Jump to content
IGNORED

HQ Player


Recommended Posts

7 minutes ago, kelvinwsy said:

Hi I ran into some cases when I shut down as many processes using Process Lasso ~ < 1000/ 40 processes I can barely run Closed Fast setting in filter and oversampling 

but convolution not enabled in over -add mode

Then I tried over-save mode ... Seem to work

BUT still touch and go

what is the difference in the Convolution Engine between ADD versus SAVE mode??? What

seems to be different ???

i also find sensitivity difference in the WAV file exported from REW .. one recent file created In 

32bit mono 192khz

The closed form fast Will not play 

the older REW WAV filed in 24bit seem to be ok for Closed Fast 

why is this so? Number of EQ frequency points is about the same except for the 24bit vs 32 bit

Thx

I usually Enable multi core And auto mode And CUDA 

Has been working well for me up to 3.17 

what is the difference if I do OR do not tick auto family ... yes I can see the freq classes on the pull down menus above ... my question is about CPU and CUDA Offload difference 

thx for the hardwork

3.18.1 sounds better and better but is like a fickle stallion ... you ride it and it is a great ride but you get thrown off when it acts up

sorry my cpu is 6700k skylake 

Link to comment
47 minutes ago, kelvinwsy said:

I usually Enable multi core And auto mode And CUDA 

Has been working well for me up to 3.17 

what is the difference if I do OR do not tick auto family ... yes I can see the freq classes on the pull down menus above ... my question is about CPU and CUDA Offload difference 

thx for the hardwork

3.18.1 sounds better and better but is like a fickle stallion ... you ride it and it is a great ride but you get thrown off when it acts up

sorry my cpu is 6700k skylake 

Before updating GTX 970 video card driver the cpu usage spiked to 80% or more 

after driver upgrade I think it came down to 40-50%

CPU Temp ? Not sure but pc just powered up and could not play Hqplayer on these higher settings ... I don’t remember the cpu fan speeding up so I don’t think cpu temp was high  ???

but the recent experiment I think ( on the filter and oversampling) settings vs Multicore Automode And Convolution Engine Settings is now more relevant to my problem getting closed form fast to work

latest experiments and tweaking brings the cpu back down to 22-24% with initialization peak maybe at 40 %

thx

 

Link to comment
1 hour ago, kelvinwsy said:

what is the difference in the Convolution Engine between ADD versus SAVE mode??? What

seems to be different ???

 

Algorithm, "save" is heavier than "add", but there's no real quality difference.

 

1 hour ago, kelvinwsy said:

i also find sensitivity difference in the WAV file exported from REW .. one recent file created In 

32bit mono 192khz

 

Yes, small changes to the filter can make quite a bit of difference.

 

what is the difference if I do OR do not tick auto family

 

Doing conversion between rate families is more CPU/RAM intensive. But changing the rate family between tracks also means  stopping and restarting the audio interface/DAC which means it is not gapless.

 

Quote

Before updating GTX 970 video card driver the cpu usage spiked to 80% or more 

after driver upgrade I think it came down to 40-50%

 

If you had too old driver, then CUDA offload likely didn't work and it fell back to CPU processing.

 

Signalyst - Developer of HQPlayer

Pulse & Fidelity - Software Defined Amplifiers

Link to comment

Just a quick note for anyone who was having that problem, as I have been, with play stopping at the end of a dsd track if the next track up was pcm (this is Roon-->HQPlayer). Roon just released 1.3, build 269, and this was one of the bugs fixed. I just tested on my setup and it went without a hitch on my test tracks.

--Jerry

Link to comment
19 minutes ago, Outlaw said:

Has anyone installed the latest windows fall creators update 1709 ? Any issues.

 

Microsoft hasn't pushed it to any of my test machines yet... I'll test it ones it appears on one of those...

 

As usual, do not rush on the updates. Microsoft will push the update to you once they think it'll work fine on your hardware + software system.

 

Signalyst - Developer of HQPlayer

Pulse & Fidelity - Software Defined Amplifiers

Link to comment

Just watching gpu percentage.For all filters except the xtr filters the gpu load is around 9 % and cpu load around 16 % this is for upsampling to dsd 256.When I try with the xtr filters after about 18 seconds get popping sounds.For the xtr filters gpu load at around 85% and cpu is around 21%.Why would the gpu load be so high for the xtr filters ?

Link to comment
3 hours ago, Outlaw said:

Just watching gpu percentage.For all filters except the xtr filters the gpu load is around 9 % and cpu load around 16 % this is for upsampling to dsd 256.When I try with the xtr filters after about 18 seconds get popping sounds.For the xtr filters gpu load at around 85% and cpu is around 21%.Why would the gpu load be so high for the xtr filters ?

 

Because xtr filter is, as name says, so extreme... GTX 1060 on my Linux desktop manages to run it without problems at 82% GPU load.

Signalyst - Developer of HQPlayer

Pulse & Fidelity - Software Defined Amplifiers

Link to comment

Hey guys, I'm new to the forum as well as to using HQPlayer. I'm planing on switching from Audirvana to HQPlayer. My setup: MacBookPro (15" 2017, 3.1GHz i7) and iFi micro iDSD. I have problems choosing the most suitable preferences for my setup. I'm sure there is somebody here how can help me with that. Find attached what it looks like at the moment. 

 

 

 

Bildschirmfoto 2017-10-26 um 16.49.10.png

Bildschirmfoto 2017-10-26 um 16.50.54.png

Link to comment

RamUwe, welcome to CA.

 

Don;t consider finding the "right filter" to be a problem, but rather a preference.  Everyone's setup is a little different, based on tastes, room, ears, music styles (and of course, specs....your dac/server combo can't do DSD512, etc etc).  So just try a few PCM filters (I'd try TPDF dither and stick to it while changing filters, at least at first) and hear what you like best.   What upsampling are you trying to achieve?  Everything to DSDx?  PCM to PCM? 

Link to comment

Ok, so only concentrate on right side of settings page (since PCM to PCM is not involved).  First set bit rate to (44 or 48) x 256, not 512.  Try a 256 modulator and try mp vs non-mp versions of filters to see what sounds best with your dac at DSD256 (and what tolerances you have...i.e at what filter level or modulator type will your dac/server handshake begin to stutter or dropouts occur).  The -2s versions of filters are easier on that.

Link to comment
1 hour ago, RamUwe said:

Hey guys, I'm new to the forum as well as to using HQPlayer. I'm planing on switching from Audirvana to HQPlayer. My setup: MacBookPro (15" 2017, 3.1GHz i7) and iFi micro iDSD. I have problems choosing the most suitable preferences for my setup. I'm sure there is somebody here how can help me with that. Find attached what it looks like at the moment. 

 

 

 

Bildschirmfoto 2017-10-26 um 16.49.10.png

Bildschirmfoto 2017-10-26 um 16.50.54.png

 

Hi, I use mac mini with micro iDSD. For couple of years I was using USB (with regen) and upsampled everything to DSD. Now I use optical connection and play everything in PCM. As per now I like the sound with optical connection better, it is not so analytical and micro-detailed as with USB, but more soft, lush, full-body and less fatiguing. Also, optical is more forgiving to not so succesfully digitalized old albums from 60-70s than USB. Settings are: poly-sync-xtr-mp, TPDIF, 768M. 

 

I also moved from Audirvana (and Amarra before it) to HQ Player. 

Link to comment
2 hours ago, ted_b said:

Ok, so only concentrate on right side of settings page (since PCM to PCM is not involved).  First set bit rate to (44 or 48) x 256, not 512.  Try a 256 modulator and try mp vs non-mp versions of filters to see what sounds best with your dac at DSD256 (and what tolerances you have...i.e at what filter level or modulator type will your dac/server handshake begin to stutter or dropouts occur).  The -2s versions of filters are easier on that.

 

Ok. I'll give it a try. Thanks :) Maybe Miska has some suggestions as well ;)

 

1 hour ago, AnotherSpin said:

 

Hi, I use mac mini with micro iDSD. For couple of years I was using USB (with regen) and upsampled everything to DSD. Now I use optical connection and play everything in PCM. As per now I like the sound with optical connection better, it is not so analytical and micro-detailed as with USB, but more soft, lush, full-body and less fatiguing. Also, optical is more forgiving to not so succesfully digitalized old albums from 60-70s than USB. Settings are: poly-sync-xtr-mp, TPDIF, 768M. 

 

I also moved from Audirvana (and Amarra before it) to HQ Player. 

 

Thanks for your experience report. An optical connection isn't possible with my setup.

Link to comment
40 minutes ago, RamUwe said:

An optical connection isn't possible with my setup.

 

Didn't know Mac Book Pro has no optical out anymore. Ok, for USB connection I recommend to start with xtr-mp or xtr-mp-2s and find out which modulator sounds better for you. -2s permits more variety and higher bit rate limit.

Link to comment

He meant “DoP” not drop.  It stands for DSD over PCM, a vehicle to stream true DSD (not pcm) to dacs and OS platforms that don’t have proper native drivers.  All is well, but it won’t do DSD512.

Link to comment
10 hours ago, RamUwe said:

 

Everything to DSD256 because DSD512 doesn't work with my setup except for when I install Win10 with ASIO. I don't want to do that.

Looking at your settings, I have a few other things to suggest.

- Set Vol Max to -3dB (or -4dB) to avoid digital clipping;

- If your filter of choice allows it, set Auto rate family;

- in SDM defaults set 48 x 256 for Bit rate (so your filter will follow 44.1 and 48 families);

- if your DAC chip is 32 bit (I do not remember exactly), set DAC bits at 32;

- as you upsample also DSD64/128 to DSD256, in DSDIFF/DSF settings enable 6 dB gain.

Link to comment
Just now, bibo01 said:

I would try to uniform output with 6 dB gain with PCM and DSD input, as OP applies DSP to DSD signal anyhow.

 

Per Jussi, the +6 dB gain setting is only for DSD -> PCM conversion. That setting does nothing otherwise. I just found this out a few months ago myself.

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...