Trailfork hero - in reality likely a zero- Mtbr.com
Results 1 to 25 of 25
  1. #1
    mtbr member
    Reputation: RRauzer's Avatar
    Join Date
    Jul 2019
    Posts
    20

    Trailfork hero - in reality likely a zero

    So today, the weather was good so I did this:



    Now, I've noticed the Climb/Decent on Trailforks have seemingly been far off from what I presume is reality on certain rides of mine before, but this seems like a doozy today. I use the app on my iPhone which I keep in my pocket.

    Has anyone done this route with a fancier gps that they logged and can tell me the climb? Since it was straight out and back on three complete sections of trails, is adding up the total numbers in TF even accurate?

    When I got back into mountain biking last September after a multi-decade (yup) gap I honestly didn't think I'd care care about any metrics, but... having lots of fun now and interested in my progress on the various fronts.

    Also, to-date, I've only ridden alone so have had no one to gauge myself stamina-wise by. I just found/signed up for a local Meetup group because it'd be fun to go with others at times and case in point, there is a ride down in that S.C. demonstration forest tomorrow that said 2500ft and "being comfortable w/ long climbs". I would of gone but didn't want to potentially be a boat anchor, hence interested in a better understanding of what I'm actually doing.
    Attached Thumbnails Attached Thumbnails Trailfork hero - in reality likely a zero-img_6082.jpg  


  2. #2
    mtbr member
    Reputation:
    Join Date
    Aug 2015
    Posts
    150
    If you throw in Peter's Creek loop it's 5k feet, so this is should be less that 4k.

  3. #3
    mtbr member
    Reputation:
    Join Date
    Sep 2013
    Posts
    46
    I have noticed that trailforks climb/descent can be way off, also ride time vs total time, which throws off your avg mph

  4. #4
    mtbr member
    Reputation: SikeMo's Avatar
    Join Date
    Mar 2013
    Posts
    580
    Trailforks has nice little snippets about trails, but if you want accurate information, use ridewithgps.com

  5. #5
    mtbr member
    Reputation:
    Join Date
    Mar 2014
    Posts
    490
    I did that ride this winter, but dropped down into Stevens Canyon and climbed up Charcoal Rd. I think it was about 4500 ft. according to Strava. It was beautiful out there today!
    Ride fast. Huck bravely. Waste no beer. Safety third.

  6. #6
    mtbr member
    Reputation: mstrkrft69's Avatar
    Join Date
    Jan 2018
    Posts
    194
    I've noticed that Trailforks bones my elevation on a pretty regular basis. Bit of a bummer since I'd like to get accurate info on my rides.

    Case-in-point. . .

    I'd love to think think the standard loop at China Camp involves 4500' of climbing, but.
    . . . . uuuhhhh No. So is TF recording every time I stand up and sit down on my saddle?? I mean I dig accuracy like that but I'd like to get a fair estimate of my days. MTB project lists that ride as 800' of climbing. So how does TF get so drastically off? It also recorded the ride 1 mile longer than the MTB project stats.

    So what works the best?? Or is my pixel just a piece of sh!t? Not really interested in using two devices on my rides, ala Garmin, but maybe that's what needs to happen.

    Sent from my Pixel 2 XL using Tapatalk

  7. #7
    mtbr member
    Reputation: Sid Duffman's Avatar
    Join Date
    Oct 2015
    Posts
    445
    Never done that exact route, but I ride from hwy 9/35 intersection, skyline trail, JNT, down to the reservoir, and back, and itís a shade under 3000ft elevation gain according to Strava tracking.


    Sent from my iPhone using Tapatalk

  8. #8
    Moderator Moderator
    Reputation: 5k bike 50cent legs's Avatar
    Join Date
    Oct 2016
    Posts
    5,527
    Iím guessing that ride is about 2000 ft of climbing, if Iím reading your track correctly. The good news is if you can do this ride you can ride the Ďusualí loop Buzzards to Flow to Hihns Mill etc at Demo with no problem.

  9. #9
    Sweater
    Reputation: Buzzaro's Avatar
    Join Date
    Jan 2008
    Posts
    1,932
    Quote Originally Posted by mstrkrft69 View Post
    I've noticed that Trailforks bones my elevation on a pretty regular basis. Bit of a bummer since I'd like to get accurate info on my rides.

    Case-in-point. . .

    I'd love to think think the standard loop at China Camp involves 4500' of climbing, but.
    . . . . uuuhhhh No. So is TF recording every time I stand up and sit down on my saddle?? I mean I dig accuracy like that but I'd like to get a fair estimate of my days. MTB project lists that ride as 800' of climbing. So how does TF get so drastically off? It also recorded the ride 1 mile longer than the MTB project stats.

    So what works the best?? Or is my pixel just a piece of sh!t? Not really interested in using two devices on my rides, ala Garmin, but maybe that's what needs to happen.

    Sent from my Pixel 2 XL using Tapatalk
    I especially like how it thinks 4500ft/hr is normal.
    All out of S**** and down to my last F***

  10. #10
    Moderator Moderator
    Reputation: 5k bike 50cent legs's Avatar
    Join Date
    Oct 2016
    Posts
    5,527
    Quote Originally Posted by Buzzaro View Post
    I especially like how it thinks 4500ft/hr is normal.
    I wonder how many people have used this as a fitness gauge, only to go somewhere and get in way over their heads. Old Growth Classic, 55 miles, 8000 ft of climbing, pffft that's nothing, I climb 4500ft in less than an hour, no problem!

  11. #11
    mtbr member
    Reputation: boomn's Avatar
    Join Date
    Jul 2007
    Posts
    10,035
    I too have found that Trailforks seems to sum up every last bit of elevation change along the way. I sync my rides to Strava as well and the same GPS recording processed through their calculations provides a much more "normal" calculation.

  12. #12
    mtbr member
    Reputation:
    Join Date
    Aug 2011
    Posts
    62
    Another source of misinformation is gaiagps. If you plan a route using that tool, it gives you a very high elevation number, compared to other sources.

    Of course, there is no real truth to the matter - it depends on how detailed your elevation model is. If you add up every pebble, you could end up with a very high number... Some standard would be nice, so people can share and compare data effectively.

  13. #13
    mtbr member
    Reputation: paintballeerXC's Avatar
    Join Date
    Jun 2005
    Posts
    2,000
    Does no one really use actual GPS devices anymore? Garmin or Wahoo?

  14. #14
    mtbr member
    Reputation: RBoardman's Avatar
    Join Date
    Dec 2014
    Posts
    2,505
    Quote Originally Posted by paintballeerXC View Post
    Does no one really use actual GPS devices anymore? Garmin or Wahoo?
    Strava iPhone app is usually fairly accurate to my Garmin computer. Trail forks is good for navigation and elevation profiles, but useless for elevation gain as shown above.

    OP: That ride is about 3550' elevation gain for the route you did.

  15. #15
    mtbr member
    Reputation:
    Join Date
    Feb 2004
    Posts
    198
    There's been conversation on the Trailforks bug topic on that other forum about this. They say they're working on it.

  16. #16
    poser Administrator
    Reputation: rockcrusher's Avatar
    Join Date
    Aug 2003
    Posts
    9,700
    Quote Originally Posted by 5k bike 50cent legs View Post
    I wonder how many people have used this as a fitness gauge, only to go somewhere and get in way over their heads. Old Growth Classic, 55 miles, 8000 ft of climbing, pffft that's nothing, I climb 4500ft in less than an hour, no problem!
    Lol!

    I use mapmyride, usually fairly accurate for elevations here around the seattles.
    MTBR Posting Guidelines
    calories>electrons

  17. #17
    mtbr member
    Reputation: RRauzer's Avatar
    Join Date
    Jul 2019
    Posts
    20
    Quote Originally Posted by paintballeerXC View Post
    Does no one really use actual GPS devices anymore? Garmin or Wahoo?
    I'm just starting (as in the last 24 hrs) to research some of the low/mid tier ones, along with the other apps mentioned in the thread. I think I'm like how a lot of people start out continue onwards: Get a bike, have fun, want to find trails, someone mentions TF, use it, have more fun, and then get the itch to "level up" by wanting to know some of your metrics.

  18. #18
    mtbr member
    Reputation: boomn's Avatar
    Join Date
    Jul 2007
    Posts
    10,035
    Quote Originally Posted by paintballeerXC View Post
    Does no one really use actual GPS devices anymore? Garmin or Wahoo?
    Even with a dedicated GPS device, if you upload the recording to TF you will get a very optimistic sum of your elevation changes. And if you upload the file to Strava you will get a completely different number. There isn't a standard for how it should be calculated

    Also, I believe both of these sites might be doing some adjustments to your elevation recorded in GPS against topo maps and/or history other riders to try to help error correction and less accurate devices. Does anyone know more about that or if its even true?

  19. #19
    mtbr member
    Reputation: aliikane's Avatar
    Join Date
    Nov 2005
    Posts
    1,434
    I have done that exact route. Black Road to Saratoga Gap up and back. It is about 20 miles and 3000K vertical feet.

  20. #20
    Rides all the bikes!
    Reputation: Sidewalk's Avatar
    Join Date
    May 2015
    Posts
    3,302
    Any ride not recorded with a barometer enabled GPS, I ignore. iPhone, Android, older GPS, etc. Climbing data is meaningless to me. Baro isn't perfect, but close enough.

    I have done races that were advertised as having WAY more climbing than reality, because they mapped it out instead of riding it. Then you see everyone comment on their Strava "Wow, my GPS is way off on the climbing!"

    Quote Originally Posted by aliikane View Post
    It is about 20 miles and 3000K vertical feet.
    That's half way to low earth orbit!

  21. #21
    Formerly of Kent
    Reputation: Le Duke's Avatar
    Join Date
    Mar 2009
    Posts
    10,690
    Quote Originally Posted by boomn View Post
    Even with a dedicated GPS device, if you upload the recording to TF you will get a very optimistic sum of your elevation changes. And if you upload the file to Strava you will get a completely different number. There isn't a standard for how it should be calculated

    Also, I believe both of these sites might be doing some adjustments to your elevation recorded in GPS against topo maps and/or history other riders to try to help error correction and less accurate devices. Does anyone know more about that or if its even true?
    So, as I have worked as a surveyor, have a MS in GIS/remote sensing and now teach military students various software packages that involve the use of digital elevation models, I'll take a stab here.

    The "earth", as Trailforks and other websites see it, is composed of a bunch of pixels, each of which have an inherent elevation associated with them. Collectively, they are called a DEM, a digital elevation model. Depending on the area, the DEM could be composed of 10m, 30m, 3m, etc., square pixels. But, because the surface of the earth is not, in reality, composed of a grid of square pixels, those pixels are of course an estimate (average of) the elevation of the 30x30m, 10x10m or whatever size square they lay on. Depending on the source of the data, those pixels could be from the Shuttle Radar Topography (SRTM) Synthetic Aperture Radar (SAR) mission, LiDAR flown by a state-paid contractor, or other sources. I don't know what Trailforks uses. In the worst case scenario, which is actually very possible, they are using what is called "first returns" from the SRTM SAR data, which is basically the average of the tops of the vegetation within the pixel (the first thing radar "beams" hit and reflect off of) and not the ground suface itself.

    Now, in a perfect world, your route would go from the center of one pixel to the center of another pixel 30m away, and the elevation change and lateral run between the two of them would be easily calculated. However, because no trail happens to align perfectly with a 30x30m grid of pixels, and we don't know the sampling rate, it could very well extracting data from the edge of one pixel, the center of another, and the far side of that same pixel. To put it in simpler terms: they might be over-extracting data and creating false "steps" within each climb or descent that are adding to the total amount of actual climbing or descending done on that trail.

    Basically, there are a LOT of potential sources for error here, and without knowing their data sources, sampling rate, etc., it's going to be really, really hard to pin down the cause.
    Death from Below.

  22. #22
    mtbr member
    Reputation: dchester's Avatar
    Join Date
    Nov 2005
    Posts
    690
    Part of TF problem is how trails are created by users. You have to draw them out or import a recorded Strava or gpx file. Connecting dots/drawing isn't going to give you proper elevation or mileage. I think TF does handle positioning just fine though.

    I'd say try uploading and crop a gpx file to TF for one single trail then compare how TF renders that gpx.

  23. #23
    mtbr member
    Reputation: aliikane's Avatar
    Join Date
    Nov 2005
    Posts
    1,434
    Quote Originally Posted by Sidewalk View Post
    Any ride not recorded with a barometer enabled GPS, I ignore. iPhone, Android, older GPS, etc. Climbing data is meaningless to me. Baro isn't perfect, but close enough.

    I have done races that were advertised as having WAY more climbing than reality, because they mapped it out instead of riding it. Then you see everyone comment on their Strava "Wow, my GPS is way off on the climbing!"



    That's half way to low earth orbit!
    Strava can be off sometimes, but generally it is pretty good. There will be tolerances because GPS can be affected by trees, obstacles, etc.

  24. #24
    Formerly of Kent
    Reputation: Le Duke's Avatar
    Join Date
    Mar 2009
    Posts
    10,690
    Quote Originally Posted by dchester View Post
    Part of TF problem is how trails are created by users. You have to draw them out or import a recorded Strava or gpx file. Connecting dots/drawing isn't going to give you proper elevation or mileage. I think TF does handle positioning just fine though.

    I'd say try uploading and crop a gpx file to TF for one single trail then compare how TF renders that gpx.
    Sort of.

    Trailforks and MTBproject both use OSM data for a trail's XY coordinates.

    Within OSM, they use heat maps and basically take a rolling average of user data to adjust the trail layout. With enough data, the error of the average effectively drops to zero, unless there is some sort of underlying factor resulting in consistently bad GPS coverage in that area.

    TF/MTBP will also adjust a user created layout over time to reflect OSM.
    Death from Below.

  25. #25
    mtbr member
    Reputation: b0bg's Avatar
    Join Date
    Apr 2013
    Posts
    335
    The Trailforks admins/developers have acknowledged this long standing issue on their message board. Shortly after that there've been some reports it's gotten really bad, so it looks like they're working on it. Hopefully they'll fix this issue as it's one of the few remaining complaints I have about the app. On the whole, I'm pretty impressed with the app, and they're pretty responsive about getting things fixed. This one has lingered on for a while though, it's probably pretty challenging to fix.

Similar Threads

  1. Replies: 4
    Last Post: 11-04-2018, 09:57 AM
  2. GoPro Hero Session Vs Hero 5
    By fiveo in forum Videos and POV Cameras
    Replies: 7
    Last Post: 12-28-2016, 08:26 PM
  3. Replies: 0
    Last Post: 06-21-2016, 11:41 AM
  4. Which hero/anti-hero would ride your steed?
    By targnik in forum All Mountain
    Replies: 1
    Last Post: 07-25-2015, 03:35 PM
  5. Blackview Hero 1 and Blackview Hero 2
    By danielhuang in forum Videos and POV Cameras
    Replies: 0
    Last Post: 05-25-2015, 01:24 AM

Members who have read this thread: 301

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

THE SITE

ABOUT MTBR

VISIT US AT

© Copyright 2019 VerticalScope Inc. All rights reserved.