Powered by OPF
OPF WIKI STATIC ARCHIVE
2,681 pages · 153 spaces · 776 tags · 4,025 history records · 96.2% of the original wiki recovered
Archived copy. This page was recovered from the Internet Archive snapshot of /display/AQuA/Audio Auditing Script taken on 2013-07-27. The original wiki at wiki.opf-labs.org is being decommissioned.

Audio Auditing Script

Created by Julie Allinson on Apr 12, 2011 · last edited by Paul Wheatley · on Aug 10, 2011 (view history) · 18 versions
One line summary A script to check a collection of audio recordings for 1) expected files, 2) expected specification, and 3) provenance.                                                                                                         
Detailed description audio_audit: Audit a series of Wave files.
OPTIONS:
-h  Show this message.
-i Input path (mandatory).
-f Perform fingerprint verification of Master/Listening pairs.
-v Display version.
 
This is a simple bash script which uses two main tools:
  • BWF MetaEdit: for extracting file-specific metadata (sample rate, sample bit-rate).
  • jHears: for audio fingerprinting (which also relies on SoX). Both client and server software are required.
     
    It is used to check a series of paired Wave files (a Master and Listening copy) against certain criteria:
    # That for each Master there is a corresponding listening copy.# That each Wave file has the correct sample rate and sample bit-rate.# Optionally, that each Listening copy has a sufficiently similar audio fingerprint to that of the Master.
Solution champion  Roger
Git link  https://github.com/openplanets/AQuA/tree/master/audio-audit
Group Evaluation Notes
  • Fingerprinting tested on sample collection but mixed results at matching pair. Masters and service files may not be close enough for fingerprinting to work
  • Further investigation needed on results and tweaking the fingerprinting parameters
  • Would be interesting to try this on other
  • Would other tools/techniques work better. Eg. AudioScout? http://freshmeat.net/projects/audioscout or libofa
  • Potentially very useful to issue champion, especially spec checking
  • Could check on submission
  • Organisation could quite easily utilise the script
  • If wrapped as windows usable tool could be used by those generating the recordings
  • Potential to investigate further metadata extraction
  • Potential to take results away, trial it, generate more requirements and then bring to London AQuA event
  • Might be able to bring in additional sound expertise for London (BL Sound and Vision?)
Detailed Evaluation
 
Tool (link)
Issue Audit audio batch against criteria