That does look quite noisy but it probably doesn’t matter so long as the average power is accurate. Intervals.icu uses a 5s moving average over the power data when looking for intervals and so on.
You could change the power display on your head unit to show a 3 second average. That might help. I have a PowerTap hub:
Hi, regarding fluctuation, yes this is a problem seen on some devices…you can probably ignore it or change the averaging period as David mentions. Of course its only safe to ignore if its fundamentally accurate. So regarding power meter calibration… you can calibrate your power meter in a few ways:
The reason #4 works is because the more steep the climb the less aero and CRR helps, so basically those are eliminated and they are even shown on the graph under the calculator
Regards
Alex
I don’t need an exact absolute value. I need a repeatable value so I don’t need to calibrate the meter. I am interested in being able to keep my pedaling power steady and if the indications fluctuate strongly it is difficult.
P. S.
The link is not working.
P. S. 2
I apologize for my English, I am supporting Google translator 
I’m sorry for maybe misunderstanding, but if you have a left only (crank) PM, keeping the power steady is entirely up to you. There is no way that you can do an ERG workout and ‘set’ your power to an exact / steady value…
EDIT: Having read that again, with ‘the indications fluctuate’ you mean grades, going up- or downhill? Keeping the power steady is still up to you, but I would understand your point better.
With fluctuating grades, the challenge is to gear up and down to be able to keep the pedals turning and maintain more or less the same power.
if i am understanding your question correctly then i believe adjusting your capture rate of power from your PM to a longer interval should smooth this out. try 10 or 30 seconds
I would have feature request regarding pm accuracy. I have two bikes equipped with a pm. The devices differ in their accuracy significantly, as I figured out by extensive testing. Is there a way to normalize the values using your platform, i.e. adjusting the readings from one bike/pm to match the other? I am aware that one will never achieve 100 % accuracy. However, if the difference between two devices is too big, the analysis done by intervals.icu becomes compromised.
You can use the “Fix data” option under the ride timeline chart to scale the power for a ride. I am going to add support for different FTP for different bikes soon and will look at also adding “y=mx+c” function to each bike for power at the same time.
Just out of curiosity: why? If you ride any of your bikes equally often, I don’t see why there would be a difference.
I suppose that if you hardly ride one, you may have a difference, for instance between a TT and a ‘normal’ bike, but otherwise…
The TT vs normal bike is one reason. Another is to handle power meters returning different numbers. Having different FTP’s gets the intensity and training load right without having to change the power data.
Which is why I have easy to swap pedals - if you have two or more bikes, why have different power meters.
Out- versus indoor, i.e. with a smart trainer, that would make more sense, although I frequently swap the pedals to that bike too, just to check accuracy between them.
Having a different FTP for a different bike, or for indoor training, is actually not that much related to you actually having another FTP.
It’s more about other factors, different circumstances, for which you can train to even those out.
Anyway, as always, this is going to be another option that one can either use, or not 
Plenty of reasons: You might like Speedplay, Shimano or Time pedals, or ride a mountain bike where crank or hub based power meters are currently the only real option.
Mike
I have a mountain bike and I’m using the same pedals - I get it that most do not favor Look or SPD-SL for a MTB, but I got used to that too.
Then again, I don’t MTB that much and probably not on the very serious courses that I sometimes see others ride…
Fantastic!
Mike_Lauder: actually there is an assioma hack that enables you to fit the pedal power meter to mtb pedals. See TR forum, or gplama.
I’m aware of the assioma hack as I was the one that came up with it. Still not a great solution for real mountain biking though as the pods are a bit exposed on the ends of the cranks. Fine for gravel I guess, although I tend to use road pedals if I know I’m not going to be walking.
Mike
BTW, is there a way to discriminate between actual coasting (0 W data points) and loss of connection between the PM and the recording device? It seems my Rotor INspider loses the connection rather frequently for a few seconds up to half a minute. For Strava, these events are recorded as 0 W/coasting and hence the avg power is reduced. In the attached screenshot, the cursor is positioned in one of these 30 s windows with no connection to the pm. I know about the data correction tool on intervals.icu. But how do I know if I actually stopped pedaling or the connection was just lost? (In this case, it was easy, since the route was flat. And I can hardly go 38 kph for 30s without pedaling.) 
Is there no difference between null and 0 on the graph? Null being ‘no data’ and 0 meaning 0 Watts 
In the fit file, there is a “0”. Interestingly, the first 3 seconds when connection was lost, there are still non-zero cadence values in the file (99 rpm). This seems to be a pattern. I found another 20 s gap in the same file. Speed an HR are actually increasing, while the power and cadence values are 0. Also note the non-integer HR values.
EDIT:
They should indeed change the standard for recording data. If the connection to a a sensor drops, there should be a -1 or a string instead of a zero. This would simplify data processing/curation.
Please post a link to that activity on Intervals.icu and I can have a look. I have seen data from Strava with “null” for missing data points rather than zero. Intervals.icu does distinguish in its “fix data” function (“missing data points” are the nulls).
