Having pursued the best perceived transient response, I have now stayed a while with ASDM7EC+512fs-fast, but only at max DSD256 DoP to my RME DAC. The filter is minphaseFIR, but due to CPU stuttering, I cannot use IIR2, it is a too hard strain on CPU, while the minphaseFIR seem a lot less CPU intense, perhaps more GPU support. They sound a little alike in my rig. I have ticked the 20kHz filter, too. It is not much of difference, though. Even my brother, who focus on rhythm and beats now think my rig is very pleasant to audition. With the latest USB optic galvanically decoupling the DAC from my NAA, it now also plays exceptionally clean from any harshness or sibilance, even at really high output volume. I can enjoy a number of almost criminally poor remasters of 80th rock music from Qobuz, without ripping my ears of and without sacrificing the sought after, immediate, lightning fast, transient response, or the resolution I desire. Finally I managed to repair the big mistake I made selling my Matrix Audio Element H PCIE USB card. Puhhh ā¦ Interesting to read the comments about really high DSD āspeedsā in comparison to the āoldā DSD256ā¦ I have auditioned other DACs at DSD512, but I am a bit hesitate to say that it is better.
I am using HQPlayer with my Gustard X26 Pro; I have just added the Intona 7055-C USB isolator into my system (using USB 2.0 cables at the moment) and things are not working well.
It wonāt play in the DSD Direct mode for pre-upsampled files (does so no problem when my PC is connected directly to the DAC).
It only goes up to 3.072M even when I select DSD512.
It only goes up to 96kHZ for PCM.
Changing the Output setting to either of the Network Audio Adapter settings does not work, as in no processing takes place.
As this device is plug-and-play, I donāt think that there are any settings for the Intona itself. Am I doing something wrong? Is there a setting on the HQPlayer that requires to be changed? Or is it the USB 2.0 cables that are slowing things down (but theyāve been perfectly fine for sending DSD512 directly to my DAC so far).
Thanks
EDIT - it seems like the issue is not limited to the HQP, Windows mixer only offers up to 96kHz upsampling also when the Intona USB isolator is in my chain. Strange. Might need to try different cables.
With 7055-C, always connect it to USB3.x port on your computer, using a proper USB3 SuperSpeed Certified cable (look for the official certification badge).
Do not connect anything to the AUX power input on the Intona.
If your DAC has USB2 interface, make use you use a USB2 HiSpeed Certified cable (again, check for the official certification badge).
P.S. If you use Roon + HQPlayer, did you remember to check that the DAC is disabled under Roon Settings / Audio?
Iāve used USB 3.0 ports on my PC but a USB 2.0 cable. On paper, with the right handshake between the devices, this should give me speeds up to 480MBps. Iāve in the meantime found a short manual for the device and judging by the blinking speed when the device is connected, it only reaches āfull speedā, not the āhigh speedā mode. These names may be a bit confusing but it seems that full speed is only 12mbps, which would probably explain the sample rate limitation. Iāve tried two different cables, one of them a Supra - same result.
Iāll get a hold of a decent USB 3.0 cable and hopefully this will solve the issue.
I have on this occasion used FLAC files directly from my PC. I initially thought it was an HQPlayer issue, but later realised that Windows also automatically limited the available sample rates, so there was clearly a handshake issue.
Yeah, but there are other aspects as well. USB2 can provide max 500 mA current, while USB3 can provide max 900 mA current. As an exampleā¦ USB3 B-cable has that extra bulge because it runs separately isolated power wires. It depends to some extent on the computer USB port implementation how agressive peak current limiting it has, so USB2 port could cause the 7055-C + DAC combination from not properly powering up if it cuts a bit too aggressively at or before 500 mA.
It is not an issue to use USB2 from the 7055-C to the device, but I would strongly recommend a proper USB3 SS for the host side.
Full Speed is USB1, thatās why you are limited to 96/24 PCM in non-asynchronous mode which is maximum UAC1 can do.
It could also have something to do with your DAC. But Iām using 7055-C for example between my development machine and T+A HA200. USB3.2 connected to the 7055-C using a proper big blue USB3 SuperSpeed cable. And HA200 connected using a regular USB2 HiSpeed cable.
I almost sent the device back, even changing the cable did not help. Then I suddenly remembered that there had been a firmware update for my DAC years ago which I had never downloaded. I installed it and voila - everything is working fine, regardless of which cable I am using (USB 2.0 or 3.0). So it seems that my DACās USB protocol was for some reason triggering a āsafe modeā or something of that kind.
Initially when I used the 3.0 cable for the first time, the blinking speed increased which I interpreted as the device entering its high-speed mode. But it made no difference to the max sample rate. Now the blinking is extremely fast, so I had wrongly identified the initial blinking speed - it wasnāt even in āfast modeā but the slowest mode possible.
The device has made a very significant difference in my system, much bigger than expected. Highly recommended, especially when one is upsampling and paying attention to small details (although itās also very apparent on the macroscale).
Hi everyone, Iām using a Philips TDA1543-based DAC without oversampling and output filtering, which can handle up to 16/44.1 material at most. Does it make sense to use HQPlayer with it, and if yes, what settings?
Hello everyone.
Iām using an old ESI Juli@ audio card.
Does it make sense for me to listen to music in HQPlayer?
If so, which filters would be more preferable for my audio card?
You can try and see what you think about the result, you can set HQPlayer to output fixed 192k rate, 24-bit TPDF or Gauss1 dither. The default filters are a good general purpose choice I recommend to get started with.
In the same line of idea, in a recent Stereophile review of Hifi Rose dac (defeatable oversampling btw) Herb Reichert once again advocates against upsampling but he uses reconstruction filter. For whatever reason it does not work in hqp with fir or polynomial but plays fine with recommended filters or personnal favorite psg xtr short (mp or lp, depending). How bad is it objectively with a Holo May ? Subjectively I confess it.sounds nice to me, especially with lowest res (16/44.1). Strangelly (or not?} with modern hires 256 sdm ouput is a better contender. I also found more realism in sacd rip played without upsampling but using asdm7ec fast
Any DAC without oversampling is objectively horrible at 44.1/48k rates. Thereās no way around that. Plus you get about -3 dB roll-off between 10 and 20 kHz.
Only reconstruction filter you can have in such case is the analog post-filter in the DAC. And for such rates it is impossible to do it properly and good quality with analog filter. With 44.1 kHz you have 2.05 kHz worth of bandwidth to reach -96 dB attenuation (for 16-bit source), or -144 dB attenuation (for 24-bit source), if you want proper reconstruction. Hence you need oversampling.
poly-sinc filters can still convert for example from 44.1k to 48k or from 48k to 44.1k. And also within same rate. And do apodizing corrections. Something like polynomial would do absolutely nothing at all in 1:1 conversion, so it is pointless there.
I also donāt recommend playing anything out at DSD64 rates, it is just not good enough for D/A conversion. Not as bad as playing 44.1/48k PCM without oversampling, but more like playing 88.2/96k without oversampling, but of course without the correlated images (distortion) that there PCM rates produce.
I generated a 1000 Hz test tone in Reaper, played it through HQPlayer with various filters, and analyzed the results in REW. The best combination for my old ESI Juli@ turned out to be poly_sinc_xtr_mp + Gauss1.
Maybe Iām wrong or just fooling myself, but to my ears this combination also sounds better.
It is probably worth trying other filters for different musical genres.
@jussi_laako I am wondering if it would be at all technically possible to create a non-apodising version of the Sinc-Mx filter?
Sinc-L and Sinc-Mx are my go-to filters for streaming music; Sinc-L is more transparent but Sinc-Mx has more weight/tactility and depending on the individual track I often prefer the latter.
My personal preference is for non-apodising filters; even the tracks that contain apodising errors usually sound better to me with a non-apodising filter (e.g. when comparing the two versions of the Sinc-Mg filter). Itās good to have a choice of the two versions as some people will prefer the other one.
I have tried the non-apodising version of the Mg filter but I still prefer the tactiliy of the Mx. It would be great if it had a non-apodising sibling.
