HQPlayer kick start guide & feature requests

[Miska]

Any option to start HQPlayer in full screen mode at windows start up?

 

Yes, you can edit the shortcut to give --fullscreen command line option to the executable and then add it to the Startup-folder.

 

Note, to get truly fullscreen view instead of maximized, enable the "Fullscreen" option in Settings dialog.

[planner]

Thanks Miska. Works like charm.

[bibo01]

Feature request -

 

I would like Miska to add the possibility of saving to disk conversion of files PCM>DSD64. In my view conversion to DSD64 would be enough because the heavy computing becomes off-line and it still allows on-the-fly upsampling to DSD128 or more which is much less cpu intensive. Furthermore, file size is not too big.

[hifi25nl]

I have been using a lot recently both HQlayer and HQplayer embedded and there is something that I would like to be implemented:

 

 

1) MinringFIR and polynomial filters in HQplayer. I would like some automatism maybe as an option that can be selected or not. If I am using one of these filters and bit rate limit is set to 192000 for example, I would like an automatic switch to the immediately lower frequency that is compatible with these filters. Example: 48k multiple should use 192k but 44,1k should use instead 176,4k. If the limit is set to 192k the 192k files should not be resampled i (in PCM, eventually only dithered).

 

 

2) I agree with bibo01 suggestion. I would push it further. Why not give the opportunity to save all upsampled music to a file? Or eventually a different software for file conversion..?

 

 

3) Similar to 2) but without the saving, I would like an option for caching in memory seconds or minutes or the full processed upsampled file. Yes, this would take maybe a long time before playing occurs (and a lot of memory) but it would permit to users of less powerful computer to test for example DSD512 without problems.

[Miska]

1) MinringFIR and polynomial filters in HQplayer. I would like some automatism maybe as an option that can be selected or not. If I am using one of these filters and bit rate limit is set to 192000 for example, I would like an automatic switch to the immediately lower frequency that is compatible with these filters.

 

This already happens...

 

Example: 48k multiple should use 192k but 44,1k should use instead 176,4k. If the limit is set to 192k the 192k files should not be resampled i (in PCM, eventually only dithered).

 

For the case where filter cannot do 1:1 conversion, you need to select "none" as filter to not resample. Otherwise the GUI would be in conflict with actual operation.

 

2) I agree with bibo01 suggestion. I would push it further. Why not give the opportunity to save all upsampled music to a file? Or eventually a different software for file conversion..?

 

It should be another product altogether. You would probably also want output in separate files and be able to name the files somehow, contain metadata, etc. I've been thinking about making a separate conversion utility, but it takes so much time and I'm already busy with HQPlayer. So I would like to do it, but just don't have time.

 

I have such things for internal algorithm testing purposes, but those are far from being nice to use.

 

3) Similar to 2) but without the saving, I would like an option for caching in memory seconds or minutes or the full processed upsampled file. Yes, this would take maybe a long time before playing occurs (and a lot of memory) but it would permit to users of less powerful computer to test for example DSD512 without problems.

 

I'm not sure it would be practical for the purpose. Let's say you would preprocess a 60-minutes long album. If the computer is not fast enough to perform processing in realtime, upsampling such file would thus take more than 60 minutes to process. And 60 minutes of DSD512 would take 151 gigabytes of RAM.

[Miska]

I would like Miska to add the possibility of saving to disk conversion of files PCM>DSD64. In my view conversion to DSD64 would be enough because the heavy computing becomes off-line and it still allows on-the-fly upsampling to DSD128 or more which is much less cpu intensive. Furthermore, file size is not too big.

 

Note, doing PCM -> DSD64 and then to DSD128 is completely different thing than going straight PCM -> DSD128...

[hifi25nl]

Originally Posted by hifi25nl

1) MinringFIR and polynomial filters in HQplayer. I would like some automatism maybe as an option that can be selected or not. If I am using one of these filters and bit rate limit is set to 192000 for example, I would like an automatic switch to the immediately lower frequency that is compatible with these filters.

 

This already happens...

 

Sorry, I have written HQplayer but I should have written HQplayer embedded. I have made a lot of configuration files with a corresponding shortcut. This way I can switch easly different configurations for HQplayer embedded. In PCM-->DSD64 for example, 44.1k files need 2822400, 48k 3072000. I guess this is by purpose. Maybe HQplayer embedded leave to the programmer of the interfaces the choice of an automatism.

[Jabs1542]

Note, doing PCM -> DSD64 and then to DSD128 is completely different thing than going straight PCM -> DSD128...

 

That's interesting - please explain.

[bibo01]

Note, doing PCM -> DSD64 and then to DSD128 is completely different thing than going straight PCM -> DSD128...

Are you saying that a two stage conversion PCM->DSD64->DSD128 is not as good as yours PCM->DSD128?

[Miska]

Are you saying that a two stage conversion PCM->DSD64->DSD128 is not as good as yours PCM->DSD128?

 

Yes... The "-2s" two stage poly-sinc conversion is technically pretty much as good as the non-2s poly-sinc (difference is very small and depends on which way you look at things), but the intermediate is minimum 8x rate at 64-bit floating point. And has been designed specifically as a two stage conversion.

 

DSD64 has less resolution, in both SNR and bandwidth terms, compared to DSD128. The intermediate format should never have less resolution than the final format. PCM->DSD256->DSD128 would give you much better result.

[Esprit]

It depends on a DAC. With iFi iDSD Micro, DSD256 is the limit at the moment. I will update things soon to add support for the fresh new non-DoP support for iFi iDSD Nano/Micro and then the limit will be DSD256 and DSD512 respectively.

Micro DSD256 is with or w/out DoP?

[Miska]

Micro DSD256 is with or w/out DoP?

 

With Micro DSD256 with DoP and DSD512 without DoP.

 

Nano can do DSD128 with DoP and DSD256 without.

[dallasjustice]

I noticed a recent HQP update which allows for higher than 192khz sample rates over ASIO Ravenna for the Merging Hapi. DSD is not documented by Merging over Ravenna ASIO using third party s/w. Jussi, can you explain more about this? See page 5 here:

http://www.merging.com/uploads/assets/Merging_pdfs/Pyramix_8/Merging_Ravenna_ASIO_and_CoreAudio_Guide.pdf

[Miska]

I noticed a recent HQP update which allows for higher than 192khz sample rates over ASIO Ravenna for the Merging Hapi. DSD is not documented by Merging over Ravenna ASIO using third party s/w. Jussi, can you explain more about this?

 

I don't know about Ravenna, but I got success report for Hapi with ASIO after one setting change (that was hidden for a while, but always existed). I don't own any Merging gear myself, so I cannot comment in further detail.

[dallasjustice]

I don't know about Ravenna, but I got success report for Hapi with ASIO after one setting change (that was hidden for a while, but always existed). I don't own any Merging gear myself, so I cannot comment in further detail.

 

Thanks Jussi. I have a few questions about HQP convolution engine and measurement:

1. I use the Acourate logsweep recorder to take measurements. It requires that I select an ASIO device in the output. I have 4CH. All 4 have respective crossovers, delay and speaker correction in place for measurement. Currently, I use Acourate ASIO and set my filters and routing inside the Acourate Convolver matrix which works well. How could I send my logsweep signal through the HQP convolution engine instead of AC? Does HQP have an ASIO driver to loop signals back through it's DSP engine?

2. I read the manual and it seems like only one filter is required for all sample rates. If I select expand HF, then HQP converts those lower sample rate .wav filters into the required higher sample rate?

 

Thanks,

Michael.

[Miska]

Thanks Jussi. I have a few questions about HQP convolution engine and measurement:

1. I use the Acourate logsweep recorder to take measurements. It requires that I select an ASIO device in the output. I have 4CH. All 4 have respective crossovers, delay and speaker correction in place for measurement. Currently, I use Acourate ASIO and set my filters and routing inside the Acourate Convolver matrix which works well. How could I send my logsweep signal through the HQP convolution engine instead of AC? Does HQP have an ASIO driver to loop signals back through it's DSP engine?

 

For cross-overs where you need copy of one input channel to multiple output channels is not yet supported on HQPlayer. So this channel-copy routing part is missing. So you can for example run low-pass for the LFE channel and high-pass for other channels, but channels are statically routed 1:1 at the moment.

 

This is next item I am going to work on. Part of the change went in at HQPlayer 3.5 release and I hope to get next iteration in at 3.6.

 

Of course you can still play to Acourate's ASIO convolver at the moment, which allows you to play back DSD files through PCM conversion.

 

You can perform full measurement with Acourate tools and just use the final filter on HQPlayer. Convolution itself is just a standard mathematical operation and the result should be pretty much the same regardless of convolution engine used.

 

2. I read the manual and it seems like only one filter is required for all sample rates. If I select expand HF, then HQP converts those lower sample rate .wav filters into the required higher sample rate?

 

It can convert filters both up and down. It is better to convert filters down than up. Only thing Expand HF changes is that when converting up, it extends the frequency response as flat from the original fs/2 frequency to the new fs/2 frequency. But when possible it is better to create filters with extended frequency response from the beginning, instead of doing it afterwards. Acourate can perform this expansion already at filter creation time. If you can get filter at 352.8/384k (DXD) rate, that would be optimal, because then you don't ever need Expand HF. But already filter at 192k is fine for most content without Expand HF (because source content rarely contains relevant frequencies above 96 kHz)...

[dallasjustice]

You can perform full measurement with Acourate tools and just use the final filter on HQPlayer. Convolution itself is just a standard mathematical operation and the result should be pretty much the same regardless of convolution engine used.

My question about the Hapi is related to these questions concerning HQP's convolution engine. I currently am not using my HQP and I don't think my version is being updated anymore. The Hapi doesn't work with Acourate Convolver due to its unusual Ravenna ASIO driver, I am told by an owner.

 

I was considering using HQplayer to play native DSD with Hapi and still use my convolution filters. I think it would be very interesting to hear that, especially with some of my DSD files. I know it's still possible to create filters using my Lynx Hilo or other device and use those filters with the playback on Hapi or Exasound e28. As you know, I believe it's best to measure and playback on the same device. I know you disagree with me on that. :-)

 

When you do your convolution update, do you think you will write an ASIO driver to allow the logsweep recorder to send the sweep through HQP's convolution engine? That would solve this problem and allow me to use HQP's convolution engine with Hapi, should I buy one. Merging Hapi/Horus are the only devices I know of which can pass >2 channels over Ethernet, other than some HT gear. I sent an email to Merging and didn't get a response. This is the only documentation I could find concerning the Ravenna ASIO driver. It's interesting that page 5 says that the Ravenna ASIO driver only supports up to 192khz. And you have a customer that's passed DSD through it using HQP?

 

Michael.

[junker]

Is there SACD .ISO support?

[Geoffrey Armstrong]

I have a new request. I keep running into a problem from time to time where the selection of DSD settings won't stick. With a DSD capable DAC using either proprietary ASIO drivers or a DAC which only supports DOP, the same problem occurs.

 

I want to use DSD settings by default; but after HQP is quit and re-started the settings often revert to the PCM settings.

 

I usually get the DSD settings to be retained in between restarts after a few attempts. It seems that if I play music using DSD settings, after a few times they will be remembered.

 

What I'd like to see is a switch in the preferences to be able to select whether the DSD settings should always be used or the PCM settings.

 

Since HQP allows saving of settings prefs to an XML file, it would be easy to create one settings file which defaults to DSD and another which defaults to PCM if desired.

 

Miska, please advise if it would be possible for you to add this switch in the prefs.

 

Thanks,

 

geoff

[Jabs1542]

I have a new request. I keep running into a problem from time to time where the selection of DSD settings won't stick. With a DSD capable DAC using either proprietary ASIO drivers or a DAC which only supports DOP, the same problem occurs.

 

I want to use DSD settings by default; but after HQP is quit and re-started the settings often revert to the PCM settings.

 

I usually get the DSD settings to be retained in between restarts after a few attempts. It seems that if I play music using DSD settings, after a few times they will be remembered.

 

What I'd like to see is a switch in the preferences to be able to select whether the DSD settings should always be used or the PCM settings.

 

Since HQP allows saving of settings prefs to an XML file, it would be easy to create one settings file which defaults to DSD and another which defaults to PCM if desired.

 

Miska, please advise if it would be possible for you to add this switch in the prefs.

 

Thanks,

 

geoff

 

I have had similar experiences. It seems that if HQP can't "see" the DSD DAC when it first starts it will default to PCM. I actually turned this indicator into a feature, turns out it's a nice way to see that your DSD DAC (or in my case the NAA) did not initialize properly.

[Geoffrey Armstrong]

I have had similar experiences. It seems that if HQP can't "see" the DSD DAC when it first starts it will default to PCM. I actually turned this indicator into a feature, turns out it's a nice way to see that your DSD DAC (or in my case the NAA) did not initialize properly.

 

That's not the case for me. HQPLAYER is definitely "seeing" the DAC, because I can change from PCM to DSD in the main window. I want it to ALWAYS have DSD(SDM) selected in that front window if HQPLAYER is talking to a DSD capable DAC. My suggestion of a switch in the preferences settings would allow it to always use either PCM or SDM depending on the user's choice.

[Jabs1542]

That's not the case for me. HQPLAYER is definitely "seeing" the DAC, because I can change from PCM to DSD in the main window. I want it to ALWAYS have DSD(SDM) selected in that front window if HQPLAYER is talking to a DSD capable DAC. My suggestion of a switch in the preferences settings would allow it to always use either PCM or SDM depending on the user's choice.

 

Ok, my HQP basically behaves the way you want. If I have my NAA and DAC enabled before I launch HQP, and it was in DSD mode the last time I shut it down, it will start in DSD mode. Note that I am running 3.5.1 for Windows.

[Geoffrey Armstrong]

Ok, my HQP basically behaves the way you want. If I have my NAA and DAC enabled before I launch HQP, and it was in DSD mode the last time I shut it down, it will start in DSD mode. Note that I am running 3.5.1 for Windows.

 

I use a number of different systems here with HQP. Some with NAA some without. I don't know why I sometimes have problems with it defaulting to PCM with DSD capable DAC and that is when the DAC and NAA (where used) are enabled for HQP beforehand.

 

In any case, I still think it would be very useful to have an explicit switch in the preferences. After all these are the "defaults" so a user ought to be able to choose whether they want HQP to default to SDM(DSD) or PCM when they have a dac capable of both.

 

Hopefully this shouldn't be too difficult for Miska to apply.

[john925]

Miska, please check your pm, many thanks.

[satbaba]

A bit more clarification; for an UAC2 DAC that doesn't have ASIO driver, best choice is between Linux and OS X where said Linux distribution offers better overall performance.

 

Hi Miska,

 

I installed Ubuntu studio 14.04 to macbook pro and on the hqplayer, but no sound

Dac: Matrix x- sabre DSD Dac

 

Is there a solution?