[Failed Engineer]

I added a M.2 SSD and new PCI network card to my Linux HQP Embedded machine and it looks like my key reverted to trial. Does that change the fingerprint? I tried re-uploading my license file and still on trial. How can I get it updated?

 

[amrutmhatre90]

Interesting video, has a hqp file attached to try out.

 

[bogi]

does that mean better not use convolution and matrix at the same time?

yes

I added a M.2 SSD and new PCI network card to my Linux HQP Embedded machine and it looks like my key reverted to trial. Does that change the fingerprint? I tried re-uploading my license file and still on trial. How can I get it updated?

To solve a license issue it may work faster to send a mail to info @ signalyst . com

 

[m-i-c-k-e-y]

You could save your settings in an XML file in HQPlayer (File>Export Settings). This is very helpful if you want to try various different settings (File>Import Settings).

Good tutorial for those who wants Roon + HQP. Are you the author?

For DSD upsampling, I however would recommend, say using A26 (a more dedicated DAC for DSD), as others are doing.

 

[Ghoostknight]

yes

i meant pipeline and convolution at the same time, it seems to work tho

 

[bogi]

i meant pipeline and convolution at the same time, it seems to work tho

I didn't tell that it is not in function in parallel. Just that Jussi warned not to use it this way because of possible unpredictable behavior:

If you use matrix, put everything there and disable the simple convolution engine from the convolution dialog. Otherwise it becomes confusing and error prone.

So it may work for you but take it as an unsupported case.

Since your usual DSP has a format of WAV convolution file (otherwise you couldn't use it in Convolution dialog), you can enter that WAV file in Matrix pipeline dialog together with your IIR filter. If you don't check the IIR to FIR filter checkbox, convolution will remain IIR and WAV is always processed by FIR engine. As Jussi showed on an example:

You can have something like "iir:type=hp;f=5,MyFancyDRC.wav". This will run the IIR as IIR (unless you ask for something else) and the FIR WAV as such.

 

[jlaako]

1. Did you already thought about VST support? (or atleast some linux supported format?) that could open up some interesting features ... like different crossfeed plugins, oldschool equalizer (if someone wishes so), compressors (which i sometimes like mostly for movies...) or SonarWorks perhaps? tho i have to admit most of the plugins would be kinda gimmicky and you probably wanna avoid running audio trough a couple of badly made plugins with all kind of resampling etc going on

This is asked every now and then, but it poses way too many restrictions. VST requires application to have a GUI (doesn't work for HQPlayer Embedded), it doesn't support DSD, plugins pose limitations on supported sampling rates, it doesn't support GPU offloaded DSP pipelines, it doesn't play together with matrix profiles, doesn't work natively on all the supported platforms, etc, etc.

2. Please make it a proper "player" .. like with album view, library managment, playlist managment that is less cumbersome and stuff like this, it feels kinda stupid to just search the files and drop them into hqplayer each time i wanna play them, it might be me, im kinda lazy

This is already supported? Please check out the PDF manual, or at least the brief quick start guide here:
https://signalyst.com/quickstart-guide/

HQPlayer Desktop application is just the server component for configuration and minimal lightweight interface. HQPlayer Client is the one for dealing with music playback.

i guess this would also make my feature request 2 obsolete since i could use each player i want and let the audio route trough hqplayer...

You can make HQPlayer look like a USB DAC using RPi4 as an input NAA... Then anything that can play to USB output works.

 

[jlaako]

i meant pipeline and convolution at the same time, it seems to work tho

Yes, matrix includes convolution as one of the plugins and avoids confusion and conflicts. So the recommendation is to use solely matrix if you need it for anything and disable the simple convolution engine.

This also allows you to use matrix profiles that you can switch on the fly during playback, to compare different EQ approaches, etc.

 

[Ghoostknight]

Ahh... a few lightbulbs went on haha

I didn't tell that it is not in function in parallel. Just that Jussi warned not to use it this way because of possible unpredictable behavior:

Since your usual DSP has a format of WAV convolution file (otherwise you couldn't use it in Convolution dialog), you can enter that WAV file in Matrix pipeline dialog together with your IIR filter. If you don't check the IIR to FIR filter checkbox, convolution will remain IIR and WAV is always processed by FIR engine. As Jussi showed on an example:

You can have something like "iir:type=hp;f=5,MyFancyDRC.wav". This will run the IIR as IIR (unless you ask for something else) and the FIR WAV as such.

Ok, so either Convolution or Pipeline...

So, my current pipeline just has this line : "/home/username/Documents/hqplayer eq/Eris_E8_HouseCurve.txt"

making it: "iir:type=hp;f=5,/home/username/Documents/hqplayer eq/Eris_E8_HouseCurve.txt" would add a IIR Highpass, even if i set IIR to FIR processing on the top?

or would it be something like this because of my eq file: "iir:type=hp;f=5,fir:/home/username/Documents/hqplayer eq/Eris_E8_HouseCurve.txt" (mind the "fir:" or something else maybe?)

This is asked every now and then, but it poses way too many restrictions. VST requires application to have a GUI (doesn't work for HQPlayer Embedded), it doesn't support DSD, plugins pose limitations on supported sampling rates, it doesn't support GPU offloaded DSP pipelines, it doesn't play together with matrix profiles, doesn't work natively on all the supported platforms, etc, etc.

ah... alot of limitations i didnt thought about

This is already supported? Please check out the PDF manual, or at least the brief quick start guide here:
https://signalyst.com/quickstart-guide/

HQPlayer Desktop application is just the server component for configuration and minimal lightweight interface. HQPlayer Client is the one for dealing with music playback.

there was something... i forgot about the client application, lmao, thanks

You can make HQPlayer look like a USB DAC using RPi4 as an input NAA... Then anything that can play to USB output works.

hmm that was one solution i thought about, i had also a rpi setup running with camillaDSP for a while in a similar setup but wouldnt the raspberry pi be quiet heavly limited by processing power? or does this somehow work with a additional server? probably not because of latency

if i wanna follow my current plan i need atleast DSD256... better DSD512 because of much less intensive noise shaping, which cant even handle my x86 machine...
and the thing keeping me from upgrading is this:
Linux doesnt like nvidia but you need nvidia for hqplayer to make offloading work (correct me if im wrong but cuda offloading is nvidia only i think), so probably better to invest in a beafy cpu for combatibility.. ( i kinda made it a budget pc around 1 year ago with a 12400, i dont play heavy games and i didnt thought about hqplayer of course.. )

tho we also have to admit... hqplayer "wastes" alot of energy on higher settings... thats why i kinda looking right now into doing a DIY dac with hardware resampler for the majority of things... you can still use hqplayer if you want tho ... also the DSC1 to DSC3 got me curious about how these dacs handle DSD differently than delta sigma chips...

Yes, matrix includes convolution as one of the plugins and avoids confusion and conflicts. So the recommendation is to use solely matrix if you need it for anything and disable the simple convolution engine.

This also allows you to use matrix profiles that you can switch on the fly during playback, to compare different EQ approaches, etc.

ok, got it! yea the profiles are quiet handy, just the one line configuration is a bit confusing at first

 

[bogi]

making it: "iir:type=hp;f=5,/home/username/Documents/hqplayer eq/Eris_E8_HouseCurve.txt" would add a IIR Highpass, even if i set IIR to FIR processing on the top?

No, a I wrote above, don't check the IIR to FIR checkbox (set it to be unchecked). That checkbox applies only to IIR PEQ settings. WAV convolution files are always FIR. Syntactically just follow Jussi's example.

hmm that was one solution i thought about, i had also a rpi setup running with camillaDSP for a while in a similar setup but wouldnt the raspberry pi be quiet heavly limited by processing power? or does this somehow work with a additional server? probably not because of latency

No. NAA does no processing on data. Input NAA is like output NAA, just data flows in the opposite direction. NAA just receives, buffers and sends raw PCM or DSD data between audio interface and network interface. Pi4 is OK to run both as input and output NAA at once. With nothing more running on Pi4, passive cooling using some amount of heatsink is sufficient. Metal case is recommended but not a must.

Linux doesnt like nvidia but you need nvidia for hqplayer to make offloading work

Linux likes nVidia on low latency kernel. Install Debian, Ubuntu or Fedora and then nVidia driver and HQPlayer. What is not supported is nVidia on realtine kernel, which si used for HQPlayer OS (bootable Linux image with HQPlayer).

 

[Ghoostknight]

No, a I wrote above, don't check the IIR to FIR checkbox (set it to be unchecked). That checkbox applies only to IIR PEQ settings.

ok so it converts all IIR eq's to FIR and there probably isnt a setting to make a filter run in FIR mode seperately...

it does leave the phaseshift of convolution files untouched and doesnt correct afterwords for it with IIR to FIR right?
i guess i will just do it this way -> "/home/username/Documents/hqplayer eq/Eris_E8_HouseCurve.txt,subsonicfilter.wav"
convolution file is subsonic in IIR and every other PEQ in the eq file gets converted to FIR...

i kinda hate convolution for EQ... for every edit you make you need to generate a new one, its very tedious so i rather just have the subsonic as IIR in the convolution file...
the only thing where it really becomes mandatory is if you phase align your speakers

WAV convolution files are always FIR.

they are not, they can have whatever phaseresponse you want, if you export REW filter settings as convolution .wav then the phase gets shifted as usual with IIR eq... you need to modify the .wav file with rePhase to make it "FIR" or linear phase

No. NAA does no processing on data. Input NAA is like output NAA, just data flows in the opposite direction. NAA just receives, buffers and sends raw PCM or DSD data between audio interface and network interface. Pi4 is OK to run both as input and output NAA at once. With nothing more running on Pi4, passive cooling using some amount of heatsink is sufficient. Metal case is recommended but not a must.

ok so NAA is a endpoint device.... you need the server and can send audio from the server to the endpoint (NAA) so far i know the scheme but im still confused
so, the NAA endpoint can have usb c input audio data (i guess it just gets detected as audio device on the host pc), then send audio to the server and server sents audio upsampled back for playback for the usb attached dac on the endpoint?
or is the NAA input only a passtrough to the connected usb dac?

Linux likes nVidia on low latency kernel. Install Debian, Ubuntu or Fedora and then nVidia driver and HQPlayer. What is not supported is nVidia on realtine kernel, which si used for HQPlayer OS (bootable Linux image with HQPlayer).

im running debian with liquorix kernel, it works but nvidia can still be hit and miss with drivers depending on your setup

 

[bogi]

ok so it converts all IIR eq's to FIR and there probably isnt a setting to make a filter run in FIR mode seperately...

You still don't get what I and Jussi told you repeatedly. IIR to FIR checkbox cleared means no IIR to FIR conversion for PEQ. PEQ in this context means inline syntax (see PDF manual) or supported TXT file format.

they are not, they can have whatever phaseresponse you want, if you export REW filter settings as convolution .wav then the phase gets shifted as usual with IIR eq... you need to modify the .wav file with rePhase to make it "FIR" or linear phase

You may be right, I just never heard about IIR WAV file processed by HQPlayer. Jussi wrote: "You can have something like "iir:type=hp;f=5,MyFancyDRC.wav". This will run the IIR as IIR (unless you ask for something else) and the FIR WAV as such.". Hopefilly @jlaako will make it clear.

According to what Miska wrote previously IIR to FIR checkbox does not apply to convolution WAV files but only to PEQ settings either in inline syntax (starting with iir: ) or in TXT file form.

so, the NAA endpoint can have usb c input audio data (i guess it just gets detected as audio device on the host pc), then send audio to the server and server sents audio upsampled back for playback for the usb attached dac on the endpoint?

Yes, you can use NAA both for input and output. Or only for input. Or only for output. HQPlayer works also without NAA endpoint device. I posted about the USB input capability in this thread less than a week ago, you may search for it.

or is the NAA input only a passtrough to the connected usb dac?

No, that's impossible and such a passthrough does not make sense - you could directly connect a DAC without NAA.
NAA always communicates with HQPlayer. It enables ot have HQPlayer computer placed in another room (potentially headless) or to play to more rooms (not at once).

 

[VNandor]

Using convolution for filtering always yields a finite impulse response filter (FIR filter) but just because a filter is FIR, it does not have to be linear phase. It's only linear phase if the impulse is symmetric. You could create your own wav file by recording the output of any minimum phase IIR filter fed by an impulse. Then use that wav for convolving. The result would be a minimum phase FIR filter.

 

[jlaako]

i guess i will just do it this way -> "/home/username/Documents/hqplayer eq/Eris_E8_HouseCurve.txt,subsonicfilter.wav"
convolution file is subsonic in IIR and every other PEQ in the eq file gets converted to FIR...

.wav = convolution = FIR.
.txt = PEQ = IIR

I believe you would want to have the subsonic as IIR (minimum phase). Second order is gentle enough not to really ring, but enough to stop DC kind of things. And possibly rest of the EQ as FIR. REW always gives you minimum-phase FIRs, thus you are better off just running all as IIR (just subsonic plus .txt).

For the txt/PEQ you have option to choose to run those as IIR or FIR. Running IIR or minimum-phase FIR is the same thing from practical point of view. But using the HQPlayer's IIR to FIR conversion can help especially when performing these operations on DSD sources. HQPlayer gives you all three different possible ways to run parametric EQ. So you can also choose to convert those to linear-phase FIR.

i kinda hate convolution for EQ... for every edit you make you need to generate a new one, its very tedious so i rather just have the subsonic as IIR in the convolution file...
the only thing where it really becomes mandatory is if you phase align your speakers

Yes, you can keep these separate for practical purposes. And I also recommend to utilize the profiles functionality.

So you can also have for example subsonic + convolution FIR:
"iir:type=lp;f=5,/home/jussi/convolution.wav"

Or subsonic + parametric IIR:
"iir:type=lp;f=5,/home/jussi/parametric.txt"

This syntax is explained in the manual PDF (can be found from the desktop Start-menu).

so, the NAA endpoint can have usb c input audio data (i guess it just gets detected as audio device on the host pc), then send audio to the server and server sents audio upsampled back for playback for the usb attached dac on the endpoint?

Exactly...

 

[Ghoostknight]

Using convolution for filtering always yields a finite impulse response filter (FIR filter) but just because a filter is FIR, it does not have to be linear phase. It's only linear phase if the impulse is symmetric. You could create your own wav file by recording the output of any minimum phase IIR filter fed by an impulse. Then use that wav for convolving. The result would be a minimum phase FIR filter.

Ah you again! thanks for correcting me once again

thanks guys,
1. i use these settings now: "/home/username/Documents/hqplayer eq/Eris_E8_HouseCurve.txt,/home/username/Documents/hqplayer eq/subsonicfilter.wav"
IIR to FIR checked so all EQs run in FIR mode (linear phase) while the subsonic convolution file is (still FIR) but minimum phase (because REW doesnt linearise phase with convolution files)!
i dont have to touch the subsonic convolution file often and can still easly change all other eq settings and they are converted to FIR on the fly... (instead of the tedious rePhase way), so this setup is perfect i think, thanks again!


this works and *i think* it even sounds better than using simple convolution engine aditionally, somehow it sounds more coherent

2. i really have to consider HqPLayer embedded now... whats the latency with this setup? are movies possible with lighter filter settings?