All of the gear arrived on schedule from its various locations around the country. I found out that it costs about 420 bucks to ship a Phantom cross country overnight. Seeing as the camera rents for north of 3 grand per day, the cheaper option is to rent one day and ship overnight on both ends.
What’s in the box? In our case we had the Phantom v640, 2x 256 GB cinemags, a Zacuto shoulder mount rig, Arri dovetail plate, Anton Bauer power connectivity so we could shoot off the shoulder, a remote trigger and a Macbook Pro running Vision Research Phantom software on Windows XP via bootcamp.
Tom took the camera home on arrival day to get acquainted with this version of the camera. Testing would be very important because the rental manager at the rental house made a scary comment during the last conversation we had before receiving the camera. Here is how the conversation went.
Rental Manager – “Who is your Phantom tech on this job?”
Me – “My whosee whatee?”
Rental Manager – “We require all of our Phantom rentals to have a Phantom tech”
Me- “I wish I was told before 2 days prior to the shoot.”
Rental Manager – “Sorry we are really busy, it must have slipped my mind”
Me- “Sounds expensive. What is an average day rate?”
Rental Manager – “About 800 bucks, they’re pretty rare”
Me- “So what do we do? Why is it required”
Rental Manager – “Well I strongly suggest you call [insert freelance Phantom tech here]. If you can’t we will let it go this time. We require this because they are complicated camera systems and we cannot be held responsible if you don’t get footage because you didn’t know how to use the camera. We do not want to get phone calls and have to walk people through menus.”
Luckily Tom did have some experience with their software. You cannot make many changes on the camera itself, and must tether it to the computer for many of its settings.
He shot some test footage to make sure everything was in order. He was pretty confident that his knowledge of the software was enough and it proved to be true. He called me up during his tests to tell me how it was going. He said “Something is funny about this camera. The middle of the shot looks to be a different color than the edges.” I dismissed it saying he probably did not have enough light. He sent a frame, which showed how the footage came off the camera and I color corrected it to see if it was going to batch-fixable in post. It appeared that it would in fact be easy to color correct. After actually going through the process though, I found that there was considerably less color information in the middle of the picture than was necessary to be able to fix it easily. This would take some serious work in post to make it usable for our purposes. You can see the shot in the images below before and after. You can see the color corrected version actually makes the plume much more noticeable.
The next day, we loaded in and built the set with plans to shoot a test subject at the end of the day. We moved in without incident and Jeff filled the room up with all kinds of lighting fixtures.
The main sources were going to be (4) 5k tungsten fresnels. He used two in the rear corners of the set to wrap the subjects body in warm light and the other two behind some massive diffusion to key them off camera left. There were added tungsten lights for fill and a half dozen 1ks for the set it self. The whole rig was wired up to dimmers so that we could have two intensity settings. The higher intensity for the very light hungry high speed camera and then the lower setting for the standard speed F800. He added cooling gels to the wrap (5k’s in the corners) because big tungsten filaments go very warm when dimmed down below 50%. This causes the shots between the two cameras not to match in color temperature which would have been unacceptable in this instance. Smart guy… really.
We set up an area for the Phantom computer behind the DOP in sight of the main subjects. We had the F800 on the main set with dolly track running parallel to the front of the set for side to side moves because we would not have time to repeat and nail follow focus on shots where they camera moved toward or away from the subject. We also elevated the dolly track on 24 inch Wenger Decks to get the camera up higher for a more pleasing composition. We wanted to show as much of that beautiful temporary hardwood floor as possible. We had the Phantom in front of that rig on the ground level so Tom could grab it easily and shoot handheld with the players.
TESTING WITH A LIVE SUBJECT:
Due to the fact that failure is not an option with pro athletes involved, we called in a favor to have a young basketball player come in and bounce a ball for us on camera so we could block out some shots and make sure everything was going to work. Practice practice practice! You can see the test footage below shot with the Phantom.
WHAT DID WE LEARN THROUGH OUR TESTS?
1. The Phantom was going to need to be tethered to multiple cables. I needed to see Tom’s shot on a monitor for one. Secondly the Anton Bauer Hytron 140’s were lasting about 20 minutes with that camera. It was very power hungry. We were going to have to plug it in to the wall. Third, because the camera was light sensitive enough, I wanted to shoot 960 fps instead of the planned 480. Therefore we needed someone else to manage the media and run the post trigger once we thought we filled the 10 second buffer. They would then send the command from the software for the camera to dump the image sequence out of its buffer to the cinemag. This workflow can be seen in part in the below video from the actual shoot. You can hear Tom calling for the trigger at the end of his count.
2. We were going to have to pay close attention to where talent was standing in relation to the light. The inverse square law always applies, but when dealing with these intensities, that dropoff is FAST when you back away from the light.
3. That weird spot in the center of the image, now infamously known in our circle as “the plume” was not going away. We shot some test footage here to show the rental house our problems (even though they said not to call them with Phantom problems without a tech on site).
4. Because I decided to shoot 960fps, we were not going to have enough storage. Always rent more than you need.
5. Lighting, as always, is crucial. Don’t ever bring your 5dMkII attitude to a Phantom shoot and think you are going to shoot available light unless you are on the Sun.
6. Things that we thought might look cool in high speed did not necessarily look cool. On the other side, things that we did not expect to like were surprisingly pleasing to watch. The point here is to experiment as much as possible.
We were now much better equipped to shoot the following day with the knowledge that was gained from the various test sessions.
Missed the first installment? You can read Phantom v640 Pro Basketball Shoot Part One: Planning by clicking here.