Film Testing Priority October 2024

Colorado 005_Million Dollar Highway1

Starts a bit dark. Likely just the time of day and position of the sun. 

A couple of times I thought the gradient was "off". For example, at km23, the road appears to descend but it shows (and feels) like a slight uphill. This one could be an optical illusion caused by the cliffs to the left though? 

As a workout it should be excellent - nice flat lead into a constant, but easy, hill - but the first 13km are, to be blunt, a bit boring :-) 

Colorado 005_Million Dollar Highway2

Better colour/footage and scenery.

Another couple of occcasions where I thought the video was out of sync with the resistance. Noticeable at the top of the Coal Bank Pass - passing the road sign, the road obviously descends but the trainer gradient carries on at 4.5% for a couple of hundred metres - and after the wee kicker on the Molas Pass where the road descends to a big bend but resistance doesn't reduce to zero until well into the bend and then is late to match the video showning a climb again. Possibly 200 metres or so. 

FWIW, I loved the unexpected gravel section just before this big bend - which definitely synchronised well with the video. 

I think its not a grade sync issue as it looks fine in other places with obvious changes but more just generally poorer data mainly due to the higher filming speed, (Its only GPS built in barometer with car-filmed routes).  Ive gone through it a bit more & made some changed based more on what it looks like ;-) v2 should be there tomorrow. 

If any of you can we do want some feedback on 

I have found a very little issue for Sherman Pass 1: The video stops 60m before the end.

Thomas said:

I have found a very little issue for Sherman Pass 1: The video stops 60m before the end.

v2 has this issue of a little too much data fixed, so marking this as ok for the rest

If possible can someone ride 

A2024.25

004_Grand Mesa Scenic Byway (Colorado)

100/100%

v3 Million Dollar Highway 1 - All OK! 

v2 Million Dollar Highway 2 - only 20.42 km from start. 

7.04 & 8.68 & 10.19 km - visually the gradient seems larger than on the screen 1-2%. 

I'm still recovering from illness. That's why I only rode 21.13 km from the start. The video is beautiful and the gradient is OK. Everything is fine. I plan to finish 004_Grand Mesa Scenic Byway (Colorado) & Colorado 005_Million Dollar Highway 2 on Saturday or Sunday.

Phil70 said:

If possible can someone ride A2024.25 004_Grand Mesa Scenic Byway (Colorado) 

Sure!

Great, just finished writing a whole review and it's gone after hitting submit. This forum is as beta as the testing were doing, lol.

Anyway, here's the short version
v3 004 Grand Mesa Scenic Byway (Colorado)

I'll be blunt.
I like the surroundings and can appreciate a hard or long climb. This one was 30km and 1700hm.
The video was filmed at too high speed resulting in a not so smooth video (low framerate)
Other vehicles and falling leaves are moving too slow of this. Looked very weird

The gradients are more off (too high) than ok. I doubt the climb is 1700hm in real life. Of course I might be wrong. Straight from the start the gradient being reported is high. More examples of this can be found at 1,66, 1,90, 2,15, 3,40km. I then stopped taking notes. The worst example can be found at 20,72. Road looks flat, gradient goes up to like 8%
The video finishes strong with -8%. Where the *** did the data come from :)

I really think this does not meet the high standards. I doubt manually correcting the gradients is worth the effort. The filming speed can't be corrected. I would consider not publishing this at all.

Maybe another tester can test the first 10km and give their opinion? I can imagine not everyone want to do the entire 40km.

Peter Scheuten said:

The gradients are more off (too high) than ok. I doubt the climb is 1700hm in real life.

After reading I was also unsure if it was a global error in the Alt data, its tough to have anything too odd happen with all the rides data going through my filters the same way.  But even though it looks more like a kind of hilly terrain it is actually a major HC cat mountain that gains 1600m on Strava.  I think this could be more the terrain just looking less steep then it actually is.

For the low frame rate, running it at 200watts looks ok to me starts to get a little jerky at 150w. Im going to walk through the data though, there will be a v5 tomorrow hopefully.