HQ Player

[tboooe]

10 minutes ago, zorntel said:

Jussi is obviously the end all on this...but the math to do DSD to DSD is much easier than the math to do PCM to DSD from what I understand.

 

Robert

That makes total sense to me which is why I am so perplexed about the performance of my system.  When I try to convert DSD64 or 128 to DSD256 I get some stuttering.  

[cmarin]

43 minutes ago, zorntel said:

To my ears on my system, the xtr filters are more analog like than the straight polysinc which feels more digital to me.  Before switching to the xtr I was using polysinc short mp which is also more analog like for me...but a little softer than the xtr which seems to just hit the sweet spot.   I wish the non-2s would work with PCM files but I'm not going to break the bank trying to get there.  The machines Larry and I built to do this stuff are about the same cost as just Jussi's processor that will do the non-2s. The sound I'm getting with the xtr-2s and DSD512 is the best sound I've ever heard on my system...I don't want to stop listening...which in the end if the point of all of this!

 

Zorntel,

 

Couldn't agree more about the xtr filters with HQPlayer 3.16.0. It raised the sound quality level of several DACs I had coupled with the SGM 2015 server in my system by a very significant amount. Amazing improvement in sound quality at a basically zero marginal cost.

[simonklp]

1 hour ago, tboooe said:

Sorry to get off topic.  In my system when i convert DSD64 or 128 to DSD256 my cpu utilization jumps to 80% but when converting pcm to DSD256 my cpu reaches about 30%.  Is this normal?  It seems from this post that your cpu handles dsd to dsd conversion more easily than pcm to dsd.

 

Hi Tboooe,

 

You have raised a good question. Initially, I thought that DSD64 upsamples to DSD512 consumes much less CPU processing power than that of converting Redbook PCM to DSD512. This may be shown by the start-up time of playing DSD64 to DSD512 (several seconds) is much less  than that of Redbook to DSD512 (about 70 sec. for my case).

 

However, after noting your question, I have taken some time to monitor the CPU utilization shown in the Task Manager. I find that the "average" utilization rate for DSD64 to DSD512 is about 21%, while that rate for Redbook to DSD512 is about 22%. There is not much difference between them.

 

But it is quite strange about your case that the CPU utilization rate for DSD64 to  DSD256 is so high at 80%.

[louisxiawei]

Hey Miska,

 

Which format do you recommend for HQplayer upsampling files to DSD512? DXD or DSD if one albume have them both?

 

 

Here is one album released on nativeDSD. It has DXD version and DSD64-DSD256 ones.

 

Now I have three version in my Roon library.

 

1. Tidal 44.1kHz 16bit

2. DXD (PCM 352.8 kHz 24 bit) purchased from NativeDSD

3. DSD 64 purchased from NativeDSD

 

I observed some interesting/weird things happened when HQplayer was upsampling these files. I use T+A DAC8 DSD for the test. 

 

1. When I use HQplayer upsampling DSD64 and 44.1kHz Tidal to DSD512.

 

HQplayer SDM setting: 

• poly-sinc-xtr-2s,
• AMSDM7 512+fs,
• 48K*512,
• auto family checked,
• cuda offload checked,
• mulicore DSP checked.
• DirectSDM unchecked

 

Result:

I find there are some funny noise at the beginning of certain tracks. In this album (Mendelssohn & Schumann by Philippe Graffin), it's 1st track (Allegro molto appassionato), 5th track (In kraftigem, nich zu schnellen tempo)  and 6th track (Langsam).

 

I think you can easily hear the noise from Tidal streaming if you don't have DSD64 file.

 

However, there is no such noise in the DXD version at all.

 

2. When I use HQplayer upsampling DXD, DSD64 and 44.1kHz Tidal to DSD256.

 

HQplayer SDM setting: 

• poly-sinc-xtr-2s,
• AMSDM7 512+fs,
• 48K*256,
• auto family checked,
• cuda offload checked,
• mulicore DSP checked.
• DirectSDM unchecked

Result: There is no funny noise of all these three formats.

 

3. When I use HQplayer playback DSD64 with DirectSDM checked.

 

There is no funny noise detected even in DSD64 file.

 

I'm not quite sure whether this funny noise is due to the DSD64 files' problem or HQplayer's.

 

Since you also mentioned that when HQplayer upsampling DSD files, they are converted to intermediate formats ( 2.8 MHz DSD64 content is converted to 176.4 kHz PCM, 5.6 MHz DSD128 content to 352.8 kHz PCM, 11.2 MHz DSD256 content to 706. kHz). 

 

However, I also talked to the Jonas from NativeDSD. he told me that although this album was recorded at DSD64. It was then converted to DXD for editing, and their final DSD64, 128, 256 for sale are all converted from edited master DXD version. 

 

So from my understanding, DXD will be better than DSD file for HQplayer upsampling to DSD512 since there will be no DSD -> PCM conversion but DXD ->DSD  during up-sampling?

 

Any idea of these noise I mentioned above? 

 

Any help would be appreciated.

 

 

 

 

 

 

[bibo01]

21 minutes ago, louisxiawei said:

Hey Miska,

...

Since you also mentioned that when HQplayer upsampling DSD files, they are converted to intermediate formats ( 2.8 MHz DSD64 content is converted to 176.4 kHz PCM, 5.6 MHz DSD128 content to 352.8 kHz PCM, 11.2 MHz DSD256 content to 706. kHz). 

...

Where Miska mentioned the above?! I sincerely doubt it.

 

Re. your issue of format choice, I would go for DXD master.

[louisxiawei]

5 minutes ago, bibo01 said:

Where Miska mentioned the above?! I sincerely doubt it.

 

Re. your issue of format choice, I would go for DXD master.

Here it is the link as Miska mentioned it.

 

Thanks for your opinion. I need to check DSD256 file further from nativeDSD. 

 

[bibo01]

2 minutes ago, louisxiawei said:

Here it is the link as Miska mentioned it.

 

Thanks for your opinion. I need to check DSD256 file further from nativeDSD. 

 

In that quote Miska was referring to a conversion DSD->PCM. Nothing to do with your case - you want to output to DSD256 or 512.

[louisxiawei]

Just now, bibo01 said:

In that quote Miska was referring to a conversion DSD->PCM. Nothing to do with your case - you want to output to DSD256 or 512.

Exactly. But what if my input file is DSD signal and want to get an DSD512 output? Let's say DSD64.

 

Then from my point of understanding, must be DSD64 - PCM - DSD512.

 

So that conversion DSD->PCM also happens during DSD64->DSD512 conversion. Am I understanding correctly?

[bibo01]

3 minutes ago, louisxiawei said:

Exactly. But what if my input file is DSD signal and want to get an DSD512 output? Let's say DSD64.

 

Then from my point of understanding, must be DSD64 - PCM - DSD512.

 

So that conversion DSD->PCM also happens during DSD64->DSD512 conversion. Am I understanding correctly?

No, you are not.

There is NO midway conversion to PCM - it is DSD64->DSD512 direct.

This is a major selling point of HQPlayer, even for convolution! 

[jimdukey]

Everyone ELSE does it that way, NOT HQ!

JRiver  and AV+ covert to DSD to PCM, say there is no other way.

But there IS! 

HQ.

[louisxiawei]

4 minutes ago, bibo01 said:

No, you are not.

There is NO midway conversion to PCM - it is DSD64->DSD512 direct.

This is a major selling point of HQPlayer, even for convolution! 

Ah, I feel embarrassed. I misunderstood it the whole time until now.

 

Thanks a lot! 

[hifial]

bibo01 is 100% correct!

 

louisxiawei, where did you get this from? What would be the point of this with HQP?

Well, there might be a benefit sound wise under certain conditions but that is not the main point of HQP.

[Silly goose]

6 minutes ago, bibo01 said:

No, you are not.

There is NO midway conversion to PCM - it is DSD64->DSD512 direct.

This is a major selling point of HQPlayer, even for convolution! 

Roon has DSD direct up sampling now I hear. I've been chatting to an old forum favourite and he has me intrigued about up sampling. 

[Silly goose]

Has anyone compared the latest DSD direct up sampling in Roon to Hqplayer up sampling? Blizzard says Daphile sounds better but I am going to need multichannel for my active speakers and Daphile won't do multichannel. 

[craighartley]

45 minutes ago, jimdukey said:

Everyone ELSE does it that way, NOT HQ!

JRiver  and AV+ covert to DSD to PCM, say there is no other way.

But there IS! 

HQ.

[bibo01]

1 hour ago, Silly goose said:

Has anyone compared the latest DSD direct up sampling in Roon to Hqplayer up sampling? Blizzard says Daphile sounds better but I am going to need multichannel for my active speakers and Daphile won't do multichannel. 

Quite a few have compared Roon upsampling to HQPlayer. I have not found one that preferred the former (not even Blizzard), but you better test it yourself.

There are various reasons why Blizzard prefers Daphile. However, a personal preference is...personal

Anyhow, I told him that I measured Mansr's modulator (Daphile's DSD upsampler/converter) and it's rather basic/coarse. I published the results here (you can google translate).

[Silly goose]

4 minutes ago, bibo01 said:

Quite a few have compared Roon upsampling to HQPlayer. I have not found one that preferred the former (not even Blizzard), but you better test it yourself.

There are various reasons why Blizzard prefers Daphile. However, a personal preference is...personal

Anyhow, I told him that I measured Mansr's modulator (Daphile's DSD upsampler/converter) and it's rather basic/coarse. I published the results here (you can google translate).

2-3 days ago a brand new Roon up sampling was released with lots of improvements. Blizzard says they implemented lots of the requests he asked for. But he said he hasn't had a chance to listen to it yet. On the Roon forum some people are saying it's a vast improvement over the first generation Roon up sampling. But Blizzard also hasn't heard the Hqplayer 2.16 update either. But I do like the idea of 1 simple software so I'm hoping Roon on it's own is good. Regarding Daphile, he said both him and every one of his clients prefer it over Hqplayer. But that was before 2.16 so who know's now. 

[bibo01]

18 minutes ago, Silly goose said:

2-3 days ago a brand new Roon up sampling was released with lots of improvements. Blizzard says they implemented lots of the requests he asked for. But he said he hasn't had a chance to listen to it yet. On the Roon forum some people are saying it's a vast improvement over the first generation Roon up sampling. But Blizzard also hasn't heard the Hqplayer 2.16 update either. But I do like the idea of 1 simple software so I'm hoping Roon on it's own is good. Regarding Daphile, he said both him and every one of his clients prefer it over Hqplayer. But that was before 2.16 so who know's now. 

I am aware of Roon upgrade. That's why I told you to test listen yourself. Convenience is good.

For the rest, it says it all!

[jhwalker]

6 hours ago, jimdukey said:

Everyone ELSE does it that way, NOT HQ!

JRiver  and AV+ covert to DSD to PCM, say there is no other way.

But there IS! 

HQ.

 

Just to be clear - the newest version of Roon (v1.3, build 216) also resamples DSD > DSD without going through PCM, as well.

 

Don't know if it sounds as good as HQPlayer (though I've listened the last several days without HQPlayer in the loop and haven't missed it), but cool they are able to do this.

[ted_b]

10 hours ago, zorntel said:

Can you hear a difference between the 2s and non-2s XTR filter at DSD512 with the DSD files?  I can't.  I can hear the difference at DSD128 and 256, but not at 512.  Either filter sounds unbelievable at 512 through...so I'm guessing that we might not be missing anything with not being able to get the non-2s filter to work on PCM at 512.

+1  I agree that the latest 2s filters mimic the regular ones by about 99.9% at DSD512, and I wouldn't want to blind test them; I'd fail.  If your DAC has DSD512 as its clear sweetspot, then I wouldn't bother spending big dollars to get to non-2s, especially if you can do the xtr-2s family.  My $.02.

[lmitche]

57 minutes ago, ted_b said:

+1  I agree that the latest 2s filters mimic the regular ones by about 99.9% at DSD512, and I wouldn't want to blind test them; I'd fail.  If your DAC has DSD512 as its clear sweetspot, then I wouldn't bother spending big dollars to get to non-2s, especially if you can do the xtr-2s family.  My $.02.

Many thanks to Ted, Zorntel, Jussi and everyone else for contributing answers here on the limits of polysinc-xtr and DSD512 hardware.  I'll going to play around for with the new graphics card for another day or two and then likely return it.

 

Ted, your clarification on PCM is helpful.

 

Zorntel, thanks for testing with your son's graphics card.  May he never know you stole it for a few hours!

 

Jussi, thanks for the explanation of polysic-xtr limits and your monster machine.

 

If anyone has a i7-7700k and a Z270 motherboard, it would be useful to know if polysic-xtr pcm upsampling to dsd512 is possible with a GPU.

 

Thanks again. Larry

[Dr Tone]

6 hours ago, Silly goose said:

2-3 days ago a brand new Roon up sampling was released with lots of improvements. Blizzard says they implemented lots of the requests he asked for. But he said he hasn't had a chance to listen to it yet. On the Roon forum some people are saying it's a vast improvement over the first generation Roon up sampling. But Blizzard also hasn't heard the Hqplayer 2.16 update either. But I do like the idea of 1 simple software so I'm hoping Roon on it's own is good. Regarding Daphile, he said both him and every one of his clients prefer it over Hqplayer. But that was before 2.16 so who know's now. 

 

I still much prefer the sound of HQPlayer poly-sinc-xtr ASDM7 to DSD256 over any configuration of Roon's filters and modulator.

 

For those of us with temperamental drivers and DAC firmware, the thing Roon does do better is it eliminates the Native DSD popping on track change and allows for a resync delay to deal with the lost 500msec of audio that is normal with McIntosh DACs.  I still think HQPlayer can improve in this area @Miska.

[jimdukey]

I have a Mac D150.

Only when using Finale Playback does the Mac clip off the beginning a note.

So I use the Mytek for that.

But for any other File Playback, it's fine, no clipped-off notes, no noise/clicks.

I listen in DSD 128.

If the Mac did that for normal general playback, I would have returned it.

I'm surprised others are having that problem w/Mac D150.

I have an iMac/Sierra, use IFI USB products.

I use Poly-sinc Extr mp, ASDM7,and min/max Volume at -3db of upsamling everything to DSD 128.

Even DSD 128 files, rather that Direct SDM!

I'm liking it a LOT!

AV+/Sierra is a no go anymore.

Apple yanked the rug out from under Damien, removing something that defeated Direct Mode, which is now disabled by default, and refusing to put it back.

Thank goodness for Miska/HQ!

[jhwalker]

27 minutes ago, Dr Tone said:

 

I still much prefer the sound of HQPlayer poly-sinc-xtr ASDM7 to DSD256 over any configuration of Roon's filters and modulator.

 

For those of us with temperamental drivers and DAC firmware, the thing Roon does do better is it eliminates the Native DSD popping on track change and allows for a resync delay to deal with the lost 500msec of audio that is normal with McIntosh DACs.  I still think HQPlayer can improve in this area @Miska.

 

I did some careful listening this morning and discovered (in my system, at least) poly-sinc-xtr drastically cuts treble content - which does sound more "analog" and "warm / smooth"  but not sure why I'd want to cut off all the high frequencies in my listening.

 

Going back to poly-sinc / poly-sinc-shrt restored the high frequencies . . . but now it sounds almost exactly like Roon's Precise / Linear Phase, 7th order processing. 

 

So I'm thinking using optimal processing for each, they sound pretty much the same since the Roon update, at least to me.

[Dr Tone]

25 minutes ago, jimdukey said:

I have a Mac D150.

Only when using Finale Playback does the Mac clip off the beginning a note.

So I use the Mytek for that.

But for any other File Playback, it's fine, no clipped-off notes, no noise/clicks.

I listen in DSD 128.

If the Mac did that for normal general playback, I would have returned it.

I'm surprised others are having that problem w/Mac D150.

 

DSD 128 DOP is perfectly fine.  I'm talking Native DSD.