>>64211645
my mistake i started writing the post about the starlink then i forgot about the image and added the image of the pixhawk orange cube.

the communication in this drone goes starlink -> raspberry pi 5 -> orange cube where the orange cube has a gyroscope and generates commands to control the power to each motor.

pic related is the starlink from the X thread i linked in op
>>64211682
are the versions of it for F405 worth a try? i have a 7 inch walksnail moonlight quad i could test it on ive only tried betaflight and inav on it.
>>64211816
yes its inherent to the material, in general the stiffer the better and carbon fiber is a good combination of strength, stiffness and light weight thats why virtually all quadcopters use CF for the frame. if you're interested look into this youtube channel
https://www.youtube.com/watch?v=-W7RzZ66eRo
this guy makes general FPV gear testing videos but he's known for his quadcopter frame company where all the frames are optimized for low/easily filtered vibrations if you want very smooth flying. all the vibration types in the video are happening at the same time during flight but are very weak (obviously exaggerated in the video). carbon fiber frames make all these vibrations higher frequency but very weak/low amplitude, you generally dont care if the arm flexes like 3 microns 900 times a second but if you use a plastic quadcopter frame the vibrations will be low frequency (say sub 100hz) and high amplitude (maybe in the milimeters) and the vibrations 'look' more similar to regular movements/turns you can make intentionally so its harder to filer out what's intentional movement and what's noise which makes the quadcopter go crazy.