On Tue, May 10, 2011 at 01:28:04PM +0000, Fons Adriaensen wrote : > On Tue, May 10, 2011 at 03:04:22PM +0200, Aurelien wrote: > . > > Then, I would have one question: why can I come to different results for > > several exports of the same session? > > I mean, I already had cases in which my master finally end up to +1.0 > > dB, and the export after to -0.1dB (without changing anything), and so on. > > I'm not speaking of big differences, but enough to make it hard to > > optimize my dynamic before mastering. > > I really thought that it was linked to realtime processing, but if not, > > I'd be interested in understanding what is behind all of this. > > If you export to an integer file format anything above 0dB will > be clipped. Yeah, for sure. > > OTOH, if the export includes resampling then the peaks in the > exported file will be higher than those at the output of the > master strip. The difference will be small in most cases, but > in theory it is unbounded. Better keep a dB or so of headroom. For the cases I spoke about, I didn't. But, by the way, I was speaking about the master vu-meter, which can be observed during the export, and which is -for me- the point to which I pay attention when exporting. This master can give a range of values for max peak different for several exports of the very same session. I've to say I work in studio, and use a lot of LADSPA/LV2, and work with a 75-80% DSP load. Don't know if that influe, but... > > Ciao, > > -- > FA > > > > > > > _______________________________________________ > Linux-audio-user mailing list > Linux-audio-user@xxxxxxxxxxxxxxxxxxxx > http://lists.linuxaudio.org/listinfo/linux-audio-user -- Aurélien _______________________________________________ Linux-audio-user mailing list Linux-audio-user@xxxxxxxxxxxxxxxxxxxx http://lists.linuxaudio.org/listinfo/linux-audio-user
