so anyway i finally got my filters somewhat dialed in a few weeks ago and everything was feeling pretty good, props out wasn't bad, motors running warmish but not hot, basically acceptable. then i updated to the latest betaflight and now it feels like i went backwards somehow. its not awful but theres this kind of mushy feeling in fast rolls and the motor temps went up a bit which doesnt make a ton of sense to me since i didnt change the PIDs at all.

im running RPM filter enabled with bidirectional dshot, was working fine before. i checked and bidirectional is still on in the CLI. the thing is i cant tell if its actually the update that changed something or if i accidentally loaded a default config somewhere along the way because i did have to do a fresh flash to update.

its a 4" build with 1404 motors if that matters, been flying this thing for a while and know how it should feel pretty well. anyone seen weird filter behavior after updating, or should i just start from scratch on the filter sliders and retune.

fresh flash will absolutely wipe your filter settings back to defaults unless you saved a diff all beforehand, thats probably your whole problem right there. the default RPM filter notch count in newer BF versions is different from what some people had manually configured before, so if it loaded defaults you might be running more filtering than you were or the notch placement is off for your specific motor/prop combo.

i'd pull a blackbox log and look at the gyro traces compared to what you had before if you kept old logs. otherwise yeah just re-run the filter sliders carefully ? start with RPM filter on and everything else pretty conservative, then back off the dynamic notch since it can stack weird with RPM filter and make that mushy feeling you're describing. 1404s are sensitive to over-filtering, they're light enough that you can feel latency in the tune pretty easily.

had almost the exact same thing happen last year on a 3.5" and it drove me nuts for like two weekends. turned out i had dynamic notch width set too high on top of the RPM filter and it was basically eating half my gyro signal. motor temps were up too which was the weird part, more filtering shouldnt cause that but i think the motors were just working harder because the tune felt off and i was overcorrecting on the sticks without realizing it.

also double check your dshot beacon settings didnt reset, sometimes a fresh flash toggles stuff you werent paying attention to and bidirectional can get weird if beacons are doing something conflicting. might not be it but worth a 30 second check in the CLI before you go deep into filter tuning again.