Thanks in part to the information from Thierry, I have finally figured out why bwf2qt is not working for me but appears to work for others and even works for me with the sample files:

It depends on when the BWF file was created!

The broadcast-audio-extension chunk (which is what turns a WAVE into a BWF) includes the date and time that the file was created (recorded). The date is 10 characters in the format yyyy-mm-dd and the time is 8 characters in the format hh-mm-ss.

The string for a file created about now would be "2005-01-2021-32-56" but the 744T records it as "2005-1-20.21-32-56" [where the '.' is actually the NULL character (a byte of zero)]. Similarly, when the hour is less than 10, the leading 0 is dropped. Either error causes bwf2qt to fail.

So, any file recorded last month would work, as long as it was recorded after 10 am. [All my tests in December were before 10 am.] Nothing recorded this month will work.

I verified this by taking one of my failed files and fixing the date/time format. It then worked.

BTW: I found a few other formatting errors in the 744T's BWF output, but none of the other errors cause bwf2qt to fail.



Thierry wrote:

> Ed Anson wrote:
>
>> Thanks for the idea. I tried the Fostex tool on another system and it
>
>
>
>> worked. Of course, that doesn't necessarily mean my computer is
>
>
> corrupt
>
>> -- just that there is some incompatibility. The other system I tried
>
>
> had
>
>> a different OS, and my system seems to be in excellent health in that
>
>
>
>> everything else works perfectly.
>
>
>
> maybe something went wrong with the file transfers to your computer?
> Were you reading the files from the same disk when you tested on the
> other computer? The fact that the Fostex app also crashes on your 744
> files means probably that the files on the medium you are reading are
> wrong.
>
>
>> So I still have a bit of a mystery. Bwf2qt works on my system for
>
>
> some
>
>> BWF files, but not the ones that matter to me.
>
>
>
> A couple of weeks ago I helped setting up a system for a documentary
> shoot, consisting of:
> New Sony harddisk camera with DV quality video
> 744
> Ambient transmitter for sending timecode recrun to 744
>
> After a phonecall to Ambient for fixing the timecode level (a little
> screw inside the transmitter was set to low and dropped the timecode
> level about 14dB), everything worked almost flawlessly.
>
> Timecode is sent from camera to 744. Camera runs recrun, because we
> want to avoid having rushes with timecode jumps (offline and online
> editors hate that and I can understand why :-)
>
> 744 was in continuous jam.
>
> First remark:
> the 744 jams only one time, namely the first time you connect the TC
> (and by eache connection-re-connection of the TC wires). If the camera
> and 744 are stopped and a 2nd take is started you have to re-jam the
> 744 manually (which makes sense). There is however no seperate jam
> button on the machine (only in the menu structure which is too many
> button presses in the heat of the moment). The best solution we came up
> with, is by cutting the power of the Ambient TC receiver briefly,
> before pushing the rec button on the 744.
>
> Second remark that was the big issue I wanted to test in the first
> place:
> the 744 has to be started AFTER the camera runs. Simple reason, if the
> TC is jammed while a take is recorded, the TC is visually jammed on the
> 744 LCD, but NOT in the BWF file!!!
> Somebody told me that an Aaton Cantar lets you decide what to do in
> this case (write the jammed or the original timecode to the file),
> which would be a very nice feature for the 744 as well.
>
> We then brought the rushes (copied from disk to DV) and the 744 to our
> company and transfered everything to a Mac Powerbook connected FW disk
> (FW connection to a G5 didn't work, but that's a known issue that
> apparently should be fixed with rev1.15).
>
> The BWF mono files were imported on a G5 in Nuendo v2 and PT6. Nuendo
> is much quicker a dropping the files on the timeline at their recorded
> TC (PT is manually spotting one by one, big disadvantage).
>
> The BWF files were also translated with Bwf2qt and imported to Final
> Cut Pro.
>
> Everything synced up fine.
> Only remark is that FCP doesn't have nice TC facilities such as Nuendo
> and FCP can't import omf files (major bummer).
> The picture editor left and I asked him to make a test cut sequence and
> omf that back to me, which he did, only I haven't had time since to
> check that again in PT, but it should work fine.
>
> Just to say that Bwf2qt really works with a 744 :-)
> Hope this helps.
>
> Greetings,
>
> Thierry
> www.a-sound.com
>