It used to be that a cinematographer needed a ton of filters in their kit to create the images they needed.
I remember our cine teacher in graduate school rolling up with a packed SUV where his entire payload was just case after case of filters, all with his name engraved on them. It was how you crafted your image.
With the digital revolution, we have a lot more power in post, and while it's still great to get physical glass in front of the lens if you possibly can, you can do a lot of what filters used to do in post if you can't spend thousands building your filter kit.
There is still one set of filters you absolutely need on set and just can't recreate in post. Neutral density filters, or ND filters, are ones you still want in a filter tray. Especially infrared control (IR) ND filters.
ND filters give you the ability to choose your T-stop. If it's a bright sunny day, sure, your lens might stop all the way down to a T16, but if you want a shallow depth of field and to get the best of your lens, you want to pop on an ND 1.8 and get yourself back to a normal aperture.A good ND filter gives you consistent color reproduction, and for digital shooting, they block infrared light (we'll talk more about that later).
When NiSi started making ND filters at around half or sometimes even only a third the price of competitors, with a one-year warranty and a reputation for good customer service, we were intrigued and wanted to put them through the paces.
There are a ton of highly technical reviews out there on these filters already that go very, very deep on measuring with data how they perform (and they perform well, especially for the price). That is what got us interested in testing. But what we really wanted to see was a coverage test where we saw if we could match our coverage.
The reason for this is that, unlike still shooters, filmmakers are always covering scenes. And indie filmmakers are always covering scenes within limitations. I've done countless three-camera shoots where we only had one set of ND filters, and we had to put an ND 1.8 on the widest lens and only an ND 9 on the middle lens, so what we wanted to see is if we could intercut different ND powers together with these filters in an edit in post.
The other issue to talk about here is infrared radiation. Digital sensors are more sensitive to IR than film was, and when you start stacking up a lot of ND filters, one issue you will often see is IR issues in your blacks, especially with synthetic fabrics.
To solve this problem, on a digital shoot you need to be sure you are either adding an IR filter to your filter stack, or using natively IR blocking ND filters, sometimes called "hot mirror" filters since they block the "hot" IR light.
Poor-quality IRND filters sometimes affect skintones, since the IR spectrum lays quite close to the red spectrum where a lot of skintones live, and if they are overly aggressive in cutting IR, they cut into the red light and skintones shift. This is one reason why many manufacturers don't use IRND filters for internal NDs.
The Blackmagic Pocket 6K Pro, for instance, chose to go with normal NDs for internal filtration to focus on skintones, which are quite pleasing on that camera, but leave you open to the possibility of some infrared issues when stacking up big ND levels.
As you can see, there is clearly a red shift and infrared radiation kicking up in the black polyester facemask that is controlled using the NiSi hot mirror. Of course, that IR contamination can often be fixed with some tweaking in post, but that adds time to a process that might better be spent elsewhere.
In addition, if you are working with fashion and the black polyester item is the product, which will be scrutinized to death, avoiding any contamination of your image to begin with is vital.
Our main goal in the testing was to shoot different angles on the scene and see how they compare at different strengths of ND, since that is such a common scenario on independent projects where you roll out with one set of NDs.
To make it trickier, we worked with their graduated filter for ND 9, which in our experience with other brands often leads to work color reproduction than full-coverage filters.
These are just surprisingly good filters for the price point.
Honestly, in terms of grading the footage, there just wasn't a color cast at all between clean, ND 9, and ND 2.1. We would feel great mixing and matching these filters all day. We've worked on jobs where the ND filter gave such a color cast it made the color grade much longer, having to build a node specifically for each ND and checking the slate on takes to see which camera got which ND.
This is definitely something we wouldn't do here. There is a tiny hint of a cast on the graduated filters, but that was quick to tweak.The reverse angle shows slightly more color shift, which is a great reminder that you always need to test in a variety of situations. In the wide shot, the filters matched exceptionally, but every shot is different, and the tones in this shot react slightly differently to the light. Well within grading range, and we still have pleasing skintones.
We also felt like the accuracy of the filters in terms of exposure was spot-on.
You can mix and match levels and quite easily recover a balance in post-production. While on a run-and-gun shoot, the internal NDs you see on cameras like the C70 and the Blackmagic 6K Pro are really quite useful, but for any kind of studio shoot where I have the time to rig a matte box and put on a full-sized filter, my preference would still be to work with something like the NiSi IRND over internal filters.
With matte boxes like the SmallRig getting more affordable, having access to quality IRND filters really makes a difference in shooting any bright light setup.