Why doesn't 2+2 = 4 in print exposure?

[36cm2]

Can someone please explain to me why you don't get the same print results when you expose photo paper continuously for 16 seconds as opposed to exposing it twice for 8 seconds? I haven't tested papers yet, but am looking to standardize my process and read about this somewhere in my preparations.

Thanks,
Leo

 

[Kevin Kehler]

The key is the light drop-off and the timer. First, you are assuming your timer is 100% accurate which it rarely is, especially if it is a mechanical (as opposed to an electrical) timer. So the timer might be counting wrong. Second, the lamp takes part of a second to warm up and it shuts off part of a second either early or late, depending on your enlarger. So you might be getting more or less light than your timer says you are getting. I know my set-up has 2+2+2+2= 9 seconds uninterrupted exposure.

 

[ruilourosa]

the enlarger system takes some time to inicialize, so... even if it´s few, take the time to the lamp to get full power, the transformator to send current, and the clock to make the passage and you will get less time in divided exposures

 

[L Gebhardt]

Depends on your enlarger. I have a color head which uses a shutter and multiple short exposures do add up to one long one. My other enlarger, using the Ilford 500 head, also seems to add up to the same (based on test strips). I'm sure there is some difference, but it isn't visible. So you really need to test.

With an older condenser system I used there was a noticeable difference, which I attributed to either the lamp warming up, or the timer (older, but digital).

 

[Worker 11811]

Pre-strike your lamp, let it warm up and stabilize color temperature then use a dowser to cut the light on and off. You can make a dowser out of a solenoid and some bicycle cable attached to a metal shutter. There may be pre-made assemblies you can buy, as well.

 

[RalphLambrecht]

Can someone please explain to me why you don't get the same print results when you expose photo paper continuously for 16 seconds as opposed to exposing it twice for 8 seconds? I haven't tested papers yet, but am looking to standardize my process and read about this somewhere in my preparations.

Thanks,
Leo

Timers, yes, and the 'Intermittency Effect'.

http://jbhphoto.com/articles/intermit/intermit1.htm
http://books.google.de/books?id=Mbl...CDgU#v=onepage&q=Intermittency Effect&f=false

The one reason why I don't like the Fred Picker way of making test prints.

 

[36cm2]

...the 'Intermittency Effect'.

http://jbhphoto.com/articles/intermit/intermit1.htm

Bingo. Thanks very much for the answers, guys. It's been very enlightening.

 

[MattKing]

It's probably worth noting that while 2 + 2 isn't the same thing as 4, 22 + 22 is awfully close to 44.

 

[2F/2F]

If you were to use a shutter and - less importantly, but still a factor - a voltage regulator, 2+2 would be much, much closer to 4.

I don't use a shutter (though I do use a voltage regulator for all my darkroom equipment now that I have the one that came with my big enlarger). This is why I don't do stepped test strips. It was clear to me my first time printing, having had a little bit of electronics knowledge from working in power plants, that slop in mechanical timers and lamps ramping up and ramping down made it so that the best increment on a test strip never matched a full print at that time. So, I get a feel for what times are common for well exposed negs at certain enlargements, adjust things on the enlarger based on how far away from well exposed I think my neg is, and do a test strip all at one time. Sometimes I will do the same thing three times at three different times or three different apertures if I am less sure of what to expect.

I also try to use exposures that are in the 15 - 45 second range, to minimize the effect that variations in light output throughout the exposure time have on the final print. A lamp's ramping up and ramping down time are larger and larger percentages of the total exposure time the shorter exposures get. It also makes burns and dodges more repeatable, for the same reason. You can never do a burn or dodge exactly the same every time. There will always be some differences in the motion of your hands or other go-betweens, and in the exact timing of their placement. These differences have a significantly greater effect the shorter exposure times become.

 

[Dan Henderson]

just for fun one day I made a test strip in the usual way of uncovering sections of paper on one half of a sheet with the other half covered, i.e. 2, 2, 2, 2, 2. On the other side I gave the same amount of time to each to the corresponding section of paper through black cardboard with a strip cut in: 2, 4, 6, 8, 10. The difference was remarkably visible and showed me that is the best way to make test strips. (I described the process using even time increments for illustrative purposes only. In practice I use the f/stop timing method: 2, 2.8, 4, 5.6, 8, 11, 16, 22 and 32 seconds.)

 

[BBarlow690]

Has anybody actually tried it? I did. Not telling what I found. You'll learn a lot in two, or maybe even a single sheet of paper.

 

[RalphLambrecht]

Has anybody actually tried it? I did. Not telling what I found. You'll learn a lot in two, or maybe even a single sheet of paper.

Yes, I have.

 

[Nicholas Lindan]

2 + 2 does equal 4; heck, 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 = 20 -- if the timer is accurate and you compensate for any lamp warm-up time. However, if you have a cold light head without a shutter, or the least an integrating/compensating timer, it probably isn't possible.

The following application note on the Darkroom Automation web site gives a method how to determine lamp warm up time:
http://www.darkroomautomation.com/support/AppNotePH212LampDelay.pdf

 

[RalphLambrecht]

2 + 2 does equal 4; heck, 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 = 20 -- if the timer is accurate and you compensate for any lamp warm-up time. However, if you have a cold light head without a shutter, or the least an integrating/compensating timer, it probably isn't possible.

The following application note on the Darkroom Automation web site gives a method how to determine lamp warm up time:
http://www.darkroomautomation.com/support/AppNotePH212LampDelay.pdf

Nicholas

Are you saying: the intermittency effect does not exist?

 

[Ian Grant]

No he's saying you can have an "inaccurate timer" that takes it into account

i.e. has a additional time added to compensate. So 2 secs is 2 + enough milliseconds to compensate for the ramp up/down time, but it would need to be calibrated for different light sources.

Ian

 

[Nicholas Lindan]

Are you saying: the intermittency effect does not exist?

The 'intermittency effect' is an effect of the quantum response of a silver salt crystal to a photon of light. In conventional emulsions a crystal has to be hit at least twice with a photon within a certain period of time. If too much time passes between photon hits then the crystal 'relaxes' and is not activated (exposed) when the second photon hits. This is the principle behind reciprocity failure, where the photon flux is so low that too much time passes between photon hits and therefore extra exposure time is required to statistically assure a double-hit. It is also the principle behind intermittency where an exposure is paused and the pause time is long enough to allow the crystal to relax and become 'unexposed' again. The critical photon inter-arrival period varies between emulsions - as evidenced by the different reciprocity and intermittency factors for different films and papers.

The lack of interimittency is the reason for the excitement about 'single photon' emulsions - it's not so much that they are twice as sensitive as that there is no reciprocity failure.

Intermittency has little to no effect in the range of exposure times encountered in printing with printing paper emulsions. See the experimental results in the application note. Or make your own instrumented or shuttered experiments and draw your own conclusions. Or, See Meese, et. al., again....

Intermittency effect has nothing to do with lamp warm-up and stabilization time.

 

[RalphLambrecht]

... Intermittency effect has nothing to do with lamp warm-up and stabilization time.

Nicholas

I agree with your statement above, but I don't think anybody claimed that. In my experiments, the intermittency effect is strong enough to through a regular test strip off by up to 1/12 stop.

 

[Nicholas Lindan]

but I don't think anybody claimed that

I think it was in one of the references you gave, last two paragraphs:

http://jbhphoto.com/articles/intermit/intermit1.htm

In my experiments, the intermittency effect is strong enough to through a regular test strip off by up to 1/12 stop.

What was your test protocol?

In my experiments a 1 second error accumulated from 20 x 1 second exposures - that is 5% or 0.07 stops or 1/14th of a stop. Adding a 50mSec correction to each exposure removed the error. It may have corrected intermittency but as with the correction 30 x 1 = 1 x 30 and 10 x 1 = 1 x 10 the effect seems to be entirely lamp warm-up time, as confirmed by the measurement of the turn-on and turn-off light output measurements shown in the application note. Additionally, the required correction time changes if a ferroresonant stabilizing transformer is added to the system.

 

[Anscojohn]

All things being equal (if they ever are!?), it is reciprocity failure.

 

[RalphLambrecht]

I think it was in one of the references you gave, last two paragraphs:

http://jbhphoto.com/articles/intermit/intermit1.htm...

Nicholas

Cut me some slack here. I referenced that article for the OP so he could read up on the intermittency effect, but I never claimed that this effect has anything to do with lamp warm-up. They are totally unrelated.

... What was your test protocol?

In my experiments a 1 second error accumulated from 20 x 1 second exposures - that is 5% or 0.07 stops or 1/14th of a stop. Adding a 50mSec correction to each exposure removed the error. It may have corrected intermittency but as with the correction 30 x 1 = 1 x 30 and 10 x 1 = 1 x 10 the effect seems to be entirely lamp warm-up time, as confirmed by the measurement of the turn-on and turn-off light output measurements shown in the application note. Additionally, the required correction time changes if a ferroresonant stabilizing transformer is added to the system.

Fair enough. You are right. There are several effect at work here: intermittency, lamp ramp-up/down and stabilized-transformer delay.

I can adjust my timer to compensate for it. My transformer delay was measured at 400 ms, and I'm not fully aware of my lamp ramp-up/down. I will measure it and ask Phillips for the data sheet to compare. Unless that's done, I can't really measure the effect of intermittency.

I have modified my enlarger to mount a large-format lens and shutter, but LF shutters are not accurate enough (+/- 1/6 stop) to reliably measure intermittency either.

 

[Christopher Walrath]

There's one way around all this. If one would like to reduce intermittency, but not to its entirety, then follow Matt's advice. If you make a test strip at f/8 exposing 32-16-8-4-2-1-1. You could expose at f/16 for 128-64-32-16-8-4-4. At the longer times the amount of affectation on the halides would be the same, however much less noticeable.

Given that this would be a stupidly long exposure time to make a test print, don't do it.

Ralph is pointing out information to get the OP down the correct path to learning about his topic if questioning.

 

[Nicholas Lindan]

Cut me some slack here. I referenced that article for the OP so he could read up on the intermittency effect, but I never claimed that this effect has anything to do with lamp warm-up. They are totally unrelated.

Didn't mean to imply that you said it. It was just in the reference you cited - where the author seems to be quite confused as what he talks about is lamp warm-up though he calls it intermittency.

There are, as you state, many, many factors to account for. A stabilized LED source can be timed very accurately and would let one explore the issue in more depth. But frankly, if adding 50mSec to the exposure time makes the problem go away that's about all the effort that needs to expended at this time.

 

[RalphLambrecht]

... But frankly, if adding 50mSec to the exposure time makes the problem go away that's about all the effort that needs to expended at this time.

Nicholas

Now you've got me going. I'm currently setting up my equipment to measure the light ramp-up/down of my Durst L1200 halogen bulb.

There is no stopping me now!

 

[Kirk Keyes]

So 2.05 + 2.05 = 4.

 

[Nicholas Lindan]

So 2.05 + 2.05 = 4.

For sufficiently large values of 4.