[RNLD] Links between publication and sound corpus

[Nick Thieberger]

Hi Steffen,

In my grammar of South Efate I used textual references to media
chunks, of the form filename, start time, end time. These references
came from my corpus of the language, developed in Audiamus. I wrote
Audiamus to allow me to select and cite the examples and to clip that
citation and insert it into my work and then also used Audiamus as the
playback system in a dvd that accompanied the published grammar. That
was some time ago and now I think I would use an archival version of
the media as the streaming source and have HTML5 calls to the
timecodes. These can be online (both via the web or served locally) or
embedded in a pdf document, and we do this with media examples in
articles in the journal Language Documentation & Conservation.
Unfortunately, there is still no easy way to select a textual example
together with its timecode and (citable, archival) media filename and
insert that into the dissertation. Elan doesn't do it.

If you do have to chop a wav file up into many small bits (and I would
discourage this method as it leads to a large management problem) then
Audacity has a very good way of chopping up an audio file based on
labels (discussed here:
http://www.paradisec.org.au/blog/2011/05/books-html-audio-images-falling-out-from-fieldwork/).
These label can be imported from an Elan export. Ideally each such
media chunk should have a trace within it indicating where it came
from.

Nick