[shwnwllms]

Thanks so much, very useful info!

I should add you’ll need a computer powerful enough to do DSD512 with your filter / modulator settings. You can always back it down to DSD256 if you get stuttering etc

Depending on filter choice you’ll need a powerful PC to reach DSD512

 

[tdx]

I should add you’ll need a computer powerful enough to do DSD512 with your filter / modulator settings. You can always back it down to DSD256 if you get stuttering etc

Depending on filter choice you’ll need a powerful PC to reach DSD512

I have a 13900K/4090 PC so no prob on the power side
However i found out after playing around with filters that while my computer has no issues with almost any filter when converting 44.1 PCM to DSD, high-res files (24 bit, 192khz etc) make it choke when using any Sinc based filter for PCM->DSD. So right now I'm using poly-sinc filters which seem to do just fine even with hi-res PCM sources. I'll keep playing around with settings to see if i can get the computer to use with sinc filters with hi-res files, since i quite like the sound of those filters...

 

[shwnwllms]

I have a 13900K/4090 PC so no prob on the power side
However i found out after playing around with filters that while my computer has no issues with almost any filter when converting 44.1 PCM to DSD, high-res files (24 bit, 192khz etc) make it choke when using any Sinc based filter for PCM->DSD. So right now I'm using poly-sinc filters which seem to do just fine even with hi-res PCM sources. I'll keep playing around with settings to see if i can get the computer to use with sinc filters with hi-res files, since i quite like the sound of those filters...

13900K should be able to handle most filter combinations even without CUDA offload. Which modulator are you using?

Extremely long filters such with EC (Extended Compensation) modulators will choke at hi res. Be sure to watch your apodising counter as filters like Sinc-L since are non-apodising so you’ll want Sinc-Mx or Sinc-Mga etc which are shorter but have the benefit of being apodising on poorly mastered material

 

[ra990]

I bought an SMSL D400EX, which theoretically supports Direct DSD mode at 512 rates. Unfortunately, I'm getting a strong hiss in the background of the music at DSD 512 rates. At DSD 256 rates, the hiss is still there in the background, but faint. I can't seem to find any combination of options that gives me the clarity of PCM with no background hiss. Disabling Direct DSD mode on the D400EX fixes the issue, but I believe that signal gets converted to PCM internally when direct DSD is disabled, so no point in that. Might as well just do PCM upsampling.

Has anyone here successfully used the SMSL D400EX with HQPlayer in direct DSD mode?

EDIT: Looks like the hiss is significantly reduced if I select Sound Color 3 or Sound Color 4, both of which run the chip at 11M vs Sound Color 1 and 2, which run the chip at 5.6M. Weird, but glad I found it. Still not totally clean though.

EDIT2: OK, finally found that with Sound Color 3 and 4, DSD rate 22.5792 MHz, I can't tell the difference with Direct DSD Mode (Fixed DSD Bypass Pre Mode) on or off in terms of background noise. Using 44.1k x 512 bit rate /limit, 48k DSD option unchecked, with ASDM7EC-super 512+fs, poly-sinc-gauss-long for 1x and poly-sinc-gauss-hires-lp for Nx. Sounds fantastic.

 

[shwnwllms]

I bought an SMSL D400EX, which theoretically supports Direct DSD mode at 512 rates. Unfortunately, I'm getting a strong hiss in the background of the music at DSD 512 rates. At DSD 256 rates, the hiss is still there in the background, but faint. I can't seem to find any combination of options that gives me the clarity of PCM with no background hiss. Disabling Direct DSD mode on the D400EX fixes the issue, but I believe that signal gets converted to PCM internally when direct DSD is disabled, so no point in that. Might as well just do PCM upsampling.

Has anyone here successfully used the SMSL D400EX with HQPlayer in direct DSD mode?

EDIT: Looks like the hiss is significantly reduced if I select Sound Color 3 or Sound Color 4, both of which run the chip at 11M vs Sound Color 1 and 2, which run the chip at 5.6M. Weird, but glad I found it. Still not totally clean though.

EDIT2: OK, finally found that with Sound Color 3 and 4, DSD rate 22.5792 MHz, I can't tell the difference with Direct DSD Mode (Fixed DSD Bypass Pre Mode) on or off in terms of background noise. Using 44.1k x 512 bit rate /limit, 48k DSD option unchecked, with ASDM7EC-super 512+fs, poly-sinc-gauss-long for 1x and poly-sinc-gauss-hires-lp for Nx. Sounds fantastic.

You definitely want to use the DSD Direct mode. You shouldn’t choose any DSP on the DAC if you are using HQPlayer

 

[KrauserX91]

Is there any filter combination which is good to focus on vocals?

 

[sup27606]

Is there any filter combination which is good to focus on vocals?

I think, all the higher end filters with high number of ‘taps’ such as sinc-l, sinc-m, sinc-mga etc lead to improved texture and imaging, characteristics that ultimately benefit the realism from vocals. I do think, filters like sinc-mga and sinc-medium slightly improve the bass impact (benefits lower pitch male voices) compared to sinc-long, which is more focused towards treble extension and classical music.

The filters tend to synnerize with the headphones also to some extent. I felt, the hd800 sounded awesome with sinc-long when playing classical music, but sounded a bit flat with jazz and rock music with sinc-long. Changing to sinc-medium sounded better with jazz. However, LCD-3 wasn’t that filter-picky, and sounded fine with both sinc-long and sinc-medium for any musical genres. So, a bit of trial and error testing is recommended.

 

[Barnabas 224]

Hello, long time reader, first time writer.

What type of dither are you using with those new filters? (PCM output, obviously)
Still the popular LNS15?

 

[sup27606]

.

 

[sup27606]

Hello, long time reader, first time writer.

What type of dither are you using with those new filters? (PCM output, obviously)
Still the popular LNS15?

For me, LNS15 it is.

 

[b0bb]

I have a 13900K/4090 PC so no prob on the power side
However i found out after playing around with filters that while my computer has no issues with almost any filter when converting 44.1 PCM to DSD, high-res files (24 bit, 192khz etc) make it choke when using any Sinc based filter for PCM->DSD. So right now I'm using poly-sinc filters which seem to do just fine even with hi-res PCM sources. I'll keep playing around with settings to see if i can get the computer to use with sinc filters with hi-res files, since i quite like the sound of those filters...

If your DAC can handle DSD based on 48kHz sampling rates, set the bit rate limit to 48k x512 in the SDM tab.
In the Outputs tab, tick the Adaptive rate box.

Steps above for Windows HQPlayer5
Try the above and see if it works converting 192k PCM to DSD512-48k without choking

 

[b0bb]

This is HQP5 SincMG converting PCM@192kHz to DSD512x48
SincMG is a good compromise between adequate filter length and not having excessive loading delays

'

Load trace, RTX3080ti GPU load 17%

 

[b0bb]

This is HQP5's notorious Sinc-L converting PCM@192kHz to DSD512x48
Excessive loading delays, up to several mins depending on sampling rate

Load trace, RTX3080ti GPU load 17%

Ryzen 7950X3D CPU is relatively unloaded.
HQP pins the 2 main threads to separate CCDs

 

[sup27606]

I have now tested Src-dx and also a rpi4 with Hqplayer. I think, the benefit of Src-dx (or any DDC for that matter) is dependent on the usb implementation of the source. In my case, I hear no difference between direct usb vs Src-dx dual bnc, where the source is a two year old MacBook Pro.

I also found out that for upscaling 16/44.1 kHz files to 24/768 kHz, rpi4 would not work with all the higher end filters such as sinc-M, sinc-Mga, etc. I even confirmed with the Hqplayer developer on audiophilestyle forum, that these filters would not work for 16x upsampling with rpi4. These filters do however work with 24/96 kHz and higher. Also, rpi4 can work as a great network audio adapter for Hqplayer running on a powerful pc nearby. I will therefore stick to the MacBook Pro for now.

 

[Rayon]

Just FYI, 13900K will require a decent graphics card for ASDM7ECV3 @ 1024 rates, at least in my setup it will not do it on the igpu alone.
Not an issue for me as I found I preferred AMSDM7EC 512+fs and hires lp filter without the need of external gpu.

Indeed GPU is needed and still I need to OC 13900k a little bit. And even then my cpu can only oversample to 44.1 family DSD1024 (no 48k). I've disabled auto family rating and oversampling cross-rate with the filter. Luckily my GPU can still oversample gauss-xl from 48k family to 44.1k family and gauss-xl support cross-family oversampling. When I listen to 48k based hi-res material, both my GPU and CPU are barely making it. It's kind of satisfying though, to suck the last drop of benefit from them. Not to mention the sound