital preamp solution HQ Player should be one of the world`s best alternatives but I wonder what Miska says himself to that topic.

As far as I remember there are quite a few statements by him concerning this topic and also he has published own measurements (digital vs. analog volume attenuation in a DAC).

 

...

P.S.: I have a Lynx Hilo DAC which has an own analog preamp and 4 outputs (2 with analog volume controll and two line out XLR which only offer digital attenuation but are better equipped (2 DACs instead of 1 per channel). I chose a lowest volume gain as possible (+ 0 dBV) and highest digital output (0 DBFS) so I need relatively little digital attenuation (-30). In this setup the Lynx sounds better than with analog volume controll but I don`t know if this is due to the better equipped line outs).

 

 

Go with what sounds best to your ears ...

In my opinion, however, HQP's volume control is best used as a fine control.

[ebrandon]

I have a problem getting HQPLayer/NAA working with a new DAC I just got today.

 

Background: I have no problem getting HQPlayer working on my Mac Mini, connected via Ethernet to another Mac Mini running NAA, connected via USB to an OPPO 105D.

 

Now I am trying to get the same HQPlayer on the Mac Mini, via wifi to a MacBook Air running NAA, connected via USB to a Woo Audio WA8 Eclipse DAC/Headphone amp.

 

HQPlayer "sees" the Woo Audio under "NetworkAudioAdapter" in HQPlayer preferences. It lets me select it.

But HQplayer does not let me add any music to the player.

 

And the NAA on the MacBook Air reports immediate connection & disconnection.

[/Applications/networkaudiod] (555): discovery from 192.168.2.227:52398

[/Applications/networkaudiod] (555): connection from 192.168.2.227:56058

[/Applications/networkaudiod] (555): CoreAudio default device id: 55

[/Applications/networkaudiod] (555): CoreAudio devices: 3

[/Applications/networkaudiod] (555): CoreAudio device: Built-in Microphone (48) - Apple Inc.

[/Applications/networkaudiod] (555): CoreAudio device: Built-in Output (55) - Apple Inc.

[/Applications/networkaudiod] (555): CoreAudio device: WooAudio (39) - XMOS

[/Applications/networkaudiod] (555): begin disconnection

[/Applications/networkaudiod] (555): disconnected 192.168.2.227:56058

 

Also, the port reported by NAA for connect & disconnect is different each time.

 

Is this a firewall problem? In my router?

 

Thanks!!

 

[uPDATE] The connection between the MacBook Air the Woo Audio DAC/Amp works. I can play locally stored music just fine. The problem is between HQPlayer and the NAA, I think.

[satbox]

Hi,

When I upsample to SDM the speaker setup in HQ has no effect (-4dB on the left speaker). On the other hand, when I upsample to PCM, it's OK. Is it normal?

Good morning,

Have you ever experienced this problem? What can I do in order to have the -4dB on the left speaker applied when I upsample to SDM?

[HuaHin]

Hi Miska,

I am new in Computer Audio and found HQP an excellent DSD player. I need your help for setup parameters:

What would be your options liveliness, vividness setting for the MSB Analog DAC using HQP?

[tboooe]

@Miska, when I create the wav files in REW, at what sample rate should I select for best flexibility and sound quality? I normally upsample everything to DSD256. In this case, does it matter what sample rate I use? If I choose to not upsample my redbook PCM, should I upsample the wav to multiples of 44.1? If so, I guess I am confused what I should do when I listen to PCM that I have at 48 and 96?

 

Do I need to enable Extend HF?

[Miska]

@Miska, when I create the wav files in REW, at what sample rate should I select for best flexibility and sound quality? I normally upsample everything to DSD256. In this case, does it matter what sample rate I use? If I choose to not upsample my redbook PCM, should I upsample the wav to multiples of 44.1? If so, I guess I am confused what I should do when I listen to PCM that I have at 48 and 96?

 

Do I need to enable Extend HF?

 

Best option is to output highest rate the filter design tool supports. Acourate for example can do 384k these days and with similar option like "Expand HF". Most of the time if you create filters for 192k rate or higher, you don't need HF extension. If you create filters for example for 44.1k rate, you'd want to have to have the HF extension for playing hires content (96k and such).

[Miska]

Good morning,

Have you ever experienced this problem? What can I do in order to have the -4dB on the left speaker applied when I upsample to SDM?

 

You need to have the Direct SDM unchecked (DSDIFF/DSF Settings dialog)...

[Miska]

@Nikhil: thanks a lot for your comments. But can you explain then why HQ Player offers "dither - None" separately from upsampling?

 

Dither is needed whenever there is more precision in the data than the output word length can represent. The "none" dither setting is mostly useful if one wants to play the source file bit-perfect for some reason...

[Miska]

I upsample every file up to 192 khz (using "poly-sinc") and I use for dithering "NS9" - are the files also converted from 16 to 24 bit (on the main Signalyst page this is stated:

 

Yes, internal precision is 64-bit floating point which is then converted to the specified (or detected) integer output resolution.

 

I use Acourate DRC - either with online convolution or offline convolution.

If I do offline convolution the flac files are all converted from 16 bit to 24 bit and also dithered and normalized in advance and then stored.

If they are dithered in advance: is it sensible to use another dithering afterwards in HQ Player (like NS9) or do I have to use that dithering in HQ Player anyways as long as I use its digital volume controll?

 

Yes, you need dither if you upsample and/or use the digital volume control in HQPlayer.

 

But overall it is better to export the filter as WAV from Acourate and then configure it in HQPlayer's convolution engine. This way you retain much better precision (64-bit FP) throughout the processing chain without intermediate reduction like with offline processing.

[nikander]

thanks a lot for your explanations, Miska!

As usual best service!

[satbox]

You need to have the Direct SDM unchecked (DSDIFF/DSF Settings dialog)...

Thank you for your help, it's OK now.

[udis]

I've read so many great things about HQ player and the improvement NAA delivers with Roon and mRendu. I'm currently using Roon on a SonicTranporter and a mRendu. I'm thinking of getting a server capable of driving HQ player but I have a few questions. I'm using a Bricasti M1 SE DAC. The Bricasti stipulates not to do up-sampling. In addition, it has the minimum and linear filter setting choices built in. So most likely I'll be using HQPlayer just for the NAA Roon configuration. My questions are: should I expect a significant SQ improvement in changing the Roon configuration? Also, since I likely won't need the horsepower to perform upsampling, what type of server hardware should I invest in? Thanks for any help...

 

I am on a similar boat. Looking to replace my Aries ( feeding m1 se) with hqplayer and mR. Please share your findings if and when you take the plunge!

[aeo12]

Love the sound of HQPlayer on my Mac, but not the library management. Right now I'm dragging and dropping from JRiver Media Player 21 and that seems to give me most of what I want-- most essentially to organize both Flac and DSD by album artist, with albums in order by year. Any other suggestions for a good Mac music library management tool to try?