Transcription technology?

Fri May 14 22:22:53 UTC 2010

Hi,

I personally find Elan to have a high learning curve (though I'm not that
familiar with it), and found trying to set it up to allow for searchable
interlinear gloss was rather difficult and frustrating. But people who
aren't interested in having that function seem happy with it. And that
doesn't seem to be an impediment for the MPI folks I know (though they get
on site tech support). It really does seem wonderful once you have set it up
for your needs.

I have my RAs use Transana with a pedal, by the way, though I'm considering
Atlas.ti. From what i understand you can export Transana files into it and
it allows for easier coding.

One concern is what format your transcripts can be exported in (other
transcription programs, atlas, even word or open office), whether the
software is proprietary and whether support for older versions or older
formats of data will be supported in the long run. Both Transana and ELAN
recognize that this is an issue for academics and archives, so I think their
future seems more promising than some of the commercial or "court reporter"
designed software out there. (Of the latter, I have had my Talk Radio and
Discourses of the American Right students (working with English talk radio
data that is not multi-party) use transcriber. It's free, multi-platform and
easy to use and does link sound and text).

My admittedly not terribly recent experiences with PRAAT indicate that this
is not the best program for transcribing long (i.e., longer than 5 minute)
segments. It's great for phonetic analysis, obviously, but did not meet my
needs for transcribing multi-party conversations or my large sound files.

I use Audacity (nice because it's cross platform and also free) largely for
digitizing, editing and  saving/converting formats, but not transcription.

Best,
Robin
______________________________________

Robin Shoaps

2009-2010 Faculty Fellow, Centre for Ethics, University of Toronto

Assistant Professor of Anthropology
The University of Chicago
1126 E. 59th St.
Chicago, IL 60637

Office: Haskell 301
Lab phone: 702-5533
Fax: 702-4503

_______________________________________

On Fri, May 14, 2010 at 5:59 PM, Sarah Wagner <swagner at mail.utexas.edu>wrote:

> Thanks for this information.  I noticed no one mentioned ELAN.  Is that
> because y'all haven't used it, or because you find other programs better?
> In particular, I'm transcribing multi-party audio/video conversations, and
> find ELAN to have a nice set up for that.  I haven't looked at all the
> others, though and wonder if I should change, since ELAN is a bit
> cumbersome
> in other respects. Thoughts?
>
> Sarah
>
> On Wed, May 12, 2010 at 6:10 PM, Steven Talmy <steven.talmy at ubc.ca> wrote:
>
> > I sent this a few hours ago, but it doesn't seem like it made it through.
> >
> >
> > -------- Original Message --------
> > Subject:        Re: Transcription technology?
> > Date:   Wed, 12 May 2010 13:26:36 -0700
> > From:   Steven Talmy <steven.talmy at ubc.ca>
> > To:     galey modan <gmodan at GMAIL.COM>
> > CC:     LINGANTH at LISTSERV.LINGUISTLIST.ORG
> >
> >
> >
> > Hello everyone,
> >
> > As far as I know, the short answer to Steve's question is no, there is
> > no transcription software that one can use to transcribe interviews, at
> > least automatically (i.e., plug an mp3 file into it and it spits out a
> > transcript). However, there are many other options, as I'll describe in
> > a moment. In answer to Galey's question, yes, there is software that
> > allows you to plug a foot pedal into a USB port on your computer:
> > ExpressScribe<http://www.nch.com.au/scribe/index.html>
> >
> >
> > More on Steve's question:
> >
> > In my experience, there is no substitute for an experienced transcriber
> > sitting down, listening to the audio/video record, and transcribing it.
> > If you have lots of money, I would suggest hiring a research assistant,
> > and training them in the theory and practicalities/conventions of
> > transcription that you are after. You could also hire a professional
> > transcriber, but the ones I've hired in the past have been very
> > expensive, and usually "tidied up" the data, e.g. glossed over
> > micro-details such as pauses, false starts, hesitations, etc. It's
> > definitely not the case that all transcripts must include such details,
> > but if that's what you're after, you'll need to train someone.
> > Regardless, if you have someone else transcribe your data, I think it's
> > important to go over what they wind up producing to check it.
> >
> > I use Audacity (freeware) for my transcriptions, and type right into
> > Word.<http://audacity.sourceforge.net/?lang=en>. Here are a few other
> > alternatives. This is by no means an exhaustive list, so please add to
> it!
> >
> > Transcriber (freeware) is good, though I haven't used it much recently:
> > http://trans.sourceforge.net/en/presentation.php
> >
> > F4 (freeware) is also good, and can be used with video:
> > http://www.audiotranskription.de/english/f4.htm
> >
> >  CLAN (freeware) has been developed explicitly for CA:
> > http://www.paultenhave.nl/clan.html
> >
> > Transana ($50) is very good for video (also audio):
> > http://www.transana.org/
> >
> > Dragon Naturally Speaking ($200 for Preferred version, now in version
> > 10) is probably closest to what Steve is looking for. It's an excellent
> > (really impressive) talk to text application: you speak into a USB mic
> > connected to your computer and it transcribes what you say. It requires
> > quite a bit of training to your voice, and prefers very clear input with
> > little background noise, which makes it pretty much unworkable in terms
> > of plugging in a sound file of naturally occurring talk and having it
> > transcribe it. I've tried it on interview recordings and classroom
> > interactions, and the results have been laughable: pretty much
> > gibberish. Maybe in a few more years...? Regardless, I have had students
> > tell me they use it to do preliminary transcripts: they listen to the
> > audio of an interview, e.g., then speak what was said on the recording
> > into the mic, and DNS converts that to text. That still takes quite a
> > lot of time though, and requires further editing for details of
> > delivery, turn-taking etc. I personally only use DNS for transcribing
> > fieldnotes, and for making comments on students' papers. I don't like
> > using it for much more than that.
> > <
> >
> http://www.dragon-medical-transcription.com/dragon_naturally_speaking_Preferred.html
> > >.
> >
> > This is a really incomplete list, so I'd love to hear what others use.
> >
> > All the best,
> >
> > Steven
> >
> > ..................................................................
> > Steven Talmy, Ph.D.
> > Assistant Professor
> > Department of Language&  Literacy Education
> > University of British Columbia
> > 2125 Main Mall
> > Vancouver, BC  V6T 1Z5
> > Canada
> >
> > Office:
> > Ponderosa E 221
> > 604.822.2353
> > 604.822.3154 (fax)
> >
> > LLED website:<http://www.lled.educ.ubc.ca>
> > personal website:<http://educ.ubc.ca/faculty/talmy>
> >
> >
> >
> >
> > On 5/12/2010 12:31 PM, galey modan wrote:
> >
> >>  Further to Steve's question, I'd be interested to know if anyone knows
> >>  of any good transcription technology which allows you to use a foot
> >>  pedal so you can keep your fingers on the keyboard and stop and
> >>  backspace with your foot, like the old-fashioned cassette transcribing
> >>  machine way.
> >>
> >>  Galey
> >>
> >>  2010/5/12 Leila Monaghan<leila.monaghan at gmail.com>:
> >>
> >>   From Steve Bialosok, please reply to him--
> >>>  Steve Bialostok<stevebialostok at yahoo.com>   (University of Wyoming)
> >>>  10:39am
> >>>  I am looking for some sort of transcription technology. My work
> involves
> >>>  both interviews and classroom discourse.  I am unaware of
> sophisticated
> >>>  software other than the type you can buy between $100-200.00 at Best
> Buy
> >>> or
> >>>  the Apple Store. I have been given a substantial amount of money (far
> >>> more
> >>>  than $200.00) to buy something that works well. And since I'm
> >>> techno-stupid,
> >>>  I'm hoping that it doesn't take a genius to make it work. Anyone out
> >>> there
> >>>  have any ideas or product names. (And to all those techno-savy folk
> out
> >>>  there, what is "easy" and "simple" for you may be brain surgery for
> me.)
> >>>  Steve
> >>>
> >>>  --
> >>>  Leila Monaghan, PhD
> >>>  Department of Anthropology
> >>>  University of Wyoming
> >>>  Laramie, Wyoming
> >>>
> >>>
> >>>
> >>
>