AeroPod, PowerPod and Newton

When I do this I always get a 10 to 20% change in average and normalized watts, plus other related changes. Just wondering what this function really does and if it should be used all the time or just in certain situations. What am I to trust, on ride data (unanalyzed) or after Isaac manipulation? Isaac manual doesn't have much to say about this function.

John has generally said here that for a 'well dialed-in' unit there should be no need for the 'analyze route' function.

I prefer to use it myself for most rides, although the differences (with/without) are far less than you report.
I generally find that although rides start and end in the same place, there is some difference in start and end altitude recorded.
Also I prefer not to accept the default assumed equal wind for out-and-back rides (we often have a building wind)
It would be unusual for me not to ride in a group or at least in company of some randoms, so i like to select 'drafting'.

At the end of the day, I probably am in the position now where I use it mostly to ensure consistent data treatment between rides.
I don't bother to use it for lower intensity rides, as a don't really care about a few TSS points here or there and there will be no 'pb' efforts.

As I have said here before, if my preferred selections (above) could be set as the defaults for ride uploads, then there would be no need to use it.

Please post a ride file that illustrates your question.

More correctly I should say the phenomenon occurs regularly on harder, longer rides, not easy days where the difference is 5% or less. Here's a recent day with an 11% change from the original download to the attached analyzed file.

But my question is as much about when this particular analysis should be used as it as about the difference caused by the analysis. All the time, some specific situations, which ones? For consistency I've been doing it on every ride, is that wrong?

The "check calibration" feature is intended only for use with short, out and back rides.

This ride was 51 miles long. On a ride as long as this, you aren't really doing an out-and-back, and it is too long to assure that wind speed has not changed.

I would not use this feature on long rides such as this.

I find that when I adjust the elevation from what the barometer thought it was to what I enter as true that the normalized power almost always changes. No other changes, just enter proper elevation. So for instance, if it reads even 40 meters out from reality there will be a noticeable change in the NP. I just accept it. I even did a new calibration ride yesterday and fond even then with the new calibration that this effect still occurred.

Velocomp wrote:The "check calibration" feature is intended only for use with short, out and back rides.

This ride was 51 miles long. On a ride as long as this, you aren't really doing an out-and-back, and it is too long to assure that wind speed has not changed.

I would not use this feature on long rides such as this.

I wasn't asking about "check calibration", I was asking about "analyze route". What does that do, when should it be used, why does it sometimes change the raw iBike data so significantly as it did in the file I uploaded?

So no help available for these questions from iBike?

brookside,

Actually this has been discussed in the forum if you search a bit...
Anyhow, gist is analyze route (without using the check cal) will, if you start and end at the same spot or elevation, or if you care to specify the beginning and end elevations if known.... will adjust the entire elevation profile end to end for a barometric pressure drift during your ride which would cause a systemic and more or less steady (steady assumed) rate of pressure change. Or restated, will correct the ending elevation to the know end elevation.

In addition, unless you check the 'wind changed' box the analyze route will also 'correct' the wind closer to your average speed, perhaps not exactly to match depending on the statistic sauce used. If your cal ride is correct and your wind zero is good, normally you would check this box.

If you are convinced that you need some wind correction, with the wind changed box checked you could go back into analyze wind and apply a partial correction value, if you know enough about the wind conditions to actually improve accuracy.

That said, Racefern who is a respected adviser on this forum is a strong advocate for leaving, at least the wind correction, alone as his nearly perfectly performing Newton and profile are to be relied on as more accurate. (possibly at risk of putting words into his keyboard )

Hope this helps,
Russ

Russ, quick point the elevation correction seems to have what I would suggest is a significant impact on power. I do not see the reason for it. The change can be as little os 30-40 meters and the NP will be affected by at least 5%. This is rather obvious that when it says I finished my ride below sea level and I have pretty good evidence this is not true and that the elevation was say 8 meters. Why it has this impact is bothersome since it does not change the total elevation gains, slopes etc since it is just an offset. So that is my question why the significant change and why they can not like other software companies correct elevation against USGS databases etc? If I am in a place that is new to me I do not know the true elevation as a starting point, so I can not enter this as a start up of device nor can I correct after the ride?
so my issues are 1- why the software can not properly correct the start and end elevations from a database and 2 why when you change such a small elevation offset (to be clear both start and end by similar amounts) the NP changes, this one bugs me, since the power it tells me on the road is different from the power when I download and correct, I know this because my Garmin records what the Newton was outputting during the ride then I can compare to the corrected data after the fact.

ronpei,

Sorry, most of your questions are out of my realm, ie software functions are Velocomp's job.
We frequently suggest changes and occasionally John includes them in releases.

Perhaps automatic elevation checks could be added to the gps related functionality....

Russ

Russ wrote:ronpei,

Sorry, most of your questions are out of my realm, ie software functions are Velocomp's job.
We frequently suggest changes and occasionally John includes them in releases.

Perhaps automatic elevation checks could be added to the gps related functionality....

Russ

yes, Russ that was one of my points, that when you use the tracker, then you have GPS data that can be used to correct the elevation to known values automatically. Other software has implemented that. Also it is weird to me that a small elevation change can have an impact on normalized power when it was a simple offset, not something that should have changed the slope/ profile. I truly do not believe that the elevation determined air density can have that large an impact on power, but if I had the time knowledge, and inclination I guess I could run the data to see.

We've been traveling now for nearly 3 weeks, and internet access has been spotty. Just got back to FL yesterday...

When the "Analyze Route..." feature is used, Isaac attempts to find an out-and-back portion of the ride, then it applies wind scaling corrections based on its analysis.

In this ride, you can see that Isaac "found" an out-and-back...except, it really wasn't a true out and back (elevation is not symmetric about the "midpoint" cross-hair). The wind correction is applied to the entire ride. In the case of this particular ride file, I question the validity of the correction, due to the fact that it isn't really an out and back.

As I said in a previous post, Analyze Route should only be used on relatively short rides, with a true out and back.

Absolute elevation does not affect this correction. You can prove this to yourself by using the Analyze/Check Wind... feature, then setting the wind level to zero. The result is within 2 watts of the Analyze Route feature.

Honestly, I would just go out and ride and not tweak things afterwards, unless you have truly done an out and back ride that is not too long.

so John then what you are saying is that the analyze route function does more of a wind correction than an elevation correction? I find telling my Newton what the elevation is at start up a pain, and sometimes I do not know the elevation when I start, however there may be points along the ride where elevation is obvious for me because I ride near the sea so sea level is there. So is there a way to enter the correct elevation into the ride file with out tweaking the whole ride because I really use analyze route to correct the elevation and I was concerned that the power change meant something else was going on. If it is a wind correction then why have analyze route and analyze wind both? I suspect the math is different but from a user interface point of view this is not obvious.

So bottom line what is the best way to enter the correct elevation to a ride file after the ride?

ronpei wrote:So John then what you are saying is that the analyze route function does more of a wind correction than an elevation correction? I find telling my Newton what the elevation is at start up a pain, and sometimes I do not know the elevation when I start, however there may be points along the ride where elevation is obvious for me because I ride near the sea so sea level is there. So is there a way to enter the correct elevation into the ride file with out tweaking the whole ride because I really use analyze route to correct the elevation and I was concerned that the power change meant something else was going on. If it is a wind correction then why have analyze route and analyze wind both? I suspect the math is different but from a user interface point of view this is not obvious.

So bottom line what is the best way to enter the correct elevation to a ride file after the ride?

For out-and-back routes, "Analyze Route" corrects any discrepancies so that the starting and ending elevations the same. This correction does not, however, affect the power calculations.

Yes, after the ride you can adjust the elevation in Isaac using Analyze/Analyze Newton Settings

Well now I have found the right answer, both via the manual (gasp) and John`s reply. One thought is if it would be possible to have the analyze settings allow one to enter both starting and ending elevations in this same window for the odd occasions when barometric pressure changes enough to throw off the final elevation?

Many rides are NOT out and back (O&B) yet they are rides which do return to start. E.g. all loop rides.

Does Analyze Route only work with an O&B, rather than a return to start. The Analyze Route asks if the ride returns to start.
I have been using Analyze Route to correct for altimeter drift which always changes power to more reasonable numbers.

Or are we confusing the actions of Analyze Route with those of Calibration Check?

Also on the topic of Out and Back Rides - if we merge gps tracks with the ibr, shouldn't the gps track be used to accurately find the O&B sections rather than using elevation points?

Check Calibration and Analyze Route are very similar in impact.

Check Calibration actively looks for an out-and-back route, whereas Analyze Route allows a bit more flexibility.

For either of them to work at full potential, it's best that it be an out-and-back ride.

The big assumption used is that ground wind speed does not change during the ride, and that it is "symmetric", that is, that the wind speed on the return portion of the route is the reverse of that on the out portion.

If you ride long rides, or if wind changes during your ride, or if your return route is substantially different from your out route, such that wind speed is not symmetric, then these two functions may make things worse rather than better.