HQ Player

[ericuco]

I am using HQ Player Desktop on my Mini. When using SDM, I get a loud pop when the last song finishes playing.

 

This happens with all DSD files whether native DSD files (DoP) or converted PCM files (.aiff & .flac). I don't have this problem when using PCM.

 

All of my files are either CD rips using dbPoweramp into .aiff or purchased hi-rez off of one the stores (e.g. HDTracks, Acoustic Sounds). Point being all my files should be fairly standard stuff, not ripped SACD files.

 

Any thoughts as to a remedy? Really like using SDM at all times so would appreciate any thoughts to resolve this.

[ericuco]

Jay - I think you are looking for the drop down box located under the volume control knob on the main player panel. That box controls which method (SDM or PCM) to use. The other panels are typically for setting defaults that can be overridden on the main panel.

[ericuco]

Miska - any thoughts on my "popping" issue (#230)?

[ericuco]

Any of you OSX guys upgrade to Yosemite yet & run HQP? Any comments on whether to upgrade from Mavericks or not.

 

Edit: sorry just saw prior post on this.

[ericuco]

Set the max and min volume -3db, I rather have this at 0db, but this description is in the manual not to overload the SD modulator with dsd files.

 

I believe the Max volume should be -3 db and the Min volume set to something like -60 db. If you have both set to the same value then that might be causing your issue with PCM playback. These values are not used when using SDM from what I understand.

 

There is still some confusion on my part exactly how to set the volume controls overall, meaning these max & min values in combination with the volume control knob so hopefully Miska will chime in.

[ericuco]

Miska - in my setup, my DAC has volume control so I do not want to adjust the volume in HQP regardless of playback settings (PCM > DSD or native DSD). Currently, I convert all PCM to DSD for playback. I also have a few DSD64 (DSF format) albums.

 

As such, I have the volume knob in HQP set to max (fully clockwise) with the max setting at -3 db and min setting at -60 db (default).

 

If I understand your comments:

 

1) Then the min/max volume settings are in play (used) for PCM > DSD file conversions. The DirectSDM setting does not come into play regardless of being checked or unchecked.

 

2) However, if the DirectSDM is checked AND playing DSD file, then the min/max settings are not in play. Also, there is no upsampling of DSD (to a higher DSD format) regardless of the sample rate setting.

 

3) Lastly, if the DirectSDM is unchecked AND converting DSD file to higher DSD, then the min/max settings are in play.

 

Sorry for being thickheaded on this and respectfully ask for your patience.

[ericuco]

Hi audiotone. Welcome to the HQ Player club!

 

We are converging to similar setups. I already have the Cubox-i in place as a NAA and waiting on Small Green Computers to build me a scaled down CAPS Zuma (quad i5 vs i7, no SOtM USB card) running Ubuntu Studio.

 

[Any things I should consider as preparation? I guess the NAA file from HQplayer is a complete image wich will have to be installed on the CuBox?

 

Correct the NAA file is complete. Just install the microSD card into the Cubox-i and power up. At the end of the boot cycle, you will be prompted for log-in & password but you can ignore this.

 

One suggestion is to get decent ethernet cables (e.g. Cat6a). I know that Chris used the Audioquest cables (pricey) in a recent install at his home. Here in the US, a good (less expensive) source is Blue Jean Cables which are made to order at about $2/foot.

[ericuco]

How did you installed the new image on CuBox? usb drive? usb stick?

 

The Cubox boots from a microSD card so you need to create a bootable image using the image provided by Miska. I downloaded the image to my Win7 laptop then used Win32 Disk Imager program to write the image to the microSD card. Easy peasy.

 

Just insert the microSD card into the Cubox and power up. Nothing else to do from there on the Cubox side. Obviously you need to do the proper connections - Ethernet to your network & USB to DAC.

[ericuco]

What would be the best setup to install NAA component?

 

Miska has a CuBox-i NAA image that you can load onto a micro SD card, plug the card into the Cubox-i and be up & running with a NAA configuration. I am using a CuBox-i4Pro. Very easy to setup.

 

In general, all the heavy lifting (e.g. upsampling, PCM to DSD conversion) is done on the server so the NAA box can be very modest.

[ericuco]

As to your second question, if you go into the settings menu and select dsd/dop and 2xdsd then that is what you will get for every input file.

 

This is not totally correct for DSD file playback. You still need to uncheck the "DirectSDM" box in the DSDIFF/DSF setting panel. Otherwise w/ box checked, DSD files play at their native rate regardless of the other setting. At least that is my understanding.

[ericuco]

Are you talking about the SDM Pack setting in the SETTINGS menu, or the SDM Box (drop down) on the Desktop Screen???? I have not seen a setting for DirectSMD mentioned anywhere.

 

No. There are two separate "settings" panels under the "File" menu - "Settings" & "DSDIFF/DSF Settings".

 

Under the "DSDIFF/DSF Settings" panel, you need to uncheck the "DirectSDM" box in order to upsample DSD files (DSD64 > DSD128). Once "DirectSDM" is unchecked, then the DSD sample rate specified under the "Settings" panel take effect (or temporarily overridden by what is specified on the Main Desktop Screen).

 

There is an official User Manual that should have been included with the HQP download. There is also a supplemental HQPlayer Kick-Start Guide by Geoffrey Armstrong (do a Google search) that is very helpful.

[ericuco]

SuperDad did a comparison between a Mac mini running Linux and a Cubox. Both were pretty tricked out, and the Mac mini left the Cubox in the proverbial dust.

 

I keep seeing this reference to Superdad's comparison pop up & thought I would throw in my 2 cents.

 

When I first included a Cubox as a NAA with my Mac Mini (2012, i5), I got similar results to Superdad. It sounded really flat so I removed it from my system. This was around the same time as Superdad's comparisons so we were using the same version of HQP.

 

I would also like to mention that Superdad has had some network issues (aired here on CA) in using the NAA setup so I think that needs to be considered as well. In other words, his setup is not a standard "plug & play" one.

 

Since that time, I have moved to a CAPS-like Linux computer (i5, quad core, 4 Mb RAM, SSD, no fan) in conjunction with the Cubox (NAA) and it sounds wonderful. The Cubox is using an inexpensive LPS although I have one of the LPS from Small Green computers on order to power both the Linux computer & the Cubox. The Linux computer and Cubox are connected using Cat6a cables running through a Cisco gigabit switch. Music files are stored on a Synology NAS also connected to the Cisco switch. The NAA setup has worked without a hitch.

 

I have since sold the Mac Mini because of all of the "tweaking" required (SD card, optimization scripts, etc.). The Linux computer sounds great without any "tweaking". Not trying to start a debate (Linux vs Mac) and if you want to do all of the tweaking, by all means, knock yourselves out. It just got to be a bit tiresome to me after a while.

[ericuco]

Anyone else seeing this?

 

With DSD7, typically no issues. However if I switch to ASDM5 then I occasionally get slight pops (clicks) between tracks as well as a stutter at the beginning of the track (1st second or so). It is a bit sporadic, not consistent or predictable.

 

I have tried lots of other settings (e.g. Pipeline) to try & narrow it down & everything points to ASDM5 although to have not tried other modulator settings besides DSD7 & ASDM5.

 

I have even monitored computer & NAS performance & neither were no where near being max'd out in terms of CPU & RAM.

 

DAC is XMOS & ESS.

[ericuco]

And grep-i firewire /lib/udev/rules.d/* shows the following. I wonder a little bit, that the ffado driver is mentioned. Maybe a conflict between snd_dice and ffado?

 

/lib/udev/rules.d/70-udev-acl.rules:# ffado is an userspace driver for firewire sound cards

/lib/udev/rules.d/70-udev-acl.rules:SUBSYSTEM=="firewire", ENV{ID_FFADO}=="1", TAG+="udev-acl"

/lib/udev/rules.d/70-udev-acl.rules:SUBSYSTEM=="firewire", ATTR{units}=="*0x00a02d:0x00010*", TAG+="udev-acl"

/lib/udev/rules.d/70-udev-acl.rules:SUBSYSTEM=="firewire", ATTR{units}=="*0x00b09d:0x00010*", TAG+="udev-acl"

/lib/udev/rules.d/70-udev-acl.rules:SUBSYSTEM=="firewire", ATTR{units}=="*0x00a02d:0x010001*", TAG+="udev-acl"

/lib/udev/rules.d/70-udev-acl.rules:SUBSYSTEM=="firewire", ATTR{units}=="*0x00a02d:0x014001*", TAG+="udev-acl"

/lib/udev/rules.d/75-persistent-net-generator.rules: ENV{COMMENT}="Firewire device $attr{host_id}"

/lib/udev/rules.d/91-permissions.rules:SUBSYSTEM=="firewire", ATTR{units}=="*0x00a02d:0x00010*", GROUP="video"

/lib/udev/rules.d/91-permissions.rules:SUBSYSTEM=="firewire", ATTR{units}=="*0x00b09d:0x00010*", GROUP="video"

/lib/udev/rules.d/91-permissions.rules:SUBSYSTEM=="firewire", ATTR{units}=="*0x00a02d:0x010001*", GROUP="video"

/lib/udev/rules.d/91-permissions.rules:SUBSYSTEM=="firewire", ATTR{units}=="*0x00a02d:0x014001*", GROUP="video"

 

Could you guys take this discussion off-line?

 

I am sure it is meaningful to you guys but not so much to the rest of us. It is good to know that a problem exists and the fix (if there eventually is one) but the diagnosis process is generally not helpful.

 

Thanks

[ericuco]

I may start using this same method also on Linux.

 

Please do include Linux. I am still getting periodic clicks between tracks when using ASDM5(?).

[ericuco]

Could someone be so kind as explain in CA for Dummies fashion, step by step, what this would involve...like, could I use my 2010 unit to play and by another Mini to convert? What would "hook them up?"

 

Yes, you can use your 2010 Mac Mini as the NAA but it is probably overkill in terms of required processing power. I am using a Cubox-i as my NAA. Miska has an image specifically for the Cubox-i so it is easy to get up and running (copy the image to a microSD card & you are off & running).

 

Another possible NAA platform being mentioned is the fitlet-b which will be released soon.

 

The reason for the Cubox-i & fitlet computers, from what I can tell, is that they are fanless, boot from SD card (no HDD or SSD required), low processing chips (less noise).

 

The computers (server & NAA) are connected via Ethernet. In my case, both computers are connected to a Cisco gigabit switch. Some people are directly connecting the computers via Ethernet (w/o a switch) but using a switch is the usual connection.

 

There is a NAA thread here on CA.

[ericuco]

 

Step 1 - You need these parts

 

 

Two computers - server & NAA, ethernet cables to connect the computers either through a switch (default configuration) or directly, usb cable.

 

The server is usually something with some decent amount of processing power (e.g. Mac Mini). The HQ Player software resides on the server. It is the server you typically use to select & play music. The only connection is the ethernet cable, at least for playing music. You can obviously have external USB HDD, NAS, etc. connected.

 

The NAA can be a minimal computer (e.g. Cubox-i). It is basically a black box so no need to connect monitor, keyboard, mouse, etc. as there is no interaction needed once the NAA boots up. The only connections are the ethernet cable and a USB cable that runs to your DAC (assumes DAC with USB input). So, the signal comes in via ethernet and out via USB.

 

 

Step 2 - Then you hook them up in this order with these cables

 

 

As noted above, ethernet cable out from the server to switch (or NAA directly), ethernet cable from switch to NAA, USB cable from NAA to DAC.

 

 

Step 3 - Then you install this software here and here

 

 

The HQ Player software is installed on the server.

 

The Network Audio Daemon (available from the Signalyst website) is installed on the NAA. This might take some computer knowledge to do this.

 

As I mentioned, Miska has a Cubox-i image (basically a ready to go OS) which is Linux based. You can download the image from Miska. Use something like Win32 Disk Imager to write the image to a microSD card. Install the microSD card into the Cubox-i and power up. Connect ethernet & USB cables and you are done. Select the NAA in the HQP Preference panel on the server.

 

 

Step 4 - Then you play unconverted CD rips in DSD128, and it sounds better than my Mini sweating it's balls off doing all that math on the fly whilst its sending data to my Mytek

 

Correct. The load on the Mac Mini should be same (or nearly so) but all (most) of the noise is eliminated by using ethernet connection to the NAA which does not do any processing, just feeds your DAC.

[ericuco]

 

Correct. The load on the Mac Mini should be same (or nearly so) but all (most) of the noise is eliminated by using ethernet connection to the NAA which does not do any processing, just feeds your DAC.

 

Just to clarify my above statement, Jud is correct. If you are mainly looking to reduce the load on your Mini, the "-2s" options,, etc. are the way to accomplish this. The NAA does not really do anything to reduce the load on the server. The NAA is used to reduce noise, not load.

[ericuco]

Okay...what about the 3 filters on the micro I leave it set to standard?

 

Suggest you guys move this discussion over to the other thread: HQ player settings with iFI Micro DSD.

 

At some point you might want to actually listen to the various settings for yourself & decide what sounds best.

[ericuco]

Sorry, but I don't check the settings as my native language is not English. Please could someone tell me how to do so that my Mytek DAC plays DSD64 and DSD128 files native without up- or downsampling ???

 

File > DSDIFF/DSF settings > DirectSDM (checked)

 

No processing (up sampling or down sampling) of DSD files.

[ericuco]

Thank you, and which settings in Preferences and main window?

 

Since the DSD files are being played without any processing (DirectSDM checked), the SDM default settings (Preference & Main window) are bypassed and have no affect.

 

Have you read the HQ Player manual (located in the installation directory) as well as the HQPlayer Kick-Start Guide by Geoffrey Armstrong? They might answer a lot of your basic questions. Admittedly, a lot of these things are a bit confusing but I suggest playing around with the various settings to see what happens. And there are no "best" settings because of all of the various components (OS, computer, amps, DAC, individual hearing & preferences, ...) in play. You just need to spend time listening and judge for yourself what sounds "best" and even that may change over time.

[ericuco]

also I want all resolutions to remain as it is. No up sampling for PCM of DSd please

If this is your goal then don't bother with HQP.

[ericuco]

On classical tracks with long lead outs that become "digital black" after the fade, the clicks (seems like they're louder in the right channel?) come. If I flip the upsampling from DSD to PCM, no more clicks.

 

Which modulator are you using for DSD (PCM > DSD)? I get clicks at the end of tracks with ASDM5 & ASDM7 settings but not DSD7. All other settings being the same.

[ericuco]

Thank you. Just spent the last 15 minutes trying to make this happen in Linux, and I'm not able to. Perhaps I was mistaken about Linux.

 

As I mentioned previously, on Linux, I consistently get clicks using ASDM5 & ASDM7 but not DSD7. No idea why this is but it is very predictable.

[ericuco]

When I finally get my Lampi Big 7 DAC up and running I will be looking to upsample my native 24/192 recordings to DSD. I can go to 256 DSD and this is what I would like to achieve. What are the best HQ Player settings, especially filter setting? Also, does anybody know if HQ Player can use the 12.228MHz sampling frequency which is exactly 256 times 48kHz (or 16 times 192kHz), rather than 11.289MHz which is not a whole number multiple of either and therefore may be computationally more demanding and consequentially of lower sound quality than the former?

 

There are no "best settings". It is a matter of preference but in general people seem to like the "poly-sinc" options for oversampling filters and the ASDM5, ASDM7, DSD5 & DSD7 modulators. Miska seems to generally recommend the "poly-sinc" filters as well.

 

As for the DSD rates, it seems that HQP "talks" to the USB XMOS board in your DAC and will only allow rates that the board can use. In my case, DSD64 (2.8224MHz and 3.0720MHz) and DSD128 (5.6448MHz and 6.1440MHz) using DoP. These are multiples of the 44.1 & 48 kHz clocks (I think). For me, the best is DSD128 (6.1440MHz) but it requires me to use the "-2s" filters.

 

To get to DSD256, you will not considerable compute resources. With my i5 quad core Linux computer, to get DSD128 (6.1440MHz) with the "-2s" filters, all four processors are running at 25 - 30%.