Do some measurements see for example archimagos excellent blog I would
also recommend audidiffmaker to filter forth differences.

The key here is to filter out the first round of duds . If the
electrical output is unchanged the sound waves hitting your ear will be
exactly the same .

The remaining tweaks that did something did they improve the performance
measurable ? Or did you tweak it until it broke, I think this is very
common ,(see all geniuses making thier own transcoding settings for SoX
) . Different is very often interpreted as better . Closest to the
original source is always best , you can never know what you listen to

Round 3 do ABX listening test on the one with positive results , which
where audible .
This is key , these differences is so small that the bias from "sighted"
tests is invalid i.e. By sighted I mean knowing anything about what A vs
B is . The idea here is can you tell a difference .

No no one buys the argument that "differences are so obvious to me that
ABX is not needed" that's is actually a sure sign that you are biased !
Modern digital sources do not produce huge differences in to any modern
DAC unless something is broken .

Now a subjective evaluation can take place what the remaining audible
tweaks that also improved the measured performance sounded like .

A random walk in OS and ALSA settings would be very futile and a waste
of everyone's time including your own time :)

Historically, there has been no really successful attempts  to prove
that any digital source conected to a decent DAC makes a real
difference. If the input signal is bitperfect and without really gross
jitter (several 1000 times more than sources usually have ) and equally
gross electrical disturbances conected trough the wire every source is
the same .

However with the rise of computer audio there is potential for things
going wrong , but then its obviously wrong .
The OS mixer and resampler in windows should probably be avoided and
some other things to . But thats why we have squeezeboxes.

So what can really be done except have sane settings in the audio system

And I suppose PIcore players is equally turn key in the respect that you
get bit perfect output in respect to the input file if you want and
don't need to worry to much

Main hifi: Touch + CIA PS +MeridianG68J MeridianHD621 MeridianG98DH 2 x
MeridianDSP5200 MeridianDSP5200HC 2 xMeridianDSP3100 +Rel Stadium 3
Bedroom/Office: Boom
Kitchen: Touch + powered Fostex PM0.4
Misc use: Radio (with battery)
iPad1 with iPengHD & SqueezePad
(spares Touch, SB3, reciever ,controller )
server HP proliant micro server N36L with ClearOS Linux
Mnyb's Profile:
View this thread:

unix mailing list

Reply via email to