HQ Player

[Superdad]

Makes sense, since A+ depends more sensitively on computer performance, while the NAA negates that for HQPlayer.

 

Actually Jud, I am saying this is the case with HQ Player as the player on my SD-booted, headless i7 mini. Not with in the NAA configuration. (I have more to report on NAA tweaks, but that is a separate post.)

[Jud]

Actually Jud, I am saying this is the case with HQ Player as the player on my SD-booted, headless i7 mini. Not with in the NAA configuration. (I have more to report on NAA tweaks, but that is a separate post.)

 

Flip side of the same conclusion: Again, Miska does not depend on minimizing use of computer resources (other than enabling use with streamlined Linux distros) but on his filter and modulator designs; and on isolation from the computer if using an NAA.

 

 

Sent from my iPhone using Tapatalk

[Superdad]

Flip side of the same conclusion: Again, Miska does not depend on minimizing use of computer resources (other than enabling use with streamlined Linux distros) but on his filter and modulator designs; and on isolation from the computer if using an NAA.

 

Well the one test I have not bothered to do yet is to reboot my music mini from a dead-stock Mavericks partition (such is one of several types on the internal HD I never use) and listen to HQ Player that way to see just how immune it is.

I don't think it will be entirely, since with my CAD script slimmed Mavericks SD and no active drives I am getting SQ equal to the same i7 used as an NAA with Miska-provided minimal-Linux/NAA boot.

This deserves a longer report--I should not leak this out just yet. As long as I am teasing though, I will say that with NAA the Ethernet cable matters and I think I would get best NAA results if I could get HQP to recognize the NAA with a direct, Mac-to-Mac connection (it improved when I plugged the desktop machine into the same DSL router as the NAA instead of my Cisco switch; as previously reported, the NAA has to go into my DSL router as it is not patient enough to wait for a DHCP IP lease through the switch at boot-up).

The simplest solution would be if Miska would make it so that I can enter an IP address into HQP Desktop to tell it where to find my NAA. The screen of the NAA machine reports its IP address at boot time, and it even self-assigns one to itself if not DHCP server is found. Right now this is my dearest wish for the wonderful sounding HQ Player.

[kurb1980]

Okay so I took the plunge after all HQplayer sounds pretty awesome. I like up sampling everything to DSD. It's a poor mans PS Audio Direct Stream. I like the way DSD sounds on this player compared to Audirvanna. It's more real and life like.

[realmassy]

Not sure whether to post here or in the SOTM SMS100 thread...anyway, I've finally decided to try HQ Player and use the SMS100 as NAA, and I have to say I'm pretty impressed. Sound quality is great, although I'm missing a bit the convenience of JRemote/JRiver.

I have some highly customised Gizmo views and browsing the library with VNC was a PITA. I came Up with this workaround: I browse my music with JRemote and I add the music to the player 'Playing Now' playlist, only add the music without starting the player. Then I use VNC to drag and drop the playlist on HQ Player. It works well.

 

Now back to the SQ: I'm using a Metrum Hex and initially didn't like at all HQP with the default upsampling filter (polysync I think), it was like listening to any of my previous DACs, the natural timbre of the Hex was gone.

Strangely, with no filter applied, leaving the source sample rate, HQPlayer was better than JRiver acting as DLNA server for the SOTM SMS100 renderer, something I can't explain TBH, it should be a simple PCM stream in both cases (I decode the FLACS on JRiver, not on the SOTM), shouldn't be? HQP has better transparency and even better instrument separation.

 

Now I'm playing with the MP filters and I'm getting better results, the sound is still very different from the standard Hex sound, but in a good way. Any Hex owner caring to share their settings?

[sonata]

... I think I would get best NAA results if I could get HQP to recognize the NAA with a direct, Mac-to-Mac connection ... The simplest solution would be if Miska would make it so that I can enter an IP address into HQP Desktop to tell it where to find my NAA. The screen of the NAA machine reports its IP address at boot time, and it even self-assigns one to itself if not DHCP server is found. Right now this is my dearest wish for the wonderful sounding HQ Player.

 

Alex, I have setup a Ubuntu Linux NAA, and HQP on another Ubuntu PC. I configured a static IP on the NAA and the Ubuntu PC, as well as a DHCP server on the Ubuntu PC. Then I connected them directly with an ethernet cable - eliminating any switch. I found that HQP would discover the NAA, but not my DAC. After a lot of mucking around I abandoned the idea. Then later while doing some more experimenting, I discovered that HQP (at least in Evaluation Mode) will never discover the DAC, unless there is a connection to the Internet. I assume I would need a second NIC on the HQP PC, that would indirectly provide an Internet connection to the NAA - but as yet I have not tested this.

 

BTW I have also connected to this NAA from my 2012 Mac Mini. I have seen you state elsewhere that you find the SQ from your Mac Mini to your Linux NAA on another Mac Mini, as being superior to A+ on the 2012 Mac Mini. My experience has been the reverse, but could be explained by the hardware I am using for the NAA (a Dell i3 PC). There is not a lot of difference, but I find A+ 2.0.2 to be warmer and more pleasing, although HQP with the poly-sinc-shrt filter comes closer than the poly-sinc filter (this is only with PCM files). Given that the interface of A+ is better than HQP, I think I will stay with A+ for the moment, and await further developments of both players.

 

I have also noticed references to beta versions of HQP, but do not know where to download from. I did see Miska link to a 3.4.0 beta, but the link is no longer active. Can anyone advise where to download a current beta version from?

 

Rob

[EuroDriver]

3.4.1 is the fully released, and available to down load from the Signalyst website

 

on my system sounds noticeably better than 3.3 ! HQP has increased the sound quality gap over the other players

[realmassy]

Alex, I have setup a Ubuntu Linux NAA, and HQP on another Ubuntu PC. I configured a static IP on the NAA and the Ubuntu PC, as well as a DHCP server on the Ubuntu PC. Then I connected them directly with an ethernet cable - eliminating any switch. I found that HQP would discover the NAA, but not my DAC. After a lot of mucking around I abandoned the idea. Then later while doing some more experimenting, I discovered that HQP (at least in Evaluation Mode) will never discover the DAC, unless there is a connection to the Internet. I assume I would need a second NIC on the HQP PC, that would indirectly provide an Internet connection to the NAA - but as yet I have not tested this.

 

I think I have a similar setup, but works fine. My SOtM SSM100 (acting as NAA) is connected to the PC running HQPlayer using a CAT cable, no switch in my case either. The PC is connected to the main network with a wireless dongle. The PC is also acting as DHCP server. HQPlayer correctly finds the NAA and the DAC attached to it.

[Superdad]

Alex, I have setup a Ubuntu Linux NAA, and HQP on another Ubuntu PC. I configured a static IP on the NAA and the Ubuntu PC, as well as a DHCP server on the Ubuntu PC. Then I connected them directly with an ethernet cable - eliminating any switch. I found that HQP would discover the NAA, but not my DAC. After a lot of mucking around I abandoned the idea. Then later while doing some more experimenting, I discovered that HQP (at least in Evaluation Mode) will never discover the DAC, unless there is a connection to the Internet. I assume I would need a second NIC on the HQP PC, that would indirectly provide an Internet connection to the NAA - but as yet I have not tested this.

 

BTW I have also connected to this NAA from my 2012 Mac Mini. I have seen you state elsewhere that you find the SQ from your Mac Mini to your Linux NAA on another Mac Mini, as being superior to A+ on the 2012 Mac Mini. My experience has been the reverse, but could be explained by the hardware I am using for the NAA (a Dell i3 PC). There is not a lot of difference, but I find A+ 2.0.2 to be warmer and more pleasing, although HQP with the poly-sinc-shrt filter comes closer than the poly-sinc filter (this is only with PCM files). Given that the interface of A+ is better than HQP, I think I will stay with A+ for the moment, and await further developments of both players.

 

I have also noticed references to beta versions of HQP, but do not know where to download from. I did see Miska link to a 3.4.0 beta, but the link is no longer active. Can anyone advise where to download a current beta version from?

 

Rob

 

Hi Rob: Great to see someone doing similar experiments with NAA and getting similar results regarding HPQ finding the NAA when using a direct, machine-to-machine connection. I left a lot of detail and steps that I tried out of my short post, but I am quite sure we are encountering the same issue.

 

Using either a script to enable bootpd (OS X's built-in DHCP server) or Internet Sharing (which probably is just a configurable GUI that uses bootpd), my DAC-connected i7 music Mac mini--booted with Miska-provided minimal-Linux/NAA USB boot stick--is able to get an IP from my desktop Mac with a direct Ethernet cable (and I know it is not just a self-assigned IP). In the case of Internet Sharing, I recall also having to have the desktop machine running it's second EN connection--to my Cisco switch or to my DSL router via Thunderbolt>Ethernet Adapter--in order for the NAA to get a proper lease.

 

But either way, HQP desktop will not recognize the DAC/NAA. I put it this way as a way to comment on your assumptions and to show you that:

a) Having the desktop with a second live connection to an external DHCP server (my DSL router), does not help the NAA (and yes, I tried various static IPs to but all connections onto the same domain with only the last IP digits differing). With booted I can already tell my desktop Mac to assign any IP I choose to the direct-connected NAA, so I know that part is working.

 

b) I think you are incorrect to assume that HQP "discovers the NAA" just because you see NetworkAudioAdapter as choice on the Backend drop-down list. It is always a choice there--just like ASIO--whether or not any compatible devices exist on your connections. You are only making progress (success) when you choose NetworkAudioAdapter and then see your DAC on the Device drop-down.

 

And just to be clear, I am successfully able to use the NAA only when I have it (music mini or CuBox-i4Pro) connected directly to my ZyXEL DSL router. At that point, the desktop player MAc can be attached either to my Cisco switch or to the ZyXEL (the latter sounding a little bit better.

 

As mentioned, what I wish Miska would add is an option in the preferences window to manually tell HQP what IP address to look for for the NAA. He says he want's the NAA to just work on the average person's network, but I'm sorry, anybody using HQP and an NAA is not an "average" person.

 

---------

 

Okay, on to my SQ experience--commenting mostly in relation to what you said:

a) I find that the current release of HQP, 3.4.1 sounds better than both the first release (3.4.0) and better than the last beta I had (3.4 beta 6). As mentioned, differences were not as large as differences between A+ versions, but HQP current release 3.4.1 is really the best so just go with that.

 

b) Listening at length to HQP 3.4.1 (either DAC-connected, LPS supplied, i7 mini running player OR same machine as NAA), and comparing to my favorite version of A+ (1.5.10), I am now fully convinced that HQP is the more musically truthful player. And this leaving aside the better filters of HQP as I made a lot of comparison between A+/HQP with all SRC turned off (playing both Redbook and hi-res tracks). With the same comparison earlier last month (with 3.4 beta 6) I was feeling like A+ was doing a somewhat better job "bypassing" CoreAudio and that HQP had a little bit of a "veil" to it. Maybe it is the version change or maybe I am just starting to hear and understand more, but I don't feel that way anymore. (I still wish for a XMOS compatible ASIO driver so we can use HQP with ASIO as back-end.) I do need to listen to the release versions of A+ 2.x. My last listen to any newer-gen A+ build was to the 2.x beta called 1.9.0.3.

Once I turn on HQP's Polysinc-short filter, all need for comparison (to A+) ends for me.

 

c) Despite what Miska implies, the machine running the NAA makes a HUGE difference in my system. I have elsewhere reported that my i7 mini with linear PS trounces my new CuBox-i4Pro (even powered by the same LPS as i7 mini) as an NAA. More coherence, less congestion, vastly better piano, bass, etc.

And these comparisons are with Miska's own Linux/NAA boot images.

 

d) Having just said that, I am getting such fantastic results from HQP running directly on my tweaked 16GB i7 mini (booted from very slimmed Mavericks SD card, CAD script, no drives mounted or connected, files from direct-connected Ethernet share) that the same machine as NAA is NOT clearly better. There are SQ differences between it with local player or NAA, but for the moment I am favoring it as a local player. Maybe finding a way to not have to run the NAA into my DSL router will tip the scales back towards NAA.

 

I can say that the NAA setup took a noticeable leap forward when I realized I had the desktop player Mac (also a 2012 i7 with 16GB, but running Mountain Lion) connected to the switch via the Thunderbolt>Ethernet Adapter (because when playing without the NAA, that's how I get my work computer on the Internet--the build in EN has the BlueJeans/Belden going direct to the music Mac). Connecting the work/player machine via its built-in Ethernet port--either to the switch or to the DSL router where the NAA must connect to--made a really nice SQ difference.

 

 

A couple of other last observations I'll share:

1) Despite the NAA responding to the Ethernet connection (of the desktop player on the end) as explained above, nothing was gained in SQ when I booted my work/desktop machine with the same super-optimized SD card that I normally use for my music computer (it was of course available since the music machine was being booting with the Linux/NAA USB stick).

 

2) Whether run as local (HQP>DAC) or as NAA, playing tracks from RAM disk seems to do nothing for SQ. That is quite different from A+ which responds greatly to location of tracks at playback time. (Both HQP and A+ permit very fast, semi-blind A/B/A/B switching to listen for this since whatever the location of the tracks when you drag them into a playlist/transport window is where they will be retrieved from at playback time.)

 

Since HQP seems immune to RAM disk, the final (?) test I plan to try is to boot my music mini from one of it's normal, "factory-Mavericks" partitions and compare HQP (local mode) versus booted from the SD I have been using and are so deeply familiar with.

 

 

Now you all know how I wasted my weekend!

 

Cheers,

--Alex C.

 

P.S. I was about to try Miska's latest CuBox-i Linux/NAA build (newer kernel and all newer versions of services), not so much for SQ difference (though I would listen for that), but to see if it changed its ability to wait for an IP address from my slow DSL modem/router through my switch. But I forgot to switch the voltage from my LPS from 12V to 5V and I completely fried my brand new CuBox-i4Pro. It's trash now (though I did take it apart to see how they squeeze everything into that tiny cube.

[Jabs1542]

Not sure whether to post here or in the SOTM SMS100 thread...anyway, I've finally decided to try HQ Player and use the SMS100 as NAA, and I have to say I'm pretty impressed.

 

I am considering a similar path, are you running DSD through the SOtM NAA to your DAC? Did you have any driver problems?

 

thanks,

Jay

[kurb1980]

I have tried it on both my primary and my sd card stripped down OS the difference in SQ is very small IMHO. Although I do prefer listening from SD.

[realmassy]

I am considering a similar path, are you running DSD through the SOtM NAA to your DAC? Did you have any driver problems?

 

thanks,

Jay

My DAC doesn't support DSD, and no I had no driver problems, my Hex has been working flawlessly since day one.

[AnotherSpin]

 

<...> I am now fully convinced that HQP is the more musically truthful player. And this leaving aside the better filters of HQP as I made a lot of comparison between A+/HQP with all SRC turned off (playing both Redbook and hi-res tracks). With the same comparison earlier last month (with 3.4 beta 6) I was feeling like A+ was doing a somewhat better job "bypassing" CoreAudio and that HQP had a little bit of a "veil" to it. Maybe it is the version change or maybe I am just starting to hear and understand more, but I don't feel that way anymore. (I still wish for a XMOS compatible ASIO driver so we can use HQP with ASIO as back-end.) I do need to listen to the release versions of A+ 2.x. My last listen to any newer-gen A+ build was to the 2.x beta called 1.9.0.3.

Once I turn on HQP's Polysinc-short filter, all need for comparison (to A+) ends for me.

 

 

Fully agree. I had quite similar assessment when first tried HQP beta sometime ago. The sound was pleasant, but... "veiled". Nevertheless, something was immediately planted deeply within me, and I've been returning to HQP again and again. Its sound was strangely attractive, and, simultaneuosly, "not right" in comparison with both Amarra and A+. Now, after HQP became my reference and first choice player for many weeks, both Amarra and A+ started to sound... not right. Probably, as it was said by Alex: "maybe I am just starting to hear and understand more".

 

Well, my suggestion to those who will try HQP for the first time. Stop listening to other players, forget HOW they sound, wash you ears and purge your perception thoroughly. Listen to HQP and stay clean from ANY comparisons. Give its sound a time to grow within you. You will be glad you did.

[freesteve]

I wish Miska would incorporate a left and right balance.

[jhwalker]

I'm really enjoying my trial of this software, and think I'm hearing a very nice SQ upsampling all my content to DSD128. Still trying the various filter combinations, etc., but was wondering: is there a reason Apple Lossless (ALAC) is not supported? Since I use iTunes for library management, *all* my tracks (save my small DSD collections) are in Apple Lossless format, and it's a bit of a pain to convert to AIFF (or whatever) to get HQPlayer to load them.

 

Any possibility ALAC support will be added in future? Or is there a specific design / philosophical constraint against supporting it?

 

Forgive me, but a "bump" on this one: I never saw an answer, and I'd really like to understand if there are plans to add ALAC support in the future? I finally bit the bullet and purchased this great player, but that's the one fly in the ointment for me

[Miska]

Forgive me, but a "bump" on this one: I never saw an answer, and I'd really like to understand if there are plans to add ALAC support in the future? I finally bit the bullet and purchased this great player, but that's the one fly in the ointment for me

 

I don't have a good answer... I've looked into it, but the Apple documentation is a bit vague so I'm not sure if it will work the way I think it will until I try it out. And it would be OS X specific feature not available on other platforms which makes it lower priority.

 

So maybe at some point. As I have a long list of things to be done and limited amount of time, things happen in some order. Meanwhile I can only recommend using one of the cross-platform containers, my favorite being FLAC, but for iTunes compatibility AIFF serves the purpose.

 

(the ALAC code release from Apple wasn't really useful, they released the codec code, but not the code dealing with the container where ALAC is stored)

[jhwalker]

I don't have a good answer... I've looked into it, but the Apple documentation is a bit vague so I'm not sure if it will work the way I think it will until I try it out. And it would be OS X specific feature not available on other platforms which makes it lower priority.

 

So maybe at some point. As I have a long list of things to be done and limited amount of time, things happen in some order. Meanwhile I can only recommend using one of the cross-platform containers, my favorite being FLAC, but for iTunes compatibility AIFF serves the purpose.

 

(the ALAC code release from Apple wasn't really useful, they released the codec code, but not the code dealing with the container where ALAC is stored)

 

Thank you for the quick answer. Not what I wanted to hear as my library (like a lot of other Apple folks, I'm sure) is in ALAC format to save space - I don't have space to expand my 1.5TB library to 3.0TB+, along with doubling that for backup, etc.

 

But . . . I'm really enjoying the player and I will figure something out. Maybe I can preconvert individual albums to AIFF before playing

[Pocobear]

I use XLD a converter that's free for MAC, it will convert, even in batch mode, all ALAC files to AIFF very quickly. I don't know how much bigger, if at all, the converted files would be, but I didn't notice too much difference. But I have terabytes available so I probably wouldn't.

[4est]

AIFF is full size, adding say 40%. I've used XLD to transcode over a terabyte worth of files at a time.

I use XLD a converter that's free for MAC, it will convert, even in batch mode, all ALAC files to AIFF very quickly. I don't know how much bigger, if at all, the converted files would be, but I didn't notice too much difference. But I have terabytes available so I probably wouldn't.

[Pocobear]

AIFF is full size, adding say 40%. I've used XLD to transcode over a terabyte worth of files at a time.

 

+40%, wow, Note to self, pay closer attention idiot! Oh well, storage is cheap now days.

[Bob Stern]

I don't have a good answer... I've looked into it, but the Apple documentation is a bit vague so I'm not sure if it will work the way I think it will until I try it out. And it would be OS X specific feature not available on other platforms which makes it lower priority.

 

[snip]

 

(the ALAC code release from Apple wasn't really useful, they released the codec code, but not the code dealing with the container where ALAC is stored)

Miska, I think you're losing a large chunk of the Mac market if you don't support ALAC. If you look at the discussions of Audirvana 2.0 on this forum, you can see that iTunes library sync/import is a huge concern of many users, which means they have a lot of music in ALAC. Pure Music is entirely dependent on iTunes. Many prospective users won't want to spend hours converting their entire library to FLAC or AIFF.

 

I'm not a programmer, but I don't see why you need to know the ALAC file format or M4A container format to read audio data from an ALAC file. Core Audio can do the conversion for you. Even though you are not using Core Audio to perform your DSP algorithms or even to send audio to the output device, it seems you should be able to use Core Audio (e.g., AudioConverterFillComplexBuffer) to read audio from an ALAC M4A file, convert it to CAF in memory, then convert it to 64-bit float or whatever other format you need for your DSP.

 

(I found the AudioConverterFillComplexBuffer method in the book "Learning Core Audio" by Chris Adamson & Kevin Avila.)

[Miska]

I'm not a programmer, but I don't see why you need to know the ALAC file format or M4A container format to read audio data from an ALAC file.

 

Because that's the only way to make it cross-platform... Whenever something is specific to a certain OS it immediately get's 1/3 of the priority because I support three platforms simultaneously and prefer to keep feature sets aligned as much as possible.

 

Core Audio can do the conversion for you.

 

Yes, I've looked into that. What is vague is what format I would get if I don't ask anything specifically. I mean I want to get bit-perfect decompression of the content to what ever is the native format and then go on myself from there. I specifically want to avoid any other conversion than decompression.

 

CoreAudio sort of works wrong way around, because it expects me to tell what I want. But I want CoreAudio to tell what was the source format before compression and give me that one.

[Bob Stern]

What is vague is what format I would get if I don't ask anything specifically. I mean I want to get bit-perfect decompression of the content to what ever is the native format and then go on myself from there. I specifically want to avoid any other conversion than decompression.

 

CoreAudio sort of works wrong way around, because it expects me to tell what I want. But I want CoreAudio to tell what was the source format before compression and give me that one.

Can you open the source file only to read its properties, then use that to specify the audio format for Core Audio to use when you subsequently tell it to convert the source file to CAF?

[One and a half]

For years we had to bend to Apple's tunnel vision of supporting formats of their own choosing, and ignoring ISO, FLAC and countless other codecs. Workarounds were possible, messy hit and miss, but Apple didn't listen, and iTunes mangles library support like no other. Lucky it is provided at zero cost, who would pay for such a mess, although at one time you had to pay...

 

Time for payback. If you want to play ALAC on HQPlayer, convert to FLAC like a lot of us had to do if we wanted to play nicely with the iTunes world. There are quite a few tools that convert for OSX and for Windows.

[LewinskiH01]

Hello Miska.

 

A few pages back I saw you posted you were going to experiment with an e28 and direct-driving 3 amps per stereo side, then time align the drivers, etc. I actually also have this in mind.

 

I asked Uli, though, and he hasn't been able to get the e28 to work with 24/192 material with Acourate because of drivers (he did get 24/96 to work). Was wondering if you got the e28 working in multichannel and if you are upconverting PCM to DSD, or how are you doing it?

 

I have an e22 that sounds very good when fed DSD upconverted by HQPlayer from a 16/44 file, but it doesn't sound good when fed 16/44...so I can see myself foregoing my stereo pre and venturing into digital crossovers done at the server and feeding an e28 only if this can be done in DSD, as I extrapolate from my experinece with the e22.

 

Cheers