First Lap/Interval Offset - Coros

Hei,
it looks like there is an offset in time for laps recorded with a Coros Pace 3, that are used in an activity screen. Subsequent laps/intervals have the correct duration.

Using the Actions → Use Laps button maps all the recorded laps to intervals.
This mornings run had a first lap being 15 minutes long, but the interval is only showing up with 13m 34s. This introduces an offset for all subsequent intervals.
The missing 1m 26s might be the amount of time between choosing an activity on the watch and hitting the start button after it found the GNSS satellites and other sensors.

I’ve been observing that behaviour for many activities that are based on the Coros device, but don’t use (auto) laps that often. So it hasn’t been bothering me too much. :wink:
Can I provide you with more information to find the cause of the bug?

Thank you for a nice tool to visualise all the activities, highly appreciated!

Did you press the START button before all sensors and satellites are obtained?

If you did, and the 1st 1m26s didn’t have any GPS data / Velocity / power etc, then it might be the reason why it’s regarded as “non-moving” time and ignored.

You can try these

I waited until everything was ready. So my assumption is that this patience is accumulating the missing time.
Maybe Coros is recording/writing something like a “recording start time” and all laps are related to that point in time and not the actual “I’ve hit the start button time” start of the activity which initiates collecting data from all sensors. :thinking:

interesting… perhaps take a look at the raw FIT file via fitfileviewer and see what sort of data is there/ is missing on the first 1+min?

Interesting tool, thanks for the pointer. :slight_smile:

When I’m looking at the fit file downloaded from the source I get the following insights:
The activity row looks as follows:

timestamp total timer time(s) num sessions type event event type local timestamp
8/23/2024 7:23:54 AM 38:23.220 1 manual activity stop 8/23/2024 9:23:54 AM

The event rows:

timestamp(s) event event type event group
8/23/2024 7:23:54 AM timer start 0
8:02:18 AM timer stop all 0

And the first few Lap rows:

timestamp(s) start time total elapsed time(s) total timer time(s) total distance(m) total calories(kcal) avg heart rate(bpm) max heart rate(bpm) avg cadence(rpm) max cadence(rpm) avg power(watts) total ascent(m) total descent(m) sport avg temperature(C) min heart rate(bpm) avg vertical oscillation(mm) avg stance time percent(percent) avg stance time(ms) enhanced avg speed(m/s) enhanced max speed(m/s) avg vertical ratio(percent) avg step length(mm) Effort PaceIQ(m/s)
8/23/2024 7:38:54 AM 8/23/2024 7:23:54 AM 15:00.000 15:00.000 2,328.66 157 139 155 83 88 201 70 17 running 22 108 0.0 0.00 0.0 2.587 3.226 0.00 920.0 2.881000
7:39:24 AM 7:38:54 AM 30.000 30.000 144.78 6 156 163 92 98 264 0 7 running 20 148 0.0 0.00 0.0 4.826 5.076 0.00 1,570.0 4.716000
7:39:54 AM 7:39:24 AM 30.000 30.000 84.09 7 161 165 87 93 205 0 1 running 19 157 0.0 0.00 0.0 2.803 4.926 0.00 960.0 2.590000

Moreover, there are record rows that start with the start time of the first lap.

However, the first interval is shown with the following data:

Elapsed Time Pace GAP Avg Cadence Avg HR Max HR Avg Gradient Zone Distance Altitude Gain Stride Avg Power
13m34s 6:32 5:45 83 138 155 2.2% 2 2.07km 62m 0.92 197
30s 5:37 5:20 87 146 147 1.1% 2 89m 0m 1.03 208
30s 5:49 5:19 87 149 151 2.3% 2 86m 2m 0.99 219

So to me it looks like the *.fit file has both, elapsed time and combination of start time and timestamps that fit. The only odd thing that I’m spotting is that the local timestamp is for some reason in UTC+4 whilst the time stamps are in UTC+2. But that is probably some different aspect.

Looking at the fit file that I can download from intervals.icu, I see that the first lap shows the interval duration in accordance with what is shown in the visualisation. Interestingly, the intervals.icu export shows an activity timestamp that describes the end of the activity, whilst the Coros file shows the start of the activity.

So maybe something is happening during the initial sync/import into intervals.icu that mangles the first lap?

intervals.icu doesn’t use the acitiy LAP data as is from the FIT file. It will go thru the second by second data in the “RECORDS” portion of the data.

That’s where need to look at it… Perhaps you can see if there is GPS data, or velocity and stuffs like that.

1 Like

I see your activities are getting to Intervals.icu via Strava. Do you have the Strava privacy circle enabled? Unfortunately that often causes Strava to return bad lap data from the API. It’s a longstanding bug on their side.

You could try connecting Coros as well as Strava. Intervals.icu will favour the activity from Coros but will keep segments from Strava.

1 Like

That is the explanation! I have the privacy circles enabled and that took part from the first interval.
Usually I take the hidden start bit away when it doesn’t matter to me, but thats after the sync cascade has happened. So reprocessing the activities did the trick already. :slight_smile:

Thank you both for your time!

Short summary, if somebody stumbles over that thread later:
Stravas privacy circles mess around with using laps for interval detection. Reprocessing the activity after removing the hidden start on Strava fixes the issue.

1 Like