HQ Player

[Le Concombre Masqué]

if it might help : if the track is n48, I can not set PCM poly sinc ext to output to 384 (not listed), coming from SDM ; if I switch to regular poly sinc, 384 is offered but it usually does not start ; if it does, then sometimes retrying ps ext works...

I'll check if I can reproduce the crashes, then I could fix those...

[Le Concombre Masqué]

?? oh yes it (poly sinc ext) can go from 48 to 384 ; just NOT when coming from SDM

Yes, poly-sinc-ext doesn't allow those conversion factors. Regular poly-sinc can do practically any ratio, but depending on the output rate and ratios it may take a while to initialize when you start playback. Depending on machine something like 20 seconds is not unusual.

[Le Concombre Masqué]

I know, I knew that. There's a bug, not a misuse here

It is not about the absolute numbers, but about the ratio. So yes it can convert 44100 to 352800 and 48000 to 384000, but cannot be used to convert 44100 to 384000 or 48000 to 352800 unlike for example poly-sinc. Different filters have different rule set how those can be used.

[Le Concombre Masqué]

can not play pcm through poly sinc ext when I just outed DSD trough Closed Form/SDM before

I still don't get what bug you are talking about...

[Le Concombre Masqué]

GIVE PCM 384 a try through poly sinc ext

Bold is mine.

 

+1

 

It's is a long time I listen everything under DSD.

 

Roch

[Le Concombre Masqué]

Please report. I have to admit I'm lost...

 

I'm back to SDMing everything, as I should considering the chip inside my DAC, back to poly sinc as well for n48 files (96 192 384) since there's only one crystal, 44 based, in the DAC and I prefer to let HQP do the conversion job.

 

However, for everything n44 would it be PCM or DSD files, I prefer the rate exclusive and new filters.

 

Trouble is : which one ? it seems very file quality dependent. While listening to the gorgeous new rendering of Kind of Blue by MFSL (sacd then ripped) closed form was and is a very obvious choice to my ears. That is also true of i.e. Meddle, another MFSL production, but CD : much better though Closed Form. In almost every other instance I prefer poly sinc ext. poly sinc ext is apodizing while Closed form ; can it be the explanation and the basis of guidelines ?

 

Never mind!

 

Power supply to Regens was not turned on!

 

All is well, will be trying Closed form filters tomorrow.

[Le Concombre Masqué]

refresh :

 

sdm everything

favor integer filters

favor apodizing filters but for MFSL . googling reveals that MFSL has a history of using very specific ADC and algorithms. If anyone gets consistently better results while using Closed form (non anodizing filter) with another label, please report.

Please report. I have to admit I'm lost...

 

I'm back to SDMing everything, as I should considering the chip inside my DAC, back to poly sinc as well for n48 files (96 192 384) since there's only one crystal, 44 based, in the DAC and I prefer to let HQP do the conversion job.

 

However, for everything n44 would it be PCM or DSD files, I prefer the rate exclusive and new filters.

 

Trouble is : which one ? it seems very file quality dependent. While listening to the gorgeous new rendering of Kind of Blue by MFSL (sacd then ripped) closed form was and is a very obvious choice to my ears. That is also true of i.e. Meddle, another MFSL production, but CD : much better though Closed Form. In almost every other instance I prefer poly sinc ext. poly sinc ext is apodizing while Closed form ; can it be the explanation and the basis of guidelines ?

[Le Concombre Masqué]

I read some use DSD7 with hires, I do with DSD and usually opt for ASDM7 when the source is PCM

 

what is your rule of thumb when it comes to choose modulators ?

Yes, that's the difference, optimizing SNR in 20 kHz audio band by utilizing wider bandwidth of DSD256 and above.

 

The adaptive modulators ASDM5 and ASDM7 automatically change internal parameters based on properties of the source signal.

[Le Concombre Masqué]

GREAT news ! but...

 

beta won't play dff files

 

no offload to my NVIDIA GeForce GT 650M even if I use gfxCard to force the system to use it rather than the Intel (and yes I installed CUDIA),

Tried with regular El Capitan and slightly trimmed Yosemite on SD card (in the latter case there's been a change in System preferences with the choice to favor performance or battery life in place of the choice to tick or not automatic permutation of GPUs

From FB HQPlayer page:

 

HQPlayer Desktop 3.12.0beta1 now available for download.

HQPlayer now includes nVidia CUDA GPU offload support. To utilize this offload nVidia GPU at minimum Compute Capability 3.0, 1 GB of graphics memory and nVidia driver is required. Latest CUDA driver for Mac can be found here. If suitable GPU is not found, offload is not performed even if requested.

 

Signalyst - Installing HQPlayer beta

[Le Concombre Masqué]

CUDA is supposed to work with NVIDIA GeForce GT 650M ; any idea why offload does not work?

Yes, to some extent, but only on Windows... Maybe I'll do something small extra on 3.13 for Intel GPUs.

 

CUDA is the only one that works decently on all three platforms, although Macs are a bit tricky on that front because Apple has been using all three GPU brands on their products and obviously those cannot be changed or extended because they don't make machines with PCIe slots anymore. Apple's newest lineup seems to be Intel and AMD graphics (cheaper?).

[Le Concombre Masqué]

thank you ; I get this :

Regarding nVidia CUDA:

 

My experience with Windows is that the nVidia driver installer (not the generic graphics driver from Microsoft) also installs the CUDA driver, which is not the same thing as the graphics display driver. CUDA opens up the GPU to be used for rendering.

 

On a Mac, the CUDA software needs to be installed separately.

 

The best way to test your system to verify that CUDA drivers are present is to use CUDA-Z. It's freeware and works on Windows, Mac, and Linux.

[Le Concombre Masqué]

thank you. Trouble is : haven't changed anything since when I tried. Will wait till the next beta now

Looks good. Assuming the Mac version has the CUDA support, you should be good to go now.

[Le Concombre Masqué]

I would list the tonal accuracy as first benefit and not by a slight margin. The sonic tradeoff being a leaner sound vs the poly sinc family. Closed form thus led me to consider that half an hour of continuous playing is required when I Corning/Green Regen connect my MBPr (everything else being 24/7 on and I had no noticeable warming period when using a regular hub instead of the Regen). If I don't jump on comparing stuff I get great listening sessions outside of classical and MFSL mastered recordings which first appeared as clear benefiters of closed form filter.

 

I also find myself using a single modulator , ASDM7, and not fiddling when using cff

I will give you a reply in non-technical sense. I find that closed form presents a wider/deeper sound stage with greater separation of instruments. To be more specific, there is space between instruments that simply is not there with other filters. It also seems to have some effect on tonal accuracy although that is slight. I have experimented with upsampling 16/44 using closed form to DSD 128 and DSD 256. I honestly find little difference in the results although there is some. I cannot test DSD 512 as my dac supports DSD 512 but my computer is overwhelmed at that level of computation even with the CUDA offload. I have tested DSD 512 vs 256 with other filters and I find little difference.

 

Closed form fast brings nearly the same results but the difference is noticeable. I think that would depend on the resolution of your entire system as to if it would matter to you. Closed form fast requires far less computing power. Interestingly I do find that using closed form brings less difference between the choice of modulator provided you choose one that can handle high oversampling.

[Le Concombre Masqué]

see my sig (one machine NAA)

Is there a work around for this? I am still experiencing infrequent dropouts.

 

I am able to change the buffer size in JRiver, so using it more than HQPlayer. But, would prefer to use HQP as my primary player.

[Le Concombre Masqué]

omg !! kudos and thanks ! redbook closed form 128 dsded with less than 12% cpu load thanks to cuda offload sounds much better than at 27% load

 

however b3 is still beta : polysinc family stuttering with 96K source (dsded ; works when pcm up sampled) after playing 16/44 though I took care to stop and switch to polysinc before launching 96K program + unsmooth transition when coming back to redbook : need to play a bit through polysinc before being to play through closed form...

 

OSX Yosemite

[Le Concombre Masqué]

Thank you Santa Claus Miska !

[Le Concombre Masqué]

I'm sorry to report stuttering with 3.12 poly sinc family

 

Yosemite

[Le Concombre Masqué]

I have no cpu issue with Redbook closed form upsampling to dsd 128 as low as 12%

 

HOWEVER whatever poly sinc filter and modulator, 3.12, as its betas, stutters

 

 

 

Yosemite

 

There was a bug in initialization of ASDM7, now it has been fixed and as a result it will consume some more CPU time than before. Probably something like 20% more. This may be enough to exhaust CPU if it was close to the edge before.

 

Using -2s filter variant has less impact on the result than the bug had.

 

With 3.12, poly-sinc filter at DSD256 on my i5-6600T consumes about 30% per core (Pipeline SDM enabled) to exaSound e28 when playing stereo.

[Le Concombre Masqué]

here are my screenshots. Polysinc stutters... I once disabled compressed memory by sudo nvram boot-args="vm_compressor=2" ; could it be related?

I think it should work for you, but I don't know why it doesn't. There are no changes in poly-sinc filters, but Pipeline SDM has now less overhead than before. Can you setup OS X activity monitor similar way as in my screenshot with services of all users included on the process list plus CPU histogram display?

 

Do you also have the updated exaSound Yosemite driver (version 5.1.6)?

[Le Concombre Masqué]

Yes, I run a trimmed OS X. thing is... everything worked perfectly until 3.12 and its betas. As of networkaudiod running on the same machine, I reported positive results long ago and wondered why you never reacted... Solved my dropouts issue + sounds smoother. Might be a disturbing thought but I personally came to doubt of the point to run a NAA machine. As I suggested once, using my Time capsule and a wifi network to run NAA and desktop on the same machine might bring the same kind of benefits than the famous Regen (reclocking...) On the other hand it's true that CUDA offload is in line with the dual machines scheme and brings benefits... Besides the compressed memory thing, I can not see which service I disabled is relevant to get 3.12 working. Is everybody with stuttering issue running a trimmed OS X ??? My trimming process has been : [TABLE]

[TR]

[TD=bgcolor: #fafafa]Right way to disable the "Sleep" image forever:

View the sleep image size

Type in the Terminal: ls -lh /private/var/vm/sleepimage

Disable sleep mode:

Type in the Terminal: sudo pmset -a hibernatemode 0

Remove the image:

Type in the Terminal: sudo rm /private/var/vm/sleepimage

Create a 0-byte file:

Type in the Terminal: sudo touch /private/var/vm/sleepimage

Make the file not writable:

Type in the Terminal: sudo chflags uchg /private/var/vm/sleepimage

Run these commands one by one and reboot. After reboot, re-run the first command to see the sleep image size. It should look like this (0B):

google_ad_section_end

/ message sig

__________________

supp swap (even if ram>8 be cautious but works for me…)[/TD]

[/TR]

[/TABLE]

sudo nvram boot-args="vm_compressor=2"

debug dropout :

sudo nvram boot-args="dart=0x0"

limit writings

sudo mount -vuwo noatime /

mount | grep noatime

defaults write com.apple.NetworkBrowser DisableAirDrop -bool YES

sudo pmset sleep 0 2>/dev/null

sudo pmset disksleep 0 2>/dev/null

sudo pmset displaysleep 0 2>/dev/null

 

sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.softwareupdate_firstrun_tasks.plist 2>/dev/null

sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.softwareupdated.plist 2>/dev/null

sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.usbmuxd.plist 2>/dev/null

sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.usbd.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.bird.plist 2>/dev/null

sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.blued.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.CallHistoryPluginHelper.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.CallHistorySyncHelper.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.iTunesLibraryService.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.soagent.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.telephonyutilities.callservicesd.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.AddressBook.abd.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.AddressBook.AssistantService.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.AddressBook.SourceSync.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.SocialPushAgent.plist 2>/dev/null

launchctl unload -w /System/Library/LaunchAgents/com.apple.imagent.plist

launchctl unload -w /System/Library/LaunchAgents/com.apple.CalendarAgent.plist

sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.apsd.plist

sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.cmio.VDCAssistant.plist

first drag servers you call in Finder’s lateral bar

launchctl unload -w /System/Library/LaunchAgents/com.apple.sharingd.plist 2>/dev/null

(but when I use a NAA :

launchctl load -w /System/Library/LaunchAgents/com.apple.sharingd.plist 2>/dev/null)

defaults write com.apple.finder QuitMenuItem 1 2>/dev/null

defaults write -g NSDisableAutomaticTermination -bool YES 2>/dev/null

sudo defaults write /System/Library/LaunchAgents/com.apple.notificationcenterui KeepAlive -bool false 2>/dev/null

sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.metadata.mds.plist 2>/dev/null

sudo chmod 600 /System/Library/CoreServices/Search.bundle/Contents/MacOS/Search

killall SystemUIServer

killall Dock

CAD :

defaults write com.apple.finder AnimateWindowZoom -bool false

defaults write com.apple.finder DisableAllAnimations -bool true

defaults write com.apple.systemsound "com.apple.sound.uiaudio.enabled" -int 0

sudo defaults write /System/Library/LaunchAgents/com.apple.notificationcenterui KeepAlive -bool false

launchctl unload -w /System/Library/LaunchAgents/com.apple.usernoted.plist

defaults write NSGlobalDomain NSAutomaticWindowAnimationsEnabled -bool NO

defaults write -g NSScrollViewRubberbanding -bool false

sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.locationd.plist

l

Looks normal, although for some reason suspiciously low number of services on the process list with some services missing. Same as with Windows, modifying the OS always has a risk that it doesn't perform as expected, because Apple/Microsoft don't test their OS in such configuration. Also enable AppNap column and check that hqplayer/coreaudiod/networkaudiod are not subject to AppNap.

 

But why are you running networkaudiod on the same machine? Please use the output straight from HQPlayer instead of routing it through network back to the same machine.

[Le Concombre Masqué]

I have tried about anything I could think of including playing from regular El Capitan straight from ssd, no sd card no nothing and I still get stutter

 

btw occurs when I go to DSD 128 ; PCM whatever rate to DSD 64 works

 

There are no such changes in 3.12 that should cause this, quite the contrary because CPU load has been reduced (apart from slight increase in ASDM7). But some Apple's Xcode update could have made some invisible change, wouldn't be a first time.

 

 

 

I remember that you've had intermittent drop out issues, would be nice to get figured out why it happens...

 

NAA makes sense when running alone in a low noise machine dedicated only for that job, with minimal OS. Or in cases like when you need to get HQPlayer output to another room.

 

 

 

I noticed for example fseventsd, notifyd, coreservicesd, etc missing from your process list, which seems wrong to me. For example Apple's documentation says "coreservicesd provides various services for CoreServices frameworks. It is not meant to be invoked directly and it must not be terminated."

[Le Concombre Masqué]

OOPS poly sinc stutters when going 64 too just not so immediate... cpu load up to 50%

 

btw who has 3.12 working on OS X ? Any clue ?

[Le Concombre Masqué]

" In fact I prefer the VM on MBP as it is very stable with no stuttering".... not exactly a report that 3.12 is flawless on os X... straight from coreaudio poly sinc stutters too here...

Streaming through Roon. I am running 3.12 on Yosemite, MBP '14 (cuda enabled). I use poly-sinc-*-2s, OS on SSD, HQplayer on RAM Disk, NAA on Audiolinux. Stutters at DSD256 or above but In most cases It works if I keep HQplayer in full screen and in the foreground and I am not running any other app. I feel it may be a OS prioritization issue. If HQplayer is in the background, stuttering starts.

 

I also tried to use renice and elevated the priority of HQplayer to real-time, it did not help much, which again makes me think this could be a network issue. It could also be related to NAA (maybe network i/o priority?). I am going to experiment with that tonight without NAA. CPU usage is typically around 150% for Redbook -> DSD256, but it could stutter at DSD128 with 100% CPU usage if not in the foreground and another app is open.

 

It is defenitely not a CPU resource issue. On the same MBR, I have a VM running windows 7.1 (with all 4 cores made available to the VM + 12GB RAM made available). HQplayer can do Redbook -> poly-sinc-*-2s up to DSD512 with no stuttering! (CUDA cannot be made available to VMs so only pipeline SDM is available). CPU hits 340%. In fact I prefer the VM on MBP as it is very stable with no stuttering.

[Le Concombre Masqué]

yes thank you ; El Capitan is unmodified on my mid 12 MBPr 's SSD and it doesn't work with poly sinc while I believe I tried any combination

Works for me on both of my Macs to DSD128:

 

Mac Mini (Mid 2011), Yosemite, TEAC UD-501 (poly-sinc, ASDM7, PipelineSDM disabled since this is dual-core)

iMac 27" (Late 2013), El Capitan, Mytek Stereo192-DSD (poly-sinc, ASDM7, PipelineSDM enabled since this is quad-core)

 

Unmodified OS, also running anti-virus software on both.

 

My spouse has MacBook Air, I can try on that one too.

[Le Concombre Masqué]

I deeply admire your work so please indulge this daring question : are you sure you're testing the very same build (december 23 23.55) that's on line ? You suggested to check app nap ; there's no prevent app nap box and sometimes it shows yes sometimes no... does not seem to relate with the stuttering but I gonna disable it system wide though...

Now I tried also on this third machine, using iFi iDSD Nano and playing FLAC content over WiFi from a file server (smb:// mount). No problems... So I don't really know what is going wrong for you...

 

[ATTACH=CONFIG]23048[/ATTACH]