On Sun, 2010-11-07 at 03:37 -0800, Patrick Shirkey wrote: > It think it is a bit more subtle. I have listened to the mix on several > systems now and have found that the overall levels are not exactly what I > am hearing when I listen with ardour. That could be partly related to the > exporter, my sound card which I am mixing on being a high quality audio > device and the general levels being just not quite right. I am working on > toning down the tracks that have become distorted in the export process. The subtlety of the differences is < -120dB. This is what the test just proved beyond any doubt (if in doubt, read DSP basics). The difference is inaudible and is created only by the imperfections of floating point calculations (you always loose some precision when operating on floating point numbers). The difference you hear is caused by your playback system (maybe the soundcard?) distorting at higher playback levels. Or, if you are monitoring the two mixes at different overall volumes, the difference can be attributed at least in part to how the human hearing is non-linear: http://en.wikipedia.org/wiki/Equal-loudness_contours You might even be imagining the differences. Don't be offended, it's really easy to hear differences between two things when you're looking for (expecting to hear) a difference. This is has been researched and documented countless times. We all suffer from that, nobody is exempt. In your shoes, I'd measure the playback system and try to pinpoint the part which is causing the distortion. If you tweak your mix to sound good with the distortion your current system exhibits, the mix will not sound like that anywhere else. So to make your mix sound good everywhere, you have to have a clean setup as a reference. (Note the word measure: this is the only way to get any sort of objective results. ) Sampo _______________________________________________ Linux-audio-user mailing list Linux-audio-user@xxxxxxxxxxxxxxxxxxxx http://lists.linuxaudio.org/listinfo/linux-audio-user
