Opened 7 years ago

Closed 4 years ago

#11 closed enhancement (wontfix)

Audio normalization support

Reported by: olivier Owned by: yomguy
Priority: major Milestone: 1.0
Component: Export Version:
Keywords: Cc:

Description

This is about normalizing audio files. Before we implement this, I'd like to clarify the following point:

Should the file be normalized when the user imports it, and then stored into the source media files directory, once normalized? Isn't this dangerous/intrusive? Isn't there a risk for the digital normalization to slightly damage this precious source audio data?

Isn't it better to perform normalization when exporting the file, thus leaving the source file untouched?

Change History (11)

comment:1 Changed 7 years ago by yomguy

  • Owner set to yomguy
  • Status changed from new to assigned

The normalizing function, which is equivalent to a real gain related to the peak (max) value, is already implemented in Telemeta. But, at the moment, this function is applied to both the backuped file and the exported ones. I think we should add two options for the two cases in the import page:

  1. Normalize during import
  2. Normalize during export

By the way, a normalizing operator can be considered at a second level on the security scale - the copying or synchronizing functions would be for example at the first level - because the applied gain value can be stored in the metadata dictionnary in order to retrieve the original file a posteriori.

Thus, this situation approves the add of the "option" (dictionnary) 4th argument of the process() method in the Exporter api. All process parameters should be transfer by the "channel". When a process return some data which could be considered important for the backup, these data should be saved in the metadata dict, as a list for example...

comment:2 Changed 7 years ago by yomguy

  • Milestone changed from Web UI & refactoring to Add lame support to the new exporter

comment:3 Changed 7 years ago by yomguy

  • Milestone changed from Add lame support to the new exporter to Web UI & refactoring

comment:4 Changed 7 years ago by olivier

I'm updating this ticket after our off-line discussion:

We'll leave the original file untouched. That means that we won't normalize the audio source file after it gets imported by a maintainer.

Instead we'll let maintainers choose whether they want the file to be normalized or not. They will have this choice when they import the file, however the file will only get normalized when it is exported.

comment:5 Changed 7 years ago by olivier

  • Milestone Version 0.3 deleted

not for 0.3

comment:6 Changed 7 years ago by olivier

  • Milestone set to Version 0.3

comment:7 Changed 7 years ago by olivier

  • Milestone Version 0.3 deleted

comment:8 Changed 5 years ago by yomguy

  • Milestone set to 0.5

comment:9 Changed 5 years ago by anonymous

  • Component changed from Miscellaneous to Export

comment:10 Changed 4 years ago by olivier

  • Milestone changed from 0.5 to 1.0

If we really need to implement this, then it surely will be possible using the new TimeSide pipe-based api.

But that won't happen for Telemeta 0.5.

comment:11 Changed 4 years ago by yomguy

  • Resolution set to wontfix
  • Status changed from assigned to closed

Indeed, this feature will be added to the TimeSide of the project and not in Telemeta.
The ticket will be then moved to :

http://code.google.com/p/timeside/issues/list

Note: See TracTickets for help on using tickets.