Jump to content
IGNORED

HQ Player


Recommended Posts

Hi Miska,

 

I played NY's 24/192 Cowgirl in the sand twice with the Classic Records Quiex vinyl in between and felt unhappy for the vinyl was far from clean and quiet while the 24/192 offered very frustrating drums (it's brighter too but I wouldn't say harsh and thought that this could have been debated as a matter of taste.) Then I pulled HQP and felt the analogue feel right while, honestly, I concluded from my previous tests that it was too soft by some kind of DSPing.... Well, I reconsider ! And apologize for being redundant :

 

so with a TEAC 501 I should convert to DSD 2x, right, cause it does it anyway, right ? what are your current best setting ?

 

I was proud so far of my 55 processes OS X Mavericks on a SD card. But HQP converting to dsd 2X is ressource intensive and I'd better operate from the SSD, correct ?

 

RAM disks are fine ?

 

Is there any script or at least process cutting steps that you do approve (I remember you don't approve CAD's and I went even further...) ? (At least I will open a dedicated session)

 

how far from a NAA included (seems scaringly complicated + means further expense) combo's SQ will I be with HQP desktop ? Plus I don't feel comfortable with the idea of having a cheap computer connected to the DAC (i.e. the right usb port of my mbp retina sounds way better than the left) or of having a cheap computer doing the heavy lifting with overloaded cpus

 

thanks in advance

 

 

 

 

The left-most plots of 0 - 20 kHz are most important as it tells about the audio band performance. First plot, "test1", has full scale 1 kHz sine on one channel and silence on the other one. Check that noise floor level and smoothness without discrete tones. If there are tones, check the level and how many of those are. For example for full level 1 kHz and 3rd harmonic at 3 kHz being -150 dB and 5th harmonic at 5 kHz being -170 dB results in (0.000003 + 0.0000003) -> THD=0.0000033%, if there are more than one, distortion is sum of those. Also check that the noise floors between channels are closely matching in shape and level. Dips in the noise floor are normal and there by design for shaping the noise floor response, flattening the slope for audio band. (for comparison with all the results, ESS lists THD = -120 dB = 0.0001% for their best chip)

 

Middle plot gives overall picture of the modulator order and noise shaping. For DSD64, 50 kHz is the specified usable band, for DSD128 it is 100 kHz. For these bands, check same as above.

 

Right-most plot is more of a curiosity, since most of the band is cut out by the analog reconstruction filter in DAC. Mostly useful for people doing fine tuning of modulators.

 

Second plot, "test2", now has -70 dB and -130 dB 1 kHz tones on left and right channels respectively. Same checks as above and you can also determine whether processing has enough resolution for reproducing -130 dB tone cleanly. Also check that the noise floors match within 50 kHz band for DSD64 and 100 kHz band for DSD128, also compare to test1.

 

 

This is kind of easy test, there are other tests that can really stress a modulator...

Link to comment

ok, here I stand :

 

I trial HQPDesktop in a standard but dedicated session vs Amarra 3 and A+ on a 55 processes trimmed Mavericks SD card and consider this fair for I can’t use HQP on that SD card.

 

I can copy/paste Stereophile’s comments on PS audio DSD converting DAC : « consistently smooth and slightly laid-back sound/ lacked the excessive bite I've heard through other gear (my comment : on certain harsh files) /sound was considerably less forward/ I often found myself inching up the volume control, hoping for just a bit (my comment : minus « a bit » afaimc)more impact. »

 

I haven’t decided yet if I’m intoxicated and addicted to an edge that brings more slam, impact, presence (and sometimes, quite rarely, unpleasant bite)or if HQP mellows things. HQP (everything processed through polysync short mp/dsd7/5.6 M to a TEAC 501) is very different : if it’s better it is by such a margin than I’m not ready to claim its victory yet.

 

Frankly I'm puzzled they write this of a DSD converting DAC as opposed to other high end DAC that all, if I read you correctly, are supposed to ALL indeed dsd convert internally....

 

Questions remain :

what are the current best settings for a natural time domain respectful rendition ?

what for NAA : Cubitruck would be the best/no need to spend more choice ? what is available batteries powered ?

where is the “usual place” to find Debian for it and a dummy proof install?

I own a usb to Ethernet adaptor from a defunct MB air ; I understand I need a thunderbolt to Ethernet adaptor for my MBPro R ; is it as kosher as a native Ethernet port in terms of isolation and whatever benefits you expect ?

Merry Christmas

 

 

 

Hi Miska,

 

I played NY's 24/192 Cowgirl in the sand twice with the Classic Records Quiex vinyl in between and felt unhappy for the vinyl was far from clean and quiet while the 24/192 offered very frustrating drums (it's brighter too but I wouldn't say harsh and thought that this could have been debated as a matter of taste.) Then I pulled HQP and felt the analogue feel right while, honestly, I concluded from my previous tests that it was too soft by some kind of DSPing.... Well, I reconsider ! And apologize for being redundant :

 

so with a TEAC 501 I should convert to DSD 2x, right, cause it does it anyway, right ? what are your current best setting ?

 

I was proud so far of my 55 processes OS X Mavericks on a SD card. But HQP converting to dsd 2X is ressource intensive and I'd better operate from the SSD, correct ?

 

RAM disks are fine ?

 

Is there any script or at least process cutting steps that you do approve (I remember you don't approve CAD's and I went even further...) ? (At least I will open a dedicated session)

 

how far from a NAA included (seems scaringly complicated + means further expense) combo's SQ will I be with HQP desktop ? Plus I don't feel comfortable with the idea of having a cheap computer connected to the DAC (i.e. the right usb port of my mbp retina sounds way better than the left) or of having a cheap computer doing the heavy lifting with overloaded cpus

 

thanks in advance

Link to comment

Thank you for the tip, did not hear fuzziness, maybe more of the physicality I was looking for, the numbers of parameters I'm dealing with in the comparison is a bit mind blowing, but here I stand :

 

 

HQP works indeed on my trimmed Mavericks sd card.

 

Have to check DirectSDM in DSSDiff settings even to play pcm (otherwise it's very faint).

 

use poly short lin/IIR as integrator/low as noise filter, 0 gain

 

use polysync ASDM5 5.6M SDM via DoP 32 dac bits pipe line sdm

 

it's unclear to me if though DirectSDM has to be checked if I want to hear something I 2X dsd.

 

Anyway, I can't hear when and where use A+ in place of HQP to play DSD files. HQP makes better sense, thus better spacial enjoyable sound on any programme would it be Bowie's complex electronics or Bach by Savall sacqueboutes (antique trumpets, quite a test with several of them playing...)

 

As of PCM files, it's even harder to test : on the sd card I get awful noises if i want to come back from hqp to Amarra3. And setting a fair reference level is almost impossible. If one admits the process of optimizing the level for each player, with a different delta depending on the programme, I have to say I enjoy more and more HQP.

 

Though I have advocated against JRiver or Korg pcm to dsd conversions for I only got marshmallow, that I have advocated against amy type of up sampling be it src in A+ or offline up sampling in RX3 for it causes losses in soundstage, dimensionality etc, I have to say that, on the contrary, HQP is the player that breathes the more space (too systematically ? haven't decided yet)

 

Questions remain :

what are the current best settings for a natural time domain respectful rendition ?

what for NAA : Cubitruck would be the best/no need to spend more choice ? what is available batteries powered ?

where is the “usual place” to find Debian for it and a dummy proof install?

I own a usb to Ethernet adaptor from a defunct MB air ; I understand I need a thunderbolt to Ethernet adaptor for my MBPro R ; is it as kosher as a native Ethernet port in terms of isolation and whatever benefits you expect ?

Merry Christmas

 

 

 

Try ASDM5 when you are bothered by the SQ of certain tracks with DSD 7. I liken ASDM5 to an impressionist painting technique. The sound is slightly fuzzied up but then the essence of the auditory message is enhanced and the performance of the instruments can sing with more emotion
Link to comment

My TEAC 501 DAC Teac UD-501 Dual-Monaural PCM/DSD USB Digital to Analog Converter | AudioStream is a whole different machine, Miska owns one btw. However, since it costs about the price of the xlr cables that go to the pre, I admit that it's very possible that I can buy better. Just don't plan to hurry.

 

The Teac 501 as described in this review explains what you are hearing: TEAC AI-501DA review | Stereo amplifiers | What Hi-Fi?

 

The Teac offers a lot for the price but a better dsd dac might get you where you are trying to go. I don't think HQP is your problem. I believe that you have described exactly what the reviewer heard in the Teac.

Link to comment

Hi Miska,

 

I'm about to claim HQP victory...

 

Remember that I have advocated against JRiver or Korg pcm to dsd conversions for I only got marshmallow, that I have advocated against amy type of up sampling be it src in A+ or offline up sampling in RX3 for it causes losses in soundstage, dimensionality

 

Remember that I wrote : HQP is very different : if it’s better it is by such a margin (compared to Amarra 3, my reference for PCM files and A+, my reference for dsd files than I’m not ready to claim its victory yet.

 

+ the heavy cpu load is not exactly what I like to see coming from months of OS X optimization...

 

see : Capture d’écran 2014-12-26 à 09.53.55.jpg

 

My current settings are: Capture d’écran 2014-12-26 à 11.15.15.png don't bother looking the PCM settings, the whole idea is to use SDM

Nevertheless, the bottom line is that there's 0 SQ related reason to use A+ for DSD files and now even Amarra sounds digital in comparison though I still like it. Just had to dare crank the volume up to 6 dB to get slam and impact with HQP... Voices are gorgeous with HQP...

 

Questions remain :what best settings (including 24 (cf Superdad) vs 32 bits) do you recommend for a TEAC 501 ?rest of the system is much more ambitious, top studio grade, though 20+ years old

 

what for NAA ? I thought of Cubox for you have an image but then I read :"the Mac mini as NAA made the Cubox-i4Pro unlistenable" ? I would like to go for something you have an image for (I haven't found where it's listed as available) so I don't get too scared to do...

Is there a way to get the TEAC working with a ASIO background and a Mac ?

I own a usb to Ethernet adaptor from a defunct MB air ; I understand I need a thunderbolt to Ethernet adaptor for my MBPro R ; is it as kosher as a native Ethernet port in terms of isolation and whatever benefits you expect ?

 

Thank you and kudos

Capture d’écran 2014-12-26 à 11.14.30.png

Link to comment

thank you; Yesterday I was not happy with MFSL SACD Tapestry (it was much worse on A+ and/or with the 1999 SACD remaster) : might want to try mp but what puzzles me is that DirectSDM has to be checked, even on native Yosemite, whether with dsd or pcm files. Thus I don't think I can act on the dsd files. I"ll try poly shrt mp in the dsdiff settings. Poly sync has been my preferred filter so far

 

Your settings look OK. You can disable "Pipeline SDM" to get lower overall CPU load (it may be necessary for DSD256/DSD512 output speeds).

 

Just try out different oversampling filters of the "poly-sinc" family to find one you like the most. Sometimes it may be beneficial to switch the filter for different types of music. I recommend minimum-phase "-mp" versions for studio productions and linear-phase ones for classical and other recordings made with minimal mic'ing such as classical music. Minimum-phase tends to sound better with sharp transients such as close mic'ed percussions.

Link to comment

i'm experiencing dropouts under regular (but the sudo nvram boot-args="dart=0x0" trick) Yosemite on ssd.

I have created a HQPlayer account and had quit my regular session.

Purging RAM does not help. Would a restart rather that session closing/opening help ?

should I consider HQP SQ with lesser filters (config was as above)? can't afford a Mac Pro to play it !

 

i apply offline a Alloy/RX3 eQ curve to my pcm files and was hoping to correct every file while outputing 2X dsd thanks to convolution ; should I forget about it ? forget for dsd at all and pursue my actual process for pcm?

Link to comment

since it occurred after i unchecked pipeline i will engage it ; does it make sense ?

 

i'm experiencing dropouts under regular (but the sudo nvram boot-args="dart=0x0" trick) Yosemite on ssd.

I have created a HQPlayer account and had quit my regular session.

Purging RAM does not help. Would a restart rather that session closing/opening help ?

should I consider HQP SQ with lesser filters (config was as above)? can't afford a Mac Pro to play it !

 

i apply offline a Alloy/RX3 eQ curve to my pcm files and was hoping to correct every file while outputing 2X dsd thanks to convolution ; should I forget about it ? forget for dsd at all and pursue my actual process for pcm?

Link to comment
Capture d’écran 2014-12-27 à 16.25.59.pngI'm also experimenting strange noises that made me fear of a fried tweeter after I tried to go back to Amarra straight from HQP... don't think so though since it occurs at low levels too, strangely located hard left.... is my system revealing dirty things in the mix ?
If things work better with it, then I recommend using it. It increases overall load, but splits work to more CPU cores.
Link to comment

Excellent suggestion ! Now it seems that the nasty things, when they happen, always happen a few minutes after the launch of a dedicated session... what would be your recommended power savings ? May I/should I keep App nap ?

 

I will rework the questions in a mp I'll send you later.

 

Thanks for the answers so far and I will check my Audio midi settings though I think they were correct. The issue with Amarra is that, even though HQP is shut, exited, nada present, it has let core audio or whatever in such a state that Amarra outputs nothing but screeching sounds if one does not reboot. Anyway I feel my future will be with HQP as sole player not in comparisons !

 

 

I cannot say much about this. I don't have any CPU load issues with my Core i5 27" iMac playing to exaSound e28, iFi iDSD Micro or Mytek (connected using Thunderbolt - Firewire adapter). This is on Mavericks.

 

Some power savings could be doing tricks on laptops though. I have disabled power savings on the iMac.

 

 

 

Sorry, I have probably lost track of the questions I have not answered yet...

Link to comment

hqp was neither killed or crashed but it happened with my 55 processes Maverick on sd card : might well be due to OS X not being nominal

This shouldn't be an issue with version 3.5.1 and later, as long as it is closed cleanly. It restores all hardware settings back to the original values. I've tested this with iTunes and Spotify and neither have had problems playing correctly after closing HQPlayer.

 

If it crashes or gets killed, then it of course doesn't have a chance to clean up.

Link to comment

I experienced dropouts a few minutes inside the first dff track of a sacd rip. everything unchecked in System Preferences for Mission control, software update, Power savings, Spotlight ; no firewall, no Time machine ; no Notification ; + dropped a few lines in Terminal to turn off hibernate and Sleep image....

 

everything went fine for the next 20 minutes or so of my listening session, with gorgeous moments and others when I wondered if my system reveals dirty things in the mix, recording artifacts, or conversion artifacts... I'd like to be serene and that when I hear nasty things it's only because my system has become even better that it was before...

Please note that I can't run HQP at all if DirectSDM is unchecked (why???!!!) thus processing and cpu overload might not be the explanation. Melvin pointed Yosemite/usb issues....

 

I run HQP on a lightly optimized Yosemite and I preferred it over trimmed Mavericks on SD card...

I find myself toggling much more often the phase switch on my pre

 

I experienced dropouts (reminds me the sounds of a CD player meeting damaged sectors) again while every parameter was checked as nominal. it occurred 20+ minutes of a 24/96 harpsichord recording at a moment that did not appear special per its rhythm, forte or whatever
Link to comment

I created a Yosemite SD card erasing my optimized Mavericks on it for anyway i found it to sound inferior to native Yosemite on ssd.

 

Is it worth it ?

 

Well, we are jan the 1st and i might be slow, happy new year btw…

 

in a nutshell : i had to have 2 similar Mac ready to save boot times and refine my findings..

 

real or not, i built consistent impressions : Y on ssd was snappier, Y on sd card and ssd dismounted brought more details, was softer, more 3D but less engaging… I can’t see the rational but adding 1dB to the SD card session brought back snap and kick while preserving 3D etc. Ten I optimized the card « au feeling », thought it widened the soundstage while it should not cripple the os (see attached)

 

Am I happy to have spent a few hours of a slow anyway day ? yeah… but my system is very resolving and the whole thing has less effect than an incorrect AP

 

btw I found myself toggling a lot the AP switch on my pre since I turned to HQP and think that HQP when outing to DSD inverts polarity.

 

 

Not sure I’m ready to spend a lot on a NAA : HQP works very well already on native YOsemite directly connected to DAC and does not benefit tremendously of my relevant, though not exactly the same as putting a NAA in the chain, effort

Yosemite to DAC via USB: lot more computer noise that can affect D-A conversion. An NAA is a simpler device, the SOTM has some isolation, so overall the D-A should be better.
Link to comment

interesting ; have you selected the best usb port on the Mac ? i know i'm not very good at it myself but how would you describe the substantial SQ increase ? which power supply did you use with the SMS?

 

Several months ago we did testing of a OPPO 105D being fed by HQ Player on Mavericks direct and via a SOTM SMS-100 in NAA mode. The improvement is sound quality was substantial.

 

Recently we tested a Surface Pro 3 running Win 8.1 HQ Player running direct to USB DAC, it was very poor. However the same SP3 feeding a SMS-100 NAA over WiFi, sounded even better than HQ Player on Yosemite feeding SMS-100 NAA over CAT 5, a totally unexpected result !

 

Seems like the SMS-100 in NAA mode is somewhat sensitive to noise over CAT 5

Link to comment

thank you, you should have a look at the usb tree and choose the less crowded port

 

We did not try / compare various USB ports on the MAC

 

The SQ differences were in the class of higher / lower noise, lifting veil, more details etc. We did not notice timbre differences, edginess, bright etc

 

The test with the OPPO 105D was done with the SOTM OEM wall wart.

 

The later tests with SP3 were done with a Chinese LPS feeding the SMS-100

Link to comment

Thank you ! actually there are 2 ways of doing this

 

turn the volume knob, which I refused to follow for in my pre i trust

 

bypass the volume control by setting min & max vales to 0 in preferences

 

Are you using the HQ Player for volume control or your DAC/Preamp? If your DAC/ Preamp then you need to set the HQ Player volume control settings for external volume control.
Link to comment

i have reported dropouts (had first one this morning with my sd card ; would they actually be clipping ?I don't like the idea to alter the volume by sw

 

Unless I am missing something here Miska recommends setting volume down in settings -6db if you are using DSD output. I have tried both 0 and his -6db and I hear no difference in sound quality. I leave the volume control in HQP off (I don't use it). I do set -6db in the control panel as it is supposed to allow necessary headroom for processing according to Miska. I believe he said this will prevent any possibility of clipping with the filters. Like most things I just do what I'm told.
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...