Wombat;695169 Wrote: > I did indeed some testing :) > I ended up with this setting (using it with frontah): > > sox.exe input.wav -b 16 output.wav rate -b 92 -a -v 44100 dither -a -f > low-shibata > > I give up audio content from above 20286Hz with this. > With this setting aliasing happens but only above 20800Hz at a low > level because the response is already pretty down there my testfiles > show. Since my Transporter cuts from 20600 @44.1kHz it has no chance to > be a problem. > One advantage of this early filter is there is nearly no pre-echo but > that was no criteria for me. You may have seen the pics i linked > earlier that show the pre- and post-echoes of different approaches. > This setting also gives me much less clipping as some ultra-steep > attempts especialy by other vendors. Since clipping is a real problem > that i stumbled across all the time this is very welcome. > The sox dither options are well done also. Sox offers different noise > shaped forms. I tried all by listening them insane loud and decided for > the low Shibata. It has similar added noise i hear well balanced between > some HF noise and lower frequency stuff. The Gesemann for example > doesn´t fit me at all. > Really strong high-pitched noise i hear with that. > You may wonder why iZotope is so hyped. The noise it adds is very much > low-frequency and sounds most comfortable at first but i don´t like > it. > There are much stronger shaped noise-shapes like High-Shibata that > really seem to allow even lower noise-floors. But with these i am a bit > cautious. The only time i ever was able to abx a downsampled 24/96 file > was with some strong shaped dither. The thread is on Hydrogenaudio > somewhere. I didn´t have the originals to do my own tests and used the > given samples so this may not be exatly valid.
Interesting , to connect it back to the topic we stablished that a signal based on 16/48 or 44.1 migth indeed encode all we can hear this all good . But is their any way of actually getting there ? " yes " apparently some resamplers fit this description others don't . Is there any " gold standard " abx'ed to be inaudible that is well tested ? As you described some sound engineers may choose a resampling algorithm that is not transparent but give some desired coloration in the process ? This is a slight contradiction why not choose the one algorithm that gives the least amount of difference especially as the goal is to make it inaudible , why tweak something your are not supposed to hear ? Is this some kind of working method from when resampling was not so well developed . To reach the objective do a sound engineer chose different approaches for different projects ? Regarding the resampling . Starts to wonder if many older releases are compromised by the use of older less optimal solutions -- Mnyb -------------------------------------------------------------------- Main hifi: Touch + CIA PS +MeridianG68J MeridianHD621 MeridianG98DH 2 x MeridianDSP5200 MeridianDSP5200HC 2 xMeridianDSP3100 +Rel Stadium 3 sub. Bedroom/Office: Boom Kitchen: Touch + powered Fostex PM0.4 Misc use: Radio (with battery) iPad1 with iPengHD & SqueezePad (in storage SB3, reciever ,controller ) http://people.xiph.org/~xiphmont/demo/neil-young.html ------------------------------------------------------------------------ Mnyb's Profile: http://forums.slimdevices.com/member.php?userid=4143 View this thread: http://forums.slimdevices.com/showthread.php?t=93990
_______________________________________________ audiophiles mailing list [email protected] http://lists.slimdevices.com/mailman/listinfo/audiophiles
