Contact   Imprint   Advertising   Guidelines

Trace errors compared to Woo

forum for kitesurfers


User avatar
deniska
Frequent Poster
Posts: 333
Joined: Fri Aug 30, 2013 3:40 pm
Kiting since: 2011
Local Beach: Plumb Beach
Favorite Beaches: Cabarete
Gear: Ozone, Core
Brand Affiliation: None

Re: Trace errors compared to Woo

Postby deniska » Mon Jan 29, 2018 3:34 pm

I donno about Xensr and best accuracy claim.. A friend of mine had one some time ago and posted some unbelievable sessions.. like boosting 16m in 16kts..
He ended up returning the device and getting a woo 2.0 instead..

@fluidity: there are a bunch of white papers & algorithms, already written to isolate true vertical acceleration and filter out the noise. You can play with them on your smartphone.. As I said the bottleneck is the accuracy of consumer accelerometers. This is why neither woo,trace,piq or Xensr tell you which chip they use..

CaliRider
Medium Poster
Posts: 84
Joined: Mon Dec 19, 2016 7:17 pm
Kiting since: 2013
Local Beach: 3rd Ave, Cotote Pt., Crissy
Favorite Beaches: Kite Beach Maui
Gear: Slingshot for all!
Brand Affiliation: None

Re: Trace errors compared to Woo

Postby CaliRider » Mon Jan 29, 2018 4:03 pm

deniska wrote:
Mon Jan 29, 2018 3:34 pm
I donno about Xensr and best accuracy claim.. A friend of mine had one some time ago and posted some unbelievable sessions.. like boosting 16m in 16kts..
He ended up returning the device and getting a woo 2.0 instead..
I asked the intel dude about that very thing since thats what i heard and he said something to the effect that there were some errors in the xenser original units which gave the occasional weird result but they fixed the software and that and the results that were not weird were spot on. sounded like the intel helped them fix the problem?? but they used xensr stuff in the x games no problems. i use the sessions app which has never given me a big jump but sometimes i wish it would. :D perhaps the apple watch is better quality on the chip that makes it all work? ? i dunno.

andylc
Frequent Poster
Posts: 326
Joined: Fri Apr 07, 2017 3:26 pm
Kiting since: 2005
Local Beach: Exmouth
Favorite Beaches: Saunton Sands
Gear: Ozone Enduro V2 8m, 11m
Flysurfer Soul 12m
Lieuwe Shotgun 132
Brand Affiliation: None

Re: Trace errors compared to Woo

Postby andylc » Mon Jan 29, 2018 10:09 pm

So you'd say you think the Apple watch + Xensr sessions is a good option? I'm a bit nervous about losing my watch in a wipeout, not that it has ever fallen off my wrist before...thinking might get the Catalyst case for extra protection...

CaliRider
Medium Poster
Posts: 84
Joined: Mon Dec 19, 2016 7:17 pm
Kiting since: 2013
Local Beach: 3rd Ave, Cotote Pt., Crissy
Favorite Beaches: Kite Beach Maui
Gear: Slingshot for all!
Brand Affiliation: None

Re: Trace errors compared to Woo

Postby CaliRider » Tue Jan 30, 2018 3:01 am

andylc wrote:
Mon Jan 29, 2018 10:09 pm
So you'd say you think the Apple watch + Xensr sessions is a good option? I'm a bit nervous about losing my watch in a wipeout, not that it has ever fallen off my wrist before...thinking might get the Catalyst case for extra protection...
Its cold here in northern CA, so I'm always running long sleeves so my watch is kinda protected. I also switched to a watch band with an actual latch and not those stupid holes and pins of the apple and nike bands. those things can come off just messing around with my gf. LOL

User avatar
deniska
Frequent Poster
Posts: 333
Joined: Fri Aug 30, 2013 3:40 pm
Kiting since: 2011
Local Beach: Plumb Beach
Favorite Beaches: Cabarete
Gear: Ozone, Core
Brand Affiliation: None

Re: Trace errors compared to Woo

Postby deniska » Tue Jan 30, 2018 5:31 am

That woo reading on Kevin's KOTA jump seems fairly accurate to me
as you can stuck up 11-12 Kevins..
Most of the other guys used woo too so it's pretty easy to compare their readings with video footage...
kevin.png

joriws
Very Frequent Poster
Posts: 675
Joined: Wed Nov 24, 2010 11:03 am
Kiting since: 0
Gear: Flysurfer, HQ, LF, Nobile, North, Ozone
Brand Affiliation: None

Re: Trace errors compared to Woo

Postby joriws » Tue Jan 30, 2018 6:14 am

Yeah 2.7s hangtime ;) Very real for 21m jump. Hangtime comes from how many imu records indicate he's on air.

Kevin-body-meter is video-angle-measure which heavily depends on distance of Kevin to camera. Kevin's body position and posture. Soloshot 3 might be able to get more accurate guess of height as it knows approx distance of tag at rider, lens focal length and probably robot camera angles it films. But it is also affected by lens distortions like pillow/barrel. But it is a bit similar to tachymeter/theodolite. But still finally taking height from rider's position at camera frame brings errors.

Only accurate cheap video measure is to put many cameras on pole to create camera to horizon plane on certain heights. Break that horizon line and you are above the height of camera filming the plane break.. In screen capture Kevin is definetly higher than camera-horizon plane. How high is that camera position? Deniska's example pic and about the planes you can see cargo ship on the background. Let's say bridge is at 20m height and crane towers 30m. Below Kevin, Kevin must be above 30m. Not!

I refer earlier KOTAs with height meter-pole on sea wall. Cargo ship on right side with standard sea containers either 2.6m or 2.9m in height. The topmost container must be around 6 containers high = 15m or 17m (rounded down). So crane top must be 25-30m high. Or 15m pole is midway of the background mountain so mountain must be 30m high. Planes & angles & perspective are the reasons you cannot video measure.

Image
Last edited by joriws on Tue Jan 30, 2018 11:45 am, edited 1 time in total.

wrogu
Medium Poster
Posts: 182
Joined: Sat Sep 11, 2010 9:12 am
Kiting since: 1
Brand Affiliation: None

Re: Trace errors compared to Woo

Postby wrogu » Tue Jan 30, 2018 8:54 am

joriws wrote:
Tue Jan 30, 2018 6:14 am
Yeah 2.7s hangtime ;) Very real for 21m jump. Hangtime comes from how many imu records indicate he's on air.
Its not 2.7s Hangtime. The 2.7s means it is the 2.7th second of the jump - the pic comes from a video synced with woo.
Kevin-body-meter is video-angle-measure which heavily depends on distance of Kevin to camera. Kevin's body position and posture.
Well, Kevin's body is in the plane ehre you measure height, so there is no better way to calculate height, unless you have any other reference object in that plane - his board maybe, if angle is right. Imho the biggest error comes from wrong estimation of takeoff point. And yes, such measurements are only an estimation .... I can't understand why people want 0,1m accuracy on jumps. It is not a lab :)

fluidity
Medium Poster
Posts: 113
Joined: Sat Mar 19, 2016 11:20 pm
Kiting since: 2016
Local Beach: Ngati Toa, Plimmerton, Titahi Bay, Waikanae, Petone, Seatoun, Lyall Bay, Eastbourne, Lake Wairarapa
Style: Wave, jump
Gear: Old Flexfoil 9, Switch Element V4 11m, Ocean Rodeo Prodigy 7m 2014, Nitro14m v6, Nitro9m v6 12mSwitch Legacy2
DIY CAD ultra-concave wave twintips 1500 and 130mm with my own fin designs. Easy upwind, awesome carving.

Re: Trace errors compared to Woo

Postby fluidity » Tue Jan 30, 2018 10:24 am

deniska wrote:
Mon Jan 29, 2018 3:34 pm
@fluidity: there are a bunch of white papers & algorithms, already written to isolate true vertical acceleration and filter out the noise. You can play with them on your smartphone.. As I said the bottleneck is the accuracy of consumer accelerometers. This is why neither woo,trace,piq or Xensr tell you which chip they use..
Yes... those white papers and algorithms make it very easy. Wouldn't be surprised of most of the original woo software written is actually around the android app!

nothing2seehere
Frequent Poster
Posts: 452
Joined: Fri Jun 27, 2014 3:25 pm
Kiting since: 2012
Gear: Kitesurfing and windsurfing
Brand Affiliation: None

Re: Trace errors compared to Woo

Postby nothing2seehere » Tue Jan 30, 2018 11:00 am

bmcfiv wrote:
Mon Jan 29, 2018 12:17 pm
CaliRider wrote:
Sun Jan 28, 2018 9:25 pm
While grabbing my coffee this morning in the south bay i met a guy Geoff that works for Intel and knew all about sensers for kitesurfing. he said they did some testing of all the devices and trace ranked the worst in jump heights, pique was better but not by much, woo was pretty good but its jump heights were always high (he laughed at the woo records, said those guys are nowhere near as high as woo says they are!) and xenser was the most accurate but they don't seem to make their unit anymore. i asked how they tested and he said they did some test where all the units were on the same object and they 'jumped' it and it had some other measurement device on it that he said was like $25k (!!!). This dude was freakin' smart! He was on the team that did the X games for intel last year with snowboard data. He said the trace guys aren't supporting their product anymore and the french guys are going to run out of money soon. Who's french? He also thought woo was about to be sued because their See Em (C M?) stole the 'jump alego' from the xensr guys. Sounds like a shitload of drama to me. I told him to make the testing data available. Love to see their stuff given this thread.

Wish I paid more attention in math class. lol
Wow, it would be great to see that data and a "proper" analysis. The other items' you mentioned have some serious implications, so assuming you are right, I'll try to summarize.

Overall
Early stage industry, small market, so expect compromises
It's a toy lol, not a pacemaker, enjoy it as a partial solution
Woo is the overall choice, Trace or a Woo with second GPS device (phone) best is you need location

Woo
Good accuracy
Solely focused on kiteboarding, software is great and social
Some purchase liability with pending Xensr lawsuit (questionable)

Trace
Poor jump accuracy
GPS integration solid
Software mediocre, social ok
"Not supporting product", is this kiteboarding or everything?
Looking at Facebook and other posts, it does seem to be quiet over there

PIQ
Poor jump accuracy
You can strap it to your hand and see how hard you can hit "kiterocky" in the face
Cool display that you can't really see
Going broke

Xensr
Unknown b/c of conflicting data
Best accuracy
Manufacturing issues / not selling anymore (new website suggests otherwise)
Looks like they reached for the moon with features, but probably didn't have funding to do properly, would love to see them succeed
What happened to Android watch integration?

Apart from this chap from Intel, I don't suppose you have (heard) any other sources about the accuracy?

Personal experience led me to believe PIQ was on a par with Xensr but we only had 2 sessions before the Xensr unit died (again). The back to back tests I have read (wetestkites etc) have suggested that PIQ and Woo were showing similar results (add a bit for Woo - but in general a big jump on one was a big jump on the other and the difference in the mounting position (fore/aft/edge/centre) could account for the height difference).

Sure you can 'game' the system by moving the sensor about on your board so it registers high but whats the point? If you are going to do that you might as well start decompiling the app and not even bother going to the beach - just post a session from work.

User avatar
Toby
Very Frequent Poster
Posts: 37201
Joined: Mon Sep 24, 2001 1:00 am
Kiting since: 2000
Local Beach: Cumbuco, Brazil
Barra do Cauipe, Brazil
Style: Airstyle
Gear: Rebel 2015 18
Brand Affiliation: None.
Location: World (KF Admin)
Contact:

Re: Trace errors compared to Woo

Postby Toby » Tue Jan 30, 2018 1:34 pm

0.1 accuracy can put you in front of the leaderboard ...so it does make a difference!


Return to “Kitesurfing”

Who is online

Users browsing this forum: kite_hh and 26 guests

cron