[bogi]

Is there a tutorial somewhere that can guide me?

Download naa-500-raspberrypi4-holored.7z from https://www.signalyst.eu/bins/naa/images
Unzip it and write it to SD card using https://etcher.balena.io/
Use a separate SD card, do not overwrite your original one which came with Red.

 

[BShaw]

Download naa-500-raspberrypi4-holored.7z from https://www.signalyst.eu/bins/naa/images
Unzip it and write it to SD card using https://etcher.balena.io/
Use a separate SD card, do not overwrite your original one which came with Red.

Got it, thank you Bogi.

 

[sendler2112]

I am now in agreement that the Gustard A26 (and VMV D2R and probably all Delta-Sigma DACs) sounds better when converting my PCM files to at least DSD 256. Are there any types of DACs that sound their best when getting an HQ Player upsampled PCM signal instead? R2R DACs?

 

[shwnwllms]

I am now in agreement that the Gustard A26 (and VMV D2R and probably all Delta-Sigma DACs) sounds better when converting my PCM files to at least DSD 256. Are there any types of DACs that sound their best when getting an HQ Player upsampled PCM signal instead? R2R DACs?

Chord DACs for example as they will convert DSD back to PCM (adding all sorts of ultrasonic noise etc in the process, defeating the point of HQP upsampling to DSD). Chord DACs are recommended highest PCM input instead

Denafrips isn’t as cut and dry as they are a black box as to what is going on, but some people (myself included) prefer highest rate PCM on Denafrips DACs. My own experience is that DSD sounded a bit “soft” on the Pontus II (loss of dynamic range) which led me to think it was doing some sort of PCM conversion and wasn’t a true DSD DAC like Holo Audio, T + A etc. I sold my Pontus and picked up a May KTE instead for this reason.

Again, speculation on the Denafrips above as we don’t really know what’s going on inside but the Chord example is cut and dry

 

[bogi]

I am now in agreement that the Gustard A26 (and VMV D2R and probably all Delta-Sigma DACs) sounds better when converting my PCM files to at least DSD 256. Are there any types of DACs that sound their best when getting an HQ Player upsampled PCM signal instead? R2R DACs?

One can get most benefit out of HQPlayer with direct DSD capable DAC. They all can be considered to be delta sigma DACs, since they accept delta sigma modulated signal (what DSD is) on input and convert it directly to analog signal. A list of such devices is maintained here: https://audiophilestyle.com/forums/...tal-filtering/?do=findComment&comment=1280696
Another thread devoted to direct DSD capable DACs appears on Roon forum: https://community.roonlabs.com/t/best-native-dsd-dacs-for-use-with-hqplayer/132298/1943

So all direct DSD DACs are of delta sigma type, but most of delta sigma DACs are not capable of direct DSD path.

 

[sendler2112]

One can get most benefit out of HQPlayer with direct DSD capable DAC. They all can be considered to be delta sigma DACs, since they accept delta sigma modulated signal (what DSD is) on input and convert it directly to analog signal. A list of such devices is maintained here: https://audiophilestyle.com/forums/...tal-filtering/?do=findComment&comment=1280696
Another thread devoted to direct DSD capable DACs appears on Roon forum: https://community.roonlabs.com/t/best-native-dsd-dacs-for-use-with-hqplayer/132298/1943

So all direct DSD DACs are of delta sigma type, but most of delta sigma DACs are not capable of direct DSD path.

I am actually pretty happy with the sonics of the Gustard A26 via HQ Player conversion and direct DSD at 256 now that I have the LAN connection working. Even on the mid level HQP settings since my gaming laptop is running out of cooling to play well on the more demanding Sinc-m filters with 512+ modulators. I am considering building a super gaming computer to future proof my options for 1024 DSD direct DACs which will be no doubt more and more prolific in the coming years.

But then again I'm feeling guilty about blowing that much energy and so much heat into my room just so I can listen to a song. Which keeps blairing for some reason even after the que has ended until I hit stop. Like for hours sometimes if I had successfully fallen asleep.

I will measure the consumption of my laptop tonight with my KillAWatt meter but I have to imagine it is a few hundred Watts when the fans are screaming. And would be even more crazy with the cranked HQ Player settings in a big desktop.

The processor demand for HQP to send out PCM files that are upsampled to 768 is much less. if I can get similar sonics from a true PCM native DAC. But I am not buying anything as expensive as a Chord.

 

[plumpudding2]

I am actually pretty happy with the sonics of the Gustard A26 via HQ Player conversion and direct DSD at 256 now that I have the LAN connection working. Even on the mid level HQP settings since my gaming laptop is running out of cooling to play well on the more demanding Sinc-m filters with 512+ modulators. I am considering building a super gaming computer to future proof my options for 1024 DSD direct DACs which will be no doubt more and more prolific in the coming years.

But then again I'm feeling guilty about blowing that much energy and so much heat into my room just so I can listen to a song. Which keeps blairing for some reason even after the que has ended until I hit stop. Like for hours sometimes if I had successfully fallen asleep.

I will measure the consumption of my laptop tonight with my KillAWatt meter but I have to imagine it is a few hundred Watts when the fans are screaming. And would be even more crazy with the cranked HQ Player settings in a big desktop.

The processor demand for HQP to send out PCM files that are upsampled to 768 is much less. if I can get similar sonics from a true PCM native DAC. But I am not buying anything as expensive as a Chord.

I've undervolted both my gpu and cpu and while they do draw a combined 120 watt at DSD512 with light 512+fs the whole pc is whisper quiet (gpu temperature 32 degrees cpu temperature 55 degrees). If you don't chase super 512+fs @1024 and you undervolt and watercool you can have an extremely energy efficient HQPlayer computer that's very silent too!

 

[dericchan1]

I've undervolted both my gpu and cpu and while they do draw a combined 120 watt at DSD512 with light 512+fs the whole pc is whisper quiet (gpu temperature 32 degrees cpu temperature 55 degrees). If you don't chase super 512+fs @1024 and you undervolt and watercool you can have an extremely quiet energy efficiency HQPlayer computer that's very silent too!

There really isn’t much to chase after dsd1024 anyway, unless you are looking for added noise

 

[bogi]

I will measure the consumption of my laptop tonight with my KillAWatt meter but I have to imagine it is a few hundred Watts when the fans are screaming. And would be even more crazy with the cranked HQ Player settings in a big desktop.

Here you can see details about 44.1k to DSD512 conversion on my notebook HP ZBook G8: CPU i7-11850H max. 47W and GPU A2000 max. 24W. Of course it depends on particular hardware. No DSD1024 but I also don't have such a goal. My current DAC has a sweetspot at DSD256.

 

[sendler2112]

I made various power consumption measurements right at the plug of my laptop with a KillAWatt meter.

My laptop just coasting along with a browser open is cruising at 17 Watts, 1% CPU, 37% RAM. Briefly spiking to 37W when launching a new tab.

HQ Player streaming Qobuz via HQP Client to DSD256/ ASDM7EC -super/ p-s-g xla is 38W additional, 8% CPU, 46% RAM.

Drag and dropped Full WAV files in HQ Player sound slightly better and use 48W. additional.

Changing to DSD512/ ASDM7EC super 512+ increases the load to 67W and 14% CPU.

sinc-mga won't load for any DSD and locks up the program every time. It doesn't even try to load. The CPU goes straight to 1%. But it will play for PCM768 and sounds well rounded and natural whereas poly-sinc-gauss xla is more spacious but slightly more artificial.

PCM768/ p-s-g-xla/ TPDF sounds very close to DSD with the Gustard A26. More reverb, less robust and natural. PCM 48 to 768 is much less demanding using only 6W additional over the idle power.

I then plugged the entire power strip into the KillAWatt. The A26 is only coasting at 3W although it does get warm. The VMV P2 Headphone amp idles and plays at 12W

The whole strip is passing 110W when playing DSD256 including the router (I'm connecting to the A26 via LAN) and several vampires. So I feel better now knowing that playing PCM converted to DSD is only really costing 50-60 W additional for the conversion despite the fans in my laptop blaring.

 

[plumpudding2]

Interesting discovery: I was listening to a 96khz piano recording and if I engaged the 20khz filter the client showed it filtering from 10khz not 20:
See here a screenshot of me turning the filter on, then off, then on again.

I'm curious whether this is intended behaviour, anyone with ideas?

 

[jlaako]

Interesting discovery: I was listening to a 96khz piano recording and if I engaged the 20khz filter the client showed it filtering from 10khz not 20:
See here a screenshot of me turning the filter on, then off, then on again.

I'm curious whether this is intended behaviour, anyone with ideas?

It is either display bug or some bug in HQPlayer. I'm curious to see how to reproduce this.

 

[sendler2112]

My playback chain is laptop harddrive 44, 48, and 192 WAV files> i7 13th, nVidia 4060, 16GB DDR5 B(ut laptop cooling is obviously inferior to a liquid cooled tower)> HQ Player upsampled to 32/ 768 or DSD256> LAN RJ45 CAT6 (or USB via Supra Excalibur)> TP Link router> LAN CAT6 into the Gustards playing direct>0.3 meter DIY xlr cables> VMV P2 power amp> Balanced solid Silver cable to HiFiMan HE100SE.

The A26 sounds better when fed DSD direct over LAN from HQ Player but DSD conversion of PCM files is resource intensive. I contemplated buying a better tower with i7 14th, nVidia 4070 Super, 32 GB. Which would cost $2000. And chew up 130 Watts. Just to play a song. My laptop shows 20W at the AC cord when playing PCM at the highest settings flawlessly and 85W for DSD, with skips.

So I purchased the Gustard R26 for $1600 instead since it plays PCM 768 native. And I can get something back from selling the A26 if I want. My new laptop that I am already stuck with plays PCM768 with ease. barely raising the power consumption from 15W to 20W. Attempting PCM to DSD 256 even using the lesser sounding UL modulator is pegging something when fed with 192 files to 85W and skipping even on this relatively wimpy setting. I have to use plain poly-sinc-gauss to keep DSD from skipping with 192k files. In between the skips, or with 44.1k files that do not skip, I can still get a good, level matched comparison between the A26 playing DSD 256 direct over LAN, and the R26 playing PCM 768 NOS direct over LAN.

I'm happy to say that the R26 sounds better anyway. So I can save from buying another more expensive computer.

Now that I am a PCM person again, I had to sit and listen to all of the dithers. HQ Player is a very powerful program with a ridiculous number of options and possible combinations. I really like the Shaped dither. it is not as deep or wide as the TPDF but makes it sound kind of thin and bright and sibilant. The shaped takes the more robust sound of the LNS15 and widens it and brings back some detail and crispness without going too far. So for the Gustard R26 LAN PCM Direct NOS:
sinc-MGa at PCM768 (more balanced and musical than poly-sinc-gauss xla if not quite as wide or deep)
Shaped dither
32 bits ( anything higher than 20 bits starts sounding better up to a certain point and I don't hear any drawback to going all the way to 32 so might as well)
FFT Filter Length 4096 on the advanced tab (higher than 512 brings more transparency but too high starts to sound more etched and digital)

 

[plumpudding2]

It is either display bug or some bug in HQPlayer. I'm curious to see how to reproduce this.

I'll see if I can reproduce it

 

[plumpudding2]

It is either display bug or some bug in HQPlayer. I'm curious to see how to reproduce this.

I'm afraid I haven't been able to reproduce the issue. Started up HQP again with the exact same settings exact same track and the metering of the 20khz filter worked just fine. Unfortunately logging was not turned on so wasn't able to check that one

While i'm making bug reports: I've noticed sometimes there's an inconsistency between the limited counter and the Max value from the spectrum analyzer. In the past whenever I saw the limiting counter go up i'd also see the reported Max from the spectral analyzer be above 0.

Today they are inconsistent, To demonstrate I started a song and the meter at the same time (both reset to 0 first). 30 seconds into the song the client and main hqp window show limited count of 1 but the level meter shows maximum value of -1.71.

I'm wondering which one to trust in this scenario. To test I increased the volume from -3db to 0 to see what the level meter shows and the max jumped to +1.79 so it does seem to work, it's just at odds with the limiter counter.