HQ Player

[Achim]

Have you somehow optimized your Win 7 installation? You might be able to free some resources...

 

Thanks everyone - bibo was on the right track.

 

None of the suggested measures changed anything - until I noticed that I had set a energy saving scheme in order to reduce fan noise - put the scheme back on "max performance" and 5.6Mhz DSD plays just fine!

 

Thanks again!

Achim

[Le Concombre Masqué]

try dsd 128. Your issue probably comes from a wrong SDM pack setting. Have a look to my preferences : If my TEAC was connected it should appear as device in lieu de "default". As of the DSF settings, Direct SDM should be unchecked hence the rest doesn't matter

Hi, sure, sorry for having not answered that. There are many differences between your case and mine. Maybe your TEAC (isn't that the same company that makes Esoteric) is a better DAC than Ayre. You may be using HQPlayer to convert to DSD 5.6.

 

On the other hand, I have not tried the DSD 2.8 path (available to me) at all. In fact, when I try to play DSD files, I can hear complete silence, not sure what I'm doing wrong.

 

Since my last tests, I have got some improvement with HQPlayer, I've set the upsampling to get to 192 KHz and the filter to sinc with TPDF dither. That sounded certainly better for me than poly-sinc, go figure. Still, I need perhaps more quiet listen-time with a larger variety of input formats and genres. I'm keeping an open mind.

 

However, what rubs me the wrong way is ritual killings of very reputable alternatives, be it software or hardware. So Teac/Esoteric are a bunch of amateurs and the HQPlayer author knows best and what they are doing is stealing their customers hard-earned money? Is that really suggested here as a proven fact?

[Nikhil]

I will try this, but since I need/want to use the volume control, I will not be able to have it on permanently.

 

I don't think enabling "Direct SDM" affects volume control. It just bypasses the DSP engine for DSD content only. Everything else works as it should including volume control.

[Miska]

In any case Miska, it's not fair to judge a dac by the cost of the chip...you know better.

 

After certain price point I would expect DAC manufacturer to be able to do a converter without resorting to COTS DAC chips. IOW, actually doing converters of their own. Many manufacturers manage to do it even at lower price points.

 

Think about following for example that do their own stuff:

- Playback Designs

- Meitner / EMM Labs

- dCS

- MSB

- Chord

- T+A (for DSD path)

- Lampizator

- Metrum

... (I probably forgot things from this list)

 

Let's say a DAC costs 5k+, what are you actually paying for if it still uses the $10 COTS DAC chips?

 

Sure, as long as they use those DAC chips it's especially easy for me, because I can easily cover wide set of DACs at one shot if they use same chip.

[Miska]

When saying "HQPlayer is the best", what does this statement actually mean--scientifically or even musically?

 

Usually I take objective approach and measure things, so scientifically.

 

I don't need to even take Esoteric DAC, it's enough for me to know which DAC chip it uses to know it's technical limitations. (you can see specs of the DAC chip from it's datasheet you get from the chip manufacturer)

[Miska]

Thanks, I did that already, to no avail.

 

In combination with highest buffer settings in the Oppo Asio driver.

 

Someone had similar problems with Oppo ASIO driver and solution was to leave buffer time setting in HQPlayer to "Default" (=use size proposed by the driver) and set buffer size from the driver's control panel.

 

Also with your CPU, it is better to leave Pipeline SDM disabled. You may have better success with -2s filter variants.

[Miska]

I don't think enabling "Direct SDM" affects volume control. It just bypasses the DSP engine for DSD content only. Everything else works as it should including volume control.

 

It also locks the volume control for PCM-to-DSD case so with "Direct SDM" enabled, there's no volume control available for DSD output mode. This is to avoid nasty surprises when source content switches between PCM and DSD...

[warrior-kid]

try dsd 128. Your issue probably comes from a wrong SDM pack setting. Have a look to my preferences : If my TEAC was connected it should appear as device in lieu de "default". As of the DSF settings, Direct SDM should be unchecked hence the rest doesn't matter[ATTACH=CONFIG]27938[/ATTACH][ATTACH=CONFIG]27939[/ATTACH]

Thanks, that was helpful. I reproduced these settings and can now listen to DSD.

 

The first thing I've compared is Oyo Como Va by Santana in DSD64.

 

Here are two snapshots taken by Spectrum Analyser on iPad from near the end when the guitar comes from the center. The first photo is HQP, the second is A+. The pre-amplifier has exactly the same volume setting--I just closed HQP and started A+ to do the second playback. There are a number of differences between the two. Firstly, for alignment, the guitar starts after around 7 sec offset in the first picture (HQP) and 5 sec offset in the second (A+). One thing I can see is that there is energy above 20kHz in A+ and hardly any in HQP. The guitar above 16 kHz is much sharper in A+: all the individual frequencies look bright and are separate from each other while in HQP, they are more washed off. The second guitar riff in A+ about 2 seconds later looks almost the same as the first one while in HQP, the second riff has hardly any bright dots. HQP looks better in that the area before the guitar strikes is darker, so perhaps some preringing occurs in A+. Based on these plots, you can't really draw much conclusions but for sure, they are different. As of now, hearing it, I like the A+ version better because I am still struggling with what I can only describe as 'disjointedness' in the HQP presentation. I am feeling that the instruments are somehow unaware of each other and are ever so slightly out of sync, which compromises the impression. If anyone had heard anything similar, I'd really appreciate your thoughts on what this could be caused by.

UPDATE: One more difference I can see, there is a dark band in HQP around 2.8 kHz that is missing on A+. Also, there is definitely more energy between 16 and 20 kHz in A+ with quite a few bright dots and stripes.

[juanitox]

on my mac mini audirvana+ sounds more "analog " for me where you find it less "disjointedness" but when i'll move to my windows PC HQP player sounds far better than with the mac mini . these software are very dependent from the OS , hardware and other software and tasks running with them. i like HQPlayer running in a Ramdsik others don't, too many variables to make a rule..

[R1200CL]

In order to test HQPlayer on a QNAP, the only way is to use Container Station and install Ubuntu, right ?

[tboooe]

 

PC is a Win7 Lenovo laptop with i5-2520M CPU and 2 cores at 2,5Ghz. Hyperthreading enabled.

 

I do not think a dual core can handle DSD128. I had a dual core i3 and I would get what sounded like static unless I maxed out the buffer settings of my dac driver. Perhaps you can try that.

[tboooe]

In order to test HQPlayer on a QNAP, the only way is to use Container Station and install Ubuntu, right ?

 

Thats right. However, are you expecting to upsample to DSD? Which QNAP model do you have?

[warrior-kid]

on my mac mini audirvana+ sounds more "analog " for me where you find it less "disjointedness" but when i'll move to my windows PC HQP player sounds far better than with the mac mini . these software are very dependent from the OS , hardware and other software and tasks running with them. i like HQPlayer running in a Ramdsik others don't, too many variables to make a rule..

Many variables for sure, thanks for the advice, I'll keep on with my quest.

[lightminer]

Hey, everyone. 2 questions.

 

First, I have Maggie 3.6qr's. Would that change my 'filter' settings for DSD? The Maggie's go really high, can't remember, maybe 50kHz? Other speakers that don't go that high may be able to use different filters.

 

Second, is there a consensus on Roon/Tidal DSD 128 settings? I seem to see things ending up at poly-ext or closed form depending on the user. Can someone mention all of the settings - I.e., asdm7 vs others, filter, and primary algorithm, i.e., closed form.

 

Assume I have infinite CPU, no matter what I do I never get above 25%. I've also got a good GPU, but doesn't seem necessary to involve since I'm not over 25%. 3930k, btw, liquid cooled, 32 GB RAM, 2 SSDs, 4 HDDs, and SOTM card. I use it for video processing and other things as well as audio - it isn't a dedicated audio rig. I may do that later. Exploring a relatively new Aries Mini which won't support Roon or HQPlayer as far as I can tell.

 

So far I didn't see much or any difference between DSD128 and leaving Roon/Tidal alone. It is easy to AB, you just set up different output zones and go back and forth, that part is very nice. One thing is I'm running a 15 foot Belkin gold, with much annoying effort I can use a 0.7M Supra (have to move monitor, computer, more........). This is on a pretty resolving system, Pass Labs amp and Supratek Pre. Lowest quality piece is the DAC - Oppo105d for now. Though with vibr stabilization and great power cord, pure audio setting, etc., sounds pretty darn good for what all else it does. Will upgrade at some point.

[lightminer]

Oh - and I think it would be great to have Small Green Computer or someone sell two versions of the DSC-1 (or DSC-2 as it evolves), one pretty normal taking cost into consideration and the other letting it open a bit wider and use really good parts, separated power supplies for each section with chokes, very solid (maybe even class A) output section, etc. That second one may still not be very expensive and a bunch of us may get one just to compare and have fun with. I've built lots of stuff in the past, but no time now to do it myself.

 

Since we are HQPlayer folks here, it could be DSD 512-only. That would constrain the hardware design and maybe make it easier to have it at lower cost. That would kind of be funny - a DSD 512-only DAC. Not even 256 or anything else. Certainly gives the hardware side focus.

 

Lampizator offers DSD only and people I think use HQPlayer to drive those.

[shoom]

I do not think a dual core can handle DSD128. I had a dual core i3 and I would get what sounded like static unless I maxed out the buffer settings of my dac driver. Perhaps you can try that.

 

I’m running the same I5 laptop with W10/Process Lasso into IntelJ1900 NAA poly sinc 2s @ DSD128/IDAC2 with no problem. As you say my DAC buffersettings are maxed out but I tend to run that way and min latency set mostof the time anyway.

No pipeline but with HT.

So it’s possible.

Sounds great.

[57gold]

After reading warrior-kid's post, upgraded A+ to listen PCM unconverted to DSD128 on Mytek DSD192 and compare sonics to HQPlayer, which I have been using for last 2 years. Immediately, noticed one sonic difference with A+...fan on 2.6Ghz Quad Core Mini started humming. Never heard it when playing HQPlayer.

 

Checked Activity Monitor and 4 Cores were at 75% area. Using HQPlayer, the Activity Monitor looks like I have 8 Cores down at 12% with full filters working. Not sure what's up?

[warrior-kid]

After reading warrior-kid's post, upgraded A+ to listen PCM unconverted to DSD128 on Mytek DSD192 and compare sonics to HQPlayer, which I have been using for last 2 years. Immediately, noticed one sonic difference with A+...fan on 2.6Ghz Quad Core Mini started humming. Never heard it when playing HQPlayer.

 

Checked Activity Monitor and 4 Cores were at 75% area. Using HQPlayer, the Activity Monitor looks like I have 8 Cores down at 12% with full filters working. Not sure what's up?

I never said I used any up- or cross-format conversion in A+ though--I never do that for PCM either. In fact, given that my DAC over-samples by 16x internally in hardware, my theory goes that whatever the up-sampling on the software side, (a) the DAC will over-sample further internally and the two methods may clash in some way and (b) the way the DAC was designed, I'm sure they never fed anything but native non-upsampled tracks. So, yes, the less _may be_ more.

 

The track Oyo Como Va from Santana's Abraxas I own is a well-regarded Acoustic Sounds DSD64 version, so in the comparison I made that revealed some sonic and perhaps some spectrum differences between the final results (like a roll-off above 16KHz in HQP), I did not use any upsampling or conversion (the DAC just showed 88, remember it does not offer DSD128 or above).

[warrior-kid]

Hey, everyone. 2 questions.

 

First, I have Maggie 3.6qr's. Would that change my 'filter' settings for DSD? The Maggie's go really high, can't remember, maybe 50kHz? Other speakers that don't go that high may be able to use different filters.

Only trying to answer the first: my speakers go 28 Hz-100 kHz PL300 | Monitor Audio. I understand that simply offers a safety of sorts in that the supersonics are treated well and not filtered in an odd way into the audio band. You may have a point in that that may affect the choice of the filter used in upsampling, if we do any.

[Freann]

...I did not use any upsampling or conversion (the DAC just showed 88, remember it does not offer DSD128 or above).

 

You mean the display showed 64? Otherwise you are respampling.

[warrior-kid]

You mean the display showed 64? Otherwise you are respampling.

You are right, both HQP and A+ are showing 64. I have no idea why I wrote 88--I cannot correct my past email though.

 

UPDATE: Quite interesting really, I can measure exactly where the instruments are in Samba Pa Ti in Abraxas DSD64. In A+, the lead guitar is in the centre, just over 4 feet high (1.25 m). The percussion is a foot or more to the right of the right speaker at the same height of over 4 feet. The organ when it comes in is ridiculously wide, something like 1.25 m to the left from the left speaker and a bit lower, while the supporting guitar is immediately to the left from the left speaker, same height. I can also sense a fair bit of floor vibration from low frequencies.

In HQP, the lead guitar and the right channel percussion are about 1m high, the organ is only about two feet to the left of the left speaker, while the supporting guitar is roughly as in A+. The guitar in HQP sounds a tad harsher than A+, a bit unnerving really. Also, I cannot feel much floor vibration through the track in the HQP version.

[Achim]

I’m running the same I5 laptop with W10/Process Lasso into IntelJ1900 NAA poly sinc 2s @ DSD128/IDAC2 with no problem. As you say my DAC buffersettings are maxed out but I tend to run that way and min latency set mostof the time anyway.

No pipeline but with HT.

So it’s possible.

Sounds great.

 

Yes, I can now definitely confirm this.

 

I guess the initial trouble was caused by an outdated Oppo Asio driver.

 

After updating this to the latest, I can run hyperthreaded, without pipeline, any filter, DSD128, with USB harddrive connected (no ssd!), still with loads of ressources to spare. But energy management must be set to max performance otherwise there will be throttling that gets you clicks and stutters.

 

 

Cheers everyone,

Achim

[Le Concombre Masqué]

Maybe you should try with an actor or actress singing and check with imdb which player provides the most accurate mouth height

 

i.e. Elvis was 5' 11¾" (1,82 m) (source : imdb)

You are right, both HQP and A+ are showing 64. I have no idea why I wrote 88--I cannot correct my past email though.

 

UPDATE: Quite interesting really, I can measure exactly where the instruments are in Samba Pa Ti in Abraxas DSD64. In A+, the lead guitar is in the centre, just over 4 feet high (1.25 m). The percussion is a foot or more to the right of the right speaker at the same height of over 4 feet. The organ when it comes in is ridiculously wide, something like 1.25 m to the left from the left speaker and a bit lower, while the supporting guitar is immediately to the left from the left speaker, same height. I can also sense a fair bit of floor vibration from low frequencies.

In HQP, the lead guitar and the right channel percussion are about 1m high, the organ is only about two feet to the left of the left speaker, while the supporting guitar is roughly as in A+. The guitar in HQP sounds a tad harsher than A+, a bit unnerving really. Also, I cannot feel much floor vibration through the track in the HQP version.

[k6davis]

I recently upgraded from a Core i7 3770 to a Core i7 6700. I upsample everything to DSD256 with poly sync & DSD7. Of course, everything is faster and runs cooler with with lower resource utilization, but the load times for the various upsampling filters and algorithms didn't improve as much as I'd hoped for.

 

This is such a simple idea that I'm sure it's already been considered, but here it goes anyway.

 

Given enough computer resources, is it feasible to offer an option to load several different sets of the filters and algorithms into the computer's memory at once and switch between them as the source material requires? That way there would be one long wait at the beginning of a session, but no wait after that. If this were possible, I'd want to load filters for the following source content, all at once: Redbook, 24-96, 24-192, SACD/DSD x 1.

 

Maybe there's something that could be changed in my settings or configuration, but (particularly) when going from Hi-Res PCM back to Redbook, the filters have to load again, as if I'm starting a brand new listening session.

[57gold]

I never said I used any up- or cross-format conversion in A+ though--I never do that for PCM either. In fact, given that my DAC over-samples by 16x internally in hardware, my theory goes that whatever the up-sampling on the software side, (a) the DAC will over-sample further internally and the two methods may clash in some way and (b) the way the DAC was designed, I'm sure they never fed anything but native non-upsampled tracks. So, yes, the less _may be_ more.

 

Just assumed everyone used HQPlayer for upconversion so that they benefited form its filters...for playing PCM files in native fashion I'd use a memory player like A+ or PM for best sound.