HQ Player

[AudioDoctor]

I am just trying to figure out how the volume control works when the file being sent is already at the max sample rate? Is it happening in-between the 24 and 32 bits? Am I somehow losing resolution when I do this?

[Rune]

On 6/9/2020 at 8:00 AM, pdg540 said:

Regarding your comments on the sinc-L filter, are you able to say whether it would be possible to use the 7EC modulator, dsd256 output using i9-9900k with all cores clocked to 5ghz and 16gb ram? If not, would adding a GTX1080Ti with 11gb vid ram make it work?

 

I run HQPlayer Embedded on a i9-9900KS and no GPU, but my Desktop windows PC has a GTX 1080 TI and Ryzen 5 3600X, but the CPU is unfortunately not fast enough for  DSD256 EC modulator.

 

So I tried the non EC modulator instead and the result is it that it runs Sinc-L ASDM7 DSD256 and it takes 23 seconds to initialize.

The GPU uses 8.1 GB of memory and the load is about 14%  and CPU is 8%

 

 

 

[pdg540]

21 minutes ago, Rune said:

 

I run HQPlayer Embedded on a i9-9900KS and no GPU, but my Desktop windows PC has a GTX 1080 TI and Ryzen 5 3600X, but the CPU is unfortunately not fast enough for  DSD256 EC modulator.

 

So I tried the non EC modulator instead and the result is it that it runs Sinc-L ASDM7 DSD256 and it takes 23 seconds to initialize.

The GPU uses 8.1 GB of memory and the load is about 14%  and CPU is 8%

 

 

 

 

I can run ASDM7EC/poly sinc xtr without cuda (HQP Desktop/Windows 10). From Jussi's post, it sounds like sinc-L with EC modulators will need cuda offload for sure and it rwill take more than 8GB video ram. The 1080Ti has 11GB ram. I'm assuming the 2080Ti, which can run sinc-L/EC modulator, has better cores and faster ram than the 1080. So I guess the question is whether the 1080Ti cores and ram are adequate. I don't really know much about video cards.

 

What CPU is in your desktop and can you run ASDM7EC/poly sinc xtr on your 9900ks? This combination sounds really, really good!

 

 

[Miska]

1 hour ago, AudioDoctor said:

I am just trying to figure out how the volume control works when the file being sent is already at the max sample rate? Is it happening in-between the 24 and 32 bits? Am I somehow losing resolution when I do this?

 

8 extra bits going from 24-bit source to 32-bit output gives you 48 dB of extra headroom. NS9 noise shaper gives you a bunch more. Then take into account that recordings from analog domain (microphones and such) have less dynamic range than 24-bit data can hold. So you are well safe from any loss.

 

What is the output to? Is this normal listening volume? 50 dB is quite a lot of attenuation, so the total gain in analog domain looks a bit excessive for normal listening levels. Looks more like night time quiet listening volume setting.

 

[AudioDoctor]

3 minutes ago, Miska said:

 

8 extra bits going from 24-bit source to 32-bit output gives you 48 dB of extra headroom. NS9 noise shaper gives you a bunch more. Then take into account that recordings from analog domain (microphones and such) have less dynamic range than 24-bit data can hold. So you are well safe from any loss.

 

What is the output to? Is this normal listening volume? 50 dB is quite a lot of attenuation, so the total gain in analog domain looks a bit excessive for normal listening levels. Looks more like night time quiet listening volume setting.

 

 

Ok, now I am getting it. Thanks for explaining that to me.

 

Currently it is my DAC into my McIntosh MAC7200 with the volume turned to full. It is not loud, that's for sure. I can hear it, but not trying to crank the levels either.

[DancingSea]

On 6/8/2020 at 5:56 AM, Miska said:

there is no down side using apodizing filter for a recording that doesn't need one. But there is downside of not using apodizing filter for a recording that would need such.

 

Is there an update to list somewhere of which current filters are apodizing, and which are not?  

 

I know you are reticent to step out of the world of techno speak, but it would be very helpful to have a list that says “use these for super high quality recordings.  Use these for recordings that need help.  Use these for accosting music.  These for voice centric.  Etc.  I realize you will say no to this request, but I figure if I keep asking for many years, you may relent out of sheer pity 😂

[jimdukey]

That info is in the Manual.

Apodizing Filters are marked.

There is also info in the descriptions of the Filters.

Read it.

[asdf1000]

1 hour ago, DancingSea said:

I know you are reticent to step out of the world of techno speak, but it would be very helpful to have a list that says “use these for super high quality recordings.  Use these for recordings that need help.  Use these for accosting music.  These for voice centric.  Etc.  I realize you will say no to this request, but I figure if I keep asking for many years, you may relent out of sheer pity 😂

 

Have you tried ext2 filter?

 

It's really great for everything !

[DancingSea]

22 minutes ago, asdf1000 said:

 

Have you tried ext2 filter?

 

It's really great for everything !

 

I have.  It does sound good.  For some reason I prefer the long lp filter to ext2, which is a bit disappointing because of how popular ext2 is.  I really like Sinc-L as well, but last night I also noticed that incredible recordings sounded incredible, and still really good, but not top shelf, recordings sounded more shabby than ever.

[asdf1000]

33 minutes ago, DancingSea said:

For some reason I prefer the long lp filter to ext2, which is a bit disappointing

 

Why would that make you feel disappointed !?

 

 

[LoryWiv]

5 hours ago, pdg540 said:

 

I can run ASDM7EC/poly sinc xtr without cuda (HQP Desktop/Windows 10). From Jussi's post, it sounds like sinc-L with EC modulators will need cuda offload for sure and it rwill take more than 8GB video ram. The 1080Ti has 11GB ram. I'm assuming the 2080Ti, which can run sinc-L/EC modulator, has better cores and faster ram than the 1080. So I guess the question is whether the 1080Ti cores and ram are adequate. I don't really know much about video cards.

 

What CPU is in your desktop and can you run ASDM7EC/poly sinc xtr on your 9900ks? This combination sounds really, really good!

 

 

There must be more to it than just GPU RAM as my humble GTX 1060 w/6GB RAM on i7 cpu can run ASDM5EC with SINC-L at DSD128 or DSD256. Uses ~ 2.5 GB video RAM, It will run ASDM7EC - Sinc-L as well with similar RAM use although stutters occasionally.

[asdf1000]

On 2/20/2019 at 7:47 AM, Miska said:

 

REW can be also extended with RePhase.

 

Acourate is one nice software, Audiolense is another one. Acourate can generate directly 384k mono WAVs with brickwall extension. Audiolense at least used to generate stereo WAVs that needed splitting to two mono WAVs, but other than that it should also work. Acourate is probably most powerful tool, but also has the steepest learning curve.

 

 

New version of Audiolense out now, supporting up to 400kHz and mono WAV file output.

 

 

[Rune]

15 hours ago, pdg540 said:

 

I can run ASDM7EC/poly sinc xtr without cuda (HQP Desktop/Windows 10). From Jussi's post, it sounds like sinc-L with EC modulators will need cuda offload for sure and it rwill take more than 8GB video ram. The 1080Ti has 11GB ram. I'm assuming the 2080Ti, which can run sinc-L/EC modulator, has better cores and faster ram than the 1080. So I guess the question is whether the 1080Ti cores and ram are adequate. I don't really know much about video cards.

 

What CPU is in your desktop and can you run ASDM7EC/poly sinc xtr on your 9900ks? This combination sounds really, really good!

 

 

Yes the 9900KS can run DSD256 ASDM7EC/poly-sind-xtr. The Ryzen 3600X in my desktop cannot only manage the non EC.

If you are going to buy a new GPU go for the 2080/2080ti or wait and see how the next gen performs.

But if you already have a 1080Ti with 11GB like me I see no reason for upgrading. I am only interested in 2 channel.

[mfalcon]

1 hour ago, Rune said:

Yes the 9900KS can run DSD256 ASDM7EC/poly-sind-xtr. The Ryzen 3600X in my desktop cannot only manage the non EC.

If you are going to buy a new GPU go for the 2080/2080ti or wait and see how the next gen performs.

But if you already have a 1080Ti with 11GB like me I see no reason for upgrading. I am only interested in 2 channel.

I don’t have the KS, just the K overclocked but I would probably agree that the 2080Ti is overkill for 2 channel without room correction.  I use maybe 30 of the GPU bit 8GB of mem with Sinc-L and the 256 EC filter.  If I already had a 1080TI I wouldn’t upgrade.  Unless you want room correction or multi channel.  I would like to use room correction but keep putting it off.  I have a near field set up that I worked on for a long time years ago and like the sound I get but i know it would definitely improve the low end.  I also want to play with routing an analog setup through an ADC to HQplayer using room correction for my records.  Just don’t have the time currently.  But when I do I think the benefits of the 2080 will be seen more.

[Nikko1960]

Hi Jussi

 

I am now hopelessly addicted to sinc-L and for my ageing Mac Mini this means throttling back upsampling from DSD128 to x64. Now, google tells me that I can plug an external graphics card into the thunderbolt port of the Mac Mini. Do you think this would supply the necessary processing power I need?

 

thx

 

nikko

[jamesg11]

5 hours ago, mfalcon said:

I don’t have the KS, just the K overclocked but I would probably agree that the 2080Ti is overkill for 2 channel without room correction.  I use maybe 30 of the GPU bit 8GB of mem with Sinc-L and the 256 EC filter.  If I already had a 1080TI I wouldn’t upgrade.  Unless you want room correction or multi channel.  I would like to use room correction but keep putting it off.  I have a near field set up that I worked on for a long time years ago and like the sound I get but i know it would definitely improve the low end.  I also want to play with routing an analog setup through an ADC to HQplayer using room correction for my records.  Just don’t have the time currently.  But when I do I think the benefits of the 2080 will be seen more.

I’m going the room correction route, mainly for bass improvements, with either one of the ‘short-cut’ on-line services -

 

https://www.homeaudiofidelity.com/english/requirements/

or

https://accuratesound.ca

[pdg540]

6 hours ago, mfalcon said:

I don’t have the KS, just the K overclocked but I would probably agree that the 2080Ti is overkill for 2 channel without room correction.  I use maybe 30 of the GPU bit 8GB of mem with Sinc-L and the 256 EC filter.  If I already had a 1080TI I wouldn’t upgrade.  Unless you want room correction or multi channel.  I would like to use room correction but keep putting it off.  I have a near field set up that I worked on for a long time years ago and like the sound I get but i know it would definitely improve the low end.  I also want to play with routing an analog setup through an ADC to HQplayer using room correction for my records.  Just don’t have the time currently.  But when I do I think the benefits of the 2080 will be seen more.

 

I only use 2 channel, no convolution and actually don't have an nvidea card in my computer right now; just using 9900k alone for HQP which gets me to 7EC/xtr. Was thinking of picking up a graphic card that will get me to 7EC/sincL but there's no chance I'm dropping 2080Ti type cash  

[mfalcon]

37 minutes ago, pdg540 said:

 

I only use 2 channel, no convolution and actually don't have an nvidea card in my computer right now; just using 9900k alone for HQP which gets me to 7EC/xtr. Was thinking of picking up a graphic card that will get me to 7EC/sincL but there's no chance I'm dropping 2080Ti type cash  

I bet someone will try it with the 1080Ti and you can by it with some assurance that it will work.  Good luck, I’m definitely enjoying it with good material.  It’s a fun adventure to figure what it will sound good with as I do t know what recording used which process so I just guess.  XTR was my go to filter for all purpose listening and I still love it.

[DancingSea]

On 6/10/2020 at 6:24 PM, asdf1000 said:

 

Why would that make you feel disappointed !?

 

 

 
I’m joking a bit, That I’m disappointed I can’t be with the cool kids 😂

[asdf1000]

46 minutes ago, DancingSea said:

 
I’m joking a bit, That I’m disappointed I can’t be with the cool kids 😂

 

If you're using HQP you're already one of the cool kids.

 

After that, use HQP in whichever way sounds best to your ears ! 

 

[Hammer]

Hi, I am running HQPlayer Desktop 4 on a dedicated Windows 10 machine that is connected to my DAC via NAA.  Since I don’t listen to it all the time, the machine is set to sleep after a period of inactivity.  I find that when I wake the machine and try to play, I have to close out of HQPlayer Desktop and reopen it in order for it to connect to the NAA. Alternatively, I can also go to settings, don’t do anything hit OK and that will work as well.  Therefore, it seems HQPlayer loses connection to the NAA while in sleep mode.  I don’t have any other instances of HQPlayer which would connect to the NAA so there is no chance another HQPlayer is taking control of the NAA while the Windows 10 machine is asleep.  Would it be possible to add a feature to “refresh” the connection if it is lost?  I know when I was using HQPlayer 3 and maybe even some earlier versions of HQPlayer 4 that it will play fine without any intervention from wake.  Thank you.

[jamesg11]

Hi @Miska et al,

 

Just now updated ultraRendu to 2.8 via sd card, sonicorbiter update.

On hqp 3.latest, mac mini, no other change to anything, & rme dac is now muting with gain overload, still on dsd direct.

 

Issue? Need to now go to hqp4? (was leaving til pc upgrade).

Or is this likely a sonicorbiter sw issue?

[StreamFidelity]

On 6/11/2020 at 1:21 AM, pdg540 said:

From Jussi's post, it sounds like sinc-L with EC modulators will need cuda offload for sure and it rwill take more than 8GB video ram.

 

I did some tests. My system is a i9-9900K fanless design without CUDA/GPU.

 

With Convolution on, ASDMEC7, Sinc-L works fine with DSD 128. There are ropouts with DSD 256. Big RAM helps. 😉

 

With Convolution on, ASDMEC7, Sinc-M works very fine with DSD 256.

 

With Convolution on, ASDMEC7, Sinc-S works very fine with DSD 256.

 

[Quadman]

5 minutes ago, StreamFidelity said:

I did some tests. My system is a i9-9900K fanless design without CUDA/GPU.

Do you overclock your 9900K or rely on turboboost to automatically take care of it?

 

[StreamFidelity]

4 minutes ago, Quadman said:

Do you overclock your 9900K or rely on turboboost to automatically take care of it?

 

I set all multipliers to 41 = 4.1GHz. This permanent overclocking ensures a homogeneous, fluid sound image in my ears.