Not your problem, you paid for it. Let them suffer 
(first world problems of course)
1 Like
I went through the test and I can hear a small difference, very subtle, I may hear maybe 16.5 or 17 at best who knows
To me personally, long linear-phase filters sound excessively harsh/bright. I cannot listen such for long time. So I think we are talking about similar thing.
Yes and no. Mathematically, RedBook has timing resolution of 346 nanoseconds. So it is not necessarily transient timing accuracy. But in my opinion it is about better preserving waveform shape of the attack. This has to do with Fourier uncertainty principle too, which humans have been shown to be able to exceed[1]. In order to preserve this shape, one also needs to preserve the high frequency harmonics that define this shape. These high frequencies are (IMO) not heard separately as high frequencies, but only detected as waveform shape.
This is a little bit like frame rate in video, you cannot see the individual frames of 30 or 60 fps video as still images, but you will certainly detect smoothness of the movements, especially faster ones. When you go for VR systems, you need even higher frame rates of 90 or 120 fps. Just like audio sampling parameters, some people will have hard time arguing what frame rate (and also resolution) is sufficient for videoā¦
6 Likes
Thatās a good analogy with the video example Jussi, and like you said the harmonics stacking on waveform creating that effect, I may add and some may disagree {including you as you have stated before if I remember correctly that your preference lies on solid state amplification} thatās this same effect might be accentuated with valve amplification {also perceived as distortion}
And as always your comments and opinions are extremely valuable thank you
Thanks Jussi. Cool to read this.
Quick A/Bāing is rarely useful for me (even though we all naturally do it sometimes). Like you hinted also, the better indicator of differences for me is over long listening sessions, over a longer periods (weeks, months) and the āfeelingā. Do I feel like I want to turn the volume up still after a long listening session or do I want to turn the volume down (listening fatigue). Does something not āfeelā right.
All of the poly-sinc family sound fantastic, so this is just finding out what my own ear and brain like the most and if they can adjust to your technically optimal filter if āforcedā over a long period of time.
I think Iāll stick with poly-sinc-short-mp-2s for another 4 weeks and let my ear/brain system properly burn in / adjust.
The beauty is I can always fall back to poly-sinc-mp-2s which gives me the slightly better frequency response if I still feel I need it.
But I do believe ear/brain burn in is real so will stick with your default for a good month.
Iām very polygamous with filters. Every time a nice good looking new one comes along i just have to flirt.
Currently listening to the new poly-sinc-ext2, itās a new CD so I cannot really comment except to say things are sounding good.
.sjb
2 Likes
Hilarious. Sorry, Iām tied & faithful to poly-sinc-short-mp-2sā¦ for a few more weeks anyway, then Iāll flirt with the new filters.
I watched this Rick Beato video about hearing with interest:
After trying a few other things Iāve gravitated back to xtr-2s-mp. It just sounds ābetterā to me, which may be some kind of confirmation bias and a different thing from accurate. But it keeps me listening to music, which is the main thing.
1 Like
Andrew: Unfortunately, the Rick Beato video repeats the mistake that so many make about the difference between low and hi-res files, namely that high-res files convey more information at higher frequencies and that our ability to benefit from them requires us to hear sounds at those frequencies. While there may be a tiny bit of truth to that for those with hearing out to 20kHz or beyond, the real difference ties more to what happens as those files are converted from digital to analog and the filters that are applied during this process. The whole reason that Jussi developed the capabilities of HQ Player was to allow the development of filters that worked with very high resolution files (even if those were just upconverted from 16/44) so as to make sure that any conversion artifacts stayed outside the hearing range. Every filter involves some tradeoffs and so different listeners will prefer different filters.
Rick also assumes that his test subject was able to āpreferā 66% of the 16/44 files was because of her ability to hear higher frequencies. But this too may be wrong, as the difference between 128, 320 and 16/44 isnāt frequency focused, but focused on removing lower amplitude data ā so it has more to do with her ability to distinguish soft from loud than frequency, and thus that difference may still be relevant to older listeners.
Where I do agree with Rick is that so many of the folks producing our music are not paying attention to these factors, in large part because they are producing a commodity rather than an audiophile product. But that doesnāt mean that someone who both preserved the overall mastering quality as well as audiophile detail levels couldnāt produce a result that was consistently preferred by audiophiles.
1 Like
A noobie question. If I am playing a file (I use Roon/HQP) and am not upconverting (say playing native DXD352 files) does my filter choice (I have specifiied -xtr) kick in? I donāt see any indication whether it is being used or not? I assume if I upconvert (say 192/24 to 384, or DSD64 to DSD512) that my filter choice will kick in.
Thanks, Larry
Yes, if it is selected, it will run even with 1:1 conversion ratio.
Jussi, thanks. Larry
Any Xtr filter users tried/trying the new-ish (September) ext2 filter?
Iāve been liking ext2 a lot lately.
Just tried it today. First impression - a bit crisper than xtr-2s. Probably a good thing in my system.
1 Like
Ext2 cured me of tinkering with filters for now. Iāve been using it almost exclusively since release. The high end extension just sounds right to me. Relaxed and natural but not attenuated.
1 Like
Same impression from me.
Yes! Me too. I donāt have the itch to fiddle after using ext2.
For those of you playing at home:
Linear phase polyphase sinc filter with sharp cut-off and high stop-band attenuation for extended frequency response while completely cutting off by Nyquist frequency (non-halfband). Processing is two stages with minimum factor of 16 before applying special second stage. If difference between source and output rates is less than 32x, operates as a single stage with only the first stage.
So for RBCD, itās 1st stage up-sampling is to PCM705/768kHz, compared with the other poly-sinc filters which do 1st stage up-sampling to PCM353/384kHz (for RBCD).
Better than 240dB stop-band attenuation I believe, if anyone is interested in the stop-band attenuation numbers wars
2 Likes
Iām always late.
I recently saw this post by Jussi, which got my attention to try it. Iāve always been a sucker for jumping on (stop) bandwagons
1 Like
I have a sixth sense for HQPlayer updates
2 Likes
Hilarious 
1 Like
@lorin @andybob @jussi_laako - am I going bonkers (donāt answer that) or is there better ādepthā (or perception of depth) with ext2 also?
1 Like