I agree. I will try others occasionally, but never seem to stay with them very long.
1 Like
I changed dac and with the Gustard A26 (AKM4499EX & NAA inside), I revisit the filtersā¦ with the possibility to go up to DSD512 (but my pc canāt go beyond AMSDM7 512+fsā¦ no EC).
In these conditions, I find the ext2, still as valiant.
I was in poly gauss long, 256ECv2ā¦ but I find that the attack of the notes on the piano lacks a little punch.
My aging ears or for jazz accoustic, it is a good association (?)
1 Like
You changed from Pegasus? Do you have any feedback? (if is not too soon)
Thanks
It is difficult to answer at the moment. Itās justā¦ different and the Pegasus is excellent but the Gustrad replaces the Musician + usb cable + ZenStream. Itās more neutral, a less invasive bass, maybe widerā¦ and a little more intelligible at low levels.
But I would have to listen to the Pegasus again.
1 Like
Thanks for the reply. Pegasus is in the shortlist for the dac upgrade (hope this year). Iāll put the A26 in there.
Low levels? What was (is) your preferred way for volume control?
I use volume management via HQPlayer.
My dacs are connected directly to a BoXem purifi amp (SQM xlr).
I manage everything from Roon (or HQP client).
@jussi_laako : On my (new) Gustard A26, there are two settings for DSD :
Narrow Bandwidth (Default) or Wild Bandwidth.
Which one ?? with poly-sinc-gauss-long & 7ECv2 ? 
I use āDSD directā with these AK4499EX.
1 Like
Narrow is the recommended one. Although this setting only applies to DSD64, DSD128 and DSD256. At DSD512 this setting doesnāt have effect.
1 Like
Thanks Jussy.
For my understanding, the context is not the same but in the HQPlayer manual, the preference is for āwideā.
(I have just reread)
For example piano doesnāt contain high frequency harmonics and for
such case ānarrowā is suitable, while close miked percussions usually contain high
level high frequency content and there āwideā may be more suitable. While āXFiā is
suitable for all cases. Default is āwideā.
These are two different thingsā¦
At DSD256 the AKM chip narrow is 144 kHz and wide is 238 kHz. (without DSD Direct it is always 89 kHz for DSD256)
I consider 144 kHz best value.
3 Likes
And for (acoustic) Jazz, poly-sync-gauss & poly-sync-gauss-hires-mp are preferable to poly-sync-gauss-long & poly-sync-gauss-hires-lp ?
(filter ānarrowā and sound option āmonitoringā on the A26).
I think either set is fine and it is more up to personal preferences which one is good. As said in the manual, my estimate is that poly-sinc-gauss-short, poly-sinc-gauss, poly-sinc-gauss-long and poly-sinc-gauss-xl(a) all work with jazz. For more percussion rich tracks I would probably go with shorter side and with sax, clarinet, trumpet, etc rich tracks more to the longer side. Also depending on type of recording, IOW how much acoustics it has and whether it is multi-track or minimal mic direct to stereo.
3 Likes
Yes, and I have to find a fairly consensual setting because Iām not going to change for every album or worse, for every track depending on the playlist!
On the other hand, I still have filters that are just not playable (?).
I can do poly-sinc-gauss-hires-lp, mp, but if I choose ip. HQP doesnāt play (?)ā¦ nothing, and I have to switch to another filter.
Which version/build/OS are you on? I just tested on this MacBook Pro and it works fine for me. Maybe thereās some error in the log file when you try to play with -ip?
Iām running Win10 (i5-8400 AVX)ā¦ uhā¦ I just tried again from the āsettingsā menu, it works.
I had done it from the HQP client. Maybe a problem in the test sequence (?)
For now, Iāll leave the DSD512 = AMSDM7 512+fs and gauss / gauss-hires-ip for the Gustard A26.
Maybe a bit less crunchy than with an EC modulator but my server canāt provide it for DSD512. It might even be the sweet spot (?).
2 Likes
Have you tried Poly-sinc-ext2? Itās a very good cross genres all-rounder. (edit: I also like Poly-sinc-short-mp a lot, although I am not certain it is the best option for Jazz).
4 Likes
@jussi_laako Is there any chance that you will release an apodising version of the Sinc-L filter? Itās by far my favourite filter for clarity and vivid/punchy/ātactileā sound. Although I could switch to Sinc-S/M/M(x) when listening to songs containing lots of apodising errors (some go into hundreds/thousands!), none of them quite reach the level of Sinc-L in terms of the above-mentioned aspects of sound. It would be great to have an apodising version of Sinc-L, unless thereās a technical reason why this is not possible for this type of filter.
It is not technically possible for this filter.
Tangentially, why is asymFIR not available for SDM?
I included the ones that I considered to scale nicely for the higher conversion ratios. The ones I left out were originally developed for PCM output rates.
I understand that. I toggle between ext 2 and asymFIR, depending on what I am listening to on PCM, would just be curious to hear the comparison on SDM.