HQ Player

[Miska]

38 minutes ago, ted_b said:

I guess I would use the highest rez (352k) and go from there?  Questions: what happens in these two scenarios, using the 352.8k convolution filters..

1) I upsample PCM from 44k to 32fs

2) I upmodulate DSD64 to DSD256 (say, with ASDM7EC)

 

Yes, that would be optimal starting point with enough filter bandwidth coverage.

 

In both cases filter is processed at the source sampling rate.

 

[sledwards]

This from another post from Miska on topic:

 

Convolution is performed at the source rate. (except DSD-to-PCM conversion case, where it is performed after conversion to PCM at 1/16th of the original rate)

 

You don't need to match the sample rate, but instead provide filter at highest rate of your content if possible. Preferably at 352.8 kHz rate (DXD material). HQPlayer then scales the filter to match the source rate, what ever that will be.

[chriss0212]

2 hours ago, Miska said:

 

Both result in same processing. Essentially text file parser ends up as same kind of filter spec as you would enter manually in matrix setup.

 

Perfect...

I would like to know, if it is possible to linearize the speaker in one line for pipeline1 source ch.1 and make some house curve corrections on pipeline 3 source channel 1.

I am aware, that i have to take care, that i don't push anything above 0dB.

 

If my idea is not possible in the matrix setup: would it be a solution, to make the linearization in the convolution setup and house curve in the matrix?

 

I found this in the manual:
 

Quote

Convolution engine applicable to filter(s) defined in Process, can be selected from two choices, Overlap-add which is the default and recommended and Overlap-save which is alternative method.

 

Can you please discribe a bit more what it means?
 

Greetings

 

Christian

[Miska]

3 hours ago, chriss0212 said:

I would like to know, if it is possible to linearize the speaker in one line for pipeline1 source ch.1 and make some house curve corrections on pipeline 3 source channel 1.

 

No, both need to be on the same pipeline. Remember that pipelines are parallel. So you would have same source, processed by two different processes and then mixed together. This is not something you'd want in this case. Instead you need to perform these processes in series.

 

3 hours ago, chriss0212 said:

If my idea is not possible in the matrix setup: would it be a solution, to make the linearization in the convolution setup and house curve in the matrix?

 

No, the two are mutually exclusive. Just put both in matrix, on the same pipeline.

 

3 hours ago, chriss0212 said:

I found this in the manual:

 

Two different ways to do same thing, so I'd stick with the recommended / default one.

 

[chriss0212]

1 hour ago, Miska said:

 

No, both need to be on the same pipeline. Remember that pipelines are parallel. So you would have same source, processed by two different processes and then mixed together. This is not something you'd want in this case. Instead you need to perform these processes in series.

 

 

No, the two are mutually exclusive. Just put both in matrix, on the same pipeline.

 

 

Two different ways to do same thing, so I'd stick with the recommended / default one.

 

Ok… Good to know.

 

that the pipeline setup will not work was something i already thought… but my hope was that convolution and pipeline work in series ;)

 

whould simply help, to see the desired tone balance in the plot.

 

But load a filter in an unused pipeline channel (just to se which house curve is use) should work right?

 

Greetings

 

Christian

[Miska]

16 minutes ago, chriss0212 said:

that the pipeline setup will not work was something i already thought… but my hope was that convolution and pipeline work in series ;)

 

Pipeline includes more advanced convolution, so you don't need the simplified one.

 

16 minutes ago, chriss0212 said:

But load a filter in an unused pipeline channel (just to se which house curve is use) should work right?

 

Yes, that works for the purpose.

 

[SunYang]

22 hours ago, Miska said:

You can set number of output channels to 4 and use matrix pipeline setup to route channels.

Thanks, Miska. I am asking because if I try setup like this, HQP don’t start to play any track.  2 Channels setup works OK. Do you have some idea how can I solve it?

[Miska]

3 hours ago, SunYang said:

Thanks, Miska. I am asking because if I try setup like this, HQP don’t start to play any track.  2 Channels setup works OK. Do you have some idea how can I solve it?

 

What kind of error do you get? You have a 4-channel (or more) DAC attached? Which OS and backend?

 

[JohnDonaldson]

@MiskaWith Desktop V4.12 I can not get 24/352.8 and 24/192 PCM files to play correctly. The first few seconds play normally then change to continuous static. With V4.10.1 this does not happen. I have confirmed this on two computers. I have not tested versions between 4.10.1 and 4.12.

[Miska]

22 minutes ago, JohnDonaldson said:

@MiskaWith Desktop V4.12 I can not get 24/352.8 and 24/192 PCM files to play correctly. The first few seconds play normally then change to continuous static. With V4.10.1 this does not happen. I have confirmed this on two computers. I have not tested versions between 4.10.1 and 4.12.

 

Chord DAC? Which OS? Which source file format?

 

[JohnDonaldson]

30 minutes ago, Miska said:

 

Chord DAC? Which OS? Which source file format?

 

The DAC is a Holo May. Both computers run Windows 10 Pro.  The files I used to test are 24/192 AIFF and 24/352.8 WAV.

[blueninjasix]

2 hours ago, JohnDonaldson said:

@MiskaWith Desktop V4.12 I can not get 24/352.8 and 24/192 PCM files to play correctly. The first few seconds play normally then change to continuous static. With V4.10.1 this does not happen. I have confirmed this on two computers. I have not tested versions between 4.10.1 and 4.12.

Is it static in both channels or only the right one? I've long had a right channel  problem with Chord Qutest at 768kHz from HQPlayer desktop but never at 705.6kHz.

[pier]

On 6/11/2021 at 3:14 AM, kelvinwsy said:

I just tried the new polysinc-xtr-short-lp with ADSMEC7 and prefer it to the polysinc gauss filters..(these soften the music.. Tension in crescendo passsges in orchestra pieces gone??)

Add on ASDM7EC 512fs instead of ASDM7EC makes it even sounding better.. The decay in drums and pianos is so nicely clear until the last fade..The SQ is so airy with full punchy bass

Easier load than the full xtr-lp i like in 4.10.3

 

 

 

 

 

I can agree, this is really. Pianos are fabs. thx!

[Miska]

6 hours ago, JohnDonaldson said:

The DAC is a Holo May. Both computers run Windows 10 Pro.  The files I used to test are 24/192 AIFF and 24/352.8 WAV.

 

How are your ASIO driver buffer settings?

 

[copy_of_a]

@Miska

 

There's a little inconvenience in HQP4Desktop (Mac) when using Auto Gain in conjunction with convolution.

Not sure it has already been mentioned in this 846 pages thread ...

 

I understand convolution by design produces some delay, which of course makes sense.

Now, when you play back a playlist containing tracks with different Auto Gain settings there's a jump in volume (gain) at the end of each track. The point in time before the tracks end is about the length of the delay the convolution needs to process the filters.

Or in other words: the Auto Gain value of the subsequent track is applied to the current track while the subseqent track is preloaded/preprocessed.

Hope this is somehow reasonably explained?

Maybe you could look into this?

 

_____

 

I really like the new poly-sinc-gauss filters (especially the "long" variant).

Elegant, fluid, nicely detailed, spaceous and totally coherent sounding... I really feel they make a difference to all the other poly sinc filters. Great!

 

Many thanks!

[giordy60]

19 hours ago, Miska said:

How are your ASIO driver buffer settings?

 

Jussi how important is the asio setting on Denafrips?
I (for example) I set 256 samples (6.8 ms of latency) is it too low in your opinion?
when I opened the settings it was set to 2048 samples.
I attach image.

 

[Miska]

2 hours ago, copy_of_a said:

Hope this is somehow reasonably explained?

Maybe you could look into this?

 

Yes, I'll look into it. I think I need to move the place where the gain is applied to earlier place.

 

[Miska]

2 hours ago, giordy60 said:

Jussi how important is the asio setting on Denafrips?
I (for example) I set 256 samples (6.8 ms of latency) is it too low in your opinion?
when I opened the settings it was set to 2048 samples.
I attach image.

 

 

It is very low for Windows. For example WASAPI defaults to 10 ms and it is very likely to produce drop-outs on Windows. 100 ms is good value on Windows machines. It is best to just set the ASIO buffer size to as large as possible. That will reduce CPU load and likelihood of dropouts.

 

[giordy60]

thanks Jussi

[Luca72c]

And thanks for the gauss xla filter! It's super!

[LoryWiv]

11 hours ago, Miska said:

 

It is very low for Windows. For example WASAPI defaults to 10 ms and it is very likely to produce drop-outs on Windows. 100 ms is good value on Windows machines. It is best to just set the ASIO buffer size to as large as possible. That will reduce CPU load and likelihood of dropouts.

 

Jussi, if the buffer size is set to maximum in my Win10 ASIO driver (max. is 16384 for my Singxer SU-1 driver), should I also use high buffer setting (EG - 100ms) in the HQPlayer settings dialogue box? is there any trade-off between sound quality and high buffer size to minimize cpu usage and dropouts?

[Miska]

5 hours ago, LoryWiv said:

Jussi, if the buffer size is set to maximum in my Win10 ASIO driver (max. is 16384 for my Singxer SU-1 driver), should I also use high buffer setting (EG - 100ms) in the HQPlayer settings dialogue box? is there any trade-off between sound quality and high buffer size to minimize cpu usage and dropouts?

 

No, set HQPlayer "Buffer Time" to "Default" which means HQPlayer will use the driver setting.

 

[Mark Labbett]

Hi i am using the demo version of HQ player and feel like i need to go to university to understand the settings(yes i have read the manual) could i please get some help my computer is an i7 4790k and 16GB ram(let me know if that's not enough info) my DAC is a Eastern Electric MiniMax Supreme i would like to play my cd rips in DSD128, my hi res(varying bit depths(16-32) and sample rates(48-384)) downloads in DSD128 and all DSD to DSD128. I have it playing but even though i have picked an apodization filter the counter is still going up. If someone could give me the settings for a starting point to achieve the output stated that would be great. Also when it first starts playing i get some crackling come through my speakers that is not present using Foobar or Jriver is that normal? Also i am not sure if my DAC only does DoP i think thats the case but not sure, Thanks for the help. By the way this is a fantastic sounding player.

[k6davis]

12 minutes ago, Mark Labbett said:

Hi i am using the demo version of HQ player and feel like i need to go to university to understand the settings...

Welcome!

 

There is a lot of depth to HQP, but even using it in a basic way will give you fantastic results. I can give you some basic recommendations.

 

Your i7 4790k is older, but it is likely that it can handle most settings at DSD128. The modulators with "EC" in their names are the highest quality - and the most demanding on your CPU. ASDM7EC is the runaway favorite, but my preference is for DSD5EC. It all depends on your equipment, tastes, room, etc. When it comes to filters, poly-sinc-ext2 arguably offers the best "bang for the buck". It's not very demanding on your server yet it offers excellent sound across a variety of types of music. A few years ago, it was Jussi's (HQP's developer) favorite filter. Since then, he's added many new filter options and lots of us here prefer some of those. Other filters stand out from poly-sinc-ext2 in various ways, but I find poly-sinc-ext2 to be at minimum an excellent baseline, and it remains a favorite of mine for its overall sound. 

 

Those filter & modulator settings will offer you superb sound as you learn more and experiment to see what works for you.

 

I'll leave your question about the crackling to others who could offer better technical advice, but that's definitely not normal.

[Mark Labbett]

51 minutes ago, k6davis said:

Welcome!

 

There is a lot of depth to HQP, but even using it in a basic way will give you fantastic results. I can give you some basic recommendations.

 

Your i7 4790k is older, but it is likely that it can handle most settings at DSD128. The modulators with "EC" in their names are the highest quality - and the most demanding on your CPU. ASDM7EC is the runaway favorite, but my preference is for DSD5EC. It all depends on your equipment, tastes, room, etc. When it comes to filters, poly-sinc-ext2 arguably offers the best "bang for the buck". It's not very demanding on your server yet it offers excellent sound across a variety of types of music. A few years ago, it was Jussi's (HQP's developer) favorite filter. Since then, he's added many new filter options and lots of us here prefer some of those. Other filters stand out from poly-sinc-ext2 in various ways, but I find poly-sinc-ext2 to be at minimum an excellent baseline, and it remains a favorite of mine for its overall sound. 

 

Those filter & modulator settings will offer you superb sound as you learn more and experiment to see what works for you.

 

I'll leave your question about the crackling to others who could offer better technical advice, but that's definitely not normal.

Thank you for the reply, do i set the filters in both the DSD and PCM areas and what about DSD source settings?

I have set the filters recommended and the apod counter is still going up???