Expedition Navigation Software / Test calibration values

**Posted by Fredrik** 6/21/2022 12:37 pm

I'm not sure if this could already be done by using expedition calculations and calibrations in old log files, but if not - would it be possible to "test" new calibration values on old log files?

I do calculations and calibrations in H5000 but I'm constantly looking for ways to improve calibrations by looking at the log files in expedition. Most of the time I make a best guess from expedition, go back to the boat, recalibrate - run a new course - and back to expedition again to check - often I start over again...

What would really help would be to try different calibrations in expedition and directly see from the track and the stripchart what the current set and drift would have been leg by leg should I have used a different calibration for heading or leeway. Or similarly how would by TWD be different tack to tack in the stripchart if I had used other calibrations for AWA / TWA ....this could of course be useful for all calibrations from compass offset, to speed calibrations ... and everything else

I guess I'm looking for a feature to go back in logs and switch back and forth looking at either the data in the log file as recorded, or recalibrated using expedition calculations and calibrations instead of logged values... to play around with different values until the stripchart and data in the track looks consistent.

**Posted by Nick** 6/21/2022 9:09 pm

You might be able to do that with (eg) Njord. A lot of people are using it for analysis.

Sailing performance might be able to do it too.

**Posted by rbhun** 9/18/2022 1:39 am

@fredrik, were you able to find a way to do this? I have the same problem. Thank you.

**Posted by Fredrik** 9/19/2022 4:09 am

Sorry @rbhun I haven't found a solution for this - still do think it would be very useful to handle calibrations back home with "unlimited" time and "lots of logged data" as compared to testing it on the water and recording new data after each change.

I've been playing around with the log files to try and recalculate backwards ... but I haven´t really had the time to get a working process for it.

**Posted by rbhun** 9/19/2022 4:21 am

I agree, it sounds like such a no-brainer. We have tons of data, post-calibration sounds so easy.

my sailing performance package just ran out.. I might give Njord a try first, then ask them to add this function too.

@Nick - any reason that the calibration values are baked in the log? Shouldn't they be added later? that would create a solution for this.

**Posted by Nick** 9/19/2022 11:56 am

Yes, the log are the calibrated values. It is the most logical option.

Mostly Exp doesn't see the uncalibrated values. The main exception are mwa & mws.

**Posted by Fredrik** 9/20/2022 1:23 am

@Nick, with your better understanding of the data logged - what are your thoughts on trying to "uncalibrate" / "recalibrate" the logfile in excel ... would it even be possible? I'm think both in terms of how the logging frequency would affect calculating backwards and also if I have all the information needed. I'm for example not fully sure of how, when or what the 3D motion sensor would change the logged data and if it would be a problem.

I understand it's not super simple - but is it even worth a try?

My first go would be to be able to recalculate TWA and CURRENT SET & DRIFT after changing "simple" calibrations to AWA/MHU, TWS, BSP, heading offset and Leeway ... but obviously I would have a go at TWA calibrations as well and possibly look in to heel corrected speed and possibly other calibrations I can't even think for now :-)

I'm thinking if I load the recalibrated log in expedition, looking at currents in the track and using the stripcharts would be really helpful to evaluate any changes made.

I need to spend some time uncovering all the calculations done and I know B&G use cubic spline interpolation for TWA - but that should be possible to recreate.

I'm a statistical data analyst by trade - so as long as I have the data and can find the calculations I'm thinking I should be able to do the math ... but before I dig to deep - do you think there has been so many calculations applied by the time the data ends up in the logfile that it can't be "recalibrated" or do you think it can be done using the log file alone? Would it be worth a try to improve calibrations?

I do have a Yacht Devices Voyage Recorder YDVR-04 in the system .... I haven't really spend to much time with the logs from that as I typically use expedition .... but possibly there could be more/ other data recorded from that.

**Posted by Nick** 9/20/2022 6:43 pm

That sounds painful and you would need to know how B&G calulates twa etc from awa, aws etc. From memory, there was something odd in the roll correction for awa too.

This is all why it is very important to get bsp, heading etc calibrated first. Then the TW calibration becomes and iterative procedure.