Transcriber problem

William J Poser wjposer at LDC.UPENN.EDU
Tue Jan 30 20:06:27 UTC 2007


Just to fill other people in, the answer to the query that
Rachel posted about her system freezing on trying to open certain
sound files in Transcriber, which we discussed further by
private email, is almost certainly due to a known defect in
the library that Transcriber uses for audio i/o. When you
open an audio file in Transcriber, it attempts to read the entire
file into memory. If sufficient memory is not available, this
is impossible. In the best of circumstances Transcriber would be
unable to open the file but would fail gracefully and would
report that it was unable to allocate sufficient memory, but
at least on MS Windows it does not fail gracefully and hangs the
system.

In the short term this means that you should not try to open
huge files in Transcriber, where "huge" means on the order of
an hour. To work with such data you'll need to split the files
into smaller pieces.

The failure to fail gracefully may or may not have an intermediate
term solution - I don't know whether this is due to a bug in
the library or a defect in MS Windows memory management.
If the former, it can be fixed.

In the longer term the audio i/o in Transcriber should be replaced
so as to read only the bit of the file that it needs at any
given time, which will considerably reduce the memory requirements.
This has been discussed before but there is no definite timeline
for it.

Bill



More information about the Resource-network-linguistic-diversity mailing list