Anyone have anything they'd like to rant about?

[jamal]

I think strava's elevation data is pretty good. For the most part my garmin is pretty close but every once in awhile i'll have gps errors, mostly in wet weather, and have to use the "correct elevation" feature.

But yeah who knows where the tf stuff comes from. Possibly from less precise gps date and trying to calculate as it crosses terrain? Like riding across a canyon on a bridge, but the topo data thinks you descended and climbed back out. So that, only on a very small scale, but continuously over the whole ride.

https://support.strava.com/hc/en-us/...-Your-Activity

trailforks, i assume, is not quite as sophisticated.

[grskier]

Strava uses your A-GPS or GPS data and finds best matches for segments. I'll need to look into the vert piece. The 30' thing kind of makes sense of why you and a buddy can have a few percent difference. I know on some GPS trackers I used, then uploaded to strava you'd see a difference in vert because you could change the amount of time between each 'reading'.

GPS data is tough on these little devices. Tree cover, clouds and rain, etc all make it harder for the device to find satellites and/or towers.

You typically need at LEAST 4 touches for accurate data. You are solving for 4 equations X, Y, Z, and time.

I'm not sure how TF works, I just know I used it to log several of my rides a few years ago as I was doing a vert challenge and it was always higher than Strava.

[Terrapin Ben]

Vertical accuracy for any given GPS is usually 2-3x that of the horizontal accuracy. So if you've got 3m horizontal accuracy on your phone, the vertical accuracy is going to be around 9m. Extrapolate that across a long ride with lots of ups and downs, and it's pretty easy to write it off the accuracy of an elevation profile generated from a track log. It makes more sense to extrapolate accuracy from (more accurate) base map. You figure both TF and Strava would include this functionality.

[toast2266]

I'm not sure how TF works, I just know I used it to log several of my rides a few years ago as I was doing a vert challenge and it was always higher than Strava.

Even if you load a ride into TF from Strava, the vert is different. So with the exact same data, it just interprets it differently.

Vertical accuracy for any given GPS is usually 2-3x that of the horizontal accuracy. So if you've got 3m horizontal accuracy on your phone, the vertical accuracy is going to be around 9m. Extrapolate that across a long ride with lots of ups and downs, and it's pretty easy to write it off the accuracy of an elevation profile generated from a track log. It makes more sense to extrapolate accuracy from (more accurate) base map. You figure both TF and Strava would include this functionality.

Both TF and Strava are extrapolating vert off of a map. I don't believe either of them use the GPS altimeter at all.

[Terrapin Ben]

a GPS unit looking at 4 satellites will record a Z position based on NAVD88. No altimeter or barometer necessary, Not super accurate, but just getting the info out there.

[singlesline]

There's a more detailed description for how Strava does their elevation modelling here:
https://support.strava.com/hc/en-us/...bile_site=true

They do a whole mess of calculations combining your device's reported elevation, elevation reported by other riders on the same segment, and other sources of digital elevation data.

If your device doesn't have a barometric altimeter, I think they mostly throw away the GPS elevation data and replace it with modeled elevation as the Z axis on GPS isn't that reliable (and devices that don't have an altimeter probably also have crappier GPS).



Even barometric altimeters aren't perfect as they change with the weather.

Here's the elevation profile straight from Wahoo for an out and back ride...you can see the end of the ride is noticeably higher than the beginning

The Bolt V2 should have a decent altimeter and clean GPS signal, and the trail is obviously identical but you still see drift as the day went on and the weather changed between ascent and descent.

Trailforks reported 5189ft of climbing rather than 4600, but they actually fixed the altitude:

Hard to see without a line at 4k, but the peak point on the secondary climb is ~4020ft on both ascent and descent.

Strava reports exactly 4600 which matches Wahoo.

Strava likely has more money and data, so I'd probably trust their corrections a little more for well travelled segments--but they are also much more road focused which means their choices might not be as relevant for for MTB where something like a punchy 15-20' climb out of a tight corner sure still feels like "climbing" even if it is short.

But trailforks has some really ugly data and I don't think they have tech in place to "correct" trails over time based on ride logs. E.g. it is not uncommon to find TF trails with totally wrong elevation data because they were initially uploaded from a faulty GPS track.

EDIT: mixed, that up--first picture is from Strava app, but because the Wahoo has a barometric altimeter, Strava accepts it as correct with only minimula adjustment unless there are big outliers--so it matches what is shown in the Wahoo app.

[Terrapin Ben]

https://www.trailforks.com/help/view/90/

Why is my Ridelog Elevation Data Different Then Another App
Elevation data recorded from your phone is often very inaccurate, so we decided to NOT use it and instead generate elevation for each lat/lng point in your ride. This is what many other ride tracking sites like Strava also often do.

Note that the elevation-adding feature will erase any existing altitude data (for example, from a GPS). Often, this is desirable; profiles made with DEM data are usually "smoother" looking than GPS, and typically contain fewer gaps or suspicious readings.

We currently use our own elevation API service, similar to Google's data and is sometimes (but not always!) more accurate than the SRTM databases. We however hope to someday improve our DEM (digital elevation model) using all the ridelog data we are collecting. But until then we've decided to re-generate all ridelog elevation data, even if it came from a Strava synced activity. This way all the ridelog data on Trailforks is at least somewhat comparable with each other and to the trails mapped on Trailforks.

Ride statistics like elevation gain or descent are not a given value like distance, but something that is calculated with a formula. Each GPS app will use different criteria and formula to calculate their numbers. This often involves some cleanup of the elevation data first, smoothing out the data, then usually there is a minimum distance between GPS points for a point to be counted towards the rides cumulative climb or descent. Some apps might also have a better more accurate source of elevation data for some areas of the world.

Because of this if you record a ride on 5 different apps or devices, you will likely get 5 very different totals for climbing elevation. Trailforks has tried to make our elevation processing somewhat match those of Strava rides recorded with their app. But the results will vary.

You can read more about GPS elevation here:
- http://www.gpsvisualizer.com/elevation
- https://support.strava.com/hc/en-us/...-Your-Activity
Updated on Tue 2020-09-15 @ 11:56am.

https://support.strava.com/hc/en-us/...vation-Basemap

For devices without recognized barometric altimeters, Strava calculates your activity's elevation gain by comparing the GPS data to a known basemap of absolute elevation measurements.

More Strava info: https://support.strava.com/hc/en-us/...-Your-Activity & https://support.strava.com/hc/en-us/...on-Strava-FAQs

[jm2e]

You try a new valve core?.

Yup. Easy Peasy, works great. But seriously, now just waiting for the next random thing to pop up.


Sent from my iPhone using TGR Forums

[simple]

Toast getting the top comment on Pinkbike today. I wanna hate on Shimano also!

[Bushwacka]

Toast getting the top comment on Pinkbike today. I wanna hate on Shimano also!

The chainring moving make tons of sense to me.

But the auto shifting Nah I am good.

[evdog]

On my PNW trip last year I ran both Trailforks and my Garmin Oregon on a bunch of rides. Trailforks almost always produced higher elevation. Sometimes 10% more, sometimes 90%, probably 30% more on average. And a handful of rides where it was lower. The rides with the biggest difference were all on trails contouring along steep mountain slopes with big trees. But I also did rides in similar terrain where the difference was much smaller. I'm sure if I look at it further I'll find better accuracy on rides that are one singleclimb then descent vs rides that have repetitive up/down.

Pacific Rim Trail in Norcal had the biggest difference with Trailforks 93% higher based on what the app and device showed on screen. Garmin had 1,700ft and Trailforks 3,300ft gain/loss.

I loaded the tracks into Gaia, RideWithGPS and Topofusion to see what each of those programs showed.
Garmin track: 1,400ft, 1,600ft and 1,700ft
Trailforks track: 1,700ft, 2,250ft, and 5,600ft in Topofusion !!!

Interestingly Topofusion showed 1141 track points for Garmin and 4,976 for Trailforks. I would have thought the phone would sample fewer track points. I used Topofusion to simplify the TF track down to the same 1,141 track points which dropped the gain/loss to 3,200ft and then used the elevation smoothing function which dropped it to 1,600ft. So at least its good to know I can adjust it to get it in the same ballpark, since there were a bunch of rides where I only used TF and those stats seemed pretty suspect.

[climberevan]

Do you people not have head units with barometric altimeters? Unless you overrride it, Strava will take the data from the head unit.

GPS is nearly useless for elevation data, and topos are still largely based on data from like the 50s. Barometric data is where it's at.

[simple]

I enjoy watching folks talk about mapping and being just about completely in the dark on how it works. Cyclists love their data. My wife is that person who gets her panties in a wad when we do the same ride and our Strava data doesn't match.

It's relatively course data but that is what you should expect. My real favorite is when race promoters use distance and elevation data from a random source and then the racers get all twisted about their data not matching.

To the person saying satellites give elevation data. They do but holy shit is it low resolution. You gotta pay for the real good resolution and we know everybody wants that free dope. LiDAR data is updating elevation mapping all over the place for my work. Planes and drones yo

[Dee Hubbs]

I had a buddy do an Everesting attempt, he based the number of laps needed on a Strava file/segment.
He did an extra lap (or 2) just to be safe. After completing the challenge with over 9+ hours on the bike, he submitted it to the official Everesting site , he was actually 11 feet short in his attempt.

Apparently the Strava segment was created many years ago with a shitty phone, and was super inaccurate. And once a segment gets created, good or bad data, it's like that forever.
His wife was super pissed that she spent her whole day passing bottles and snacks, and his time didn't stand.

[Huskier]

I have sold used s*** so many, many times. On this website, craigslist, whatever. I set a cheap price, describe things clearly, quick and easy sale has always gone off without a hitch.

Then this doofus shows up today to buy a used fork for under $200, with the word "firm" on PB listing. He proceeds to talk s*** about it to try to get me to lower the already giving it away price. "Oh this is in bad shape it needs to be serviced" (it doesn't) "I thought X, I expected Y" (couldn't be more clear on the listing) yadda yadda...

I honestly just wanted it out of my garage, and the amount of time I wasted meeting this mouthbreather was worth more to me than the fork. Honestly, who the f*** does that? If you're a little short fine, in fact that's probably the best way to bargain if that's your angle. But telling someone how terrible the thing you're trying to buy from them is? I'm sorry what?

If you want an old non-boost Lyrik HMU I'll cut you a deal for not being a pain in the ass. SMH..

[panchosdad]

I had a buddy do an Everesting attempt, he based the number of laps needed on a Strava file/segment.
He did an extra lap (or 2) just to be safe. After completing the challenge with over 9+ hours on the bike, he submitted it to the official Everesting site , he was actually 11 feet short in his attempt.

Apparently the Strava segment was created many years ago with a shitty phone, and was super inaccurate. And once a segment gets created, good or bad data, it's like that forever.
His wife was super pissed that she spent her whole day passing bottles and snacks, and his time didn't stand.

Tell him he got a K2. JFC the shit people get their panties wadded up about.

End rant.

[jhyatt]

Said it before and I will say it again.
Bags of dog shit on the trail.
Don't leave it there. Not for a minute. Nor a year. I don't leave my trash on the trail. Don't leave yours.
Bags of dog shit are not an exception to littering. Even if you might pick it up on your way out.
End rant

Sent from my SM-G781U1 using Tapatalk

[Andeh]

Said it before and I will say it again.
Bags of dog shit on the trail.
Don't leave it there. Not for a minute. Nor a year. I don't leave my trash on the trail. Don't leave yours.
Bags of dog shit are not an exception to littering. Even if you might pick it up on your way out.
End rant

Sent from my SM-G781U1 using Tapatalk

My buddy said that where he grew up that was a huge problem. So someone printed out a bunch of little surveyor type flags with the black POW flag changed to be a turd emoji and "Poops on Walks" and "We Will Forget." Then stuck them next to every bag of poop. Didn't change any behavior but pretty funny at least.

[XXX-er]

I seen an artical that splained people think that if they put the shit in the bag it will biodegrade becuz the bags say biodegradable

I know, who could be that stupid ?

[toast2266]

It seems like 90% of dogs take a shit a few minutes into the walk. Busy trailheads just need to build a little 4 minute loop that circles back to the trash can. Do that loop first, let your dog shit, deposit it in the trash can, then continue with your walk on the main trail.

[mcski]

Tell him he got a K2. JFC the shit people get their panties wadded up about.

End rant.

Exactly. That’s fucking hilarious and way more memorable than actually completing some computer assignment. Tell him he won a pair of green boots

[rideit]

It seems like 90% of dogs take a shit a few minutes into the walk. Busy trailheads just need to build a little 4 minute loop that circles back to the trash can. Do that loop first, let your dog shit, deposit it in the trash can, then continue with your walk on the main trail.

Or put some shit cans a 1/4 mile in.

[toast2266]

Or put some shit cans a 1/4 mile in.

Then some poor sap has to hike 1/4 mile out with a 50 pound trash bag full of dog shit.

[rideit]

Ebikes!!!

[Shredhead]

Or put some shit cans a 1/4 mile in.

They do that here on some of the more popular dog trails.
It helps a ton!