ND Tests

BobDiaz

BobDiaz

The Professor
I came across a ND Filter Kit, Tiffen 46NDK3, and decided to give it a try. As you all know the HMC40 is 43mm, so a 43mm camera --> 46mm filter adapter is required. As it turns out I can stack up to 2 filters with the lens hood on, so this works well. I do have to take off the lens hood to add or remove filters.

The filters in the kit are:

ND 0.6 (2 stops)
ND 0.9 (3 stops)
ND 1.2 (4 stops)

Combinations allow for higher selections:

ND 0.6 + ND 0.9 (5 stops), total ND 1.8
ND 0.6 + ND 1.2 (6 stops), total ND 2.1
ND 0.9 + ND 1.2 (7 stops), total ND 2.4

I could stack all 3 and get 9 stops of ND, but this is overkill and I can't think of a reason I'd need this much ND. The lens hood will no longer mount when I put on all three ND filters.

For normal shooting at 1/60 of a second exposure, the ND 0.6 seems ideal. IF for some reason I needed 1/30 of a second exposure, the ND 0.9 would work. The ND 1.2 also works for normal shooting.

My feeling is that if you had to purchase only 1 ND filter for shooting at 1/60 of a second shutter, the ND 0.9 may be the best choice overall. However, both the ND 0.6 and the ND 1.2 do very well too, so it's a matter of personal taste.

If you want to use a long shutter speed, like 1/2 a second for timelapse OR you want to shoot as wide open as possible for shallow depth of field, ND 1.8, ND 2.1, and ND 2.4 seem to be within the right range. In general, I like the ND 2.1 as the filter of choice. Again, this is a matter of personal taste.

These tests assume full sunlight for normal outdoor shooting. I have not had a chance to go to a beach and see the impact on the lighting yet.


Bob Diaz
 
What about color cast? Are these filters really neutral? I have heard about a slight blue shift.
What happens to color when you stack more than one filters? Can you mix footage shot with bare camera with footage shot with ND filters on?
 
Weird, with a single filter, I didn't notice any color shift, but with stacked filters, it's as if the green channel was lowered in level. In the shot I'm uploading now, the trees in the background should be green, not brown. Please wait, It's still uploading....

I need to run some more tests on this.


Bob Diaz
 
In earlier posts in this forum discussing whether or not ND filters create color shifts one suggestion was to always do a white balance whenever you add ND filter. I am new to all this, but do you really need to white balance when you add an ND filter--if so does that imply ND filters do add a color shift (assuming light does not change)?

Regarding stacked ND filters which do seem to add color shift in Bob's example, would white balancing stacked filters solve color shift or is something else going on with optics?
 

This was a quick and dirty video to show what happens as I try different ND Filters.

At ND 0.6, there a slight shift in the green, BUT you have to watch for it. It's very slight.

At ND 0.9 and ND 1.2, the drop in Green becomes stronger. ND 1.2 has the strongest drop of a single filter. Both the ND 0.9 and the ND 1.2 can be fixed with a minor color adjustment in post.

Stacking filters adds to this effect; a ND 0.6 and ND 1.2 have a noticeable drop in the Green level. However, it is possible in post to adjust for this and hide the shift.

Doing a Color Balance BEFORE shooting did not solve the problem. My guess is that the Color Balance fixes Red / Blue Ratio, but the Green is untouched.

With all three filters stacked, ND 0.6, ND 0.9, and ND 1.2, there was no way I could fix this will a simple color correction in post.

In the cases where I could fix the color in post, I did NOT try to get it bang on, just close.


I assume that the more expensive filters don't have this shift OR at least it's not as strong.



Bob Diaz
 
Bob, thank you for this ND test. Lacking experience I try to accelerate learning by reading and I don't recall warnings about color shifts with ND filters. In past forum notes regarding ND filters I recall Felipe has been pretty much the lone voice warning about color shifts with ND filters. There was the occasional comment to white balance when ND filters were added. However, in Bob's video he mentioned doing a white balance and it did not seem to help with color shift.

So, for newbies like myself, what is going on? Do we want to stay below 0.9 ND filters to avoid color shift?

Clearly stacked ND filters (at least the ones Bob used) are not good at least under the conditions Bob was using (fixed shutter speed, full auto otherwise).

What cautions or camera adjustments (is it just that Bob is getting into too much gain on camera and introducing color shifts when he stacks filters?) should we use to avoid color shifts with ND filters. Forgive me if I misquote, but I believe Felipe has maintained increasing shutter speed is best way to cut down on light without impacting color instead of using ND filters--and he has cautioned of course that higher shutter speeds will change the "look" of the video.
 
Dennis High shutter speeds, work well as long as your camera support is very steady and camera movements are slow, so will not look to active.
High shutter speeds are nasty if you move too much the camera, or make fast movements.
I love how crisp the video looks when you use the right camera support with high speeds.
I used to use B+W filters but also have a slight color shift so I sold them.
Kodak gelatin filters are nice, but manufacturer says they don't last long, I bought one and only time will tell about durability.
This was shot at speeds between 1/250 and 1/500
http://video.felipehenao.com/agua.html
This was shot at speeds between 1/250 and 1/1000
http://video.felipehenao.com/San_Marcos.html
 
I suspect it may vary a lot between different brands and quality levels of filters (some manufacturers have "economy" and "deluxe" lines, as with many other products). I'm not aware of any comprehensive and objective tests out there -- anybody got a clue? Kodak used to tout the superiority of their ND filters that were made with a metal film deposited on glass rather than the traditional dyes -- these reflected, rather than absorbed, incident light. I don't know if anybody is making them that way now (or if they even were really "all that" to begin with). I generally stick with B+W and Hoya HMC series for my filter needs -- but I haven't tested them objectively and don't know how well they measure up. This sounds like it would make a valuable research project for someone! As Bob observed, white balancing in the camera only operates along the red-blue axis (holding green constant) but is unable to correct for errors in the green-magenta direction.

- Greg
 
Not only do some ND brands have a certain color cast, but also beware of IR contamination. That is likely the culprit with color shift in very high ND cases.
 
I'd post the final video in the Video section, but seeing as how we talked about it here...



The first shots were taken on a Friday after 5 PM with a strong overcast. I never seemed to get the color correct. The wide angel shots were without the ND Filter and were taken on a Sunny Saturday.


Bob Diaz
 
Bob,

While I am glad you did the work on the ND filters I have to laugh at the video of field day because it looks just like my clubs!! 2-4 working the rest watching! We had good weather too, well at least for most of the day! Central Illinois Radio Club. Got a real kick from the video in time laspe and the music!!

Jeff
KC9QQM
 
KC9QQM de KQ6WQ

We had fun. Like most field days, there are some who do nothing but work the radios and others who see this as a great time to socialize.


73,

Bob Diaz
 
That's a good question, is there any good resources for ND filters out there? I was about to order this set as well, but now I just don't know. I'd like to find a set that has a minimal color shift if possible.
 
You must log in or register to reply here.
Back
Top