Latest build does not address distance, elevation gain issues
Support told me that the issues I had with Karoo's estimation of distance and elevation gain were being addressed with the next build. Unfortunately, with said, now-delivered, build, I haven't seen a significant improvement in Karoo's discrepancies with other platforms.
After installing this latest build, I ran both the Karoo and the Strava app on a pretty flat ride for which I've now eleven records with several devices. As a baseline, average distance for the four Strava records is 23.13 miles (standard deviation 0.25). Distance today for Strava: 22.86, for Karoo: 23.80. That's 4.1% more this time for the Karoo with both devices running simultaneously and longer still than my two Karoo records under the previous build.
Strava needn't be the gold standard for devices, and I can include two records each for a Wahoo and a Garmin. Wahoo average is 22.87 miles, Garmin: 22.64. Taken together, the non-Karoo distances average 22.94. So, Karoo under the new build runs 3.7% more in estimated distance over the rest of the pack. Yet, the devices seem to share a common trackpoint interval.
But it was the exaggerated elevation gain that grabbed my attention from the first ride with the Karoo. This is a flatland ride, gently sloping down to the turnaround point. Strava app's four records average 287.2 ft (s.d. = 3.2). The last build of Karoo had a whopping max of 718.5 (s.d. = 88.4). The current build came down to 453.0. Still, compare that to the whole non-Karoo group's 302.4 feet. Karoo is still 50% over!
Related, no doubt, is the mis-calc of my wheelset's circumferences. I've a speed sensor, paired and positioned in the sensor menu above the GPS. Karoo is currently telling me that an unloaded tire of 87.37-inch circumference is, instead, 88.21.
I'd love to love this device. I'm hoping this sort of feedback will spur Support and Dev to get after these issues. In the meantime, it seems fair that the trial period for purchases be extended till such low-hanging fruit in the issues queue have been resolved.
-
@Lorne strange, only seems to be certain units affected, I'm still running the Feb build, I haven't updated yet wanted to hear feedback from the HH community. I rode 200km at the weekend through the peak district, the elevation gain was right where I expected it to be and my friends Garmin 830 was a few hundred ft higher, 9600 vs 9800ft which sounds correct due to slightly different route to meet up point, but even if we started exactly at the same point that margin of difference is common place between different garmins even the same models.
-
You might be onto something, @Pantani. Maybe the greater the steepness of the climb, the less of a discrepancy.
I did a ~20 mi., hilly ride today with the Strava app and the Karoo running. Karoo's 4.8% distance overage matches the above observation of 4.7%. But Karoo's 640 ft. elevation gain was "only" 20% more than Strava this time. That's a drop from the above flatland ride discrepancy of 50%.
Again, for comparison with other platforms, three matched rides with a Wahoo were inconsequentially different (0.05%, s.d. = 0.09) from today's Strava distance. Their estimated elevation gain ran 5% over Strava (s.d. = 2).
-
I'm experiencing significant mileage differences with the latest build (April 22 of 2021) for the Karoo 2. When I upload my ride to the Hammerhead website the map shows the correct mileage, but the unit records a much higher (and incorrect) mileage that it sends to Strava. I did a 55.4 mile ride with a friend who was using a Garmin to record the ride, which is the mileage Hammerhead displays on their website. My Karoo 2 recorded the ride at 57.5 miles...a pretty significant difference of 2.1 miles.
-
@Berry Do you use a speed sensor? As (looking at similar topics) there seem to be issues with the auto-calibration of the speed sensor, this may explain the behavior you are seeing.
I decided to get rid of my speed sensor. Was using one for years with my Wahoo Bolt, otherwise I would have a lot of speed fluctuations. Now with the K2 I hardly see any speed fluctuations. I guess because of a better GPS accuracy.
-
Interesting idea, @David. I pretty much noticed the discrepancy on every of the 30+ rides I did before I had to give up on the device at the end of the trial period. Most were done in less than 10 mi./hr wind conditions.
I would advocate for the option to disable the barometer and both extend battery life and take some of this variance out of the elevation gain readings.
Please sign in to leave a comment.
Comments
7 comments