Jump to content
UltravioletPhotography

ND filters and long exposure test


StephanN

Recommended Posts

Introduction:

 

After the point of How-To-Do-Long-Exposures was raised in this thread https://www.ultravio...96-brook-in-uv/ , I looked around a bit and this thread for example has some information about ND-filters: https://www.ultravio...-uv-capability/ (please excuse me, if I've overlooked other threads). Therefore, I did expect to have a stronger absorption of the filters in UV than in VIS.

 

I went back today to the brook of the first thread, and did some tests.

 

Camera used: Canon EOS 6D, converted by MaxMax to black & white, internal X-Nite 330C (Edited, was 330 without the C), external X-Nite BP1

Lens 1: Nikon EL-Nikkor 80mm f/5.6 (tests)

Lens 2: Steinheil Auto-Culmigon 35 mm f/3.5 (bonus)

 

I did two tests, the first is a comparison of the ND-filters I own; the second is a comparison of photos with ND-filter and long exposure time, with photos without ND-filter, massaged in-camera, in Lightroom and in Photoshop.

 

I did not test my ND-grad filter, because I'm barely using it at all, preferring HDR (https://en.wikipedia...c-range_imaging), if necessary.

 

Warning: I am all but a photoshop-wizard, if anybody has suggestions how to do this in a better way, bring them on. I'd also be willing to provide the raw-files, if somebody wants to play around with them.

 

Part 1, Test of ND-filter (all at f/5.6)

 

No filter, ISO 50, 4s:

post-176-0-94274600-1599914447.jpg

 

Hoya Pro Digital 1, NDX4 (2 stops), ISO 50, 8s (Lightroom +1 stop, as I made a mistake dialing in the time):

post-176-0-30029300-1599919394.jpg

 

Haida Slim Pro II MC, 8x (3 stops), ISO 50, 120s (Lightroom +1 stop):

post-176-0-81875700-1599919410.jpg

 

Haida Slim Pro II MC, 16x (4 stops), ISO 50, 30s:

post-176-0-19338700-1599916727.jpg

 

Hoya Pro Digital 1, NDX32 (5 stops), ISO 50, 60s:

post-176-0-23731100-1599916753.jpg

 

For the next series I had to increase the ISO quite a lot, in order to get manageable times; The quality is abysmal, of course, but for an estimate of the optical density this will be a first approximation.

 

No Filter, ISO 6400, 1/30s:

post-176-0-29185400-1599916814.jpg

 

Haida Slim Pro II MC, 64x (6 stops), ISO 6400, 30s (Lightroom +2 stops):

post-176-0-99817000-1599916927.jpg

 

Haida Slim Pro II MC, 1000x (10 stops), ISO 6400, 240s (Lightroom +2 stops):

post-176-0-51440600-1599916989.jpg

 

I also own the Haida filter for 12 stops, but refrained from using it. Already the 6 stops-filter is much, much stronger in UV than in VIS, and given the fact that so far I've never done a long-exposure of more than 2 minutes, I don't expect to need even stronger ND-filters. Doing long-exposures in the region of minutes adds also unwanted noise and pixel-problems to the photo, so nobody wants to do this.

 

This brings me to the second part of the test.

 

 

Part 2, Comparison of long-time exposure and computer-simulation

 

For this, I went back to ISO 50, and found that an exposure-time of 4s still gave a nice enough histogram, as the light hadn't changed enough to make a difference.

 

One frame, ISO 50, 4s:

post-176-0-12216300-1599917529.jpg

 

In-camera multiexposure of 9 frames, each with ISO 50, 4s, mode "average":

post-176-0-65316900-1599917563.jpg

 

Lightroom HDR-composite of 9 frames, each with ISO 50, 4s:

post-176-0-48960200-1599917653.jpg

 

Photoshop, merging of the same 9 frames in 9 layers, averaging:

post-176-0-46326500-1599917688.jpg

 

Both, in-camera and using photoshop, the resulting image looks very much like a "proper" long-time exposure. I did no test to find out what would be the minimum exposure-time per frame required to obtain this result. Possibly some members who are more at home in photoshop or galaxy-photos might know the answer to this. In theory any exposure-time might serve, but the number of frames would probably be more than I'd be willing to handle for such a simple photo, and one maybe would have to resort to a proper program used by astronomers.

 

Part 3, Bonus:

 

I switched lenses, and used the Steinheil for a couple of overview-photos of the brook, all at ISO 50, f/8, 8s. Here, the exposure time of 8s is already almost enough for a nice smooth image. In both examples I set the focus on a rock in the foreground.

 

One frame:

post-176-0-99308900-1599918219.jpg

 

In-camera average, 9 frames:

post-176-0-94439500-1599918240.jpg

 

One frame:

post-176-0-38636700-1599918257.jpg

 

In-camera average, 9 frames:

post-176-0-72780900-1599918297.jpg

Link to comment

You are definitely seeing IR with the 6 stops and 10 stops ND filters. They probably absorb UV much more.

 

Yes. Haida filters seem to absorb more UV than the Hoya ones.

Link to comment

These look great. I like the in camera ones best. The rock has more details than Photoshop and the water is silky smooth.

 

Did you try a 1/2 second exposure with additive setting?

Link to comment

These look great. I like the in camera ones best. The rock has more details than Photoshop and the water is silky smooth.

 

Did you try a 1/2 second exposure with additive setting?

 

Nope, sorry. 1/2 second might be a bit dark, so not sure if I'd be adding up unwanted noise with underexposed frames? I've got no experience with additive-mode, having used it probably 3 times in 10 years just to see if it works.

 

The in-camera average does seem to be the winner here. They do indeed look great.

 

The photoshop version is almost indistinguishable, just a touch darker but that can easily be remedied.

 

In-camera has the advantage of seeing the result immediately, and the disadvantage of only being able to handle 9 frames (not sure if MagicLantern can change this, will have to check this).

 

Photoshop has the advantages of being able to use as many frames as the PC can handle and to play around a lot, and the disadvantage of not seeing the result immediately.

 

Personally, for this kind of long-exposures I think that I'll stick to the in-camera average.

Link to comment
There's a big screen TV in my retina doctor's waiting room that was showing soothing nature scenes this morning. Some of the river scenes were time lapses composed of long duration exposures like the ones in this thread, although strictly in the visual region. The frothing ripples and eddies of the river had the same creamy smooth look except they were in motion, undulating and writhing. Perhaps it was due to varying wind activity, as the individual strands in a waterfall in the background of one were flapping back and forth horizontally. Whatever the cause, the overall effect with the added dimension of motion was quite striking and captivating.
Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...