should analysis work without an existing track map?

21 views
Skip to first unread message

toga94m

unread,
Jan 12, 2016, 7:50:10 PM1/12/16
to AutosportLabs Beta Testers
I imported some SD card logs from driving this weekend. Running the new 1.4.1 on a Win8 laptop, not trying to connect to the RCP, just standalone on the datalogs. In-car tablet dashboard showed sectors and lapcounts incrementing. This is on a road that isn't in the RCP track database.  

Do I need to create a test track with coordinates before anything will appear in the analysis tab? I get a flat line for speed&AccelX&AccelY in the lower half, nothing in the upper half. Pinching the touchpad seems to zoom in/out, but still a flat line (no vertical scaling). Scrolling shows the data values changing though, so it imported nicely.

Also, the dashboard/numbers page (default first screen I see) looks pretty sparse with no BT connection to live hardware. Maybe a default message that says "no connection" should be here, until data is available?
rcp141 analysis.png
rcp141 first start.png
RC_170.LOG

Brent Picasso

unread,
Jan 12, 2016, 7:51:38 PM1/12/16
to toga94m, AutosportLabs Beta Testers
Hi,  We'll check with your log file.  You might also want to check your metadata headers (see other post here) to make sure the values are sane.

--
You received this message because you are subscribed to the Google Groups "AutosportLabs Beta Testers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to autosportlabs-beta-...@googlegroups.com.
To post to this group, send email to autosportlabs...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/autosportlabs-beta-testers/d4bc6a3b-1078-4c44-9e6a-b41abcd1ea0a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--
Brent Picasso
Technology for Race and Street

toga94m

unread,
Jan 12, 2016, 7:59:29 PM1/12/16
to AutosportLabs Beta Testers, pint...@gmail.com
A quick import into Excel shows they're rational numbers, plus the speed channels show reasonable values in the analysis tab.

Possibly something to do with my lograte choices? There are different sample rates, so there are gaps in the data in the logfile. That makes Excel unhappy, but I assumed your app would know how to deal with gaps.


On Tuesday, January 12, 2016 at 7:51:38 PM UTC-5, Brent Picasso wrote:
Hi,  We'll check with your log file.  You might also want to check your metadata headers (see other post here) to make sure the values are sane.
On Tue, Jan 12, 2016 at 4:50 PM, toga94m <pint...@gmail.com> wrote:
I imported some SD card logs from driving this weekend. Running the new 1.4.1 on a Win8 laptop, not trying to connect to the RCP, just standalone on the datalogs. In-car tablet dashboard showed sectors and lapcounts incrementing. This is on a road that isn't in the RCP track database.  

Do I need to create a test track with coordinates before anything will appear in the analysis tab? I get a flat line for speed&AccelX&AccelY in the lower half, nothing in the upper half. Pinching the touchpad seems to zoom in/out, but still a flat line (no vertical scaling). Scrolling shows the data values changing though, so it imported nicely.

Also, the dashboard/numbers page (default first screen I see) looks pretty sparse with no BT connection to live hardware. Maybe a default message that says "no connection" should be here, until data is available?

--
You received this message because you are subscribed to the Google Groups "AutosportLabs Beta Testers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to autosportlabs-beta-testers+unsub...@googlegroups.com.

toga94m

unread,
Jan 15, 2016, 9:34:40 PM1/15/16
to AutosportLabs Beta Testers, pint...@gmail.com
I hacked together a quick track map from Google maps (attached JSON) and now I see the RCP GPS data superimposed with my track outline.

I'm also attaching two more run logs. I would drive across the marked FINISH line, then hit the hardware log button to stop logging, continue to a convenient stopsign, hit the hardware button again, and drive to/thru the START coordinates. So it's one logfile per lap/autocross run.

Testing 1.4.1 on Win8 laptop, still not able to vertically zoom the speed/AccelX traces, so they look like flat lines, maybe a 1-pixel bump occasionally.
RC_171.LOG
RC_172.LOG
54a3133b183b3aadeb000003.json

toga94m

unread,
Jan 15, 2016, 10:06:18 PM1/15/16
to AutosportLabs Beta Testers, pint...@gmail.com
Ohhhh. got it. Pinching doesn't cause vertical zoom doesn't happen in analysis - only time zooming. And somehow my min/max were set to +66 and -66 for the accel channels.. guess I didn't check that before heading out. Though, editing those min/max values in the logfile header doesn't make a difference. I turned on Yaw display and that one shows up at the bottom of the screen, so now i can see the effects of pinchzooming. Also noticed the in/out sense of zooming is inverted from expected on the track map, but I think I remember that being reported already.

tom
rcp141 log170 view.png

Brent Picasso

unread,
Jan 15, 2016, 10:09:30 PM1/15/16
to toga94m, AutosportLabs Beta Testers
Hi Tom,

Thanks for the update. Can you take a look at the metadata headers in your log files? It looks like the min/max values are not correct. 

In fact, just about all of the min/max values are off in your file.  Could you also post your RC configuration file so we can check out the settings?

The format of the metadata header is:  <Channel>|<Units>|min|max|sample_rate

example:
"Battery"|"Volts"|0.0|10.0|1,

I re-attached your log channels with updated configurations. 

-Brent


On Fri, Jan 15, 2016 at 6:34 PM, toga94m <pint...@gmail.com> wrote:
I hacked together a quick track map from Google maps (attached JSON) and now I see the RCP GPS data superimposed with my track outline.

I'm also attaching two more run logs. I would drive across the marked FINISH line, then hit the hardware log button to stop logging, continue to a convenient stopsign, hit the hardware button again, and drive to/thru the START coordinates. So it's one logfile per lap/autocross run.

Testing 1.4.1 on Win8 laptop, still not able to vertically zoom the speed/AccelX traces, so they look like flat lines, maybe a 1-pixel bump occasionally.


On Tuesday, January 12, 2016 at 7:59:29 PM UTC-5, toga94m wrote:
A quick import into Excel shows they're rational numbers, plus the speed channels show reasonable values in the analysis tab.

Possibly something to do with my lograte choices? There are different sample rates, so there are gaps in the data in the logfile. That makes Excel unhappy, but I assumed your app would know how to deal with gaps.


On Tuesday, January 12, 2016 at 7:51:38 PM UTC-5, Brent Picasso wrote:
Hi,  We'll check with your log file.  You might also want to check your metadata headers (see other post here) to make sure the values are sane.
On Tue, Jan 12, 2016 at 4:50 PM, toga94m <pint...@gmail.com> wrote:
I imported some SD card logs from driving this weekend. Running the new 1.4.1 on a Win8 laptop, not trying to connect to the RCP, just standalone on the datalogs. In-car tablet dashboard showed sectors and lapcounts incrementing. This is on a road that isn't in the RCP track database.  

Do I need to create a test track with coordinates before anything will appear in the analysis tab? I get a flat line for speed&AccelX&AccelY in the lower half, nothing in the upper half. Pinching the touchpad seems to zoom in/out, but still a flat line (no vertical scaling). Scrolling shows the data values changing though, so it imported nicely.

Also, the dashboard/numbers page (default first screen I see) looks pretty sparse with no BT connection to live hardware. Maybe a default message that says "no connection" should be here, until data is available?

--
You received this message because you are subscribed to the Google Groups "AutosportLabs Beta Testers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to autosportlabs-beta-...@googlegroups.com.



--
Brent Picasso
Technology for Race and Street

--
You received this message because you are subscribed to the Google Groups "AutosportLabs Beta Testers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to autosportlabs-beta-...@googlegroups.com.

To post to this group, send email to autosportlabs...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.
RC_172.LOG
RC_171.LOG

Brent Picasso

unread,
Jan 15, 2016, 11:00:37 PM1/15/16
to toga94m, AutosportLabs Beta Testers
Working on this as part of  1.4.2 release - https://github.com/autosportlabs/RaceCapture_App/issues/762

Side note - I see your in your speed trace the distance resets to zero, causing the line chart to go back to the beginning. We'll have to address that somehow in post process - I think that's an artifact of the logging process starting- need to trim the extra data so you're just looking at the run itself.

as an experiment you can try doing a single log file and try doing multiple runs back-to-back and see how that looks.

-Brent

toga94m

unread,
Jan 16, 2016, 2:37:22 PM1/16/16
to AutosportLabs Beta Testers, pint...@gmail.com
In real-world autocross, I'd pull up to the start line, then hit the hardware start/stop button, wait for a GO from the flagger, and start driving. Timing starts when I cross the start line beam, so that'd be the likely time for data import to start. Wouldn't that also be when SECTOR goes from -1 to 0?

Then after crossing the FINISH coordinates, I'd roll slowly back to grid, then get around to stopping the log by pressing the button again. Wait 20-50 minutes before next run. I tried to emulate that scenario during these log runs - back to back laps, not logging during the transits between finish and pre-start.

toga94m

unread,
Jan 16, 2016, 2:46:21 PM1/16/16
to AutosportLabs Beta Testers, pint...@gmail.com
Here's my MK1 config running 2.8.5, no changes since it was used to log those datafiles. I don't see in the app (1.4.1 on windows 8) how I could change the min/max values for anything other than battery voltage. So, I also have no idea how the min/max for acceleration were set to 66 and -66. They seem to be +/-3 in the config file. Speed is 0-1mph in the header, battery 0-10 in the header.

After saving that config file to my laptop via USB, I ran a manual log for a few seconds.. here's the header, still from that unchanged config file. So something's odd on 2.8.5 on my MK1, between the config file and the header export...

"Interval"|"ms"|0|0|1,"Utc"|"ms"|0|0|1,"Battery"|"Volts"|0.0|10.0|1,"AccelX"|"G"|-66.0|66.0|25,"AccelY"|"G"|-66.0|66.0|25,"AccelZ"|"G"|-66.0|66.0|25,"Yaw"|"Deg/Sec"|0.0|0.0|25,"Latitude"|"Degrees"|-1.0|1.0|10,"Longitude"|"Degrees"|-1.0|1.0|10,"Speed"|"MPH"|0.0|1.0|10,"Distance"|"Miles"|0.0|0.0|10,"Altitude"|"Feet"|0.0|0.0|10,"GPSSats"|""|0|10|10,"GPSQual"|""|0|40|10,"GPSDOP"|""|0.0|10.0|10,"LapCount"|""|0|0|10,"LapTime"|"Min"|0.0|0.0|10,"Sector"|""|0|0|10,"SectorTime"|"Min"|0.0|0.0|10,"PredTime"|"Min"|0.0|0.0|5,"ElapsedTime"|"Min"|0.0|0.0|10,"CurrentLap"|""|0|0|10

tom
skidmore config.rcp
Reply all
Reply to author
Forward
0 new messages