Disclaimer: I'm no expert at this by any means! However--
One of the problems frequently encountered with normalizing the audio to a peak level of 0 dBfs (judging from the number of questions about this subject on the Internet) is that it offers no headroom for transcoding into other formats (such as from WAV to MP3).
Many people have experienced this problem when uploading peak-normalized WAV or FLAC files to SoundCloud, for example, because when SoundCloud's servers transcode the file to lo-fi MP3 for streaming purposes, distortion, dropouts, artifacts become common, annoying and sometimes extreme. To prevent this, SoundCloud (and other music-hosting sites) often recommend that audio files have their volumes normalized to at most -1 or -2 dB below peak (below 0 dB), so that the files can be encoded/transcoded safely with waveform headroom to spare and without those artifacts and problems.
In Ardour, I personally always normalize each track individually to -2 or -2.5 dB for mixing or saving purposes, but defeat normalization (specify that it not be used) during export of the audio mix (master output) or export of any individual tracks using Export--> Stem Export.
_______________________________________________ Linux-audio-user mailing list Linux-audio-user@xxxxxxxxxxxxxxxxxxxx http://lists.linuxaudio.org/listinfo/linux-audio-user
