HQ Player

[bibo01]

...

The library management and user-friendly access and playback is the bit that is far from ideal, but why would you even want to compete with iTunes and their iPad Remote app. Combine iTunes, or a DLNA compatible version to run on a NAS with MinimServer, with your algorithms and everyone else can pack their bags.

 

Cheers

For the highlighted mode you need HQPlayer Embedded (on Linux only) and have a chain like MinimServer->Rygel->HQP Embedded. Your control point would be BubbleUPnP.

I use it successfully with AudioLinux as OS.

[bibo01]

Thank you. I saw a few older posts whereby the conclusion was that a NAS itself would not be powerful enough. I wonder if that is still the case if staying within PCM only.

 

So you have a NAS running Minim and another box running linux HQP? Are you then able to send the HQP stream output to an UPnP Renderer?

Yes, I have a Synology DS110j which is a low power NAS and I have no problem sending stream to main PC with Rygel + HQP Embedded on it and output either locally (USB) or to a NAA on the network.

I tried with both MinimServer installed on the NAS as well on the PC, playing PCM and DSD (256 too).

[bibo01]

So, what is the best DAC bits setting for an exaSound USB input...32 bits? I have it set to "default" right now.

 

32bits...probably "default" is the same

[bibo01]

Is native DSD supported with the Amanero USB board under Linux, with the Signalyst kernel and alsa patches applied?

 

I built an NAA based on Ubuntu 14.04 and applied the patches, but it only seems to work using DoP. I'm wondering if I messed something up with the build.

 

Thanks.

 

No, due to its bulk transfer, for DSD Linux ALSA supports Amanero board through DoP only, therefore it is limited to DSD128 playback.

[bibo01]

So with DSC1, where Miska demonstrates the measurement results for DSD512: Preliminary measurement results for the DSC1 DAC - Blogs - Computer Audiophile , are you saying that DSD512 is not achievable with Linux (or Linux/NAA)?

Using Amanero (so also DSC1) one cannot achieve DSD512 under Linux.

In order to achieve DSD512 playback to DSC1, one needs a Windows based main PC or a Windows based NAA.

[bibo01]

I just finished building a DSC1, but I only have a Linux NAA, so it will be DSD128 max for now I guess.

 

So what's the smallest/cheapest/low power option for a Windows NAA? Anyone tried anything smaller than an Intel NUC?

There are a few cheap thin clients around usable as Win NAA, by HP for example.

However, I can recommend Fitlet-B for smallest/low power Win NAA. I have been an early adopter of this device and it is very good. You may read my report about it in this thread (just Google translate it). You can run Win 10 on it too.

[bibo01]

OSx HQPlayer (Yosemite) + NS Supremo. Doing PCM > DSD128, if pipeline sdm is selected, Supremo displays "352"; if pipeline sdm is NOT selected, Supremo displays 5.6 MHz correctly. Why? Is anyone having the same problem?

[bibo01]

OSx HQPlayer (Yosemite) + NS Supremo. Doing PCM > DSD128, if pipeline sdm is selected, Supremo displays "352"; if pipeline sdm is NOT selected, Supremo displays 5.6 MHz correctly. Why? Is anyone having the same problem?

Could someone please test on their Mac system.

[bibo01]

Which priority level do you use? Risk with using realtime priority is that if you reach 100% CPU load on processing, then you likely won't be able to stop playback anymore from the GUI because it becomes completely unresponsive.

With AudioLinux, which has RT kernel and RTirq/app, for a high priority level I set: rtc=95, usb DAC=93, ethernet port=91, HQPlayer=89. All the modules (that I see) loaded by HQPlayer are running at that priority. All is working extremely well.

I doubt that Windows has this sort of capability.

[bibo01]

On Linux too, HQPlayer will adjust it's internal priorities... There's no such thing as running HQPlayer at any single priority...

 

Windows does have similar capabilities. Only OS X doesn't.

Indeed, I believe that, given a certain system priority, HQPlayer manages the priority of each module internally.

[bibo01]

EuroDriver,

 

Any experience with HQPlayer using a Linux OS? If so, what are you opinions about sound quality vs. Windows?

 

I'll be happy to get anyone else's opinion on this matter as well.

 

Thanks,

 

Joel

I can second Miska's view on this.

On my server I have both Windows (8.1 and WS) and Linux (AudioLinux). In Linux HQPlayer runs very smooth and fast, and I can also use HQPlayer Embedded to control on smartphone/tablet with BubleUPnP. The only downside is that I cannot playback DSD256 which my system is capable of in Windows. Only very few DACs can work at DSD256 in Linux (for the moment).

Furthermore, in my system I have a problem with remoting the desktop properly.

[bibo01]

Hammer,

confirm that both PCs reside in the same domain; be sure that no firewall is running on the NAA PC; start PC desktop, then PC NAA and DAC, then HQPlayer.

[bibo01]

Bogi, I will sure tell how I reached this, it's very easy, but JRiver seems to make some problems with this solution that I realized a little late. Every time I send a complete album to HQP JRiver creates a duplicate of the original file with the extension _EDIT, if you do it 4 times you have 4 copies named _EDIT, _EDIT(1).._EDIT(3) for every song in your folder, until now I didn't find out why JR does this and how I can switch it off, so the solution is also not final

This solution is also not restricted to complete album, but kinda more restricted than yours, you can only send complete playlist, complete album or single tracks. When sending single tracks there is also no _EDIT copy done. It seems to have to do something with "send to external program" that these copies are made...

I am now looking a little bit deeper at customizing foobar, as I saw you can do really a lot with it...

I will forget about these theater views for audio because they are not enough customizable for tags as far as I can see... It's more important for video and for that I'll use Kodi anyway but don't want to use it for audio, I only want the best solution possible for audio in terms of software and handling media library...

How did you modify the BAT file for JRiver to be able to receive a complete album?

[bibo01]

I did not modify the .BAT at all, I used it as it is. The point is to change the following settings in JRMC and Windows :

First I did a File association for .flac files to be opened with the .bat in Windows. In JMRC then I did the following in Tools -> Options -> File Types : For .flac set Playback method to External program (default), and for Audio Playlist and Media Playlist I set Playback method to External program (custom) -> setting the .bat in the custom settings.

Now for the albums I did the following : If you right-click an album, then Send To -> Send To (external) -> Add/edit programs -> now you create an entry for HQP pointing to the .bat. Then just right click the album and choose send to "HQP entry". One instance of HQP has to be running to get this to work. As I told you the sending of the files works perfectly but there is still the problem that JRMC creates these _EDIT files for some reason when doing that. So if this can't be stopped somehow this workflow is just not practical. But maybe someone knows the reason and has a solution for that...

Thanks. For album playback only the second part of your reply is needed.

As a matter of fact I had already done it, but my system behavior is very different from yours, that is why I asked. Perhaps this is due to my music library being on a NAS or simply a different setup (JRiver 20).

 

Let me tell you what happens to me:

 

1) If I select a single track in JRiver/SendTo/.BAT with HQPlayer already opened, one terminal window opens briefly and then disappears, while the track is loaded in HQPlayer's playlist windows. It's all OK.

 

2) If I select a full album in JRiver/SendTo/.BAT with HQPlayer already opened, many terminal windows are opened (I suppose as many as the number of tracks) but most of them present an error: "Server closed the connection, server disconnected." Therefore, only a bunch of tracks (2 or 3) are loaded in HQPlayer's playlist windows.

Perhaps this is due to network delays. I feel that this behavior could change if in the bat file a "waiting time" was added for loading multiple tracks.

If someone with some batch scripting language knowledge could pitch in with a suggestion, it would be great.

 

Under no circumstance I get any _EDIT files you described in the album directory.

[bibo01]

Hazard,

 

look into your DFF/DSF settings menu, you probably have DirectSDM checked. It means that DSD files are played with no DSP applied. That is why you cannot change the output rate and for a DSD64 the drop down menu has only one option - 2.8MHz.

 

Under Linux check that in your Settings menu you have DoP selected in SDM Pack.

[bibo01]

Or exaSound e28 which I'm using.

 

Neither one is DSD-only DAC though. I could make a multichannel variant of DSC1, but problem is where to get a USB input module. I'm not too eager to make one from scratch. RigiSystems USBPAL is the only multichannel-DSD OEM module available but they are not too willing to sell things in small quantities, unfortunately. Otherwise that one would be easy to support.

 

What about exaU21?

I don't know if it is still in production.

 

PS: Sorry, it is PCM only.

[bibo01]

Unfortunatley still have problems, when I uncheck the DirectSDM box, I can select a new output rate for DSD files from the drop down box. But - no music comes out. The player appears to be playing just fine. But nothing comes out. This is same problem if I play PCM files (upsampled to DSD128) or DSD files (I have tried a number of sample rates and makes no difference). As soon as I re-check the DirectSDM box then music plays fine. PCM files are still upsampeld to DSD128 and of course the DSD files now cannot be upsampled.

 

I use a DIYINHK USBtoI2S board. It requires DoP signal for DSD replay, so I have selected 'SDM pack' as DoP. And this works fine. As long as DirectSDM box is checked.

Can you please post a picture of your Settings window.

 

BTY, DIYINHK USBtoI2S board does not do DSD normally. In order to work for DSD it has to be setup in a special way. Have you done it? Are you sure that it's actually playing DSD?

[bibo01]

I have trialled HQPlayer with both Win10 and Linux and results are very pleasing when both PCM and DSD files are upsampled to DSD128 and sent to my USB board via DoP. I want to see if there is any improvement with DSD256 and/or DSD512. On the Signalyst web site there is a link for additional files required for DSD native and this leads to a couple of folders (jessie, old, trusty, vivid) with heaps of files. Which files do I actually need to install? I don't understand Linux, but I have also asked a computer geek friend who runs Linux on his home network, he is similarly perplexed.

Firstly, you have to be sure that your DAC supports non-DoP DSD. Only a handful of DACs do.

If it does, you need to install the following packages (for 64bit kernel):

• linux-image (linux-image-3.18.1naa_1_amd64.deb),
  
• linux-headers (linux-headers-3.18.1naa_1_amd64.deb),
  
• libasound2 (libasound2_1.0.28-1jl1_amd64.deb)
  
• libasound2-data (libasound2-data_1.0.28-1jl1_all.deb)
  

You find them here: Index of /src/jessie

To install them you give two comands: the first one to download it, the second one to install it.

For example:

 

1) sudo wget http://www.sonarnerd.net/src/jessie/libasound2-data_1.0.28-1jl1_all.deb

 

2) sudo dpkg-i libasound2-data_1.0.28-1jl1_all.deb

[bibo01]

OK thanks but I still have no idea which files to download.

 

 

 

Thank you, now I do know what files to download!!

 

For info, I am using Ubuntu Studio 14. I have ordered a JL sounds USBtoI2S board.From their web site:

 

DSD Audio over PCM frames. DSD64 (2.8224MHz and 3.0720MHz) and DSD128 (5.6448MHz and 6.1440MHz) by method DoP with 0x05/0xFA

markers

- DSD Native DSD64 (2.8224MHz and 3.0720MHz), DSD128 (5.6448MHz

and 6.1440MHz) and DSD256 (11.2896MHz and 12.2880MHz) on Linux

and MAC OS

 

I also have a DIYINHK USBtoI2S board, I am using it with DSD as DoP right now, but it has same XMOS chips as JLSounds so I will see if it plays native DSD. One further question - is it possible to play native DSD from Windows? It appears from above that it can only do DoP.

I would ask JLSounds directly, re Windows.

However, Linux native DSD support is still in the works as documented here:

https://github.com/lintweaker/xmos-native-dsd/issues/1

[bibo01]

Eurico, as I explained in this post, the Linux non-DoP driver for JLSounds is not ready yet. So I would not be in a hurry to upgrade CuBox-i OS.

Anyway, here I have illustrated the procedure for OS upgrading.

[bibo01]

The 64-bit custom kernel build I have for Ubuntu now has support for the JLSounds device. I'll also try to update the CuBox-i image and make a BeagleBone Black image once I get 3.8.0 release of HQPlayer out.

JLSounds is not able to write a Linux driver for its own product - you come out and do it.

Pretty amazing!

[bibo01]

The 64-bit custom kernel build I have for Ubuntu now has support for the JLSounds device. I'll also try to update the CuBox-i image and make a BeagleBone Black image once I get 3.8.0 release of HQPlayer out.

 

Did you put this 64-bit custom kernel online?

[bibo01]

JLSounds is not able to write a Linux driver for its own product - you come out and do it.

Pretty amazing!

I just read that a patch for JLSounds has been issued:

https://github.com/lintweaker/xmos-native-dsd

[bibo01]

It is just one line to tell kernel that certain USB vendor id and device id can do native DSD in this certain way that many vendors have now adopted... Nothing big.

 

 

P.S. I didn't do the patch, just picked it up and applied on Ubuntu kernel.

 

Yes, I realize it, that's why my post #2707

[bibo01]

OK my JLSounds USB board arrived today. I have installed files for natiive (non-DoP) DSD playback, as advised by bibo01. But I am confused by the additional info provided by Miska. What files am I supposed to install now? Sorry, I am a linux idiot (and not much better with Windows).

The procedure remains the same. The only difference is that for "linux-image" and "linux-headers" you use the latest versions date 12 June 2015:

http://www.sonarnerd.net/src/trusty/linux-image-3.13.0-54-lowlatency_3.13.0-54.91jl20_amd64.deb

http://www.sonarnerd.net/src/trusty/linux-headers-3.13.0-54-lowlatency_3.13.0-54.91jl20_amd64.deb