I don't think the actual planform shape (silhouette) of the fin has much effect on it at all (though I have no data to prove it and if anyone does feel free to correct me on this!). The most important factors are the airfoil being COMPLETELY symmetrical, and the fins being glued on as absolutely straight radially (90 degrees from the tube surface) as possible, and of course as absolutely straight longitudinally (parallel to the long axis of the rocket) as possible...
I'd rank those in order of importance as:
1) longitudinally straight (parallel to the long axis of the rocket)-- if the fin isn't glued on straight, but at a tiny angle, that angle is going to act like a rudder pointed off to one side-- IE instant spin once it leaves the rod, but probably a fairly regular rate, increasing as speed increases and decreasing as speed decreases. The rocket will rotate because that fin will ALWAYS be at some angle of attack to the airflow and generating lift.
2) airfoil unsymmetrical-- if the airfoil is sanded even slightly off-center, meaning usually the trailing edge is slightly wider on one side than the other, but a leading edge can be sanded slightly off center as well... the fin is going to generate aerodynamic lift perpendicular to the surface of the fin, just as an airplane wing does. This perpendicular lift will be translated to roll (spin).
3) radial alignment-- if the fins are "cockeyed" and not as close to perpendicular to the tube as possible, when an angle of attack ends up being induced due to normal flight forces, the fin is going to act on the rocket to generate roll, because the fin is at an angle to the tube, therefore some of its corrective force will act to create a roll motion due to the force being generated along the width of the fin being out of line with the centerline of the rocket...
If it were me, to maintain the absolute minimum roll possible, I'd go with unairfoiled fins left square cut (more drag, but eliminates the possibility of incorrect airfoils inducing perpendicular lift and creating roll, assuming all the fins are sanded with their edges perfectly square). Ensure that the fins are mounted EXACTLY along the fin lines, by mounting them to ONE SIDE of the fin lines on the tube (just make sure they're ALL to the right or the left, whichever side you choose) and use either a fin alignment tool or a cardboard jig to ensure that the fins all remain exactly 90 degrees to the tube surface, and either 120 degrees to each other (for three fins) or 90 degrees to each other (for four fins) while drying.
Of course, if you REALLY want rock-solid video and want to do an advanced project, I saw on Brittain Fraley's website where he did a "sun-sensor based anti-roll stability system" for his camera rockets... it utilizes several photocells feeding their input into a controller, that controller comparing the photocell outputs to each other... the photocells are arranged radially around the rocket, with the "center" cell supposed to be facing the direction of the sun, therefore generating the highest voltage signal. The other sensors placed partway around the tube at regular intervals get less and less sun, becuase they're pointing further and further away from the direction of the sun, and therefore generate a lower voltage signal to the controller. The controller compares these voltages, and when the sensors on one side of the sun sensor start generating a higher voltage than the 'center' sensor, it automatically moves a servo in the appropriate direction, which then moves a small pair of roll-control fins, moving in opposite directions, to roll the rocket in the opposite direction until the "center" sun sensor is directly in line with the sun again and generating the highest signal, and neutralizes the fins back to center. He showed test videos with the system turned off-- rocket lifts off, starts rolling, and the video goes squirrelly... then he does a second flight with the system on-- rocket takes off, starts to roll maybe 5-10 degrees clockwise, and the small fin can be seen steering to the left, instantly re-centering the rocket... any drift in roll either right or left is instantly countered by the pair of fins moving in opposite directions on opposite sides of the rocket... As the rocket slows down and nears apogee, the rocket starts to tip over from vertical, and the fins start deflecting back and forth, increasingly severely, as the rocket heels over at apogee... but of course the rocket is moving so slowly the small fins have virtually no effect even though they're virtually pegged "hard over" because there's not enough airflow over them to generate much force. The rocket ejects, and the fins again try to steer the rocket so that the sun sensors remain "centered" but of course they can't, so they sorta "flail about" on the trip back down...
VERY interesting project... might actually be one where a heli gyro might work (of course gyro drift might still make sun-sensors more desirable and render the gyro idea a non-starter...)
Later! OL JR
