[MF3D.FORUM:1479] Re: NSA MF slide show

["Bill Glickman" <bglick@xxxxxxxx>]

Tom

       I do agree with your math....the inverse square law is
indisputable... but I think this is where we may have been confusing each
other.....  With the electronic flashes, the light leaving the flash alters
based on the fl you are shooting at.   So for longer fl the beam is very
narrow, and for wider fl the beam is very wide....so if you set the beam to
wide and then did the math to further the distance to the subject I would
fully agree with your inverse square math.   But when you change the beam of
light so its very narrow, it travels much further with greater intensity,
due to less loss.   This is why a fresnel lens can be put over an electronic
flash which narrows the beam to be used for a 300mm lens at several hundred
feet!  The reason is the narrow beam...whereas we could never get any of the
calculations to prove that we can shoot at 200 ft with a 120 gn flash.  I
have used these and they do work amazingly well...except they are designed
for one fl only.

Bottom line, if a flash such as a bulb is left alone, then the your inverse
square math applies well, but when you change the pattern of the beam
through reflectors, the math is altered, and usually only best described
through testing.   Does this properly describe where we could have been
missing each other?

Bill





> I was addressing the case of multple flashes to compensate for a more
> distant subject. That could be multiple lamps fired simultaneously or a
> single lamp fired many times.
>
> At a given distance the light hitting the subject doubles when the light
> source doubles. But if you double the distance to the subject then the
> light hitting it is quartered. So the light source will need to be
> quadrupled in order to get the same exposure.
>
> For the focal length I am just using your data and more-or-less fitting
> it into the equation. It appears that the distance is proportionate to
> the fourth root (square root of the square root) of the focal length.
>
> To include N (the number of flashes) and fl (focal length), I would be
> inclined to write this as:
>
> MaxDist = GN (ft) / fStop * sqrt(ISO/100) * sqrt(N) *
> sqrt(sqrt((fl/50mm)))
>
> I like to do the messy stuff first and the division last so I would
> arrange it this way:
>
> MaxDist = sqrt(sqrt(fl/50mm)) * sqrt(N*ISO/100) * GN / fStop
>
> All of this assumes that the flash is located near the camera. That is,
> the distance from flash to subject is the same as the distance from
> camera to subject. I expect that this is often NOT true when slave
> flashes are used. Then the math is certainly more complicated.
>
> Tom Hubin
> thubin@xxxxxxxxx
>