Yes, pls PM to Alexey Galkin (HiddenPilot here). He is responsible for production and shipment. Just ask him to do all you need.
Vit,
HQPlayer on my machine Windows 10x64 is strange happening that are all settings parameters at defaults and auto, such sampling rate, filter, dither etc... . I tried to setting to other parameters but it still showing at defaults and auto. I tried to reinstall it but still get the same results. Parameters bar are showed default and auto at gray color, unable to change it. Any advise.
Thanks
HQPlayer on my machine Windows 10x64 is strange happening that are all settings parameters at defaults and auto, such sampling rate, filter, dither etc... . I tried to setting to other parameters but it still showing at defaults and auto. I tried to reinstall it but still get the same results. Parameters bar are showed default and auto at gray color, unable to change it. Any advise.
Thanks
Hi Vit,
I tried other solutions, such as RPi2 using Signalyst NAA 3.52x and my i5 laptop (using battery power only) and also running the x86_64 version of Sygnalyst NAA 3.54x image.
With these, both show up under HQ Player in IPv4 and IPv6.
I setup my HQ Player on Windows 2016 dedicated server but also installed on a windows 10 computer for diagnostics. I'm getting terrible stutter with both NAA. I know RPi2 is weak but my HP Core i5 has plenty of juice for NAA. So I turned to the HQ Player server (set windows to optimize application and not process and vice versa). HQ Player is set to use multicore DSP, don't have CUDA GPU in it. It is also using the same settings as yours. I'm using DSD 256, your settings are proper for this or optimized for 512?
When playing a PCM 44.1 file that is converted to DSD256 by HQ Player, I see 75% CPU utilization max (global all cores) and going down and up. But it stutters like crazy. Keep in ming that this is using my AK4490 DAC for now. I'm still working on the DSC 2.5.2 for now.
Any hints you may have?
Thanks
Do
I tried other solutions, such as RPi2 using Signalyst NAA 3.52x and my i5 laptop (using battery power only) and also running the x86_64 version of Sygnalyst NAA 3.54x image.
With these, both show up under HQ Player in IPv4 and IPv6.
I setup my HQ Player on Windows 2016 dedicated server but also installed on a windows 10 computer for diagnostics. I'm getting terrible stutter with both NAA. I know RPi2 is weak but my HP Core i5 has plenty of juice for NAA. So I turned to the HQ Player server (set windows to optimize application and not process and vice versa). HQ Player is set to use multicore DSP, don't have CUDA GPU in it. It is also using the same settings as yours. I'm using DSD 256, your settings are proper for this or optimized for 512?
When playing a PCM 44.1 file that is converted to DSD256 by HQ Player, I see 75% CPU utilization max (global all cores) and going down and up. But it stutters like crazy. Keep in ming that this is using my AK4490 DAC for now. I'm still working on the DSC 2.5.2 for now.
Any hints you may have?
Thanks
Do
1. Dont use poly-sinc-xtr filter at all, it's very computational power hungry. Start with poly-sinc-shrt-2s - it will work w/o slutter/drops. Then try other filters. Choose best sounding for you.
2. After experimenting with DSD256 you may try DSD512.
"Sawtooth" CPU load is ok for poly-sinc-xtr filter.
2. After experimenting with DSD256 you may try DSD512.
"Sawtooth" CPU load is ok for poly-sinc-xtr filter.
Small HOWTO for BeagleBone Black + iancanada McFIFO reclocker + DSC.
HOWTO - BBB+McFIFO.pdf - Google Drive
Photo of my build
IMG_1628.JPG - Google Drive
IMG_1627.JPG - Google Drive
IMG_1626.JPG - Google Drive
About pops:
1. I'am waiting for luchoh advice in configuring Botic kernel (/snd_soc_davinci_mcasp/parameters/amanero_mute_delay).
2. I'll put the video on YouTube so you can evaluate their loudness.
HOWTO - BBB+McFIFO.pdf - Google Drive
Photo of my build
IMG_1628.JPG - Google Drive
IMG_1627.JPG - Google Drive
IMG_1626.JPG - Google Drive
About pops:
1. I'am waiting for luchoh advice in configuring Botic kernel (/snd_soc_davinci_mcasp/parameters/amanero_mute_delay).
2. I'll put the video on YouTube so you can evaluate their loudness.
Last edited:
Link to pop recorded by iPhone 7. Slight pop is at the 2 second after start.
Recording distance is about 1 m from twitter.
Since automatic level control was used, at the beginning of recording volume level is ~2 times larger than the actual volume. So noise can be heard.
Recording distance is about 1 m from twitter.
Since automatic level control was used, at the beginning of recording volume level is ~2 times larger than the actual volume. So noise can be heard.
Hello very nice build, SQ description? what did you use previously to listen dsd?
thanks
gianluca
In the photo, I see that you have connected the secondary windings of the transformer in parallel. It is better to use only one. If the windings differ from each other at least by half the turn, this is an additional load on the transformer.
Do, my BBB is rev C. I think all will be Ok with all revisions.
Since rev C has 4GB eMMc (instead of 2GB), I'd recommend to use SD with older revisions.
Is the pop appearing at the beginning of every track or one for the whole album?
Ok cool!
I'm thinking of getting a Nvidia Quadro P400 for my server just so I can offload to CUDA. It is not too expensive and should help..? Unless not enough Cuda cores in it...
Thanks
Do
Probably a tesla k20 or at least m2090 which has good double precision performance,
Hihi you so funny!
If only I had this kind of money for a CUDA accelerator... Oh well...
The P400 is actually for hq player
Do
I am serious
, if I am not mistake a P400 has only 48 cuda cores with no DP support which is less than any of newer entry level nvidia card like GT710 (192 cores)http://la.nvidia.com/content/PDF/product-comparison/Product_Comparison_Oct_2012.pdf
I am serious
http://la.nvidia.com/content/PDF/product-comparison/Product_Comparison_Oct_2012.pdf
This spec sheet is the old 400. The P400 has 256 CUDA cores and I believe 3x DP, but since it is a real Xeon server, I’m not connecting any display on it.
Do