ND filter (extreme NDx)

[mike_lowe2]

I'm looking to run some very long exposures with my 4x5" camera...however I've

no experience in utilising extreme ND filters. Sure, a few stops is fine. But

I'm looking to incrememnt exposure by hours.

 

Ive found an NDx filter which is basically ND10000, alleged at 13stops.

 

Ive been calculating, and have approximated with something like this:

 

Example ? 2second exposure at f45

 

With NDX filter, ND10000

13stops

 

00 ? 2sec

01 ? 4sec

02 ? 8sec

03 ? 16sec

04 ? 32sec

05 ? 64sec (1min 4secs)

06 ? 128sec (2mins 8secs)

07 ? 256sec (4mins 16secs)

08 ? 512secs (8mins 32secs)

09 ? 1024secs (17mins 4secs)

10 ? 2048secs (34mins 8secs)

11 ? 4096secs (1 hour 8mins 16secs)

12 ? 8192secs (2 hours 16mins 32secs)

13 ? 16384secs (4 hours 33mins 4secs)

 

However, the whole reciprocity failure will mean I need to adjust pretty

accurately - is there an equation to assist in such calculation?

Plus a constant light reading would assist in gauging if the light failed to be

constant - again a problem.

 

I can work from neg a little overexposed perhaps, but I think my main fear is

massive underexposure and wasting many an hour. :-)

 

Some words of advice would be appreciated.

 

Thanks

[nikos peri]

<i>but I think my main fear is massive underexposure and wasting many an hour.</i><p>

Well it's a really good thing you're not into astrophotography! ;-)<br>

I don't think you'll find anything resembling a specific equation, rather a seat-of-you-pants general indication. A 4.5 hour metered exposure under variable lighting, with an unspecified film... [licks finger and sticks it in the air] ... go for dawn till dusk and adjust it from there. :-)

[jeff_schraeder1]

At an exposure time of 16384 seconds, a one quarter stop error is 43 minutes!

[Monophoto]

Mike -

 

I didn't try to follow your exposure calculations - its too early in the morning for that. But that's the easy part.

 

There is no formula for reciprocity correction. Actually, reciprocity failure characteristics differ between films, and becomes very non-linear as the exposure time increases. And more importantly, once your exposure time is out in the minutes to hours range, small errors don't matter any more.

 

If your calculated exposure is 4 hours or so, the corrected exposure to allow for reciprocity failure will be many times that - probably in the 12-16 hour range or more. And on top of that, you probably should bracket your exposure.

 

You didn't indicate what the subject is, but if your ability to do the exposure is bounded by the daylight/darkness cycle, you may want to rethink the use of that ND filter.

[allen_whittier]

What are you photographing? What's the film?

[mike_lowe2]

Well thank you for some responses!

 

I did have something of a little smile when someone mentioned bracketing! Haha... "Ok, so I'm done witht he 4 hour, now best do an 8 hour just to see if that one extra stop makes the difference!"

 

Aaaanyway - in terms of what I'm photographing. Essentially I'm hoping over a due course of time to photograph 'a day' - i.e. from dawn 'till dusk.

What I do know is that to try and do that immediately wont happen accurately - so I'm checking what kind of NDs there are - ND100, ND1000 (7 stops), ND10000 (13 stops) in order to test initially short exposures with wider apertures, then apply those exposures with smaller apertures to see if I can match an exposure but with the extension of time. That should help me to offset and gauge the different between the two, and then I can begin to make longer exposures.

I know it's going to be fiddly... and it'll mean a meagre single frame every Saturday!

Still, it's just something that I have in my head - Ive been exposing for hours at night, but that's pretty much guess work correlated to the size of the moon.

You get moon paths in 6hour exposures, I'd quite like to see some sun paths in a daytime similarity.

 

Plus digital cameras don't offer anything like the possibility of doing stuff like this...I'm a great fan of my 'T' setting. :-)

[mike_lowe2]

Also, the film - Fuji NPS 160 colour negative 4x5". Slightly out of date now too... :-(

[michael_briggs2]

If you goal is an actual exposure time of 8 to 12 hours, then your metered/calculated exposure time without taking into account reciprocity failure should be MUCH shorter. If you select a filter that gives a exposure time of 4 hours before taking into account reciprocity failure, I think that you will have massive underexposure.

 

The manufacturers don't seem to give much recommendations for color films for more than a few seconds, probably because they are concerned about the three colors behaving differently, but they publish graphs for black and white films to many seconds. These can be used for very rough guidance. For example, the datasheet for Delta 400 shows a factor of four correction of the indicated exposure at an indicated time of 20 s -- and the curve is rising steeply! So thinking that the correction might be a factor of 3 at 4 hours is very likely way off. The Ilford curve for Delta 400 ends at a corrected exposure of 200 for an indicated exposure of 35 s.

 

I've never done this, but I'd try a filter and f-stop that give an indicated exposure (before reciprocity correction) of tens of minutes. The reciprocity correction effect will vary with the brightness of items in the scene (more reciprocity failure for dim items, less for bright), so the contrast might be altered. With negative film, overexposure is much better than underexposure.

[richard baznik]

Mike:

 

Your last entry raises the interesting possibility that film that was fresh when you started the exposure could become outdated during the exposure!

 

In a more serious vein, my experience suggests that reciprocity is likely to fail more precipitously with old film than with fresh film. My experience is strictly limited to maximum exposures of three to five minutes, however.

[AaronFalkenberg]

Michael brings up some very valid points. Inherent contrast also increases the longer the exposure becomes. You would definately need to contract development. <br>

<a href="http://www.seeinglight.com/reciprocity.shtml">Here</a> is a reciprocity chart for TMax 100 (pretty good rec. characteristics for B+W). You can kind of extrapolate the chart and see that a 2h exposure is somewhere around 7hrs and it just goes up exponentially from there.

[allen_whittier]

Mike,

 

I use 64t in both Kodak and Fuji transparencies. It has much better reciprocity characteristics. Depending on the contrast needed you can go out 2 to 4 minutes without correcting. It's my primary low light film. The Polaroid 64t also works well.

 

I go along with the bracketing idea. This shot won't turn out the first time, I guarantee. You'll have time to process the film between shots so it won't take long to zero in on what you need to do. I wouldn't spend much money at the theory stage.

 

If you want sun tracks, you might try a sun filter for a telescope. The small ones aren't that expensive. You might take a photo of the surrounding area with the sun behind you then put a heavy duty filter on like the sun filter and track the sun double-exposed onto the image.

 

I do all day shots where I fire the camera at key times of the day, or at even intervals throughout the day. It gets great results. You might try taking a few hundred shots throughout the day rather than opening the lens all day. It'll save you a few f stops of nd filtering. I took one shot that I fired the lens every 15 seconds for a couple hours to soften out dappled light. It worked well. I do have to admit though I have several lenses with self cocking shutters to get people out of architectural shots using the same method. Reciprocity is still a problem that takes some effort to work around. It's different with every shot that's outside the box.

[mike_lowe2]

Aaron seems to put an interesting point out there - that 2hrs=7hours, and with an exponential increment 4hrs=14hours (if not more!).

This is why I asked here, as obviously the NDX10,000 is more dense than I want (as I don't want to stop the camera up, I want to make use of v small apertures).

Perhaps an ND1000 would, by stopping 7 stops down rather than 13 stops (of the NDX10,000), make more sense to try and chart in the first instance... I just didn't want to buy more than necessary initially.

 

As for multiple exposure over extended open shutter - I understand that this could be beneficial - problem is I don't have a lens that re-cocks itself. I'd have to physically handle the camera too much - and softness is more likely.

Also, I kinda wish to stand away from the shots without having to think so consistently for every 15seconds for hours and hours. Developing film for 30mins is one thing...exposing for hours looking at a stopwatch would hurt my brain! :-)

 

Light at this time of year is quite poor, so it's hard for me to figure when I can begin to experiment really. Need to get the correct kit into place initially and then start the experimenting.

I'd never expect this to work first time. And with the amount of time consumption, it's really going to be quite a lengthy process.

[pete_andrews10]

Mike. Take the easy way out and use a sector wheel in front of your lens! This is a black disc that's rotated in front of your lens by an electric motor, and has one or more slits in it. If the slit is 1000th of the diameter of the wheel, for example, then that gives you the equivalent of a 30D ND filter.

 

The big advantage is that you don't have to worry too much about low-intensity reciprocity failure, since the effective exposure time is also 1000th of the rotation time of the disc. Say the disc rotates once a minute, then the exposure per revolution is the equivalent of 60/1000 or 0.06 seconds. So 100 rotations will give you a total exposure of 6 seconds, but spread over a period of 100 minutes.

 

In other words, you get virtually the same effect as a continuous exposure of 1 hour and 40 minutes, except that you only have to expose the film for a total of 6 seconds, and the reciprocity failure at this time is much more manageable.

 

The sector wheel can be combined with a conventional ND filter to extend the time even further.

 

Remember it's the intensity of light falling on the film that leads to reciprocity failure, not the length of the exposure as such. Therefore low-intensity reciprocity failure not only affects the overall density of the image, but also the contrast. An overly extended exposure will be VERY contrasty, since the shadows will hardly record any exposure at all, while the highlights will be grossly overexposed.

 

Sensitometry theory says that if we give a film 10000 exposures, each of 1/1000th of a second, the resulting image will be much denser and less contrasty than if we'd reduced the intensity of the light by 10 thousand times and extended the exposure time to 10 seconds.

 

In my view a sector wheel shutter is the only sensible way of achieving terrestrial exposures that run into hours, or even days.

[bob_beard1]

Mike: Four hours is a long time to wait around! I have used ND filters for certain effects and have used exposures in the 30 minutes range for BW films, and have as a rule of thumb done about 2.5-3x the calculated times with fairly good results on Tmax100 which has a good long exposure response. Fuji Acros is apparently good for longer exposures as well but I have not used it myself for long exposures.

 

Here is a link to a site where a fellow has done quite a bit of testing regarding reciprocity failure.

 

http://home.earthlink.net/~kitathome/LunarLight/moonlight_gallery/technique/reciprocity.htm

 

He does give some formulas for extended exposures but if you are really going for 4.5 hours then you will just have to test it. Can you set up two cameras (or more!) and do the bracketing that way?

 

Can you use a larger fstop to cut down on the total exposure time?

 

Lastly, here is a link to some charts done by some very experienced photographers, so should be useful at least for B/W shots.

 

http://www.seeinglight.com/reciprocity.shtml

Good luck!