3D GPS Rides

https://www.microsoft.com/nl-be/store/p ... 40kx?rtc=1

Moderators: mcorn, malfukt, Nasz, Lunatix

User avatar
Tacx Video Production
Posts: 3832
Joined: Sun May 09, 2004 11:49 pm
Location: Belgium
Contact:

Re: 3D GPS Rides

Post by Tacx Video Production » Sat Oct 06, 2018 6:56 pm

if your using any website to make the ride then the grades will be for the most part incredibly vague, you need a barometric GPX file like from a Garmin 1000 then import that directly into the Tacx cloud & it should be a lot more accurate.

bigskypylot
Posts: 70
Joined: Wed Sep 19, 2018 4:53 pm

Re: 3D GPS Rides

Post by bigskypylot » Sat Oct 06, 2018 8:11 pm

That's kinda what I was thinking, I don't get back home as much as I would like. I guess I will just have to wait til I get back and record it again . Is there any program that can substitute for that in the meantime? Wishful thinking? :P

User avatar
Tacx Video Production
Posts: 3832
Joined: Sun May 09, 2004 11:49 pm
Location: Belgium
Contact:

Re: 3D GPS Rides

Post by Tacx Video Production » Sat Oct 06, 2018 8:29 pm

If its an area where you know others cycle a lot then I would take a Strava Premium account, you can then search the strava segments in the area of interest & then click the best times to show you the performance of the rider that posted it...from that you can actually download their GPX file. Its takes a bit of detective work if you want better data then the landsat derived grades all the websites use.

mcorn
Moderator
Posts: 7873
Joined: Fri Nov 24, 2006 8:55 pm
Location: USA, Bellingham, Washington

Re: 3D GPS Rides

Post by mcorn » Sat Oct 06, 2018 8:40 pm

Tacx Video Production wrote:
Sat Oct 06, 2018 6:56 pm
if your using any website to make the ride then the grades will be for the most part incredibly vague, you need a barometric GPX file like from a Garmin 1000 then import that directly into the Tacx cloud & it should be a lot more accurate.
Not sure that is really the issue here. Both Topofusion and 3D Route Builder show a 15-16% maximum grade for this ride, but he says that TDA showed 6% (I have not checked that personally). If what he says is true, then the TDA conversion may have done something to the route beyond normal smoothing.

bigskypylot
Posts: 70
Joined: Wed Sep 19, 2018 4:53 pm

Re: 3D GPS Rides

Post by bigskypylot » Sat Oct 06, 2018 8:49 pm

here is the screenshot

wel it says quota has been reached and won't let me attach the shot. can i send in a private message?

bigskypylot
Posts: 70
Joined: Wed Sep 19, 2018 4:53 pm

Re: 3D GPS Rides

Post by bigskypylot » Sat Oct 06, 2018 9:30 pm

Well, I just did a search on Ride-With-GPS and found the same route had been uploaded on August 31st. Now to see how that works. I uploaded and with low smoothing, it gave me 14.7%, If I recall, the Kniep Rd section might be over 20% it's been quite a while since I've ridden it. Gotta get back and ride that route again but it'll be nice to have it on the trainer for now

User avatar
Tacx Video Production
Posts: 3832
Joined: Sun May 09, 2004 11:49 pm
Location: Belgium
Contact:

Re: 3D GPS Rides

Post by Tacx Video Production » Sat Oct 06, 2018 10:53 pm

mcorn wrote:
Sat Oct 06, 2018 8:40 pm
Tacx Video Production wrote:
Sat Oct 06, 2018 6:56 pm
if your using any website to make the ride then the grades will be for the most part incredibly vague, you need a barometric GPX file like from a Garmin 1000 then import that directly into the Tacx cloud & it should be a lot more accurate.
Not sure that is really the issue here. Both Topofusion and 3D Route Builder show a 15-16% maximum grade for this ride, but he says that TDA showed 6% (I have not checked that personally). If what he says is true, then the TDA conversion may have done something to the route beyond normal smoothing.
There is more to it then just looking to see if the max grade is correct, it needs to be correct in the right place, if its landsat data the grade changes will be fairly random for most rides. If its an original barometric data derived ride there will be enough data points with the 15% grade for the filter to get the grade. Ive put our RLV ride data through it & the grades stay fairly close to the original. It needs to be fed good data. :wink:

bigskypylot
Posts: 70
Joined: Wed Sep 19, 2018 4:53 pm

Re: 3D GPS Rides

Post by bigskypylot » Sat Oct 06, 2018 11:52 pm

good to know... i'll just make do with what i can find until i make it back home. Just happy I can do a 3D ride :)

mcorn
Moderator
Posts: 7873
Joined: Fri Nov 24, 2006 8:55 pm
Location: USA, Bellingham, Washington

Re: 3D GPS Rides

Post by mcorn » Sun Oct 07, 2018 2:34 am

Tacx Video Production wrote:
Sat Oct 06, 2018 10:53 pm

There is more to it then just looking to see if the max grade is correct, it needs to be correct in the right place, if its landsat data the grade changes will be fairly random for most rides. If its an original barometric data derived ride there will be enough data points with the 15% grade for the filter to get the grade. Ive put our RLV ride data through it & the grades stay fairly close to the original. It needs to be fed good data. :wink:
You are raising the question whether the underlying altitude data are accurate, which is always a valid concern. I think there are tons of GPX files floating around that are not particularly accurate, but they can still be a fun ride. But I think if we downloaded a file that shows a grade of 16% in Topofusion or other analyzer program like 3D Routebuilder and after processing TDA shows a maximum of 6%, that would be a problem. But as I said, I have not actually checked out what TDA is doing with this particular file.

User avatar
Tacx Video Production
Posts: 3832
Joined: Sun May 09, 2004 11:49 pm
Location: Belgium
Contact:

Re: 3D GPS Rides

Post by Tacx Video Production » Sun Oct 07, 2018 8:45 am

mcorn wrote:
Sun Oct 07, 2018 2:34 am
Tacx Video Production wrote:
Sat Oct 06, 2018 10:53 pm

There is more to it then just looking to see if the max grade is correct, it needs to be correct in the right place, if its landsat data the grade changes will be fairly random for most rides. If its an original barometric data derived ride there will be enough data points with the 15% grade for the filter to get the grade. Ive put our RLV ride data through it & the grades stay fairly close to the original. It needs to be fed good data. :wink:
You are raising the question whether the underlying altitude data are accurate, which is always a valid concern. I think there are tons of GPX files floating around that are not particularly accurate, but they can still be a fun ride. But I think if we downloaded a file that shows a grade of 16% in Topofusion or other analyzer program like 3D Routebuilder and after processing TDA shows a maximum of 6%, that would be a problem. But as I said, I have not actually checked out what TDA is doing with this particular file.
No Mike thats not the case with landsat data, it could be that 1 single data point has 15% that the sample rate is not high enough to get through any level of smoothing, filtering which is vital so you do not break your knee caps going from -10% to +10% in a single second. To me Ladsat data is ok to get a rough idea of the terrain, you will climb roughly the right altitude in about the correct place. But to get any accuracy at all you need to find barometric data with a high 1sec sample rate. You can find such files on strava especially if you know the names of any local pro riders. The units are displayed also & the age the GPX files were made. Look for recent files made with a power meter attached. :wink:

mcorn
Moderator
Posts: 7873
Joined: Fri Nov 24, 2006 8:55 pm
Location: USA, Bellingham, Washington

Re: 3D GPS Rides

Post by mcorn » Sun Oct 07, 2018 5:07 pm

Tacx Video Production wrote:
Sun Oct 07, 2018 8:45 am
filtering which is vital so you do not break your knee caps going from -10% to +10% in a single second
If you look at the data points for this hill in the GPX file they are not like that. He indicates what the course is in real life, which is similar to what the GPX file shows as well.

mcorn
Moderator
Posts: 7873
Joined: Fri Nov 24, 2006 8:55 pm
Location: USA, Bellingham, Washington

Re: 3D GPS Rides

Post by mcorn » Sun Oct 07, 2018 10:03 pm

I re-imported this file to TDA with lowest smoothing and maximum slope went up from 6.2% to 7.9%, so it appears to me that TDA conversion is doing something more to decrease the slope. The decrease is not as great on lowest, but it is significantly different than the original file. It is possible that how TDA is measuring slope is different, but I was not able to check the profile on an actual TDA. I only saw the summary information for the ride.

bigskypylot
Posts: 70
Joined: Wed Sep 19, 2018 4:53 pm

Re: 3D GPS Rides

Post by bigskypylot » Tue Oct 09, 2018 3:32 pm

here are the rides, including the one i made from ride with gps. i think the steepest grade i found was 14.7% from any of these files while I know from riding it in the past, it is 18-23% if memory serves me.

https://ridewithgps.com/find#search/0/s ... ort_by%5D=

bigskypylot
Posts: 70
Joined: Wed Sep 19, 2018 4:53 pm

Re: 3D GPS Rides

Post by bigskypylot » Tue Oct 09, 2018 6:26 pm

I just looked at the same route that I uploaded to Rouvy and it shows max grade at 25% for the Kniep Rd. segment of the Alpine Valley loop. Hmmm

mcorn
Moderator
Posts: 7873
Joined: Fri Nov 24, 2006 8:55 pm
Location: USA, Bellingham, Washington

Re: 3D GPS Rides

Post by mcorn » Tue Oct 09, 2018 8:17 pm

bigskypylot wrote:
Tue Oct 09, 2018 6:26 pm
I just looked at the same route that I uploaded to Rouvy and it shows max grade at 25% for the Kniep Rd. segment of the Alpine Valley loop. Hmmm
As I said before, I think TDA smoothing is doing something more than I would expect. Even when I did the least smoothing on import into TDA, it cut it down more than it should. The purpose of smoothing is to eliminate rapid slope change anomalies, not fundamentally change the course.

User avatar
Tacx Video Production
Posts: 3832
Joined: Sun May 09, 2004 11:49 pm
Location: Belgium
Contact:

Re: 3D GPS Rides

Post by Tacx Video Production » Tue Oct 09, 2018 8:53 pm

mcorn wrote:
Tue Oct 09, 2018 8:17 pm
bigskypylot wrote:
Tue Oct 09, 2018 6:26 pm
I just looked at the same route that I uploaded to Rouvy and it shows max grade at 25% for the Kniep Rd. segment of the Alpine Valley loop. Hmmm
As I said before, I think TDA smoothing is doing something more than I would expect. Even when I did the least smoothing on import into TDA, it cut it down more than it should. The purpose of smoothing is to eliminate rapid slope change anomalies, not fundamentally change the course.
I think its tough to tell what is correct, I checked out the hills on the Alpine loop on streetview, they are hard but not 25% by the look of it to me, more like 15% max. 25% is a Koppenberg like cliff :-) The other point I said before is the sample rate this data comes from landsat. The data Ive put into it derived from known RLVs that we produced for the Video sim look solid to me, not changes in terrain, the hills are a little more smoothed but overall they are the same. Not sure if its there still but Omaha beach was in the GPX list at least in the Beta version Mike.

mcorn
Moderator
Posts: 7873
Joined: Fri Nov 24, 2006 8:55 pm
Location: USA, Bellingham, Washington

Re: 3D GPS Rides

Post by mcorn » Wed Oct 10, 2018 12:40 am

Tacx Video Production wrote:
Tue Oct 09, 2018 8:53 pm

I think its tough to tell what is correct, I checked out the hills on the Alpine loop on streetview, they are hard but not 25% by the look of it to me, more like 15% max.
The GPX files show 15% but TDS is cutting them down to 7% or 8% depending on which smoothing level is chosen, so I think the cloud is making an error on the conversion.

User avatar
Tacx Video Production
Posts: 3832
Joined: Sun May 09, 2004 11:49 pm
Location: Belgium
Contact:

Re: 3D GPS Rides

Post by Tacx Video Production » Wed Oct 10, 2018 5:59 am

If it is then it’s probably simply do to the much lower sample rate of a generated GPX. To test that you could try increasing the files sample rate, try 4x higher & then see.

bigskypylot
Posts: 70
Joined: Wed Sep 19, 2018 4:53 pm

Re: 3D GPS Rides

Post by bigskypylot » Wed Oct 10, 2018 6:36 pm

Just did a ride and everything was going well up to about 17 minutes in and then for some reason or another, it just started saying "Cycle to continue" or "press esc to exit" or something g to that effect. Bluetooth is connected and everything seems to work except for the HRM display. Any idea what's going on? So far, I really like this and would subscribe at the end of my trial period, that is if I can use it

bigskypylot
Posts: 70
Joined: Wed Sep 19, 2018 4:53 pm

Re: 3D GPS Rides

Post by bigskypylot » Wed Oct 10, 2018 6:51 pm

i tried the trainer over on my Rouvy desktop app and it doesnt work either. My laptop is close to the unit and shows the blue light on the unit. It's worked previously. It's like it just stopped working for some reason or another.

Post Reply

Return to “Tacx Desktop app (TDA)”

Who is online

Users browsing this forum: No registered users and 3 guests