I don't know if you ever saw the beautiful shot of a chameleon shooting his tongue out to catch a flying bug, a black, unlit background, but, as I recall, the photographers took and set up a flash unit, like the Vivitar 283 or 285, at the distance they wanted, and controlled the amount of flash duration by making a device(s?) to hold both ends of a fairly thick, strand of fiber optics cable from a position on the flash window and the opposite end fitted onto the Auto flash Thyristor eye on the same flash unit.
The cable was very short, calculated too only provide enough ("fast" flash) to the distance from the lizard and flying prey, no more.
Obviously, the custom cut optic cable, I think measured from the flash window to the target and back to the marked film plain, which only allowed a super short, direct path of the flash cycle to reach the controller Auto Vivitar Thyristor eye and start, ending the cycle with just enough light to travel the preset (speed of light volume) measured distance, at the speed of light back into the electrical relay, also at the speed of light.
I believe I have this remembered correctly, I only read the article once several decades ago, but if it'll help, give it a shot and let us know your method and formula calculations, please.
Cheers
PS: I believe that if you shoot this bullet (or series of bullets) in a dark enough set, even outdoors in skylight, with a properly lengthened optic cable, only the actual brief flash will illuminate your bullet, as the super short cutoff will prevent the light from traveling long enough and far enough to see and record the movement of the bullet and background.
Excess illumination from the burning cartridge propellent might be well controlled by shooting through a pair or more of light absorbing, black Vanta type, painted sheets of card stock, I believe, as well.