Variable ND filter

[REAndy]

Has anyone tried one of those variable ND filters? I think it is similar to a polarizing filter, but it would have 2 polarizing films. One fixed and the other you can rotate which would make it "darker" (you know like when you take 2 pairs of polarizing sunglasses and put one lens in front of the other and turn one 90° to the other and you can see it get darker)

My concern/question really deals with using it on a camera that does NOT have TTL metering. How would I know what "power/value" it is set at to compensate exposure. My only thought was I could an AE camera with out the filter, and note the exposure value. Then put it on the camera, rotate the filter for minimum ND (mark the filter ring); note the new exposure. Then rotate more for another stop of light loss; mark ring. Rotate for another stop of light loss; mark ring. etc. etc.

But I was wondering if someone is using them; how they work for you; and if you do not have TTL metering, what is your solution.

Thanks.

 

[Jim Jones]

If you use a handheld meter, try metering through the filter at the filter setting you intend to photograph with.

 

[Kino]

As Jim Jones indicates, meter through the filter.

Also, you can take it outside, place it in full-open or best transmission position and then scribe or mark the two lens rings to show full-open.

Note the default light loss; there will be some loss at full-open.

Then, as you close it down to maximum, mark one of the rings with 1/2 stop increments.

Now, add default density + your additional density mark and get total light reduction.

 

[AgX]

Also some TTL cameras are metering wrongly with standard polarizers. Thus the variable ND filter then needs the back polarizer to be of the circular type.

 

[MattKing]

I have one of these and they can be challenging even with a TTL meter.
One challenge is that the variable ND filters can end up lowering the light level to below the sensitivity range of the meter. I have a 77mm filter of this type, and a metering chimney finder for my RB67, and the ND filter takes the light levels below the sensitivity range of the meter.
And some/most of the variable ND filters can impart a colour cast, which is relevant to colour work.
If you are going to try to do this with a hand meter you will need to make sure that you keep extraneous light from the meter.

 

[Maris]

I've tried several variable neutral density filters based on the principle of contra-rotating polarisers and they all have bad problems:
Polarisation efficiency is not the same for all wavelengths of light. As the filter is adjusted to maximum density colour casts appear.
Polarisation efficiency is compromised by oblique rays not being polarised the same as normal (straight through) rays. Fully crossed polarisers seen from a distance might look uniformly dark but up close, like when screwed onto a lens, the dreaded "dark cross effect" appears. The ND effect is not uniform over the field of view.
The slightest strain, mechanical or thermal for example, introduces a twist in the plane of polarisation somewhere in the filter (usually near the edge) and again non-uniform ND is the result.
All of these problems are most severe at or near maximum ND. For example I tested a Variable ND filter marketed as a ND1000 (about 10 stops) and it was useable up to about 5 stops. That's when I bought a solid 10 stop ND filter...works perfectly.

 

[REAndy]

Thanks everyone for sharing their experience and insights. I think I'll pass on the variable ND and try one of those (not sure what to call them) "big square plastic filters that come with adapter rings and a 3 slot filter holder" in the ND style. I will stack-up the filters if I need max ND.
These "big square plastic filter things" look like a Cokin setup I had years ago for my 35mm (back in the late 70's), but those filters were maybe about 2.5 inches square. These "new ones" look larger and must be about 3.5 inches as it says it will fit up to 88mm.

 

[trendland]

Has anyone tried one of those variable ND filters? I think it is similar to a polarizing filter, but it would have 2 polarizing films. One fixed and the other you can rotate which would make it "darker" (you know like when you take 2 pairs of polarizing sunglasses and put one lens in front of the other and turn one 90° to the other and you can see it get darker)

My concern/question really deals with using it on a camera that does NOT have TTL metering. How would I know what "power/value" it is set at to compensate exposure. My only thought was I could an AE camera with out the filter, and note the exposure value. Then put it on the camera, rotate the filter for minimum ND (mark the filter ring); note the new exposure. Then rotate more for another stop of light loss; mark ring. Rotate for another stop of light loss; mark ring. etc. etc.

But I was wondering if someone is using them; how they work for you; and if you do not have TTL metering, what is your solution.

Thanks.

What you metioned remember me on a once a time workflow filming the sunrise on motion picture Film. It was filmed in time lapse with 1frame every 45 sec.
And with compensation of the lens lris.
A lens stop wasn't in steps it was constantly. So I marked the space between 2.8/4/5.6/8/11
lens markings in 10steps between. So I longer possible exposure duration of the full sceene via
exposure correction of 10 steps between one full stop.
In short to make it like your desciption is a bit complicate but pls. do it with that more complicate
way because you can learn much with that - that's your main profit.
with regards

PS the markings I made on my CINEMA lens came from a copied ruler milimeter on shieds of paper.
The scale had to be vary because the space onto the lens was not identical between each stop
(4/5.6/8 a.s.o) but it has to be devided with factor 10.
Last I did stick down my 10 step markings between lens markings.
PPS : No idea about if the original lens markings have been such precise that it made sense to divide it with 10 (perhaps it is just a log. function so my linear devided steps have been a failure )
but the resulting sceene was superb and I wasn't lazy - that is allways a good combination.

 

[trendland]

I have one of these and they can be challenging even with a TTL meter.
One challenge is that the variable ND filters can end up lowering the light level to below the sensitivity range of the meter. I have a 77mm filter of this type, and a metering chimney finder for my RB67, and the ND filter takes the light levels below the sensitivity range of the meter.
And some/most of the variable ND filters can impart a colour cast, which is relevant to colour work.
If you are going to try to do this with a hand meter you will need to make sure that you keep extraneous light from the meter.

But a corrected ND filter shouldn't do so (resulting color casts).
Isn't that the conception? ND = neutral density and "neutral" indicates : No colors are involved !
(in easy english )
with regards

Sure some glass of cheap ND filters isn't the best but then one might not sell it as ND !
Better marketing for such kind of stuff could be "dark lens protection filters"

 

[AgX]

Read "neutral" as near neutral. And a variable density filter based on polarisation is especially complex as Maris explained above.
A spectrally equal-density filter or equal reflectivity mirror does not exist. Look at spectrographs where available. But one can chose between different spectral characterisctics. And the spectral range of interest may fall into an equal part.