Mattenstrasse 90
2503 Biel/Bienne

Correcting IR contamination on RED Sensor

This has been a through Test, to see how CRUNCH and our Rental-customers can correct the IR contamination that happens when using ND Filters.
It is commonly known, that CMOS sensors are sensitive to IR and UV, and that by reducing light input to the sensor by adding ND Filters without IR and UV filters in front of the lens, the InfraRed-Lightwaves still come through and affect the image.
The more ND you put in front of the Lens, the more you see the IR contamination.
So there are 3 ways to correct this problem.

1. Schneider IR Cut Filter (in front of the ND Filter)
2. Rosco Tru Color Filter (in front of the ND Filter)
3. Color Correction in Post

1 and 2 are Physical corrections
3 is a digital correction.

Here are the results tested with a ND1.5 Filter. Shot at ISO 2000 1/50 on a Samyang 35mm f1.4 with White Balance set to 5600K.

First: the image untouched out of the RED camera: (left: ND1.5, middle: ND1.5+Rosco right: ND1.5+Schneider)
Click on image to view full-size

Second: the image corrected White Balance within REDCineX: (left: ND1.5, middle: ND1.5+Rosco right: ND1.5+Schneider)
Click on image to view full size

And at last, The Image without Schneider or Rosco Filters. White balance ajusted in REDCineX + slight Red Shadow curve ajusted. (left: ND1.5, right: ND1.5)
Click on image to view full size

So in the end, is it worth filtering out IR contamination with physical filters, or can you simply ajust in post?
Our conclusion is that the Schneider and Rosco Filters certainly have their validity, but in this case we find that simply ajusting the shadow curve a little can do wonders to the ir contamination.

Stay tuned for more tests.

Written by: Daniel Kunz, CEO of CRUNCH motion design

Trackbacks for this post

  1. Choosing Neutral Density Filters with Infrared Protection: Necessary or Overboard? - NoFilmSchool

Leave a Comment