ND filters
Before this current trip to Asia, I tested a number of ND filters with my Minolta Color Meter. I have 2,3,6,10 stop glass multi coated B+W neutral density filters in 58 and 78 mm sizes. I have step up/ down rings to 72, 67 and 55mm, which covers most of my lenses. For my beloved Zeiss 21 2.8, which is a 82mm front, I have a Hoya 3 stop ND and a Vari ND filter. My tests show that there is essentially no color shift with the B+W multicoated glass. I tested each of the BW filters, and the max color deflection was 100k, with most of them having no deflection at all. The Hoya was 400k and +2 green. The Fader ND was a shocking 1000k and +5 green. My 77mm SingRay showed 500k and +2 green. Bob at Singray explained that the 500k shift was intentional to offset the inherent blue cast in polarizers, which is something I had not heard of before.
The meter I use when shooting motion is the color meter, not an exposure meter. I check the exposure in the viewfinder, and only if I am unsure, do we pull out the light meter. The color however gets checked for each and every shot. I have found that color correction with h.264 is not something that is forgiving. We try as best we can to nail the look in camera. When I use the Singray, I have to compensate for the color shift by reading 500k up from where I want to be which adds an unnecessary layer of confusion that I would rather not have.
When shooting the Vari ND vs no filter, the color shift is strongly apparent. Ugly. The SingRay is not as bad, but the shift is still apparent. This says nothing of the flare that I get when shooting backlit, which completely obviates the Zeiss ability to hold shadow detail. Practically speaking, I find myself using the Singray on my 24-70 zoom only when I am in a situation of fast changing light conditions. After all, a yellowish green cast is better than not getting any shot at all. But, this is only when I have to work very fast, like today hand held on the back of a motorcycle racing down rice paddies. I think out of all the shots we did in 3 weeks, the SingRay was used maybe twice.
Hi David,
Just by reading your blog I feel more confident as a photographer.
So I only felt right sharing as well; as ND color shift issues also effected my work. What I found out was similar to the RED, the 5D needs extra IR pollution filtration when using any ND, via a “Hot Mirror” filter as the first thing the light hits. When you ND (with a filter that is only blocking visible) it increases the (relative) IR levels that your sensor is taking in. Hence the massive color shifts.
Filmtools.com has an awesome (slightly expensive) line of IRND filters from Tiffen that combine the Hot Mirror and ND into one piece of white water glass. Which is the ultimate solution (at least in my opinion
Thanks for sharing! Also anytime you need an experienced solid assistant / bts shooter (still or motion film) in Miami look no further. Would you like me to send you my resume?
Best,
Adam Reign
Miami, FL
Thank you for your kind words. Yes the IR is an additional problem, one that I have seen show up after 6 ND. The tests I did that were showing color shift were done on a color meter and the shift was actually the other way, towards green. A better way to do the test would have been directly onto a CMOS sensor. This was done very well by Shane Hurlbut, ASC and posted on his blog. I highly recommend reading Shane, super knowledgeable and a really nice guy. Good luck and thanks again for taking the time to write in. Please send your resume to david@dhstewart.com. I was just in Miami 2 weeks ago for an ad job and could have used you.