Hello all, for the past few days I have been ripping vinyl into digital format, mainly MP3 and FLAC. What I usually do when I finish recording is to normalize the waveform and then amplify it to the maximum value Audacity allows me (no clipping). Then I convert the WAV file into FLAC (24bits/96khz) or MP3 (V0). I have noticed though that the resulting .flac/.mp3 file sounds softer and quieter than the original WAV file. While the WAV file from within Audacity sounds quite loud, when I play the .flac file with VLC it sounds softer. Is this normal or is there something I am missing (maybe a setting somewhere)?.
I am using Audacity 2.0.3 with ASIO support and Windows 7.
As a general rule, conversion between formats, even compressed ones does not change the volume of a performance. What can change is what the programmer was thinking about when he wrote the decoder/player. Sometimes, your computer will get it in its craw to apply different settings depending on who’s software is being used.
Just for one example, it’s possible VLC Player is following the video pathways in the computer not the audio ones.
Windows machines can also have Special Effects applied without you knowing. The reason your shows sound like a concert hall is Windows is applying the Concert Hall effect to everything.
I’m pretty sure if you export to WAV, FLAC, and MP3, they will all have the same volume when played in VLC. Audacity has a volume control and VLC has a volume control, and they may not match. FLAC is lossless compression, so when it’s decompressed you get the original identical PCM data back. It can’t be quieter or louder because it’s the same identical data.
You can also open the FLAC and/or MP3 file in Audacity and run Amplify again to check the peaks. Amplify will default to 0dB (no change) if the peaks are still normalized to 0dBFS.
Typically, the lossy MP3 compression results in some higher peaks and some lower peaks (without changing the “loudness”). This means you’ll probably see “clipping” when you open an MP3 encoded from a 0dB normalized file. (The MP3 isn’t really clipped… MP3 isn’t hard-limited to 0dB like regular (integer) WAV files.)
Hello both, thank you for your prompt reply. I tried two ways:
1st:
I set the volume in audacity and VLC to the same (100).
I listened to the amplified export to both audacity and vlc, audacity sounds louder.
2nd:
I export a file from Audacity in the volume it was ripped.
I normalise, then amplify to suggested settings.
I export the file again.
I listen to both files in VLC, they sound the same.
But your posts gave me ideas. As you said VLC might use a different sounding device (I have ASIO set to audacity vlc might use windows direct sound) or a different algorithm. I will try to amplify the song to something way louder (I will allow clipping) and see if this makes a difference.