Quick and probably complex question…Has anyone figured out how to measure the velocity of the balls exiting the shooter your team has created???
Go into a closed room where there is no wind and fire the ball. You want to measure the distance, time and launch angle. The distance divided by time is the X component of your initial velovity. The X compenent times cosine of the angle will give you your initial velocity.
air resistance is negligible in a closed room.
*Beware of metric to standard conversion and radians to degrees!
use a radar gun 
there has been a lot of debate and discussion on this subject, esp regarding how much air resistance slows the ball down (air is still present in a room, even when its not moving)
and also the effect that spin has on the distance a ball will travel.
The two answers that seem to have the most validity are making a time-gate with two optical sensors, that the ball passes through as it exits your bot, or using a radar gun.
If you search this forum on " muzzle velocity " you will find a lot of posts on the subject.
HOw are you measuring this accurately - what instruments are you using? And do you have an idea how FIRST is going to measure this important component at the competitions??
Since our school advisor is an AP physics teacher, he has lots of interesting stuff, like photogates. These are basically optical switches, so when a ball passes one, it blocks the laser and registers as a hit. If you put a photogate right up to the cannon and measure the time it takes the ball to pass the photogate, you know the ball’s width so you know the ball’s speed. It is very accurate because the photogates have a very high refresh rate, and they’re digital.
If you have access to a high school physics room, ask around and see if anyone has a photogate, they’re pretty common.
We used a digital video camera and a piece of PVC piping marked in 1’ increments.
We were able to not only see muzzle velocity, but how consistant our muzzle velocity is even in ‘rapid file’ 
very interesting. A camcorder has a refresh rate of 16mS (60 images per second)
did you have to set the camera in ‘sports mode’ or some other ‘fast action’ mode of operation to get a clear image?
What type? I was under the impression that my miniDV camera ran at 30 fps, largely becuase in iMovie, there are 30 ticks in a second. I googled it and couldn’t find anything.
Only a few pro-level camcorders offer 60fps rates. The vast majority of camcorders (either older formats or DV) offer 30fps interlaced, 30fps progressive (non-interlaced), and/or the quickly-becoming-more-popular 24fps progressive “just like film” rate. Sixty fps camcorders aren’t rare, but they aren’t common, either.
that is exaclt what our team did. We took the Photo gates from a Kelvin “Kel-Air” track and re mounted them to tell us the balls speed.
You can use either banner sensors with reflective tape or the photoswitches from this year’s kit to do it in code. Put them across the exit point of your shooting mechanism.
Start a timer when the interrupt from the beam being broken occurs, get its value when it sees the other side again, and do some math (since the ball is 7" in diameter).
If a ball is going 12m/s, it will take about 14.8ms between these events, well within the capabilities of the RC’s timer resolution (see www.kevin.org/frc for how to use timers).
I think it is fantastic that a few teams out there are fortunate enough to have access to instruments that are accurate enough to measure the ball speed precisely. So, can you do the rest of us a HUGE favor.
Now that you know exactly how fast your ball is launching and what angle of elevation you are launching, can you please calculate just how much air resistance and spin are affecting the balls flight. In other words, does the basic equation for trajectory, or more precisely range, equal what you are seeing? And if not, by how much are you off?
ok, you are correct. The even and odd fields of the interlaced video are captured at 60 ‘fields’ per second
most camcorders, when you pause them or single step the frames will output only one of the fields (even or odd), and then it jumps to the same field of the next frame
so yes, your view-able images would be 33.33mS apart.