Does anyone has info on the format of the IDS log file that is generated when recording parameters
- I suspect this is a forlorn hope, but just asking....you never know

Currently I am having to correlate some home grown data logs (Boost, Injector Duty Cycle) with IDS logs (MAF, O2 upstream) by using rpm values, stepping through the IDS data one point at a time in the IDS screen. And it's tedious....

If only there was a way to export the IDS data, then I could pull everything into a spreadsheet and get an instant picture of how close I am to melting the engine (Avos twin screw).

 

As long as the parameters you are looking for are all standard OBDII, TorquePro and an ELM327-type device will do that and more. Cheap, too.

More to your question, the log files I have seen are in XML. Not to say that these files have actual data in them. And I use SDD with my '03, not IDS...

 

Thanks - I don't think O2 Upstream (wideband, current reading) is a standard parameter for X100, but not sure and I will check into it. Downstream (narrowband, volts) is, but its not what I'm really looking for, since it is kind of binary on AFR reading.

The .dsg file for defining the logging configuration does seem to be XML.

But the .dlg file which contain a logged run of data seems to be binary. From a quick scan, it appears, amongst a bunch of binary stuff, to contain a long list of sensor descriptions and related verbage (in ascii), and for those parameters that are actually logged, a bunch of binary after the ascii text relating to that particular sensor.

I too use SDD, but it then does direct passthrough to open up IDS for my '01 and '02 XK8 and XKR.

 

There is a quick list of standard OBDII PIDs here: OBD-II_PIDs

As a pure guess, maybe the binary data is the raw stream off of the ISO-9141 bus. Maybe a hex viewer would clarify.

 

Did some scans with a hex viewer as suggested
So it looks like the data is in pairs of 32 bit little endian values, likely a timestamp (since it continually increments) and a data value.
Time stamps have a discontinuity in the middle of the data, possibly the pre-post time trigger, with pre written after the post for some reason, if that theory is correct
I guess I will try and roll up the sleeves and create a basic parsing program to extract what I need, if and when I can find the time......

 

Well if anyone else needs something, at least for X100 IDS engine module logs, I hacked a quick parser using Processing (https://processing.org/), java based.
It works for what I needed......upstream O2 and rpm
Not sure it would work for anything else as is, but of course could be tweaked
Obviously I don't have insights into most of the log file format, but managed to figure out enough to pull the logged data samples into a .csv file for analysis using a spreadsheet.
Text file of source code attached