praat, ELAN, speech analyzer, transcriber....
Pilar Valenzuela
I just started using FLEX to work on the documentation of the Kawapanan (Cahuapanan) languages in Peruvian Amazonia. I have started a dictionary and have glossed a text. In the field I will be doing audio and video recording. I would like to get your feedback, as FLEX users, as to the pros and cons that you have found in using software to transcribe audio and video. I would like to add audio to my dictionary entries and, if possible, import the transcribed texts into FLEX without having to retyped them. Also, is it possible to add audio to the texts?
Thanks in advance for your answers. Pilar
Sarah Ogilvie
Sarah Ogilvie
--
You received this message because you are subscribed to the discussion group "FLEx list". This group is hosted by Google Groups and is open for anyone to browse.
To post to this group, send email to flex...@googlegroups.com
To unsubscribe from this group, send email to flex-list-...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/flex-list
mark
FLEx's text function is clearly geared toward short texts without
audio. It cannot be used for time-aligned transcription of video or
audio. Several users have been advocating that FLEx be made to support
bidirectional import/export with ELAN, which is probably the most
widely used time-aligned annotation program; I've also seen statements
from FLEx developers on this mailinglist to the effect that they want
to add support for export to ELAN. If you would also benefit from
this, please add your voice to one of the previous threads (http://
is.gd/5E6GD) just to make clear to the developers that there is
increasing demand for integration.
Mark D. (MPI for Psycholinguistics)
Beth
> I've also seen statements
> from FLEx developers on this mailinglist to the effect that they want
> to add support for export to ELAN.
Have you tried the "XML for ELAN" export in the Texts area, in
FieldWorks 6.0? You have to be in a text, and in a tab like Analyze
or Print View, for the option "Export Interlinear" to show up in the
File menu.
No promises about what it does or doesn't include, but you can try
out as much as they have there so far.
And I hear that version 3.8.0 of ELAN has a command File/Import/FLEx
file, for importing the results of the above command.
So some of the pieces are starting to be there. But to export from
ELAN and then import back into FLEx, that's not there yet. It is
possible to import SFM texts. (I have a script to make it easier
than is described in the FLEx documentation.)
In FLEx you can link audio or video to lexical entries or words in
text or whole clauses in text. Some people make a Note field just
for audio and put audio links there. Granted, it's not time-aligned,
but at least you have the audio with the utterance. (Add a Note
field to the sentence. Then do Insert/External link. Then navigate
to the recording of that sentence. To link it with just a word,
select the word, and then do Insert/External link.)
-Beth
Aaron Broadwell
Just to add my voice to this, Pilar. This is my current understanding,
but others should correct me if I'm getting it wrong.
It seems that at the moment the best method is to choose whatever tool
you prefer (ELAN, etc.) for the transcription part of the task. When
you've got the transcription done, you'll have to cut and paste it
from that program into FLEx, since there's currently not a way to
import these kinds of text-speech aligned files.
What I am not sure about is which of the transcription tools provides
an output that is just pure text (without time marks, etc) which would
be appropriate for copying into FLEx. Here are my rough-and-ready
impressions after having tried a few
a.) Praat is excellent for things like pitch, and you can annotate
text with the TextGrid function. However, the output was not
pretty :-) The text I wanted to import was interspersed with all kinds
of time indications and it was very messy. (Though Praat is a very
complex program and probably there was some way to do what I
wanted.)
b.) Transcriber is a fairly simple program, but its output was more
compatible with what I wanted as a FLEx input. It's not particularly
oriented toward field linguistics, however, so it doesn't do much to
support word-by-word glossing of the text.
c.) ELAN (in my very humble opinion) has a steep learning curve, and I
found segmenting and annotating the digital speech sample to be more
difficult than in comparable programs. (It is also able to do very
complex things like annotating video, so the complexity may be
justified.)
d.) EXMARaLDA is good and fairly easy to figure out. It is a small
suite of tools for aligning Sound and transcription. The main tool
which is relevant is called the Partitur Editor. The Output command
within program this will produce an RTF output in the form of a Word
table, where the text and gloss are nicely aligned with each other.
You can copy lines of the target analysis into FLEx without too much
trouble and then copy lines of the free translation for the Gloss or
Analysis tabs within FLEx.
Good luck!
Aaron