Would a turret be more efficient if I used constraints or CFrame all the movement?

[Wegny]

I’m making a turret that can pitch and yaw. I’m currently using two servos, which have their targetAngle set to make them point at the target.

However, I could also ditch using the servos and instead hand-code the entire movement of the turret using a script to control the CFrames, keeping the parts anchored the entire time and not using any physics.

Which approach would work best for players using low-end devices?

[LittleLegender]

cframe gets a bit ugly when there is too much cframing

if your using cframe to turn “Models” your models parts would start Slowly but surely fly Away from each other and this is due to a floating point issue

i think physics would work just fine for you