[pulseaudio-discuss] Audiophile newbie reporting in

[Michael Cronenworth]

Hello all,

I recently purchased a ASUS Xonar Essense STX card and I'm interested in 
tweaking pulseaudio to get the most sine wave for my buck. For those of 
you that may or may not know, the card above has a hardware chip for up 
to 24-bit/192kHz processing. I plan on using the card for music 
listening and single source recording (think vinyl records). No audio 
production here so low-latency throughput doesn't matter to me.

Google searching for a few days has driven me to this mailing list as I 
cannot find any straight answer. The PA wiki doesn't seem to have an 
audiophile area.

Background info:
Most of my music are 16-bit/44.1kHz recordings. I am expanding more and 
more of it to include 24-bit/96kHz. I have a fast, quad-core CPU so CPU 
usage will not be a problem if I need to use a resampler that is 
computationally expensive.

My limited knowledge of audio properties has made me ask myself three 
things:

1) Is Pulseaudio even the right man for the job? A friend of mine wants 
to do away with PA and use Jack. He's only using it for music listening 
(no recording) so this seems like a bad idea. Is it? I still need to be 
able to have sounds come from non-music player apps. Example: wine for 
Starcraft 2 ;)

2) default-sample-format
     Is it beneficial to set this to s32le? Any negative effects for 
16-bit tracks?

3) default-sample-rate
    Is it beneficial to set this to 96000? I realize anything beyond 
this is not able to be comprehended by our brains so I don't see a need 
to flirt with the 192kHz top range of the card. However, the same friend 
from above feels that upsampling constantly "is a bad thing" and should 
be avoided because of going from 44.1k to 96k "averages samples". Is 
this right? How else will I hear my 96k recordings if I leave PA on 44.1k?

Any other tips or comments are appreciated.

Thanks,
Michael