Reciprocity fun timez

[Jarvman]

I've just picked up a ND8 filter to try some really long exposure in landscape but aren't sure exactly how to work out the duration of exposure. I tried taking a picture underneath the local pier earlier as the tide was coming in which almost tried to swallow me and the tripod the time it took to get the chamonix set up! That was at ISO 1.5 I guess, seeing as my meter only goes down to 3, giving an exposure of 2 mins at f/16, the film is HP5. Not knowing how to compensate for reciprocity I just doubled the time exposing for 4 mins seeing as I usually just add a stop. The neg looks ok but it's pretty dark, all the detail is there though. I could've probably looked this up but people here always explain things so much easier, so any pointers would be much appreciated.

 

[keithwms]

http://mkaz.com/photo/tools/reciprocity.html

 

[Jarvman]

Ok cheers, that would seem to solve my problem! but it says for HP5 that the indicated time of 2 minutes should be adjusted to 17m 21secs. If this is correct I'd be drowned already and how come the neg looks overexposed at only 4 mins?!

 

[keithwms]

Let's check your metering and compensating procedure. Suppose that your light is sunny 16. Then with hp5+ rated at 400, correct exposure would be f/16 and 1/400. With an 8-stop ND, that means your filter-factor corrected exposure would be 2^8*(1/400)=0.64 sec. Check my math! I'd call it 1/2 sec and conclude that no recip. correction is needed.

You don't need to meter through the ND; just meter normally and add eight stops. Then adjust for recip. failure.

 

[Jarvman]

No wonder I'm confused, the chart says for moonlight photography. I'm shooting in the daytime. There must be a chart for that too yeah? Sorry, I never stated.

 

[keithwms]

Makes no difference when or what you're shooting.

What did your meter tell you to expose, without the filter?

If your metered exposure was, say, 1/8 sec at f/16, then with the ND you'd be up to ~32 sec. Then with reciprocity correction, that'd be about 2.5 mins.

 

[Jarvman]

I can't remember now but I'm think it ended up at 60 secs at f/16 for ISO 3, the slowest the meter will go which would make it 2 minutes at ISO 1.5, 8 stops beneath ISO 400 yeah? Working backwards from that it means my exposure must've been 1 sec at f/16 which seems right underneath a concrete pier on an overcast day. Perhaps I've read it wrong somehow. I'll try it again tomorrow.

 

[keithwms]

I think your confusion is coming with the ISO setting. Just leave the meter at ISO 400, and meter without the filter. Then adjust that reading, first for the filter factor and then the reciprocity failure. No reason to change ISO and all that if you're not metering through the filter.

 

[Edward Pierce]

Ok cheers, that would seem to solve my problem! but it says for HP5 that the indicated time of 2 minutes should be adjusted to 17m 21secs. If this is correct I'd be drowned already and how come the neg looks overexposed at only 4 mins?!

It is likely that you are over developing. Check the shadow areas of the negative to see if you're getting the right amount of exposure.

There is a significant increase in contrast with long exposures, which needs to be compensated for with reduced development to bring those high values down.

I've never used HP5 but I do this with Tri-X, which may be similar. For an indicated exposure of 4 minutes, I would expose for 11 minutes and expect a 2 to 3 stop increase in contrast due to the reciprocity effect. I would develop N-2, or about 25% less than normal time. This is only a starting point.

Also, when planning on reduced development like this I would reduce the speed rating on the meter by 1/3 stop to support those mid-low values with a little more exposure.

 

[Jarvman]

These negs aint contrasty, they're black black black. I think I could've exposed an extra stop by accident but that's all then processed for the reccomended time. Hmm. I would've thought that the time reccomended by the meter without compensating for reciprocity would yield a much better neg.

 

[Vaughn]

I agree with Keith -- leave the meter at the film's rated ISO! I figure you must have over-exposed by 3 stops or more.

A ISO of 3 and f/16 at 2 minutes gives an EV of 12. An EV of 12 at an ISO of 400 (HP5) yields an exposure of 1/60 sec at f/16. That is 5 stops difference. Add your unneeded Resiprocity failure addition of one stop, and then subtract 3 stops for the ND8 filter, you overexposed by 3 stops...and probably more, but your methodology is unclear from your description.

A simple "sunny 16" rule would have said 1/500 at f/16 which may indicate that you actually over-exposed by up to 6 stops.

Vaughn

 

[Jarvman]

This time I got an ambient reading of f/22 at 60secs with taking into account the ND8 filter without adjusting the meters iso. The suggested reciprocity correction is 7mins 8 secs! So I went and exposed for that amount of time. I took meter readings throughout the exposure and the sky perhaps brightened up by a stop throughout the length of it. Then processed it as standard in the developer. Will this increase in brightness be multiplied many times over the course of exposure? Is it a big no no doing long exposures in changing light, even if not dramatically? Does it have to be completely overcast and unvariable?

 

[Vaughn]

There will be less than a stop increase in the over-all exposure.

How did you meter? Again, in an average scene with sun, the exposure would be about 1/500 at f/16, or 1/250 at f/22. Add three stops of light for the ND filter and you have 1/30 at f/22. Your ambient reading is still way out of the ballpark by 11 stops. But then, under a pier is not an average scene. Are you metering just the underside of the pier?

Just to be sure that you are compensating for the filter correctly...a ND8 filter needs 8 times the amount of light, or 3 stops...NOT 8 stops. A ND256 filter(if there was one) would require 8 stops. additional light.

Vaughn

 

[Jarvman]

Ah, you could be right! I'm compensating for a whole 8 stops, then adding the reciprocity on top of that. Would be the reason I'm overexposing! I'm gonna have to find the filter case and check. It could rightly just be 3 stops. It looks a pretty dark filter though and there are B+W filters that actually reduce the intensity by 10 whole stops. I'm confused.

 

[Jarvman]

Yeah you're right, 3 stops, doh.

 

[keithwms]

Ah, yeah, there is an annoying confusion with ND labeling that comes up from time to time. Some manufacturers use the convention that the number is 2^(stops); in other words 2^3=8 and ND8 means the actual filter factor is 3 stops. In my field, we say ND0.3= 1 stop... based on log2, in other words. <sigh> If they would simply say "filter factor=... stops" then there would be no confusion.

Anyway if you just look through an ND, you should be able to tell the difference between a 3 and an 8 stop ND quite easily. The eye will compensate for the 3-stopper quite easily, if you hold it over your eye. An 8-stopper will be quite black and is a bit of a specialty item, kinda pricey, so people normally just stack the lower power ones.

 

[Jarvman]

Yeah I thought when I bought it it seemed a bit washy compared to my infrared filter. I'm going to have to get the B+W 6x and 10x I reckon

 

[Vaughn]

Glad we got that worked out!

Here is my under-the-pier image...Using a 25A (red) filter and T-Max100

(there was a url link here which no longer exists)

Vaughn

PS...the 25A also gave me a 8x filter factor. But the main reason I used it was to bring the values beneath the pier and the sunlit areas closer together -- and the long exposure added to the image by making the waves disappear and smooth out the sea.

Try stacking filters to increase the exposures -- but multiple the filter factors. A 25A stacked with a ND8 would be 8x8, a factor of 64 times the metered exposure (or 6 stop increase needed).

 

[Jarvman]

It's an often done shot aint it. Just wanted something to try it out on before rather than being creative. I Reckon I might stack the 8x with a 5x red filter. What's that be? A filter factor of 40x? How many stops is that? haha funny, I actually had your sig in my head when I was down on the beach without knowing whose it was Vaughn. Only knowing I'd seen it here.

 

[Vaughn]

Does not matter how many times a shot's been done -- it is just a matter of making it your own.

40x -- about 5 and a half stops. I like being out there...getting something on film is a bonus!

Vaughn

 

[telkwa]

If you want to convert filter factor into stops, grab a trusty scientific calculator (or use windows'). If the filter factor is x, then the number of stops is log(x) / log(2).