WatchUSeek Watch Forums banner

1 - 20 of 36 Posts

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #1 ·
My major moves are long day hikes in the Sierra Nevada. Having done a few with the Ambit2 and Suunto Trackpod, here are the accuracy results.

Hike 1--Granite Lip: 16 miles RT, 5300 ft. net elevation gain

Ambit2: Worn on right wrist
Average EHPE: 9.00 meters
Warmup time: significant (basically the whole 5 hour trip to the park)

Trackpod: Mounted on top of stick on my pack, about head height and about 6 inches from my left ear.
Average EHPE: 4.47 meters
Warmup time: about 30 minutes an hour before getting to the park, and then only 5 minutes before starting

So you can see that the Trackpod accuracy was about double that of the Ambit2, even though it had less warmup time. Obviously the favorable mounting position was the reason.


Hike 2--Kennedy Pass: 24 miles RT, 6400 net elevation gain (trailhead is only about 13 miles from the other one)

Ambit2: worn on chest strap, about 3 inches to the right of the Motorola Ant+ HR belt coupler
EHPE: 4.99 meters
Warmup: An hour or so

Trackpod: Same mounting position
EHPE: 3.87 meters
Warmup: an hour

Garmin 62sc: worn in pouch on upper left chest
EHPE: unknown, but prevailing accuracy was 8-10 feet, with as low as 24 feet under heavy tree cover. I would guess average EHPE around 10 feet or so.
Warmup: 5 hours

So the Garmin and Trackpod performed with about equal accuracy, even though the Trackpod was in a slightly better position for sat reception. Likely the Garmin has a superior antenna and would perform better under the same mounting positions.

Even though the Ambit2 accuracy was not as good, it was much improved over wearing it on my wrist.


HR dropouts on Ambit2:

Worn on wrist on Granite Hike: zero HR readings turned up in 2.6% of the xlss log

Worn on chest strap near the HR monitor: When the pack was being worn, only 0.7% dropouts occurred. However during breaks, the watch was left on the pack so I got a lot of zero readings during that time. Not sure if it's worth it to keeping moving the Ambit2 to my wrist during breaks and then putting it on my chest strap, in order to get better data, but considering I'm also getting better GPS accuracy, it's going to stay on the chest strap while moving.

Better HR monitor syncing is the reason I wore the T6 this way as well, but the problem is worse with the Ambit. Unfortunately, I couldn't use the trusty T6 to measure HR because I wanted to use the Garmin for that, as it will record track/HR/Temp every second. I'm hoping that Suunto will release an Ant+ HR belt in time for next year's hiking season. In the meantime, I'll use the Garmin + Ambit2 for HR on hikes, and the T6 + Ambit2 for runs.

Anyway, those are the results of my comparisons, FWIW. :)
 

·
Registered
Joined
·
170 Posts
The pod sure is impressive for it's size and weight, 62sc is 8 times bigger.

What do you think of the actual track accuracy compared to EHPE you get? I am asking because for me the accuracy looks a lot better than what EHPE would indicate if carried on top of hat. So kind of even with similar EHPE readings I feel like getting worse track in upper arm strap than on hat. Been thinking it might combination of the body shadow making the error more consistently to one same direcition (upper arm case) and that making a typical kalman filter less effective, as hat gps errors are more random, better for kalman filter. But that's just guessing and I may be imagining the whole thing.
 

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #3 · (Edited)
The pod sure is impressive for it's size and weight, 62sc is 8 times bigger.

What do you think of the actual track accuracy compared to EHPE you get? I am asking because for me the accuracy looks a lot better than what EHPE would indicate if carried on top of hat. So kind of even with similar EHPE readings I feel like getting worse track in upper arm strap than on hat. Been thinking it might combination of the body shadow making the error more consistently to one same direcition (upper arm case) and that making a typical kalman filter less effective, as hat gps errors are more random, better for kalman filter. But that's just guessing and I may be imagining the whole thing.
I haven't compared that yet, but it would be hard to do so on a trail unless you can actually see the trail on Google Earth. I'll see what I can find out when I get some time. One thing to check for is how close together the up/down tracks are. The closer together they tend to be, the more accurate (especially if they appear to coincide). Of course, body shadowing could cause the tracks to be further apart then they otherwise would be, but I don't mind that error so much as long as the error is consistent and so I get the correct distance and ascent/descent data (although on a steep slope it could affect the altitude readings somewhat, since GPS is used for calibration).

Probably the best place to check for that would be on a street or on an oval track though. Would be an interesting experiment.

Regarding the stated EHPE, my guess is the actual accuracy on average would be about half of what it's estimating. That is, when it estimates 3 meters, then it means the actual position is within 3 meters, not that it actually is 3 meters. So the actual error is between 0-3 meters, and probably somewhere in the middle. Anyway, EHPE is really just a measure of signal strength, which is the important thing. The better the signal, the better the track (as a general rule). The Ambit2 track suffers because it doesn't record every second but rather every 15 seconds or so (on hikes), so you get a lot of switchback cropping.
 

·
Registered
Joined
·
41 Posts
One thing that can be very useful is for Suunto to bring out the EHPE as a parameter that can be monitored on the watch. On my fenix or the other Garmin handheld devices, I don't start my exercise right after the device locks on the satellites but I keep waiting until the "accuracy" parameter reduces and flattens out (an extra 2-3 min.)
 

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #5 · (Edited)
One thing that can be very useful is for Suunto to bring out the EHPE as a parameter that can be monitored on the watch. On my fenix or the other Garmin handheld devices, I don't start my exercise right after the device locks on the satellites but I keep waiting until the "accuracy" parameter reduces and flattens out (an extra 2-3 min.)
Yes I agree it would be very useful. Also if it would give the average EPHE at the end of a move it would be a killer feature, IMO. To know how well a track recorded, or is recording, is something I think just about everybody would be interested in knowing.

However, at least Suunto has the data available in the xml file, for post analysis, should anybody want to take the trouble. It takes me about 10 minutes or so of a long move (almost a million rows in Excel!) to get the EPHE values I'm looking for. Garmin doesn't provide accuracy data in their files (at least those files I have seen).

It's possible that Suunto thinks that it's in their best interest that people don't have the mindset to wait around until the EHPE drops to a certain level and that the watch is ready to go after it beeps in two seconds. The Ambit2 especially is famous for superfast GPS syncs and throwing accuracy into the mix may not go well with that selling point. Who knows though for sure.
 

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #6 ·
After looking at the tracks on Google Earth, it looks like the Garmin track is the best, followed by the Ambit2. The Trackpod has the poorest track. Here is a sample of each...


Garmin 62sc
Garmin62sc.jpg



Ambit2
Ambit2.jpg



Trackpod
Trackpod.jpg
 

·
Registered
Joined
·
170 Posts
Hmm, there is systematic error left or right. I don't get similar error if I place my pod properly. But I do get that too if I place it "wrong", let it shoot left or right. You had it on top of stick, right? Was the antenna shooting straight up (flat part with button straight up), like it would on top of hat, or was it shooting sideways? I believe you had it shooting left or right, making vulnerable to reflections from that direction resulting systematic shift. If you remember where it was shooting, you can confirm it by taking a look at the track shift actually is to the very same direction antenna was shooting.

Take a look at typical patch antenna pattern here: Patch antenna - Wikipedia, the free encyclopedia
and imagine shooting sideways. You end up having quite bad signal strenght for the half of the sky and satellites at thet part are not be seen directly, but instad reflections are used.
26b.jpg
For the ambit and 62sc you may have had the antenna aimed more right, or at least 62sc's antenna must have been shooting straight up if carried in breast pocket. So, if it was not alerady shooting straiaght up, I'd say before claiming pod's track is poor try aimimg it up once (and perpare to get amazed ;).
 

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #9 · (Edited)
Trackpodpic.jpg

I do indeed have it pointed straight up, although from the pic it may at about 5 degree angle. But a pack or even a head is always going to be weaving or bobbing so you have to expect that.

In any case, it reads much better up there than it did on the arm strap.

I would like a better track but I surely can't expect it to compare to the Garmin 62, which has a much better antenna. You can see the pouch for the Garmin near my left shoulder. It is nowhere near pointing straight up, but at least it is getting a good view of the sky.

Keep in mind that these tracks were taken on very steep slope in a deep canyon (Kings Canyon National Park), so the conditions weren't ideal for satellites for any of the devices.
 

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #10 · (Edited)
So, if it was not alerady shooting straiaght up, I'd say before claiming pod's track is poor try aimimg it up once (and perpare to get amazed ;).
It certainly had the best view of the sky of the three devices, and yet had the poorest track. If you see anything that can be improved upon, let me know. :)
 

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #11 · (Edited)
Here's a track sample from my previous hike (same area). I hadn't purchased the Garmin 62sc yet, and had the Garmin 76map instead (which gives a horrible track with frequent dropouts, even with an external antenna. My how have GPS devices improved in the last few years). I didn't bother to post that track, but here are the other two.

The Trackpod was mounted in same location as the pic above shows. However the Ambit2 was worn on my wrist instead of on my chest strap. As a result, the two track look similar in the amount of error, rather than the Ambit2 showing a much improved track in the other hike (by virtue of its position). So you can clearly see it's much better to wear the Ambit2 in a more stable position with the antenna pointed up whenever possible.


Ambit2
Ambit2.jpg




Trackpod
Pod.jpg
 

·
Registered
Joined
·
170 Posts
The photo explains it all. Your pod is aiming 90 degrees wrong, pretty much the worst way to have it what it comes to logging accurate track. Even uside down had given me better track that that sideways rotation. No wonder you get the systematic error you get. The flat button side of the pod should be facing up. Will make big difference, just look at the pattern at the wikipedia page and imagine how you have missed half of the sky. You should it facing up like this:
pod.jpg
 

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #13 ·
Lol, this is getting interesting. According to the Suunto Trackpod manual and the Youtube video by them, the arrow should be pointing UP. That's how I have it. And now you're telling me that's the worst way to have it? I'll gladly try it that way, but it's amusing to told how wrong it is as if it's something I should have known about.
 

·
Registered
Joined
·
1,134 Posts
Lol, this is getting interesting. According to the Suunto Trackpod manual and the Youtube video by them, the arrow should be pointing UP. That's how I have it. And now you're telling me that's the worst way to have it? I'll gladly try it that way, but it's amusing to told how wrong it is as if it's something I should have known about.
Hi. This might be a "translation" issue. "Arrow Pointing up" probabaly should be "Arrow button facing up."

AFAIK...
Flat Patch antennas, like the trackpod (and handhelds without a sticky out antenna) work best when they "face" up. Helix antennas (sticky out, like in the Garmin 60/62) should "point" up.

I believe the manual generally uses the word "facing." But shows illustrations of the unit attached to arms and straps with the face not pointed at the sky.
 

·
Registered
Joined
·
170 Posts
I had to peek at the manual (I never read manuals...). It sure is confusing, most of the pictures are like you wrote, arrow pointing up!? There is one picture with button side facing down (not arrow) as an example how it should not be positioned, that's the only indication of the best way to carry it I could find. Anyway, there must be ceramic patch antenna inside and it must be at least about 1.5cm x 1.5cm (maybe bigger, has anyone opened one yet?), so arrow up can't possibly be the way to shoot straight up, there is just no room inside to have antenna there like that. Your tracks will be a lot better, but 62cs has so much more weight it should end up being a little bit better. But maybe better carrying position will compensate a bit, so who knows, it may get close.
 

·
Registered
Joined
·
170 Posts
translation, good point. I peeked at the Finnish manual, it isn't any better. It is even talking about gps sender making it sound like it sends signals to satellites.... rrrright :)
 

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #17 ·
Trackpod was designed to be either worn on the arm strap or in a jacket pocket with the arrow pointing up. Suunto makes this clear in their video, so it's hard to believe they would recommend using it in the worst possible way for reception.

I'm not a GPS antenna expert by any means, but if helix antennas are meant to be pointed up, then they should have put a helix in the Trackpod, as UP is what it was designed for.

Thanks for the info though. I'll try it by wearing a cap on my next run, and if it makes an improvement, then I'll mount it that way as well on my next hike (in 2 weeks).
 

·
Registered
Joined
·
170 Posts
On top of hat button side of the device facing up I consistently get accuracy like this. Red track, out and back two times, so there is four lines. Yellow line is 6 meter, so here tracks are mostly within +-2 meters.
podrun.jpg

If I carry it on top of shoulder track doesn't end up quite as good, but not that bad either. Button side is facing about 30 degrees wrong and my head is shadowing the device. Blue track, three tinmes out and back, so there is six lines. Yellow line is 8 meters long. As you can see you are starting to see two line packs, one for each runnig direction. Accuracy is typically about 2 meters worse than what I get if I use it as hat gps.
shoulderpod.jpg
If I carry it as the manual pictures show, buttons side facing left or right, the error gets much larger, maybe about two times bigger or a bit more. If it shoots ahead the track will not look that bad at straights, but turns will not match (track overshoots), so there is similar error but the direction is different.

Hope this helps to estimate what kind of accuracy one should be able to get in forest with this device at best. In different cirumstances like in super heavy tree canopy or in canyons it will be worse.
 

·
Registered
Joined
·
215 Posts
I think XCJagge is right, same experience here. The manual is indeed irritating, even the German Translation.
The best results you get on top of a hat facing up the whole side with the button. I believe the antenna is positioned right beneath (beside?:-s) the triangle Button under the case and not at the top were the LEDs are placed.
 

·
Registered
Joined
·
1,115 Posts
Discussion Starter · #20 ·
Well actually this is good news for me as I was never very happy with the pod tracks. I'm looking forward to seeing improved results and will post the comparison tracks of my next hike. :)
 
1 - 20 of 36 Posts
Top