My Karoo Test Journal

[Steve]

Garmin has had ten years to get it nearly right.

That's because Garmin didn't have any real competition though. HH don't have that luxury. If they want to compete with the big boys then they need to get it right an awful lot more quickly. Otherwise they're going to be out of business. HH have over-promised and under-delivered.

[drisotope]

It's been a minute since I posted in here,but today was so good an example I had to update.

Edge 520 / Karoo

Moving Time
2.09.03 / 2.09.11
Distance
35.40 miles / 35.59 miles
Avg. Speed
16.5mph / 16.5mph
Elapsed Time
2.13.05 / 2.12.51
Elevation
1,220ft / 1,227ft* (Strava corrected to 1,339ft)
Weighted Avg. Pwr
206W / 197W
Total Work
1,355kJ / 1,365kJ
Avg. Cadence
77rpm / 77rpm
Avg. Temperature
58ºF / 62ºF

This was perhaps the closest outcome between the two units that I've had... ever. The elevation, the distance, everything.

The Karoo continues to consume battery at an utterly unpredictable rate, but auto-pause works great, it's (mostly) not freezing sensors as of late, and even the thermometer isn't giving me random readings of like 100ºF just because it's in direct sunlight. I mean, sure, I did forget to power it off a few days ago and it ate all 88% of it's battery in less than 12 hours while in standby, but it could be worse.

A kilojoule readout would still be really, really nice.

[dthempel]

10w average power might be considered "significant" depending on your training and what you were doing. Might be the difference between zones you were trying to hit. Plus, the over 200W number always looks better.

[Steinello]

I use the Karoo since mid April and track my rides (additonal) with my smartphone and Locus Maps Pro.
I can´t see great differences between the Karoo and Locus and I trust in Karoo at this status
(... my experience with other Software (e.g. Komoot) shows greater differences especially at the altitude information)
I think Hammerhead has done a good job in the meanwhile - but it would be nice if they use the integrated compass,
that we can use the Karoo for orientation also when we stop driving.

[drisotope]

10w average power might be considered "significant" depending on your training and what you were doing. Might be the difference between zones you were trying to hit. Plus, the over 200W number always looks better.

Well, that comes down to whatever math the different sites are using-- Garmin Connect vs. Strava. The Garmin had me at a 175W average, and the Karoo at 176W average. I still contend that the sluggish behavior of the auto-pause makes for, umm... inaccurate accuracy? The Karoo takes 2-3 seconds to pause, and 2-3 seconds to unpause, so in the end it works out as nearly identical moving time to the Garmin, but misses the few seconds when I'm making the most power, which would explain the 10W difference in normalized power. Both units recorded a max output of 685W and a 211W 20-minute power for the ride, so the differences are getting ever smaller.

This morning was very similar:

Karoo: 172W raw avg. power, 189W normalized, 193W 20-minute, 800W max
Garmin: 172W raw avg. power, 197W normalized, 191W 20-minute, 800W max

At this point, I just take the numbers as they are. As the difference between Tuesday and Wednesday was all of 2W, the intensity (IF) and Power Stress (PSS) were almost identical per hour, but today was physically much less intense. Like the days when I'm on sandy fire roads so blown out I spend a half hour walking, and most of that time carrying the bike. Power meter gives me a big zero for that, so my raw power will be like 120W, and weighted is like 230W.

[reggieb]

10w average power might be considered "significant" depending on your training and what you were doing. Might be the difference between zones you were trying to hit. Plus, the over 200W number always looks better.

Well, that comes down to whatever math the different sites are using-- Garmin Connect vs. Strava. The Garmin had me at a 175W average, and the Karoo at 176W average. I still contend that the sluggish behavior of the auto-pause makes for, umm... inaccurate accuracy? The Karoo takes 2-3 seconds to pause, and 2-3 seconds to unpause, so in the end it works out as nearly identical moving time to the Garmin, but misses the few seconds when I'm making the most power, which would explain the 10W difference in normalized power. Both units recorded a max output of 685W and a 211W 20-minute power for the ride, so the differences are getting ever smaller.

This morning was very similar:

Karoo: 172W raw avg. power, 189W normalized, 193W 20-minute, 800W max
Garmin: 172W raw avg. power, 197W normalized, 191W 20-minute, 800W max

At this point, I just take the numbers as they are. As the difference between Tuesday and Wednesday was all of 2W, the intensity (IF) and Power Stress (PSS) were almost identical per hour, but today was physically much less intense. Like the days when I'm on sandy fire roads so blown out I spend a half hour walking, and most of that time carrying the bike. Power meter gives me a big zero for that, so my raw power will be like 120W, and weighted is like 230W.

Yeah, the weighted average is what it is, different calculations. If Strava is your system of record, then they'll calculate their own thing any way, so the raw matters more, and if that's identical, you're going to see the same thing in Strava. Do you, by the way, let them both upload to Strava and then compare apples to apples there ever? Strava actually might agree more with the Karoo, its weighted average tends to come out a touch lower than my Garmin 810 any way.

I am going for my first ride with my new Karoo (resale value on them isn't great, so I snagged one off eBay) tomorrow. It can't take the place of my Garmin yet - it will need to be able to zero my power meter first. I'd also like to see structured workouts, but I am at the point where I mostly do those on Zwift, so I guess it doesn't matter as much to me. Reading your history is encouraging, there has been an upward trend in the quality over time for sure.

I just wish they'd open a damn app store. Doing so would get some of the weight off their shoulders, outsource the feature updates and focus on making the core product better. I've been bugging them about this via support, but I don't think anything I say gets passed on any way.

[drisotope]

It's odd-- in general, Strava will prevent duplicate uploads-- but probably around once a week, I get a ride where I rode with myself, and Strava takes both the Karoo and the 520 uploads together. The numbers for power will generally be within a watt.

In true Karoo fashion, I put in 50+ miles yesterday at Bike the Coast in Oceanside-- and unlike the nonsense the Karoo did on Thursday, it was (nearly) perfect. At around the 3-hour mark of total time, the Karoo was down about 45% on battery, so I plugged it in, as I knew I had at least 2 hours of total time left, and I'm not about to let the battery get below 30%. But recording and display-wise, A-OK. No sensor drops, uploaded perfectly, even with poor cell signal.


Edge 520 / Karoo

Moving Time
3.38.19 / 3.41.02
Distance
53.84 miles / 55.13 miles
Avg. Speed
14.8mph / 15.0mph
Elapsed Time
4.42.44 / 4.42.46
Elevation
1,512ft / 1,580ft* (Strava corrected to 1,660ft)
Unweighted Pwr
122W / 142W
Weighted Avg. Pwr
172W / 153W
Total Work
1,597kJ / 1,641J
Avg. Cadence
69rpm / 69rpm
Avg. Temperature
64ºF / 72ºF

Again, it gets kooky after upload. The Edge is saying less total work, less everage power, but significantly higher weighted average-- despite both units recording the exact same max power. I also have no idea where the Edge lost the 3 minutes and 1 mile, because elapsed time would indicate that I started both units together. BTW, the final unadjusted elevation on my wife's Edge 500 was a scant 30 feet lower than Strava's correction.

Why does every GPS unit I use read so low on elevation, all the time?

[JohnJ80]

I just wish they'd open a damn app store. Doing so would get some of the weight off their shoulders, outsource the feature updates and focus on making the core product better. I've been bugging them about this via support, but I don't think anything I say gets passed on any way.

You can sideload pretty much anything you want. I wrote some directions here:

https://github.com/JohnJ80/Karoo-App-Sideloading/wiki

I run a podcast app, a weather radar app, and both Osmand and Komoot or navigation (gets you audio TbT cues).

Since in the last update HH installed a standard ANT+ stack, other apps like Trainerroad, Zwift etc.... apparently talk to it too.

J.

[Steve]

Why does every GPS unit I use read so low on elevation, all the time?

When you say "so low", what are you comparing with, and how do you know which is the "correct" value?

Elevation is notoriously difficult to calculate. On maps, we have to rely on elevation basemaps, but what happens when you go over a raised structure like a bridge, is the elevation given at ground level or at the bridge level, for instance? On GPS devices we have barometric sensors, but they are easily confused by varying weather conditions.

Unless you go along a route and physically survey it to find the elevation of every up and down bit, you're never going to know what the correct value is.

[reggieb]

I just wish they'd open a damn app store. Doing so would get some of the weight off their shoulders, outsource the feature updates and focus on making the core product better. I've been bugging them about this via support, but I don't think anything I say gets passed on any way.

You can sideload pretty much anything you want. I wrote some directions here:

https://github.com/JohnJ80/Karoo-App-Sideloading/wiki

I run a podcast app, a weather radar app, and both Osmand and Komoot or navigation (gets you audio TbT cues).

Since in the last update HH installed a standard ANT+ stack, other apps like Trainerroad, Zwift etc.... apparently talk to it too.

J.

Oh, for sure, I know I could sideload. But most users won't, from a business perspective, they could focus on the OS and base features, and crowdsource the work of creating new features, like structured workouts, etc. within their ecosystem, as opposed to a standard Android launcher and stock apps.