HQP Desktop 3.23 is out and has two new filters (you know you wanted more filters ).
Jussi describes them as closed form mega. The PCM version has 1M taps and the DSD version 16M.
Here’s some shots of my system using the DSD version:
I haven’t listened to them critically yet. Initial first impression is slightly “richer” than closed form.
For PCM705/768kHz, what’s the difference in CPU loading for you, between closed form and the new mega variant?
My CPU wouldn’t have a chance with DSD
Hardly any difference with PCM 705.6. Closed has CPU cores at 0-5% with minimal GPU (1-2 %). Closed Mega has CPU cores at 1-5% with similarly minimal GPU. Pretty sure I couldn’t pick between them blind.
But you can never have enough:
But I am looking forward to this coming to HQP Embedded, to try out.
Jussi continues to innovate. Impressive stuff.
Andy, are you using 256 out of preference? I thought you could do 512 on your system?
I have the max PCM and DSD rates set to 705.6 and 44.1x512 because something goes wrong in my system with 768/48x512. I think it’s in my DAC.
With poly-sinc filters source material in the 48k family gets output at 44.1x512, but the closed form group of filters are not poly-sinc so the output drops to the max supported rate for 48k.
Here’s some pix of 44.1k material at DSD 512:
What’s that windows performance monitor you are running?
Seems negligible impact moving to 16M
Left of the trace is closed from, right is 16M version.
I will preface this by saying I don’t really understand filters and can only go by what I hear, which could be way off base!
Closed-form mega is a very smooth sound. The body and presence of vocals sound natural.
It has been noted by many listeners that poly-sinc-xtr seems to attenuate high frequency (when I understand the measurements show the opposite? a greater HF extension.) I find in practice it actually enhances HF detail—the utter absence of edge reveals more what I have heard some call “inner detail.”
Closed-form shines, imho, in the midrange. Well-mastered recordings of small combos with vocals for example, are stellar. But I think this exceptional midrange quality can cause a perception of relative edginess in the top end. The recordings where it falls short are often very busy productions, especially modern masterings. The trademark impressively wide soundstage is still there, but it seems to a contiguous mass of sound rather than an image with distinct elements. With poly-sinc-xtr-lp there is more separation and thus clarity and detail.