3.4 Autotune PIDS same as 3.3?

If your flying 3.3 and have you copter flying very good… after you instal 3.4 and do an auto tune shouldn’t it come up with PID values that are very close to what you know worked well? Wouldn’t that be a good way to see if 3.4 is capable of doing a good PID calibration by doing that comparison? OR do the changes in 3.4 mean it might take different PIDS to fly the same as it did before?

Basically, yes.

The motor refactoring on 3.4 branch does change numbers quite a bit.

Sorry Luis, but Richard asked 3 question. To which one does your answer apply?

I thought it would be clear with the 2nd line:

So you can generally expect the PIDs to be about 10% lower for most X, V or H frame vehicles. Plus frames, TriCopters and Helicopters should see gains that are 27% higher.
Hope that helps and thanks for testing!

Thanks Randy (and Luis).
This is the info we were looking for.

Seems a clear answer, but:
Starting with the defaults at the Hexa F550 Flywheel from DJI it was difficult to control the copter.
Now, after 30 test flights during 90 min. I’m with Rate Roll/Pitch P= 0.3, I=0.15, D=0.024, Yaw P=0.4,I=0.018,D=0
Wobbel starts above Rate Roll/Pitch P=0.7 !!
Hoover ist stable, but with short, non-rythmically bumps and the mots sounds "angry"
EKF and VIB verry low, hoover with 45%. Autotune does not work (no twiches)
Quality far away of Pixhawk with AC 3.3.3 with defaults and the same frame.
Fast reaction on stick-control bumps.
Perhaps future releases or any idea?