[Corpora-List] Question concerning audio file search

[Thomas Schmidt]

Dear Briony,

> does anyone know of an existing solution using a Windows graphical interface?

You might want to have a look at the ZECKE search tool. It is a (free, GUI 
based, cross platform) concordancer designed to work on transcriptions in 
EXMARaLDA format. You type in a (regular) search expression, you get a KWIC 
concordance of matching results, and clicking on any of the results in the KWIC 
table will take you a playback of the corresponding audio file. EXMARaLDA has 
an XML format which is more powerful than the transcriber XML format, so it 
should be possible to transform the latter into the former with an XSL 
stylesheet. Take a look here:

http://www1.uni-hamburg.de/exmaralda/index-en.html

The software can handle wav files, a length of an hour or more should not cause 
any problems. Feel free to ask me for assistance if you think you might want to 
use this tool.

Kindly,

Thomas

-- 
Thomas Schmidt
SFB 538 'Mehrsprachigkeit'
http://www.rrz.uni-hamburg.de/exmaralda
http://www.kicktionary.de
http://www.exmaralda.org/annotation


Zitat von Briony Williams <b.williams at bangor.ac.uk>:

> sato hiroaki wrote:
> > I've just made a software tool for using DVD movies as a multimedia corpus.
> 
> Following the publicising of this useful-looking tool, I wonder whether any
> members of the Corpora list could help me with a related task, concerning
> audio rather than video files. I'm asking on behalf of someone else.
> 
> Is there an existing Windows-based software application that will do the
> following? (preferably free of charge, and without a large memory
> requirement):-
> 
> 1) Given: several very long .wav sound files (possibly an hour long), with
> associated text transcript files (.trs files) as produced using the
> "Transcriber" software application.
> 
> 2) User input:  User types one or more words to search for in the
> transcription files.
> 
> 3) Output: Software returns each chunk of text that contains the search
> string (where "chunk" could be a phrase, sentence, paragraph, topic, or
> larger, depending on the granularity of the transcription files).
> 
> 4) User input: User selects one of the search results.
> 
> 5) Output: Software plays back the portion of the (large) sound file
> corresponding to the chunk selected by the user.
> 
> I can think of a way to do this using Cygwin and Edinburgh Speech Tools - but
> does anyone know of an existing solution using a Windows graphical interface?
> My contact seems to prefer a point-and-click interface if possible.
> 
> Thanks in advance for any responses.
> 
> Best regards
> 
> Briony Williams
> 
> -- 
> Briony Williams
> 
> Arweinydd Tîm Technoleg Lleferydd / Speech Technology Team Leader
> Uned Technolegau Iaith            / Language Technologies Unit
> Canolfan Bedwyr                   / Canolfan Bedwyr
> Prifysgol Cymru                   / University of Wales
> Bangor                            / Bangor
> Gwynedd LL57 2EN, UK              / Gwynedd LL57 2EN, UK
> 
> E-Bost / E-Mail : b.williams at bangor.ac.uk
> Gwe (Cymraeg)   : http://www.bangor.ac.uk/ar/cb/technolegau_iaith.php.cy
> Web (English)   : http://www.bangor.ac.uk/ar/cb/technolegau_iaith.php.en
> Ffôn / Tel      : +44 (0) 1506 200862
> Rhithfro / Blog : http://murmur.bangor.ac.uk
> ....................................................................

> 
>