HQ Player

[asdf1000]

1 hour ago, juanitox said:

no , it doesn't work with high rez

 

You have the same i7-6700K CPU? On Linux like @brother love also?

 

[StreamFidelity]

@brother love

 

Many Thanks. I found the mistake with me. Crucial again is the clock frequency, which must be at least 4GHz for the EC modulator ASDM7EC and DSD 256 for my system. Since I run a fanless PC, I come in the temperature in the critical area. At the moment everything is fine, the continuous operation is crucial.

 

poly-sinc-ext2 + ASDM7EC + 44.1 / 24/2 -> DSD 256
Runs smoothly with a very good sound. The CPU utilization is 16%. The power consumption is 34W, which is not a problem for my HDPLEX 200W Linear Power Supply.

 

poly-sinc-ext2 + ASDM7EC + 96/24/2 -> DSD 256
Sensational sound. Also runs smoothly. The CPU utilization is 21%. The power consumption is 39W.

 

Conclusion

For me the EC modulator ASDM7EC is only really good with DSD 256. With DSD 128 I found it too lean.

[asdf1000]

4 minutes ago, StreamFidelity said:

Conclusion

For me the EC modulator ASDM7EC is only really good with DSD 256.

 

Can your i9-9900K do PCM192kHz to DSD256x48 with ASDM7EC modulator, with ext2 filter?

 

And which fanless case and motherboard, if you don't mind me asking?

 

[StreamFidelity]

@Em2016

Can your i9-9900K do PCM192kHz to DSD256x48 with ASDM7EC modulator, with ext2 filter? => Yes 😀

 

 

This is my Audio PC

 

On 5/25/2019 at 6:18 PM, StreamFidelity said:

After detailed research, I have the following new fanless audio PC under construction:

 

• Intel Core i9-9900K (95W TDP), 3.6GHz - 5.0GHz OctaCore - HD Graphics 630
• Streacom FC9 Alpha (Black) - Desktop - micro ATX
• ASUS ROG Maximus XI Gene
• Corsair Dominator Platinum - DDR4 - 16GB
• Samsung SSD 970 PRO 512GB, M.2
• JCAT USB Card FEMTO
• JCAT NET Card FEMTO
• HDPLEX 200W Linear Power Supply (LPS) Black
• HDPLEX 400W HiFi DC-ATX

GPU is out of the question for me because I want to have a completely fanless PC. Since I currently have no problems running DSD 512 (filter: poly-sinc-xtr-lp-2s, shaper: AMSDM7 512 + fs) with an Intel Core i5-8400T, I am not afraid without a GPU.

 

Some will wonder why I chose a gaming board. These are just very robust and equipped with good components. Pink Faun does the same. 😉I believe that good hardware is essential for a good sound.

 

Added to this are memory with very low latencies and superfast SSDs.

 

For extremely important I consider a trouble-free data transport and a clean power supply. The JCAT USB and NET cards are equipped with Femto clocks and are powered externally by the HDPLEX 200W LPS with 5V each. The same LPS in combination with the HDPLEX DC-ATX converter is designed to provide the new audio PC with clean power that is free from high-frequency noise.

 

The operating system comes Server 2019 on it, optimized with Audiophil Optimizer 3.0 (to come in late May) and Fidelizer.

 

I'm really looking forward to my new project. 😀

[asdf1000]

26 minutes ago, StreamFidelity said:

• Intel Core i9-9900K (95W TDP), 3.6GHz - 5.0GHz OctaCore - HD Graphics 630
• Streacom FC9 Alpha (Black) - Desktop - micro ATX
• ASUS ROG Maximus XI Gene

 

Thanks!

[craighartley]

16 hours ago, Miska said:

 

This runs filters on the GPU and only modulators on CPU. Since the CPU is just quad-core, it is fully busy with the modulators and such. If it stutters occasionally, ASDM5EC is probably just about lighter enough that it may work fine.

 

 

Thank you, you’re right: no stutters with ASDM5EC with these settings. Problem is that I prefer 7th order modulator! Still, it’s lovely for now. 

[Miska]

19 minutes ago, craighartley said:

Thank you, you’re right: no stutters with ASDM5EC with these settings. Problem is that I prefer 7th order modulator! Still, it’s lovely for now. 

 

I have some optimization ideas how to possibly make things faster, so it may get better in future...

 

[tboooe]

6 hours ago, AnotherSpin said:

 

It is not clear for me from re-reading your post – did you try direct ethernet cable connection between server computer and NAA endpoint (server -> endpoint) without bridging to router, or there is still bridging of the kind: router -> server -> endpoint ?

Sorry for not being more clear.  I am not bridging.  

[brother love]

@Em2016:  I checked i7-6700k music server with PCM 96k & 192k music  files just now & concur with @juanitox re: poly-sinc-ext 2 48k x 256 DSD findings:

 

ASDM7EC (CPU usage :  35%) 1-2 sec. silence after 30 sec. of play before resuming, ASDM5EC (CPU usage :  29%) plays & sounds great!  I have buffer time set to default... maybe increasing time helps?  Will audition & explore different avenues further when time allows.

 

 

[AnotherSpin]

1 hour ago, tboooe said:

Sorry for not being more clear.  I am not bridging.  

 

Ok, this is what I am trying to do as well in mac OS. I prefer this sound more than from any other config as per now.

 

Did you try various ethernet cables for direct? 

[StreamFidelity]

1 hour ago, brother love said:

ASDM7EC (CPU usage :  35%) 1-2 sec. silence after 30 sec. of play before resuming, ASDM5EC (CPU usage :  29%) plays & sounds great!  I have buffer time set to default... maybe increasing time helps? 

 

Pink Faun writes, "Well, and most importantly, with Linux, we can reach an incredibly low latency (audio latency and processor latency). Because the latency is so low in the streamer 2.16 (audio latency at 192 kHz <2 , 67 ms on AMD Ryzen processor and high quality digital audio interface. Audio latency at 44.1 kHz <1 ms depending on hardware configuration." Looks Streamer 2.16

 

I believe it and I'm not far from that. In my system (Windows 10 Pro) I make sure to keep the latencies as low as possible. I set in the Asio driver the lowest value with 128 samples = 1.5 ms.

 

[craighartley]

3 hours ago, Miska said:

 

I have some optimization ideas how to possibly make things faster, so it may get better in future...

 

🙂

[juanitox]

4 hours ago, Em2016 said:

 

You have the same i7-6700K CPU? On Linux like @brother love also?

 

 

no , i use windows server 2019 with A.O 3.0

[tboooe]

1 hour ago, AnotherSpin said:

 

Did you try various ethernet cables for direct? 

I am just using a short Blue Jean CAT 6a cable.  I am planning on trying a Ghent JSSG etherrnet cable soon though.

[AnotherSpin]

41 minutes ago, tboooe said:

I am just using a short Blue Jean CAT 6a cable.  I am planning on trying a Ghent JSSG etherrnet cable soon though.

 

Just playing today with Supra Cat7+ and Supra Cat8 in the same setting. Think I like 7+ today.

 

Added: Not sure what JSSG adds to ethernet cable, but, do you know Miska's recommendations on ethernet jacks? My Supra Cat7+ have no shielding connected to jacks.

[LoryWiv]

12 hours ago, JTS said:

Sounds great. Scratch that, sound amazing. poly-sinc-xtr-mp>ADSM7EC>RME ADI-2 DACfs DSD direct>ZDTJ>HD650 (late night configuration)

 

 

 

 

Similar preference for me:  poly-sinc-xtr-mp-2s>ADSM7EC, runs with cpu <= 20% and sounds great at 128. Stutters on my i7-7700 4.2 Ghz however when I try 256. Perhaps next release post-optimizations, but I am pretty darn happy for now (sorry for the strong language 😜)

[Miska]

4 hours ago, StreamFidelity said:

Pink Faun writes, "Well, and most importantly, with Linux, we can reach an incredibly low latency (audio latency and processor latency). Because the latency is so low in the streamer 2.16 (audio latency at 192 kHz <2 , 67 ms on AMD Ryzen processor and high quality digital audio interface. Audio latency at 44.1 kHz <1 ms depending on hardware configuration." Looks Streamer 2.16

 

I believe it and I'm not far from that. In my system (Windows 10 Pro) I make sure to keep the latencies as low as possible. I set in the Asio driver the lowest value with 128 samples = 1.5 ms.

 

Lowering latency increases overhead and CPU load. On Linux I use 100 ms buffer (latency) and it has been generally best. It is also important to keep it high enough that it doesn't enter into audio range.

 

1 ms buffer switching period means 1 kHz noise. 100 ms buffer switching period means 10 Hz noise.

 

USB Audio Class has 125 µs packet interval and that's where the 8 kHz ticking noise comes that you can see in output of some DACs.

 

P.S. Technically it shouldn't be much problem to reach that 125 µs latency on Linux if you want to, I got that kind of figures already 10+ years ago on old VIA CPU.

[tboooe]

17 minutes ago, Miska said:

It is also important to keep it high enough that it doesn't enter into audio range.

 

1 ms buffer switching period means 1 kHz noise. 100 ms buffer switching period means 10 Hz noise.

This is a fascinating comment.  I thought it was best to keep the buffer as small as possible.  But what you are saying is that we need to actually increase the buffer to at least 100ms???

[Miska]

4 minutes ago, tboooe said:

This is a fascinating comment.  I thought it was best to keep the buffer as small as possible.  But what you are saying is that we need to actually increase the buffer to at least 100ms???

 

My recommendation is 100ms. This keeps activity period outside of audio frequencies and doesn't have much overhead.

 

For NAA backend, leave it at "Default" unless you have a good reason to do otherwise. With ASIO, leave the setting at "Default" in HQPlayer and adjust the buffer size from the ASIO driver control panel. At least something that would correspond to 100 ms, or otherwise as large as possible (taking into account the sampling rate you are going to use). With CoreAudio, macOS handles all this business for you...

 

[tboooe]

6 minutes ago, Miska said:

 

My recommendation is 100ms. This keeps activity period outside of audio frequencies and doesn't have much overhead.

 

For NAA backend, leave it at "Default" unless you have a good reason to do otherwise. With ASIO, leave the setting at "Default" in HQPlayer and adjust the buffer size from the ASIO driver control panel. At least something that would correspond to 100 ms, or otherwise as large as possible (taking into account the sampling rate you are going to use). With CoreAudio, macOS handles all this business for you...

 

Ok so if I am understanding correctly I should leave buffer as default in HQP but adjust the buffer setting in my USB driver control panel that is running on my NAA pc?

[Miska]

1 minute ago, tboooe said:

Ok so if I am understanding correctly I should leave buffer as default in HQP but adjust the buffer setting in my USB driver control panel that is running on my NAA pc?

 

Yes, if that is a Windows machine using ASIO driver. If it is something else, you don't need to do anything.

 

[tboooe]

25 minutes ago, Miska said:

 

Yes, if that is a Windows machine using ASIO driver. If it is something else, you don't need to do anything.

 

Thank you...I just checked an my ASIO USB driver buffer setting maxes out at 32,768 samples.  Since I am converting to DSD512, that is no where near the 100ms buffer setting you suggest.  Should I instead use the HQP buffer settings instead?  If I do use the HQP buffer settings what should I set the ASIO buffer to or does it matter since the HQP buffer setting will override anyway?

[tboooe]

1 hour ago, tboooe said:

Thank you...I just checked an my ASIO USB driver buffer setting maxes out at 32,768 samples.  Since I am converting to DSD512, that is no where near the 100ms buffer setting you suggest.  Should I instead use the HQP buffer settings instead?  If I do use the HQP buffer settings what should I set the ASIO buffer to or does it matter since the HQP buffer setting will override anyway?

After messing around with the buffer settings in my Singxer USB control panel, I notice that when I set the buffer settings to 32,768 samples and play music at DSD512, the latency information shown by the control panel says that the input latency is approx 50ms (~1 million samples).  Is 50ms in the audible range?

[tailspn]

If 100ms is 10Hz, then 50ms is 20HZ.

[LoryWiv]

 

These 2 settings in my Singxer SU-1 produce equivalent CPU / GPU utilization. Is one sonically preferred to the other?