Volume Normalizer plugin for XMMS

Current version: 0.8.3

Similar Projects

SourceForge page
Mailing Lists
Bug Reports

Hosted by
SourceForge Logo

Tasks that need doing
  • Create a monitor that will display the current power, gain and such, possibly also show song change time and ability to dump the data to a file.
  • Make sure that the sample size is small enough, not larger than 1/100 of a second worth of data.
  • Remember the maximum power level of the song when it stops playing so that next time we will have the maximum level from the start. However, even if there is a maximum the plugin should try to find the maximum again and save the newly found maximum for next time, this is good for retraining if a song was changed from last time or if the song was not heard to its full length last time.
  • Find a way to search the power of songs from the playlist when they are not played (for example by using a timer and a low priority thread).
  • Allow the user to set the size of the queue.
  • Need to keep tab of the multiplier so that clipping will not occur.
  • Add tests to the test driver, currently it only tests that the plugin is loadable, if it's not loadable linking the test driver will fail.
  • Solicit the debian equivalent of RPM .specs from debian users. And have someone to create the .debs when doing file releases.
  • As a part of the normalization it might be applicable to use Fade-in/Fade-out to avoid startling noise at the start if a song starts loud, or a hard cut if a song ends loud. This can be expressed in duration (seconds) that it takes to change from a zero multiplier (complete silence) to no change, it is also possible to add a start multiplier so that it will not need to start from 0.
  • Possibly add a preprocessing stage that will reduce the dynamic range, this was asked for in the context of classical music where the dynamic range is wide and causes also wide changes in the perceived volume inside the song. This is not handled currently since the plugin adjusts the volume based on the maximum volume of the song so as to preserve the dynamic range inside the song. The way to do it is by using a compressor effect as a preprocessing stage. It should allow the configuration of the starting power and the rate, exact definition of how to write a compressor effect should be found.

This page was last modified on June 28, 2005.