[alt-photo] Re: Ultra Long Exposure Reciprocity Failure?

[Greg Schmitz]

Correction - brain fart, it's been a while.

Reciprocity "failure" increases contrast, so development time must be 
reduced.

Just looked at my notes for one of the films I hated more than any other 
- Kodak Extapan,  T-Max was so wonderful because of reciprocity 
characteristics that were insignificant when compared to Kodak Extapan.

In order to reign in contrast from long exposures it is necessary to 
DECREASE development time, which, of course, reduces effective film 
speed.   A "catch-22" of sorts; or perhaps a vicious circle.

Perhaps you can come up with an off the cuff equation for exposure, but 
I've always found suggested equations for adjustments to be next to 
useless unless the effects for adjusted development and the change in 
effective film speed are factored in.

Testing is the way to go I think.

--greg


On 1/20/12 12:15 AM, Greg Schmitz wrote:
>
> I've always run simple tests using a step wedge and a light source I 
> can control / or a grey scale and a camera.  It's simple to make 
> estimates for long exposures when they are compared to shorter 
> exposures.  Without looking at my notes, my recollection is that for 
> black and white there may be a decrease in contrast and developing 
> time may need to be adjusted as well as exposure.
>
> FWIW when you get into really long exposures you've got lots of wiggle 
> room.  A 50% error with a 400 second exposure is 200 seconds.
>
> --greg
>
>
> On 1/19/12 5:07 PM, Francesco Fragomeni wrote:
>> Hi all,
>>
>> I've been trying to figure something out with this for a while now 
>> and my
>> guess is that the only real answer is testing but I'm hoping that 
>> someone
>> here might be able to offer some mathematical insight that I haven't
>> thought of or seen elsewhere. I am interested in making some ultra long
>> exposures and I am looking for a way to approach calculating the rate of
>> change in reciprocity failure as the exposures get longer. I understand
>> that reciprocity failure is not linear but logically there should be 
>> a way
>> to mathematically work out the rate of change of the failure over the
>> increase in exposure. I'm probably wrong about that otherwise someone 
>> would
>> probably have solved for it by now but alas I'll still put the 
>> question out
>> there.
>