1

I am working on a research project related with wav files. I have wav files of different songs from different artist. I am going to extract some information/features from these audio files. Before extracting features, I thought it would be always better to have all the wav files in one uniform format/standard. For that I brought all my wav files into 44100Hz sampling rate and 16 bits bit depth. But I am stuck at one point. I cannot find out a way to bring all the wav files into one volume level.

When I listen to my songs, I hear that some files are in a higher level in terms of volume and some are low. So that I have to always adjust the volume once a new song(wav file) is started. I think those files have to be in a same volume level in order to be used in my project.

I searched about this and found iTunes for MAC has got the support for this as described here. But I am not sure that whether I can export wav files from iTunes or not into my required format.

I found MP3Gain is one other tool which is capable of this task. But it does not support mp3 files.

My question is can I bring my all sound files into one volume level? If it is possible, what are tools for that?

  • Without searching for this volume leveling software, can't you just apply a levelling compressor to the whole project(all songs, one file, right?)? – Marc W Sep 27 '15 at 3:31
  • Could you please explain me about that a bit. (I am actually new in this field). I need to control the loudness, Will a leveling compressor be suitable for my case? I do not need to get the parts affected which are in different volume levels according to the original song. – vigamage Sep 27 '15 at 3:58
0

Import song files into a wave editor, make the adjustment and save as a new file. A lot of wave editors even come with analysis tools to tell you loudness levels in RMS terms or peak levels in dB terms. Use the tools to analyse then use gain-control tools to lower levels and then save them as new files.

However, if you are wanting to extract melodic information then your processing algorthms shouldn't really care about loudness of A versus B.

1

First question to clarify is do you really need to align loudness level of your audio files ? It really depends on the goal of your project and what subsequent processing do you plan on these files. If you want to align them, do you need to align perceived loudness or normalize the audio signals ?

This being said, you could use FreeLCS, an open source tool for aligning perceived loudness (as defined by EBU R128) of audio files. Notice : this will modify your input files, it is not a metadata based system where the playback level is dynamicaly adjusted.

  • In my project I need to extract melody from audio. I thought to bring all the audio files into one uniform volume level in order to prevent my project getting biased to a particular audio. Say I have one audio clip with higher volume level compared to others. I am not sure that my project will get biased into that particular file. I need come out of that risk. – vigamage Sep 26 '15 at 12:32
  • And the other thing is, this is what I meant my having a uniform volume level. Say I have to songs in my playlist of say JetAudio. I set the volume of the player into one desired level and finish listening the first song. Just after that, the second level automatically get started playing with very low volume level which is not enough. I need the second song to be in the same volume level. Simply I need to not to set the volume up for the second song. – vigamage Sep 26 '15 at 12:35
  • And also, I do not need my audio files to be modified in the process. – vigamage Sep 26 '15 at 12:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.