• Wondering which camera, gear, computer, or software to buy? Ask in our Gear Guide.

ND filter and other filters - enhancement baked onto image?

Consider the below scenario:

Two identical camera of the same brand and model are setup with the same lens (let us assume the camera is an Arri Alexa or Blackmagic URSA Mini). The two cameras are shooting next to each in an outdoor daylight scene. Both cameras are in continuous/steady operation. Camera A has no lens filters however with camera B, ND filters, cool and warm filters and a series of different filters are fitted one at a time and each is left in the camera for 15 seconds while the camera is recording the scene.

The RAW footage from camera A and camera B are then played side by side and color graded by way of applying the same LUT to both footage.

As a reminder, camera A had no filter and camera B had a series of filters applied in sequence.

The questions here are as below:

1. When looking at the footage comparison, will camera B's footage look certainly different than camera A's footage especially when the various lens filters are used on camera B?

2. When a lens filter (be it ND, warm/cool or whatever) is used on a camera it therefore enhances the image in some way to make it look better or different etc. than if no lens filters were used. The question here is, when a camera is recording using a lens filter attached to the lens, is the enhancement baked permanently into the image? Is it possible via software to remove the enhancement and get 100% original picture quality as if no lens filter was used (i.e. footage than looks 100% like camera A's footage)?
 
I’ve done this exact test using one Arri Amira and swapping filters in and out.

To answer your questions:

1. Yes, but it depends on the specific filter as to how different. Theoretically a high quality ND/IRND shouldn’t affect the image much if at all other than to limit the amount of light that hits the sensor. In practice, many ND filters have a colour cast or other imperfections in the glass that do have an effect on the image, however whether it’s a large effect or one that is imperceptible would depend on the filter.

Setting up two Amiras side-by-side, you could dial in Arri’s internal FSNDs and the images would be pretty much identical other than exposure differences, but I don’t know if they’d be a 100% technical match.

2. Given the nature of a glass filter, I would suggest it would not be possible to grade Camera B to look exactly 100% like Camera A at a technical level. Depending on the specific filter, you could get extremely close - and you can certainly get close enough to intercut without issue, where the perceived difference is 0, again depending on the filter.

Some filters are very difficult to imitate in the grading suite, and undoing a filter is even harder than implementing a filtered look in some instances. For example you couldn’t shoot with a Soft FX filter in and then undo that in the suite. You could shoot clean and approximate the look of a Soft FX filter in the suite.

Theoretically, with a simple colour cast or ND filter, given a perfect filter it should be technically possible to get an almost perfect match, however practically it is generally easier to ‘bake in’ the look of the filter.

Arri’s FSNDs would be your best chance of getting a 100% match.

At the end of the day, in the realities of making movies we’re not generally looking for absolute technical perfection, and in that sense there are many situations where a filter could technically be ‘graded out’ if needed - again, of course depending on the exact filter used.

I often shoot with glass filters in for all sorts of different softening and colour FX. I also like to work in such a way that my images are close to final on the day anyway, so the grade shouldn’t take months...
 
Last edited:
Note on the NDs:
using the ND will ask for a different f-stop to get the same exposure as without ND. This will result in a difference in depth of field.

Most colour and polarising filters (among many others) also require an exposure adjustment as they cut light (again, the nature of a filter), so your depth of field will often be different. Not sure if that’s included in the ‘100% same’ criterion or simply just the look.

I’ve pulled out colour filters in the past to get more light on the sensor and then have to grade those shots to match. You can get them very close, close enough to cut together seamlessly - even close enough that there is little discernible difference. However they still won’t be a 100% match in much the same way that two of the same lens will not give you an exact 100% match.

At the end of the day, it’s all glass. At the most basic level, you’re adding a glass element to your lens so it will have an ‘undoable’ effect on the image - even using a clear filter will have an effect on the image even if it’s imperceptible. Then there’s potentially inherent imperfections in the glass, plus wear and tear....
 
Back
Top