HQ Player

[k6davis]

Any news on Roon + HQPlayer integration? The early rumours were for something to happen about now

 

Thanks

 

Following on from my Roon+HQPlayer integration question I dropped Roon onto my server today. Fun set of capabilities, but a shame about the sound quality!

 

Back to HQPlayer

+1

 

I couldn't agree more with both posts. I'm hoping for some progress with Roon+HQPlayer integration (as are a lot of people on the Roon forum) but AFAIK, there haven't been any status updates.

[Iain]

HQPlayer with Muso works pretty well. Have you tried to set it up under Tools/Options/Players?

Be sure to use latest build 01 of 2.3.22

Actually it probably will with one major exception; I can't get them to connect.

 

Using v2.3.22; database is configured and Muso now sees HDPlayer, but that's as far as it goes.

 

BTW, has anyone seen this:

http://www.computeraudiophile.com/f11-software/manage-hqplayer-playlist-foobar2000-24282/

 

Looks like some interesting possibilities, that.

[Miska]

Mac OS X Yosemite, HQP 3.9.0.1, PCM192 FLAC file.

 

I just noticed poly-sinc-ext filter doesn't work. I mean playout doesn't even start. Not sure if it's a bug or not.

 

For me, it works OK. Can you tell the exact settings that don't work? Note that for higher output sampling rates the initialization phase may take a while depending on computer.

[bibo01]

...

HQPlayer supports .m3u, .m3u8 and .pls playlist formats, .m3u8 being recommended format. HQPlayer doesn't support XML playlists (.xspf) at the moment. If your library scan folders include playlist files in recognized format, they are also added to the library, without metadata so they appear as unnamed folder in the main window view. You can also open playlist files using the corresponding toolbar button, or simply by just dropping the playlist file on HQPlayer window.

Is HQPlayer going to support .xspf playlist anytime soon? I was thinking in terms of integration with Tomahawk

[Miska]

Is HQPlayer going to support .xspf playlist anytime soon? I was thinking in terms of integration with Tomahawk

 

It shouldn't be too hard to do if there's some real need for it. I know also VLC prefers xspf playlists.

[bibo01]

Miska,

 

I discovered probably a bug.

When single repeat is inserted, clicking on Next Track does not work but Previous Track does.

It's one way or the other, I suppose.

[Miska]

I discovered probably a bug.

When single repeat is inserted, clicking on Next Track does not work but Previous Track does.

It's one way or the other, I suppose.

 

That is intentional. Clicking Next goes back to beginning of the current track, same thing happens when end of track is reached (next track is always current track in repeat single mode). Previous is not considered in this case because there's no backwards play. You can switch to any other track by clicking it in the track list.

[Flac Attack]

I'm interested in using HQPlayer for its ability to upsample 16-bit/44.1kHz files to 24-bit/192kHz (and in the future to DSD, once I get a DSD DAC). Before I install the trial version I want to make sure my hardware is up to the task, otherwise if it doesn't work as well as I hoped I won't know if the problem is the software itself or my inadequate hardware.

 

The desktop PC on which I plan to run the main HQPlayer module has two old Xeon quad-core processors at 2.4GHz (model E5530 to be exact), and 16GB of RAM. It is currently running Linux Mint, but I plan to switch to Xubuntu or Ubunto GNOME soon.

 

Then I have a 9-year-old PC to be used as an NAA, which has a dual-core AMD Athlon capable of outputting 24/192 via its HDMI video card (but not 24/88.1 or 24/176.4). That feeds an Onkyo 876 receiver which has 24/192 Burr-Brown DACs, which in turn feeds a pair of JBL LSR308.

 

Is that hardware good enough to manage upsampling from 16/44.1 to 24/192 with HQPlayer, and eventually to DSD128? Although the main PC has 8 cores which is plenty of power in aggregate, the individual cores aren't very powerful given the age of the CPUs, and on skimming through this thread I've read that many cores may not help very much because some important tasks of HQPlayer aren't parallelized.

 

How much network bandwidth is needed for a main+NAA configuration? I'm getting 40-50 Mbits/sec with powerline networking, but I can expend the effort to run gigabit ethernet wires if that's what it takes.

 

After reading through large pieces of this thread over several days, I was able to answer my own question. It looks like my old 8-core 2.4Ghz system will be good enough for upsampling redbook to 24/192 and DSD128, provided that I don't use the most CPU-intensive settings of HQPlayer.

 

Now I just have to decide on a DSD-capable DAC to make the most of HQPlayer (after getting things to work with 24/192). The TEAC UD-503 looks sweeeeeeet! I hope it's readily available in the US later this year when I'm ready to buy one.

[Chodi]

After reading through large pieces of this thread over several days, I was able to answer my own question. It looks like my old 8-core 2.4Ghz system will be good enough for upsampling redbook to 24/192 and DSD128, provided that I don't use the most CPU-intensive settings of HQPlayer.

 

Now I just have to decide on a DSD-capable DAC to make the most of HQPlayer (after getting things to work with 24/192). The TEAC UD-503 looks sweeeeeeet! I hope it's readily available in the US later this year when I'm ready to buy one.

 

Given your limitation with your computing power you may want to consider a DXD capable dac and forget DSD. Upsampling with HQPlayer to DXD standard 384kHz uses much less computing power and the results are arguably nearly as good. My computer uses only 6% cpu with DXD and it uses 16% with DSD128, 24% with DSD256. I have a reasonably high powered system with an Intel I7 cpu. You won't really lose much staying with DXD.

[Le Concombre Masqué]

I prefer not to dsd pcm nowadays. However, on my own system, invitees always prefer dsded files : I guess I want a change from an always pleasant palatable sound, taking the risk to be disturbed by a not so pleasant sound once in a while for the sake of hearing room acoustics (i.e. fast reverbs, undamped walls in rooms used to record chamber music) and so on. btw I never regretted buying my TEAC 501

Given your limitation with your computing power you may want to consider a DXD capable dac and forget DSD. Upsampling with HQPlayer to DXD standard 384kHz uses much less computing power and the results are arguably nearly as good. My computer uses only 6% cpu with DXD and it uses 16% with DSD128, 24% with DSD256. I have a reasonably high powered system with an Intel I7 cpu. You won't really lose much staying with DXD.

[Flac Attack]

Given your limitation with your computing power you may want to consider a DXD capable dac and forget DSD. Upsampling with HQPlayer to DXD standard 384kHz uses much less computing power and the results are arguably nearly as good. My computer uses only 6% cpu with DXD and it uses 16% with DSD128, 24% with DSD256. I have a reasonably high powered system with an Intel I7 cpu. You won't really lose much staying with DXD.

 

If I buy a DAC in the $1000 price range it better have DXD capability! And shortly after getting a nice DAC, I figure I'll build a powerful dedicated system to put right beside the audio components. I've seen some barebones machines with modern Intel I7 or AMD 6000-series CPUs selling for under $400 ... I could get one of those and throw in some other spare hardware (optical drive, SSD etc.) to complete the build, and that should be able to handle anything audiowise that I throw at it.

 

But I'll take it one step at a time. First let me get HQPlayer working with my existing hardware for upsampling to 24/192.

[Le Concombre Masqué]

MISKA BEWARE GLITCHES REPORT

 

1st but not least : coming from a 192 track outed at PCM 384 I played a longtime favorite by Leonard C, 44 outed 352, and his voice was higher pitched and the song was too fast. Since I was in the mood for that slow low voice song I played it SDMed. Excellent. Then I tried again PCM (poly sinc no dither 352 ) and everything was back to normal

 

2nd I experience dropouts that I had long forget about since I use NAA

 

(btw : neither Miska nor anybody has ever commented my practice of using the same MBP for desktop and NAA (Time Capsule wifi network). Obviously it's not about less electric noise on the DAC connected machine but it definitely delivers a more suave, less bland sound (maybe I should share the trick on the Regen thread where other phenomenons such as reclocking are debated) and, here I come : setting the buffer to 100 ms made my dropout issues disappear.

No more)

 

Don't know though if those dropouts relate to 3.9.1 or to the fact that I don't SDM anymore PCM files but out them to 352 or 384 (should not, the cpu usage drops to less than 3%)

[John ALSR]

The issue of how much CPU power is needed to run HQPlayer arises frequently. After reading through the entire thread at the beginning of August I downloaded the trial version of HQP; and, on September 2nd purchased it. My intend is/was to use HQP to upsample PCM files, stored in WAV file format, to x2 SDM to my new Mytek DAC.

 

In a reply to my previous post here, Miska provided the proper settings needed for x2 SDM upsampling over Firewire to the Mytek: WASPI; 2-wire checked; SDMpack-Dop; and CPU "Pipeline." Additional settings were volume min/max at -3dB; and, most critical "Buffer" at 250mSec. I will return to the buffer setting issue below.

 

The computer was based upon a "Maximum PC" magazine HDTV design of several years back, in which the then new parts sat unused until about 6 months ago. The processor is an AMD Athlon x4 615e at 2.5GHz speed. It has 16 Gb of RAM, with a small SSD for the Windows 8.1 operating system and programs. All music files are stored in WAV file format on a Samsung SSD 850 Evo 1T size. Note that using the Samsung Magician program that SSD is set for "Maximum Reliability" and the "Rapid" mode is set to convert some of the RAM to a RAM disk buffer for the SSD.

 

Can we all agree that the x4 615e CPU @ 2.5GHz speed is 'feeble'? Its claim to fame was/is its 45W power consumption; and, the internals of the computer were intended to be fanless (which is not its current situation).

 

Not being entirely sure as to proper settings for the filter and oversampling protocol (16/44 WAV to 5.6MHz or 6.1MHz SDM?) I started with:

 

Poly-sinc-shrt-mp-2s (for transient response?) / DSM7

 

With the minimum buffer setting this resulted in non-immediate dropouts, i.e. dropouts occurring later in the music. The Mytek's LED meters showed that the dropouts were associated with 'maximum red clipping.' Eventually I increased the buffer length to 250mSec after the 100mSec buffer setting evinced a single dropout after several hours of superb playback. Since increasing the buffer length to 250mSec there have been NO dropouts!

 

With my limited understanding I began to 'work' my way 'up' the ladder of oversampling filters:

 

Poly-sinc-mp-2s

Poly-sinc-short-mp

 

and finally, the BIG DADDY

 

Poly-sinc-mp

 

Along the way I changed the modulator from DSM7 to ADSM7, for no reason other than I'd read here that others thought it better than the DSM7.

 

So here's the results from the Poly-sinc-mp filter and ADSM7 modulator upsampling 16/44 PCM WAV files to x2 SDM (5.6mHz):

 

The Poly-sinc-mp filter took 20 seconds to initialize, as timed and seen on the Windows "Resource Monitor CPU" readouts (for all four cores). All four cores went to 100% maximum used during initialization. Twenty Seconds!!!

Afterward, perfect playback at:

 

CPU 0 = ~45%

CPU 1 = ~50%

CPU 2 = ~45%

CPU 3 = ~55-60%

 

The computer is used solely for music playback; and, the only other programs running when these measurements were made was the Resource Monitor and HWmonitor (CPU temperature). Once the 20 second initialization is effected, all further playback (of 16/44 PCM) occurs quickly and without tedious delay.

 

As I understand, the above settings maximize quality to the Mytek, which is limited to x2 SDM using Miska's special 2-wire Firewire method. I have noted that upsampling 16/44 PCM to SDM 6.1mHz seems to use slightly more CPU percentage of use.

 

My question for others who use the Mytek (Miska?) is what is its x2 SDM 'sweet spot' and is there any way to have 44kHz and 48kHz files distinguished for automatic playback. I'm not sure I understand all the issues involved.

 

Second question is that I'm currently using the Mytek firewire cable supplied (as part of the Mytek store's sales package). Would there be any worth in upgrading that cable; and, if so, to what cable is recommended?

 

I think it likely that my PCM file as WAV format and the EVO 850 SSD with RAM disk "Rapid" mode might influence the extremely positive results in sound quality I am now getting. Yet on the other hand, the 'feeble' AMD x4 615e CPU seems quite adequate (accepting the 20 seconds initialization time) for even the most demanding SDM filter settings. Then again, I haven't tested every filter.

 

Why not stop when the music sounds better than it ever has here at home?

 

John

[ytsejam]

My question for others who use the Mytek (Miska?) is what is its x2 SDM 'sweet spot' and is there any way to have 44kHz and 48kHz files distinguished for automatic playback. I'm not sure I understand all the issues involved.

Second question is that I'm currently using the Mytek firewire cable supplied (as part of the Mytek store's sales package). Would there be any worth in upgrading that cable; and, if so, to what cable is recommended?

John

Hello John,

currently using Mytek / HQ Player in a NAA Setup (HQP Desktop with Ubuntu on a i5 4690 PC, NAA with Windows 8.1 and Mytek Asio Driver on a CAPS Design with Firewire, Upsampling PCM / 1xDSD to 2xDSD). The Firewire Cable between the NAA and the Mytek is a Neo Oyaide 1,0. There is a small but noticeable improvement in comparison to a plain FW Cable. Before building the CAPS i used a 5m FW cable (also Oyaide) connecting the desktop pc directly and the differences to a plain cable were more obvious. The Oyaide is a good choice imho, but it seems that FW lose further ground in the market, the new Mytek Brooklyn DAC e.g. has no firewire interface anymore. Your question about the sweet spot is very interesting ( using 6144000 at the moment).

Regards, Thomas

[57gold]

How come I can not adjust the buffer in Preferences? It is greyed out.

 

Still have dropouts...Miska's only suggestion is that I have latency getting data to Mac Mini so replacing a 2X 1.5TB OWC Raid1 with USB2 out with a 4TB LaCie Thunderbolt fanless HDD. May solve dropouts and fan noise, which are both incredibly annoying.

[Le Concombre Masqué]

READ #3562 the NAA/desktop on a same machine trick solved my dropouts issues, probably by the introduction of latency though in that case the buffer setting isn't greyed in Preferences (though I doubt it, maybe hqp controls the latency between its 2 parts ; but it certainly not grants coreaudio buffer controls).

 

there is an impact on SQ. I usually tend to prefer the NAA suaveness but a glitch yesterday forced me to forget about NAA for the track I wanted to play and enjoyed playing without NAA. Later I restarted my mbp, got back NAA, compared and... preferred the piano as I just heard it without NAA

How come I can not adjust the buffer in Preferences? It is greyed out.

 

Still have dropouts...Miska's only suggestion is that I have latency getting data to Mac Mini so replacing a 2X 1.5TB OWC Raid1 with USB2 out with a 4TB LaCie Thunderbolt fanless HDD. May solve dropouts and fan noise, which are both incredibly annoying.

[Miska]

How come I can not adjust the buffer in Preferences? It is greyed out.

 

CoreAudio wants to manage buffers on it's own, so it is not available on OS X...

[vmanuelgm]

Hi again, mates.

 

I am trying the Eastern Supreme, thanks to Heinz Posingis from Germany, Audiotubes.de owner, who kindly sent it to me with Burson Singles and Duals.

 

Have it now with Sparko singles and Burson Duals, waiting for Sparko Duals which seem to come in a donkey. I am angry with Andrew Sparko, cos he publishes he uses UPS Worlwide saver for duals, and shipped by USPS Priority Mail, so 10 days more waiting more or less... I wrote to him and didn't anwer yet about it.

 

The Supreme is over Junior in terms of depth, a real bargain for the price. The Junior is also a bargain, very good DAC's, which can be tuned with opamps to achieve the sound you are looking for to match your system!!!

 

Besides, HQPlayer improves the pack, a real kickass team...

 

 

[John ALSR]

Reference #3563

 

Attempting to use SDM Poly-sinc/ASDM7 (PCM source) upsampled to x2 SDM at 6.1MHz with 16/44 PCM WAV file results in +40 seconds initialization time and completely choppy playback.

 

Apparently 16/44 PCM to x2 SDM only works for 5.6MHz upsample using the 'feeble AMD 2.5GHz CPU.' Playing back now successfully at 5.6MHz x2 SDM upsample to the Mytek.

 

John

[Chodi]

Reference #3563

 

Attempting to use SDM Poly-sinc/ASDM7 (PCM source) upsampled to x2 SDM at 6.1MHz with 16/44 PCM WAV file results in +40 seconds initialization time and completely choppy playback.

 

Apparently 16/44 PCM to x2 SDM only works for 5.6MHz upsample using the 'feeble AMD 2.5GHz CPU.' Playing back now successfully at 5.6MHz x2 SDM upsample to the Mytek.

 

John

 

The ability to use 6.1 depends on the clocks chosen in the design of your dac. Many Sabre based dac's can't do 6.1 for that reason. Nothing to do with your computer.

[Miska]

The ability to use 6.1 depends on the clocks chosen in the design of your dac. Many Sabre based dac's can't do 6.1 for that reason. Nothing to do with your computer.

 

At least Mytek, exaSound, iFi and Fostex DACs can do 48-base rates also (3.1/6.1/12.2). Upsampling 44.1k to 6.1 takes more computing resources, mainly in terms of CPU cache, than going to 5.6. Practically any DAC that can take 48-base PCM in, should also work just fine at 48-base DSD rates, but for some reason all manufacturers don't support this. (on some DACs it is limitation of the Thesycon driver on Windows, rather than the DAC itself)

 

Regarding CPU load, my Mac Mini with Core i5-2415M CPU can upsample to DSD256 using -2s filters and to DSD128 using any of the filters. DSD512 is out of question with any of the filters on that machine.

[Le Concombre Masqué]

I'm back to 3.8 after the reported glitches ; Miska, any comment ?

[Flac Attack]

Reference #3563

 

Attempting to use SDM Poly-sinc/ASDM7 (PCM source) upsampled to x2 SDM at 6.1MHz with 16/44 PCM WAV file results in +40 seconds initialization time and completely choppy playback.

 

Apparently 16/44 PCM to x2 SDM only works for 5.6MHz upsample using the 'feeble AMD 2.5GHz CPU.' Playing back now successfully at 5.6MHz x2 SDM upsample to the Mytek.

 

John

 

Try upsampling to 6.1MHz from 48kHz content, that may work better as it's the same integral ratio as going 44kHz to 5.6Mhz.

[John ALSR]

Reference Posts: 3563, 3564, 3569, 3570, 3571 (Miska), 3573

 

Thank you for all replies to my original post (3563). Coming from a university research background I can testify that communication is rather difficult. Hence the elaborate "operational definitions" required in reporting empirical test results for science purposes.

 

My hidden 'hypothesis' was "Can I use HQPlayer driven by the Feeble Computer to achieve maximum capable performance output with the Mytek DAC (original version) for DSD upsampling of PCM sources?" Somewhere along the way my 'data' wasn't sufficient.

 

Understanding that there is a computational hierarchy of the DSD upsampling filters, and that of the recommended filters, both the "poly-sinc" and "poly-sinc-mp" are the most computationally demanding, I sought to test the thesis.

 

In all cases the Mytek display showed "HdSd."

 

Upsampling 44-base PCM using "poly-sinc" at the 5.6MHz rate worked.

Upsampling 44-base PCM using "poly-sinc" at the 6.1MHz rate DID NOT WORK.

 

Today's testing showed that:

 

Upsampling 48-base PCM using "poly-sinc" at the 6.1MHz rate worked.

 

As noted in my original post, a slight increase in CPU cores load was noted:

 

CPU 0 = ~45%

CPU 1 = ~55-60%

CPU 2 = ~50%

CPU 3 = ~55-60%

 

What seems very clear is that base-44 and base-48 PCM cannot be mixed for upsampling to their respective x2 SDM rates using the most computationally intense filter types.

 

To the question of the Mytek "Sweet Spot":

 

Mixing 16/44, 24/48, 24/88, 24/96 PCM files for playback (too lazy to add 176/192KHz files) and setting the filter type to "poly-sinc-shrt-mp-2s" and setting the upsample rate to the Mytek maximum of 6.1MHz allowed for all files to be played in a mixed-format playlist situation!

 

Apples and oranges: It is hard to keep things straight while empirical testing, or in most other situations too. I am now convinced that when I want to play back PCM files at the maximum quality (assuming I could hear that) that I would use one of the "poly-sinc" filters and only be able to play back one PCM-base format file type dependent upon the upsampling rate chosen (i.e. 44Khz PCM to 5.6MHz; 48KHz to 6.1MHz.).

 

If I want to 'party' then I need to select a "-2s" filter variant with reduced computation load, and then I can mix all PCM file types. This is pretty wonderful don't you think ...

 

... using The Feeble Computer (to get the best the Mytek can do).

 

Regards,

John

 

p.s. So what is the Mytek "sweet spot" using the "-2s" filter variants, i.e. 5.6MHz or 6.1MHz?

[brightonjel]

Just downloaded 3.10 (Mac OSK 10.11) and tried out the new Closed Form Interpolator with both DSD 7 and ASDM 7, up sampling Red Book CD to 2xDSD. There's an underlying low-level popping noise that's continuously there every second or so. Anyone else hearing that? (Feeding Geek Pulse via USB)