Stopping down vs ND filter

SuperSet

Well-known member
Hey fellas, can anyone offer any advice whether it's better to stop down the lens vs using an ND filter when shooting in sunlight? I've heard that above f11, you get some diffraction?
 
I'm not even sure what Diffraction is, but the main reason to use an nd filter over just stopping down is to control the depth of field, not to mention an nd filter acts like a smooth, stepless iris ring.
 
I would highly recommend investing in a vari-ND filter to serve your lenses! It is also true that there most likely will be diffraction with most lenses when you start venturing deeper into the land of 2 digit f-stops... If you have no need to use depth of field to separate the subject or what so ever, then I would say f5.6 on most 25-30mm lenses is the sweet spot for having a deep depth of field.
 
The term diffraction is well known in photography but even here you need sometimes bigger f-stop numbers in certain situations i.e. landscape situation with a wide lens and you want to get more close to your foreground. Iam not sure how professionals see this, but since many use besides their tack sharp Panasonic lenses aswell older prime lenses which have a more soft and nice look, I bet diffraction isnt that big of an issue in videography.

Of course you can stop down IF you wish to have this look with 1/50 for example, but its probably better to use an ND Filter or ND Fader so you can use the F-stops you want. You can use aswell 2x polarizer (one linear and one circular), this works quiete well.
 
Last edited:
Good photos there on Rockwell's site.

Diffraction ruins your image sharpness, and the more you stop down your lens, the more diffraction comes into play. Also, the smaller the pixels on your camera, the earlier diffraction comes into play. And the higher-res your camera, the sooner you'll see the effects of diffraction.

So on a standard-def 2/3" camera, you could go to f/32 and not see any softening due to diffraction. But on a 1/4" HD camera, you'll see diffraction affecting the image as early as f/4!

For HD cameras, depending on the pixel count, the diffraction limits are something on the order of:
1/4" chip: f/4.
1/3" chip: f/5.6.
m4/3: about f/11 is as low as you should go.
On an APS-C, diffraction will show up at f/16.

That's one reason why it's so irritating (to me) to shoot video with these types of cameras. You simply can't do it, in daylight, properly, without adding some ND filters of some type. In daylight, even at 100 ISO, at a 1/50 shutter, your scene might require f/22 for proper exposure -- and that means your image will be all soft and blurry due to diffraction. Where you'd really want to be is around f/5.6, and that means you either have to ruin your motion by using 1/800 shutter, or put on four stops of ND filtration, or go ahead and stop down the lens and get soft images.

In general with lenses, you want to avoid the most-open stops and the most-closed stops. If you have an f/1.4 lens that goes to f/32, like that Nikon in the Rockwell article, you'd want to avoid 1.4 and 2.0 on the open side, and f/22 and 32 on the closed side, and use the aperture range in the middle. That's where most lenses are designed to perform best.
 
I'd invest in some good ND filters. Get a set, like 0.6x, 0.9x & 1.2x or 1.5x. More infos about ND filters here.
I've recently bought a good one, for a very affordable price: Cavision ND 1.5 to fit in my mattebox. Works fine and doesn't have much of a color shift (very slight and subtle blueish shift) as shown here.
That being said, with that ND 1.5 (which is equivalent to ND32 / 3.125% of transmission) you can easily use it outside with f/1.8 at ISO100. That means that tweaking the ISO and Aperture, you could also use the lens at f/4 too by adjusting the ISO to 400 or a little more. That's why a filter set may come in handy: use one that will stop less light and stop down the lens as needed. Gives you more flexibility.
That's what I'd do!
 
yeah, being to lazy to search, if anyone wants to answer, what is "diffraction"

"diffraction" --- The classical setup to observe diffraction is to set up a beam of light, then place a 'knife edge' that partially blocks the light. 'theory' suggests that the sharp edge should produce a 'sharp' edged shadow... enter in diffraction, and the edge is slightly blurred.

The leaves of the lens aperture act as 'knife edges', and so as one closes down the aperture, the diffraction blur becomes more pronounced.

The amount of diffraction is wave length dependent, and so one will have 'color' banding because of the differences in diffraction of the range of wave lengths that make up visible light.
 
So what's the sweet spot, f stop wise for the hvx200?
An HVX200 has comparatively huge pixels, as compared to a 1920x1080 1/3" camera, they're 4x as big. So its limit is about f/9.5, any lower than that will cause a substantial drop in resolution. So avoid double digits and you'll be fine.
 
Diffraction aside...

The biggest reason for using ND over F-stop is to give a consistent DoF when shooting scenes with multiple shots. I happen to use 5.6/8 quite a bit with my 14-140 lens. So I'd want to maintain that throughout my shots, which may be problematic if I'm shooting outdoors, or even indoors depending on the lighting conditions.
 
An HVX200 has comparatively huge pixels, as compared to a 1920x1080 1/3" camera, they're 4x as big. So its limit is about f/9.5, any lower than that will cause a substantial drop in resolution. So avoid double digits and you'll be fine.

Thanks...found a post from 2006 that seemed to contradict the 1/3 sensor rule stated here and wanted to see what the current thinking was.
 
Some great info in this thread so thanks again for all the contributions. I bought a variable ND filter today so this will replace my standard SOP of stopping down the lens for F11-F15 to correct exposure. I'll see if this helps and keeps it from softening the picture.
 
Problem with variable ND's is that they themselves can soften the picture, especially at telephoto focal lengths. Some of the cheap ones can soften the image horribly, worse than diffraction would. I have a Heliopan and it doesn't seem to introduce much softening, but -- test yours with a long lens (with and without the filter, on a still, at the same iris; use the shutter speed to compensate exposure); check for a substantial difference in sharpness, if you see one, you probably can't trust your variable ND.
 
Yep, I have a Genus VariND that I'm loathe to use because of the sharpness hit and also a greenish color cast. It works well when absolutely necessary, but even some cheap single-density Promaster filters give better results overall.
 
Ok, I'm using a Tiffen Variable ND on my 25mm and it's hard for me to tell if it's softening the picture at all. So I bought another one for my 14mm.
 
Back
Top