How do calibration numbers work?

Let’s say I’m riding along at a steady 220 watts, per my power meter numbers. I stop to calibrate. Let’s say before calibration the offset was showing 7, I then click “Calibrate”, and the offset changes to 12.

My question is, was a I really doing more than 220, or less?

I don’t know if the answer is same for all PMs, but if it matters I’m using Powertap P1 pedals.

It could have to do with temperature compensation as well. The calibration sets a baseline for the strain gauges to measure against. Most power meters have temperature compensation built in to account for the reading changes as a result of warmer or cooler temps. It differs by power meter but there are acceptable ranges for calibration values and acceptable changes within a ride. Unfortunately, i do not know the powertap values.

My thread title may not be clear, but I’m not asking about how the calibration works, I’m just wondering about which way the offset numbers are adjusting. Let me ask it this way: When you see the power number on the display, is that the result of the offset being added to the raw value, or subtracted from the raw value?

Depends what the calibration number says - positive or negative.

Getting to the heart of your question as asked.

Prior to the second calibration you were doing less than what your screen said. 7 to 12 is not much in terms of watts.

And if you calibrated correctly the first time you started right and just deviated - so you may have been spot on reality = value on screen for the majority of your time.

“Calibration” doesn’t actually mean “Calibration”. It means “record the residual torque that exists on the strain guage” which the power meter then uses to change the value that it sends to the headunit.

For a Quarq, 32 offset points is equal to 1Nm of torque. 1Nm of torque at 80-100 cadence is between 9-10 watts.


@Stevemz beat me to it. The strain gauges know nothing about power, only torque.

Zeroing your power meter is like taring a set of scales.



When you calibrate with a head unit or TR, does it change something in the PM, or just how the head unit/TR interprets the numbers? Like does zeroing with my Bolt before I do a TR workout on my phone (can’t zero from there) do anything?

You can’t zero from where? TR on your phone? I have garmin vector 3 and Assioma duos and both zero from TR when I use my phone.

My phone, I have an older quarq that is only ANT+, no BT.

To clarify a bit more: I have my laptop read the power meter and goes to Zwift. I ‘pair’ the PM with my Kickr to use the PM power, not the Kickr power. I use my phone to run TR and control the Kickr. Mildly complicated

I’m not 100% positive but I believe it’s stored on the power meter, not on the headunit.


Yes, it is stored in the PM. That is why you can run the Zero Offset (not really a “calibration”) on any device, and the data shared from the PM will be correct on any/all devices that read it after the ZO.

Worth a read:


To bring this excellent answer closer to an answer to the exact question that was being asked:

Assuming that initial calibration value of 7 was incorrect, and the new calibration of 12 was correct for the given temperature, etc, then your power meter would have been over-reporting your watts while the calibration value was 7 (i.e. you were doing less than 220 watts when your power meter said you were riding at 220 watts).

The inverse of this, i.e. if you had an artificially high calibration number, then your power meter would report a lower number of watts than what you’re riding.

Although this is definitely really interesting, it’s probably not something worry about in reality. The calibration value typically changes due to temperature changes, and most power meters these days have automatic handling for changes in temperature, which basically means that the calibration number you saw when you started your ride may not be the “real” calibration number that is being used internally after a couple of hours on the bike, because the temperature may have changed and the power meter will have adapted its internal calibration to this.


Zero offset numbers can change for a number of different reasons, not just temperature.

As long as it’s within the normal float range for a power meter, it should be accurate to it’s stated range (usually 1-2%).

If we assume 300w for an FTP, 300w could be 294-306w but your body doesn’t really know what power it’s putting out and training zones are a range for a reason. Threshold zone for this FTP is 290-320.

So as long as you are riding in the range, and your power meter hasn’t floated consistently in one offset range, you are getting the desired training stimulus.

Quarq says that as long as it’s +- 50 points from a stable value, it’s considered in good condition.

Mine floats somewhere between 30 on the low and 60 on the high end.

Thanks Chad, that is exactly what I was looking for

1 Like

This is not actually true if your PM has temperature compensation. When you zero offset the PM records the residual torque and temperature. Say the 7 number was at 60 degrees, then you calibrate 2 hours later and its at 75 degrees and it reads 12. The PM will adjust for this with the temperature so it reads correctly across all temps as they change throughout the day. Of course the temp compensation curve will have some inherent error in it as well so its better to have the zero offset at the temperature near the ambient temperature you’ll be riding in.

1 Like

Thanks for clarifying. This is what I was actually trying to say (I think) in the last paragraph in my reply, although you explained it a lot better here.

greid wrote:

“Assuming that initial calibration value of 7 was incorrect, and the new calibration of 12 was correct for the given temperature, etc, then your power meter would have been over-reporting your watts while the calibration value was 7”.

Thanks, that’s all I was looking for, but lots of good tangential info in this thread.

Also, to clarify the scenario: I usually wait until after the warmup to do the calibration. I understand that temperature differences as the workout goes on should theoretically be taken into account automatically if I were to calibrate at the beginning of the workout, but I figure it can’t hurt to calibrate after everything is warm (me, the PM, and the room). So, until I calibrate, the offset is whatever it was when I calibrated during the last workout, which is likely innacurate for the correct workout, and I was just curious how the values displayed during the current workout prior to calibrating are affected, given the difference between the previous offset and the new/accurate offset obtained when I calibrate after the warmup.

If I understand correctly then, the offset measures the residual strain on the gauge with no external forces (feet not clipped in). Therefore the post-warmup offset of 12 is saying there are 12 units (I use the arbitrary “units” instead of watts since it’s not clear whether it actually represents watts) of force inherent in the gauge, so it needs to adjust for that by subtracting those units from the power numbers while I’m pedaling. But prior to calibrating, it was only subtracting 7 units of force, resulting in reporting power numbers that were greater than actual.

Let me know if I have anything wrong.

Nope, still accurate before and after zero calibration.

@redlude97: I don’t think you’re understanding what I’m saying. The value before calibration is from a previous workout. Clearly the values can’t be accurate both before and after calibration (for the current workout), otherwise there would be no need for calibration. (And the offset wouldn’t have changed from 7 to 12 when I calibrated.)