[FF2]

I think many culmulative altimeters filter out small changes so that in your example, the total might have been near zero rather than 500m. I think my bike computer/altimeter only records (adds) changes over 10 feet (3 meters, more or less).

I have not read the manual for my "watch altimeter" in a while but I think it may also do some filtering.

Your method might actually be more accurate in some circumstances and it certainly helps MAGNIFY all those tiny ups/downs. Depending on where folks carry the BB's, just each stride or up/down-turn on a ski slope could raise/lower the BB's location. Skiing for instance - dip body into turn, rise up after turn (loss then gain), now repeat over and over............

[Stinsonddog]

Originally Posted by FF2

I think many culmulative altimeters filter out small changes so that in your example, the total might have been near zero rather than 500m. I think my bike computer/altimeter only records (adds) changes over 10 feet (3 meters, more or less).

I have not read the manual for my "watch altimeter" in a while but I think it may also do some filtering.

Your method might actually be more accurate in some circumstances and it certainly helps MAGNIFY all those tiny ups/downs. Depending on where folks carry the BB's, just each stride or up/down-turn on a ski slope could raise/lower the BB's location. Skiing for instance - dip body into turn, rise up after turn (loss then gain), now repeat over and over............

Interesting point. I do agree that the intervals would have to be similar to get the same readings. I don't know what my watch is. Sunnto Vector, and will have to go home and compare to what is set for this program. I can tell you in my early post it seems way to much though for a variance based on this alone.

[FF2]

Originally Posted by Stinsonddog

Interesting point. I do agree that the intervals would have to be similar to get the same readings. I don't know what my watch is. Sunnto Vector, and will have to go home and compare to what is set for this program. I can tell you in my early post it seems way to much though for a variance based on this alone.

Are you "lift assisted" or do you manually yo-yo those slopes?

[emacf1]

what do you thing about the following algorithm:

collect all points that have been recorded in 1min and calculate the average altitude from that points - and the altitudeGained/lost will be calculated then on the "average altitude" values...

what do you think?

[emacf1]

@at.rex - with the new algorithm you would had gained 552feet... (sounds that better to you)?

[FF2]

I would hope someone more knowledgeable in math/formulas/filtering than I jumps in - this probably can be quite complicated.

I am leaving momentarily on a 35 mile bike ride with approximately 1,700 feet of elevation gain (those hills do go down again). So I will see how the current version does against known facts and report back (HOURS and hours from now - I'm out of shape!!!)

[emacf1]

Originally Posted by FF2

I would hope someone more knowledgeable in math/formulas/filtering than I jumps in - this probably can be quite complicated.

I am leaving momentarily on a 35 mile bike ride with approximately 1,700 feet of elevation gain (those hills do go down again). So I will see how the current version does against known facts and report back (HOURS and hours from now - I'm out of shape!!!)

you could email me the GPL file and then I can let it run though the new implementation...

[emacf1]

Just received Stinsonddog GPL file - and with the new impl GPSLogger calculates 15882ft for this ski trip...

[FF2]

Hey, someone help me - where do you see the total elevation gain? Details page?

I have pages for path, elevation, speed on each I see the duration and the distance at the end of the data display.

I did email my bike trip.

[at.rex]

@at.rex - with the new algorithm you would had gained 552feet... (sounds that better to you)?

That is still almost 500 feet too much.

[at.rex]

Hey, someone help me - where do you see the total elevation gain? Details page?

If you have downloaded the latest version you will find details page under options when your track is highlighted.

[FF2]

Originally Posted by at.rex

If you have downloaded the latest version you will find details page under options when your track is highlighted.

Thanks. I was trying to find it when the path was active.

Hmmm, not a bad bike ride - 8,500 ft of gain over 33.8 miles - I'm ready for the Tour de France!!! (Actually, I feel like that's accurate unfortunately. Not that I'm ready but it felt like 8,500 feet)

It should be closer to 1,600 - a shade over...........

[emacf1]

Originally Posted by at.rex

That is still almost 500 feet too much.

... I have difficulties to agree with you here (after i have seen the data) - I will send you some data detail this afternoon/evening when I am back home (when you agree, I would prefer top post it here in public to be able to discuss it with a larger audience) - there you can see the average data that was recorded during your trip - and form that data the 500ft are quite reasonable...

But I can understand, that it sounds indeed quite much when I look at the area where you made that trip (I am still jealous/envious)

[at.rex]

... I have difficulties to agree with you here (after i have seen the data)

I did a 30 mile bike ride that I broke into two tracks. I used my bike computer with altimeter and also a Garmin CS60 GPS unit. On the first part of the ride I went 12.5 miles and climbed a total of 711 feet according to the two units. GPSLogger recorded 1070 feet. On the second part of the ride I went 18 miles and climbed a total of 1106 feet according to the two units. GPSLogger recorded 1711 feet.

T

[at.rex]

Hmmm, not a bad bike ride - 8,500 ft of gain over 33.8 miles - I'm ready for the Tour de France!!! (Actually, I feel like that's accurate unfortunately. Not that I'm ready but it felt like 8,500 feet)

Looks like you had a better workout than I did.

T

[emacf1]

Originally Posted by at.rex

...my bike computer with altimeter and also a Garmin CS60 GPS unit. On the first part of the ride I went 12.5 miles and climbed a total of 711 feet according to the two units. GPSLogger recorded 1070 feet. On the second part of the ride I went 18 miles and climbed a total of 1106 feet according to the two units. GPSLogger recorded 1711 feet.

It's difficult for me to express what I wanted to say - I don't doubt, that the results GPSLogger currently reports are not the values you expect...

Ride 1: 12.5 miles and climbed a total of 711 feet
Ride 2: 18 miles and climbed a total of 1106 feet

Is the first one you send me? [or are you now talking about different ride again?] I am sorry I am a bit confused

... What I wanted to say with my '500ft' post question was, that I have already locally adjusted the GPSLogger code - and with that adjustment the dialogue say: gained 552feet.

I agree, that 522ft does not equal 711ft... It's quite difficult to find a good/reasonable algorithm that reflects:

a) GPS data accuracy failures (deviation)
b) Logging interval

Concerning the 1000 point example I have given view posts above - with the Old Code we would had get 500m gained - with the new implementation we would had get more or less 0m gained... so IMHO it's getting closer...

So what I am asking you is for "good" additional ideas/algorithms that could by allied.

[at.rex]

Is the first one you send me? [or are you now talking about different ride again?] I am sorry I am a bit confused

No, this was a bike ride that I haven't sent to you. The one I sent you was of a short hike. I just posted the bike ride results so you could see the difference between a dedicated GPS and GPSLogger. I wish I could help with your efforts but I don't understand why there is such a great difference between the readings. And, I want to thank you again for all your time and effort that you are putting into this problem.

T

[emacf1]

Originally Posted by at.rex

No, this was a bike ride that I haven't sent to you.

Would it be possible for you to email me also that GPL file? [so then there is a good reference to other other devices where we "know" the alt gain values the dedicated devices have printed...] even if I hope, that you understand, that the algorithm I have used right now (and that is build in 0.4.15) is also "correct" - but (when you look at the 1000Point example) quite strange...

It's always difficult since as a human beeing you have a certain expectation [but when a computer just accumulate facts, that the picture could be totally different]...

[skyedream]

would this work on black berry 8320? and do i need maps or does it included?
and do i have to enable to data? i turn data off so in case i don't have to use data to be charged.

[FF2]

Mathias:

I sent you an email (at least, I tried) yesterday with a gpl file with a 33 mile bike ride that showed approximately 8,500 feet of elevation gain. My bike computer/altimeter (Cateye older model but pretty accurate) showed about 1,600 feet.

You can try your new method on it and report back.

As I wrote above, I can see your method picking up minor undulations in the road (up a foot, down a foot, up a small two feet, down a foot, etc) which could really add up even when the terrain feels flat. And depending on where the BB is located (I had mine in a hip pouch), each time I stand on the pedals and lower myself, could be added to elevation gain and that is probably not reflective of reality.