Discussion: Binding splits to a GPS track

I was pleasantly surprised to see that one of my GPS tracks uploaded from Garmin automagically bound itself to a set of race splits, even displaying pins at the approximate control locations on the map.

http://www.attackpoint.org/sessiondata.jsp?session...

However, the feature did not work for other races on the weekend.

What circumstances will cause the binding to occur?

Is there any way to manually bind a GPS track to a set of race splits?

Tracks and splits get matched up automatically if the time on your log entry is within 5 seconds of the time on your splits. You can force it to match by adjusting the time on your log entry.

The Start clock for the Sat CSU Qual was off by about one minute which would account for the lack of splits coordination for that.

It is no wonder why I have not seen this feature before. I am very undisciplined about starting and stopping my watch. I seem to get it started about halfway to the first control, and then stop it ten minutes after the finish!

This feature is soooo cool!

I use the SI-Timemaster to synchronise the SI punch units we have about every 2 weeks (and always right before a sprint event). I see about 1- 2 seconds drift per week . So do we need to have the start punch in particular coordinated to GPS time within 5 seconds and folks to start their GPS also within that 5 second window for this auto link to work?

I think Ken means that the elapsed time for an individual runner's log entry needs to be within 5 seconds of the elapsed time in their race splits. So the timing on the SI units themselves is irrelevant provided the start and finish units are off by within 5 seconds of each other. Or, if runners enter their log entry time using the splits printout, the time on the SI units is totally irrelevant.

Yeah, what feet says.

Thanks for the explanation. It makes a lot of sense

I had the global time in my mind after looking at winsplits on-line pack running feature (which seems to report the actual clock time in GMT-2 or Sweden for punches our event in Minnesota !) .

Ohh, so starting GPS 5s before the start and 2s after the finish is a bad idea! That would explain some things. Ahh.

I am still not getting it.

The splits uploaded for my course on Sunday give my total time as 2:05:59.
My tracks give a time of 2:06:01.
I changed my workout time to 2:05:59.
So, within the 5 second threshold?

But the splits shown on the map are Recorded not Posted splits.

Is this because I have auto splits on and/or took some splits?

Or, am I missing something obvious?

If you took your own splits it probably wouldn't work.

Having said that, I'm not Ken so have no idea really :-)

Yeah, right now the posted splits only show up on your track when you don't have any from your watch. I need to make it just another option in the splits dropdown menu (so you can pick either), but I haven't gotten to that yet.

On the Winsplits upload page there is a field for "Zero Time" and the help text reads: "The zero time of the event, entered as hh.mm, e g 10.00. The zero time is found by selecting Settings in the Event menu in OE20xx / MT20xx / OS20xx. Not needed to be filled in if you are not uploading a split time file."

Not having access to the "OE20xx / MT20xx / OS20xx."software, I have always just entered something like 00:00 or 10:00. Perhaps that is why it looks like swedish time or GMT...

Another thing that would be useful for those of us who sometimes forget to start the watch on time - or who start the watch early to avoid having to fiddle with the watch at the start line - would be to allow for a manually-entered offset between the splits start and the GPS track start.

For example, I started my West Point race without remembering to start the GPS watch. I started it about 1 min 14 into the race. I can add 1:14 to the auto-generated log entry duration from Garmin, which causes the splits from the SI card to bind to the GPS trace, but of course, they are offset by 74 seconds and so are wrong.

Please add to the wish list!

OK, thanks for confirming what I thought. I will have to take the auto split off. It is just a distraction anyway.

While we are on suggestions...I would like to see is a way of having the Splits be connected to the specific log entry, especially when there are two races in the day. Right now, they show up at the bottom of the day's log.

You can try exporting track as gpx, erasing points with notepad, or you forgot to turn it on at start you can simply change time stamp of the first point to the correct start time. then just upload back to AP. I believe it might do the trick.

Changing the time stamp of the first point does cause the track to be the right duration, and the splits still bind (so they appear in the left margin of the Attackpoint GPS track map) but for some reason the resulting points don't appear on the map (and all the distances between points show as zero in the list in the left margin).

On the Winsplits upload page there is a field for "Zero Time" and the help text reads: "The zero time of the event, entered as hh.mm, e g 10.00. The zero time is found by selecting Settings in the Event menu in OE20xx / MT20xx / OS20xx. Not needed to be filled in if you are not uploading a split time file."

Not having access to the "OE20xx / MT20xx / OS20xx."software, I have always just entered something like 00:00 or 10:00. Perhaps that is why it looks like swedish time or GMT...

When there is a mixture of SI-5 and SI-6+ cards being used, and the race runs over or starts after the Noon time period, then Zero time can be important

SI-5 cards only show time as 00:00:00-12:00:00. So, if you start a race at 1 PM, and the zero time is 0, then your start time in the computer is 1:00:00.

SI-6+ cards are 24 hour clocks, so the same 1 PM start time will show as 13:00:00.

So, in WinSplits and RouteGadget and other programs which offer a chance to look at things in terms of real time, these people have actually started 12 hours apart. In RG, you can choose Mass Start to have your guys run together, but if you want to see real time head to head with someone who started near your time using an SI-5 vs. an SI-6+, then you will be sitting there a long time waiting for the second guy to start. :)

For a race where everyone starts before 12:00:00, there isn't an issue.

The way to solve this for an afternoon race where everyone starts after 12:00:00 is to have the zero time as 12:00:00 so that the 1 PM start time is interpreted the same for both people (as 13:00:00).

For a race where some people start before 12:00:00 and some after 12:00:00, then having a zero time in the software and using the same one when uploading splits is needed for real time action. At local meets, I use 6:00:00, assuming that no one will accidentally start before 6 AM or after 6 PM. Zero time can be a bit tricky at times, so caution is required.

Yeah, right now the posted splits only show up on your track when you don't have any from your watch. I need to make it just another option in the splits dropdown menu (so you can pick either), but I haven't gotten to that yet.

I'm disappointed in you Ken.

ouch, that hurts! (the marquee, that is:)

Man,
the
weird
stuff
you
can
do
with
HTML!

There goes the neighbourhood.

I think what you meant is "Man, the weird stuff you shouldn't do with HTML!"

Blame Juffy. I couldn't remember the 'quote' tag.

Ken has once again proven he is the best! The posted splits, when available, are now what shows up with the tracks on the globe! Thank you, Ken!

You can now adjust the time offset between your posted splits and gps track. Just go to the map and drag any one of the markers into the right spot (like an obvious corner).

At some point I will also make the elapsed time display show "race time" rather than watch time. but only if you guys promise no more marquees :)

Brilliant, thanks!

I hereby take the no-marquee pledge!

Its a pretty awesome feature.
You know the way it places the split markers on the map, is there any reason why it doesn't have a value in the distance or pace column? Surely they are just pulled directly from the track?

That sounds like a bug. As you say, it should be able to get that from the track.

update: should be fixed now.

Thanks