production blog

: : L>R - 1st AC Leigh V. and 2nd AC Vince S. as we shoot the emulsion test of the 5299

At long last, I got to shoot a film test again.  Most of the work lately has been on some flavor of HD, but this feature film project that’s scheduled for later this year is just that, a film shoot.  So thanks to a responsible producer and a curious director, we’re shooting tests.  This batch focused on three things: frame rates, shutter angles, and grain structure of the 5279 and 5299 (Kodak Vision 500T and Kodak VisionHD Color Scan, respectively).  Here’s what and how we tested:

Note: We’ve only projected the 5279 so far and next week we’ll be viewing the 5299 once we find a suitable post house with a Kodak Digital Processor .  I’ll update this article then.

TEST SPECS:

Camera:
Panavision Panastar II - Primo Prime 50mm - Tiffen Filters

Lab:
Technicolor - processed normal

First Test:

Click on the image below to view a cheat sheet version of our testing notes.

[ click picture to enlarge ]

Frame rate
First we wanted to see what frame rates would look.  Leigh rotated her arms like a propeller as close to the same speed for each test from 6 to 120 frames per second. Conclusion:  14, 24, 48, and 120-fps looked the best based on this test and the speed Leigh moved her arms.  Things might look different when we view it on a monitor but as for projected, these are the speeds we liked.

Shutter
Again, having Leigh rotate her arms, we shot using a shutter angle of 40° to 180°.  Conclusion: We definitely like the 40° shutter.

Combo Test
As the cheat sheet above shows, we did a few frame rate and shutter angle combinations.  Conclusion: Here again, we love the 40° shutter at 24, 48 and 120 frames per second.

As mentioned, we have yet to view this material on a monitor.  Next week, we’re going to telecine to HDCam.  It’ll be interesting to see if we choose differently when viewing the material on a monitor.

Second Test:

Click on the image below to view a cheat sheet version of our testing notes.

[ click picture to enlarge ]

For the second part of our testing, we shot an emulsion test of two stocks, 5279 and 5299.  I haven’t work with 5279 in a very long time and 5299 is one of Kodak’s newer stocks.  Both stocks were rated at EI500.  This was a very simple over and under test.  Ultimately we were looking for grain (in this case a good thing) and that’s what we mostly looked for.  The plan was to have the lab print a one light version based on the grey card and then a timed version that’s corrected in 1/2 stop increments.  The one light didn’t get done and so we’ll be back next week to view this material again.  But we did see the timed version.  We’re also waiting to get the lab report to see where the printer lights came in at.

It was a four stop over and under test.  The lens always stayed at T2.8.  As for the lighting, here where things got a bit screwy.  The test was to be done under two 1K lights and we were told that they would be provided but with as many shows that were there (Bourne Ultimatum was prepping next door), that didn’t fully happen.  We got instead three 650w Fresnel’s - definitely not ideal especially with high speed and shutter tests).  They had to be color corrected and I metered them by adding some 1/8 and 1/4 blue to the point where it ended up within ±50 kelvin. They also suffered from hot spots and so I’m not totally sold on what we got back as far as color is concerned.  (Note to self, always bring your own lights to a camera house.)

Conclusions:  From what we did get back, we liked the grain of the projected image when it was underexposed and timed back by 1/2 stop.  Here again, what we like might all change when we view this on a monitor.

I start a feature in a week and I knew we’d be shooting HD, but I wasn’t sure which camera I’d want to use.  To figure it out, I visited Alternative Rentals this morning to do a test on an Sony CineAlta F900 with a new software package they’re promoting and it’s called YedLog - named after the one of the inventors Steve Yedlin.  The software essentially bypasses the preamp and sets all the menu setting to a zero state.  What you end up with is a Log color space in 4:4:4 (F950) or 4:2:2 (F900).

It’s a fascinating possibility.  I’ve included my camera log for what we captured.  The test was a very simple over and under exposure test to see what what can be saved when conditions are really out of wack as well as what are the most pleasing settings.  In the coming month, I’ll be back to do it again this time with a model and if time permits and Alternative Rentals is agreeable to it, I’ll take the camera out of the prep room and into real locations both day and night.

I’ll be looking at the footage next week on a Avid Nitris and report back on what I find.

: : click on image to see larger version

Here’s a bit of a tip. On a project this weekend we needed to some footage of wood floors at a new home.  No low-mode available so here’s the next best thing.  We took a sand and mounted the camera on it.  We then pushed it forward, pulled it sideways, and pushed it in an arch around one of the tiles.  One addition to this is to attach a C-stand arm to the sandbag to help you push and pull it by giving you leverage.  An other option I used on a shoot last year is to attach a rope to the sandbag.  Run the rope through a pulley and pull the rope.  It’s smooth and you can get a very long run with this method.

: : photograph from the monitor.

: : the “poor man’s” low-mode.

: : attach a rope the stand bag and run it through a pulley.