TDMS File FormatW

[TCPlomp]

We are looking into the new TDMS file format. One problem we've got is that the file structure is bad(not) documented. If  I create a file with a 'untitled' group and 'untitled' channel and a U8 data set from 0 to 99 I get the followin files

.tdms:

.tdms_index:

I can't find anywhere my data so it is propably compressed, does anyone (NI?) know the file format or is a white paper coming?

Thanks,

Ton

Message Edited by TonP on 09-15-2006 11:18 AM

[DavidJCrawford]

Ton,

Just did a search on the website and found these

http://zone.ni.com/devzone/conceptd.nsf/webmain/570CC2350FD14A0F862571840061EFAB

http://zone.ni.com/devzone/conceptd.nsf/webmain/DE71E0CD8B397817862571A2006D6C2F

I would guess you have already seen them and hasn't helped you.

Trying to help ...

David

[TCPlomp]

Yup have seen them, but I am looking for documentation on how the data is stored so that we can write a note on it, and we can read the files in 10 years or so with any software.
I like the TDMS format because it allows me to store a lot of info about the data (sensor/daq/computer) in an easy way. And it is compressed just this week I had a measurement of 9 channels 20kHz for 10 minutes (500 MEG in txt) and the binary file was much smaller but needed a conversion tool for a third party so ge sticked with the txt files....

Ton

[DavidJCrawford]

Take a look at the bottom of this page

http://zone.ni.com/devzone/conceptd.nsf/webmain/8F07969FB2FA71C1862570DF006B0190?opendocument&node=5...

Maybe the contact listed might be able to help?

David

[TCPlomp]

OK,

one error I made this data is uncompressed available (look at the top picture starting at byte x98.....

And I thougth the alphabet was some kind of mapping mechanism...
Then I tested it with an array of DBL's of 10Meg elements, yes the file was 80.000.150 bytes sized apparently if there is one channel the tdms_index file is at the top of the TDMS file

Ton