Altitude (from GPS) optimizations

Menion · April 25, 2013, 11:56:03

tommi, thanks for help. Thanks for silly questions and also for changes in manual. Whole wiki is free to use and free to edit, so feel free

to change everything you find useful. There is always way to revert changes back in case, something will be "bad".

I changed automatic calibration, so it now use SRTM data if they're available, otherwise use normal measured GPS altitude. I'm sure this will need some testing, to make sure it works correctly, so take next release versions as "Altitude testing version". There is always possibility to fill recorded track with offline altitude data (in case measured altitude will be completely wrong), so I'm not worry too much.

To some of your comments:
"This sounds like if I record a track from A to B that the correction might change during the record of the track due to different Geoid altitude of A and B." - that's correct in case you travel longer distance. Anyway changes aren't too big on bike ... check this map http://www.esri.com/news/arcuser/0703/g ... id3_lg.jpg where are visible changes all around the world. Values are from -100 to around +50 metres.

"What does "during usage" mean? During track recording?"
yes also during a long track recording. And don't worry, there should not be any huge change after calibration, but Locus should spread new calibration into next few minutes so there will be fluent movement of altitude from previous calibration value to new ... we'll see how this will work

@sherman: do you know any site that offer METAR data for the whole world? I'm not able to find any. Only some local

tommi · April 25, 2013, 14:14:35

Quote from: "menion""What does "during usage" mean? During track recording?"
yes also during a long track recording. And don't worry, there should not be any huge change after calibration, but Locus should spread new calibration into next few minutes so there will be fluent movement of altitude from previous calibration value to new ... we'll see how this will work

Just for my understanding: Is this also done during track recording in case of SRTM available and pressure sensor?

sherman · April 25, 2013, 15:41:41

Quote from: "menion"@sherman: do you know any site that offer METAR data for the whole world? I'm not able to find any. Only some local

I am using this site: http://www.aviador.es/Airports/Nearby/

But I believe there are some web services with API which should be easier to use with an android app, as on Google Play are several of them. I will do some research and will let you know.

Menion · April 25, 2013, 16:08:05

thanks Sherman!

and tommi yes ... did you tried to record track with pressure sensor with just one calibration at start? If so then you have to noticed that altitude in the end is wrong for quite a huge value (in tens of metres). So there is need for some calibration even during a day

edit: to your info, Locus check calibration (and re-calibrate if needed) every 2 hours

tommi · April 25, 2013, 16:29:42

Quote from: "menion"and tommi yes ... did you tried to record track with pressure sensor with just one calibration at start? If so then you have to noticed that altitude in the end is wrong for quite a huge value (in tens of metres). So there is need for some calibration even during a day

Due to weather related pressure change, I guess.

Quote from: "menion"edit: to your info, Locus check calibration (and re-calibrate if needed) every 2 hours

Ah, thank you. I added this info to the docu page.
Don't forget to link the page from http://docs.locusmap.eu/doku.php/manual ... gs:sensors when you release the new Locus version! (But I think you intentionally didn't do it up to now because it' s not matching the current official Locus version.

Menion · April 25, 2013, 16:47:18

hmm I again forget on link on manual, thanks. I'll add this into release version into top action bar in Altitude manager.

sherman · April 26, 2013, 16:20:35

Hi,

here is a webservice for METAR data: http://www.aviationweather.gov/adds/dat ... amples.php

In the request you can specify for example most recent data from all airports within a given radius from a given coordinates:

http://www.aviationweather.gov/adds/dat ... ation=true

Christian · May 02, 2013, 14:04:06

I 'm surpriced of the new altitude manager. great work!
So i checked most of my tracks to re-calculate the altitudes. But i'm not sure... after re-calculation all the values are 0 (NULL).
:-(
Is this a bug or a feature?
All settings in altitude manager are on default.
Christian

Menion · May 02, 2013, 14:48:00

it's a feature. I decided that best what I can do for you, is set all altitudes to 0. Only this will make Locus really useful

)

ehm, just kidding, please update to 2.11.1 version and sorry for a confusing

Christian · May 03, 2013, 16:55:52

you are funny, man :-)

But... the reason why we - especially you - started the thread still exists: values of 0 for the flat. Means only values for uphill and downhill and these are wrong :-(
Any suggestions?

Christian · May 07, 2013, 17:23:57

Anybody out there?

Menion · May 07, 2013, 17:37:16

hi, sure. sorry ...

So back to topic.

First part is done I think. Maybe there is place for improvements, but core "Altitude manager" is published and I think, it brings benefit to all that wants "Improved altitude". So way to improve values itself, is done ... (agree?)

here comes second part. Some computations with values and statistics all wants to see

.

here is a nice example https://getsatisfaction.com/locus/topic ... _from_srtm

so question si simple. How to compute it. I expect that there are no "exact variables" that's are used on all devices. So how to calibrate computations? On Garmin device?

Christian · May 09, 2013, 16:48:55

Hi Menion,
Thanx for your answer. Sorry, this was my fault. I thought the computation problem ist already solved together with the Altitude Manager (nice piece!!!)

I'm not a mathematician but if i have values from the SRTM files and the track- what's the problem with the computation of uphill, downhill and flat ? Am i wrong?

tommi · May 09, 2013, 18:01:53

Quote from: "Christian"I'm not a mathematician but if i have values from the SRTM files and the track- what's the problem with the computation of uphill, downhill and flat ? Am i wrong?

I'm not as well but I think the higher values for up and down come from much more noise in the graph, every little peak counts...

Menion · May 15, 2013, 19:00:44

yes as tommi wrote ... "every little peak counts" .. that's a general problem.

I was thinking about it and I think that best should be to find some optimal value thanks to other programs. You for sure (christian mainly) have some GPX files and programs that show uphill/downhill values with which you're satisfied right? I have idea that if you give me let's say 5 different GPX files with tracks together with screenshots from various programs where will be visible values you want to see also in Locus, I'll try to find a optimal parameters to achieve it, what you say?