A confusing problem with lens f stops and timer f stops

[BobD]

I've yet to come across an incandescent bulb that does this. I frankly find it a far-fetched idea

It's a common occurrence for bulbs to get brighter or have inconsistent output before failing so it is hardly "far-fetched." The filament deteriorates providing less resistance and therefore more current flows through it. The heat produced further affects the lamp's output. You might not notice the difference but the paper will.

Google it if you don't believe me. There are plenty of references for this. And, you might want to look at some references for good manners as well before calling someone's well-intentioned suggestions "far-fetched," especially when it turns out the bulbs appear to be the issue here as I suggested.

 

[snusmumriken]

Ok so what I have done is this;

First I set the the lens at f/4, timer at 16s, focused the lens, gave the exposure on a scrap paper with Stouffer 21 step Wedge tablet, then I have lowered it to f/4 at 8s and finally to f/4 at 4s, because each time, I thought I could give shorted exposure. Results are below. Arrows are the steps I cannot see the difference with the following black part.

View attachment 348742

Then I thought, ok because f/4 at 4 seconds should be equal amount of light exposure to the papers with f/8 at 16 seconds (2 stop closing aperture, 2 stops giving more light), I should print my contact sheet at f/8 at 16 seconds. Result below. As you can see it is very dark, not something I have expected based on the above test. Because even at f/4 @16 seconds the image on the top right have some whites.

View attachment 348743

At this point I thought something is wrong so I have printed at f/4 and 4 seconds. Result below. Now that seems more correct. The images on the top are overexposed but the images on the below should be correct prints at Grade 2 (paper is Fomavariant 312, filter setting is 5M)

View attachment 348744

Now I have confused and thought, ok f/8 @16s gives a very dark image, how about f/8 @8 seconds. Then I thought still it is too dark, so I have closed the aperture 1 more stop to f/11 and printed with 8 seconds. Now the result is more or less the same to f/4 at 4 seconds.


View attachment 348745


Please correct me at any stage.

PS we are not discussing my photography skills, and because I have forgotten to put the glass on the negatives, they are blurred, I have corrected that.

Many thanks,
Fatih

Thanks for explaining. That is indeed bizarre. No wonder you were puzzled!

Maybe other people using those bulbs will have some ideas, but obtaining fresh ones seems a very sound step at this stage.

Then again, I have no experience with that Fomavariant paper. If this is actually an example of reciprocity failure, it’s a characteristic of the paper, so trying a different paper might help you to pin down the cause.

Do keep us posted how it turns out!

 

[koraks]

And, you might want to look at some references for good manners as well before calling someone's well-intentioned suggestions "far-fetched."

Calling an idea far-fetched is not an instance of bad manners. I'm sorry if you feel that way, but it's not intended as such, nor do I acknowledge that it's easily mistaken as such.

especially when it turns out the bulbs appear to be the issue here as I suggested.

The jury's still out on that one. See @Ian C's response below; it's more likely IMO to be a faulty voltage stabilizer circuit. These tend to work with phase control and something as simple as an old and leaky capacitor could easily result in the exact kind of behavior we see here.

I agree that bulbs can get brighter momentarily before failing, but this still does not explain the seemingly consistent and slow warm-up time of these particular bulbs, involving an apparently repeatable and slow increase in light output. That is not very consistent with a failing bulb, nor with the failure mode you proposed. I still think it's far-fetched on those grounds.

 

[Ian C]

The Lamp for the DeVere 504 Dichromat head uses conventional quartz-halogen lamps, which are a type of tungsten filament lamp. These attain full output within a few milliseconds, assuming, that the power supply and any voltage stabilizer (if one is used here) work properly.

It might be a good idea to check the contacts of the lamps. About once every year or two, I remove and insert the lamps several times in succession. The idea is that this lightly abrades the contacts of the lamps and sockets to ensure good conductivity.

My 504 and 5108 Dichromat heads appear to reach full brightness immediately when switched on. Yours should too.

 

[Fatih Ayoglu]

What I need is a continues light meter, which will output its results into an excel so I can understand how long exactly it takes to reach full output. Any such a device known?

 

[koraks]

Any such a device known?

@dkonigs is working on something that would probably do this (if you asked him nicely...). Maybe one of @Nicholas Lindan's products can be coaxed into doing the same.
An alternative for the electronically-inclined would be a light sensor hooked up to an Arduino (or some other microcontroller board), which e.g. relays continuous measurement data to a PC. I've played around with exactly this a couple of times for photographic applications. It works well for the purpose you have in mind, but it requires some aptitude in the field of electronics and microcontroller programming. The parts aren't very expensive.

 

[Fatih Ayoglu]

@dkonigs is working on something that would probably do this (if you asked him nicely...). Maybe one of @Nicholas Lindan's products can be coaxed into doing the same.
An alternative for the electronically-inclined would be a light sensor hooked up to an Arduino (or some other microcontroller board), which e.g. relays continuous measurement data to a PC. I've played around with exactly this a couple of times for photographic applications. It works well for the purpose you have in mind, but it requires some aptitude in the field of electronics and microcontroller programming. The parts aren't very expensive.

I should have listened my lectures at the university much much better I will run test strips with wedge table to see what I do get and will change the light bulbs as well.

 

[Sharktooth]

If your DeVere has a color head, the problem could be something simple like forgetting to set the filtering for some exposures.

I'm not familiar with the Devere, but all the color heads I've worked with have some sort of switch or lever to move the color filters out of the way to get brighter light for focusing. If you forget to move the filters back in place for the exposure, then the image will be too dark.

 

[dkonigs]

@dkonigs is working on something that would probably do this (if you asked him nicely...). Maybe one of @Nicholas Lindan's products can be coaxed into doing the same.
An alternative for the electronically-inclined would be a light sensor hooked up to an Arduino (or some other microcontroller board), which e.g. relays continuous measurement data to a PC. I've played around with exactly this a couple of times for photographic applications. It works well for the purpose you have in mind, but it requires some aptitude in the field of electronics and microcontroller programming. The parts aren't very expensive.

And my own motivation for doing it was really getting more accurate exposures when doing incremental test strips and maybe burn/dodge. Warm-up/cool-down time very much has a measurable effect when you're making a sequence of very short exposures. But it really shouldn't have such a noticeable effect when doing a longer print exposure. We're talking about something that maybe adds a second, between warm-up/cool-down, at worst.

My first course of investigation in this case wouldn't be to try to build an accurate time-vs-light profile. Rather, I'd simply stick a meter below the enlarger and chart how much light it projects at the different f-stops. And if the meter gives a continuous reading, maybe simply observe if the number noticeably changes over time.

 

[koraks]

Rather, I'd simply stick a meter below the enlarger and chart how much light it projects at the different f-stops. And if the meter gives a continuous reading, maybe simply observe if the number noticeably changes over time.

Yes, that's very sensible, and indeed what @Fatih Ayoglu has done in post #19 (and apologies for parachuting you in the middle of a thread full with options, evidence etc.) I think that his findings give rise to a reasonable question of how and in particular why light intensity seems to ramp up remarkably slowly.

 

[Fatih Ayoglu]

So I will do some tests tonight. The setup will be;
Brand new 2x 250W bulbs, expose Stouffer 21 step wedge tablet on 1x5 inch strips as f4 at 4s, f5.6 at 8s, f/8 at 16s, f/11 at 32s and f/16 at 64s. I will use a fresh batch Ilford MGRCV paper for this test and will do in complete darkness (although I know my red light is safe and I don't have a leak anywhere).

Theoretically, the printed densities should be exactly same on each print, so we will see.

 

[Fatih Ayoglu]

If your DeVere has a color head, the problem could be something simple like forgetting to set the filtering for some exposures.

I'm not familiar with the Devere, but all the color heads I've worked with have some sort of switch or lever to move the color filters out of the way to get brighter light for focusing. If you forget to move the filters back in place for the exposure, then the image will be too dark.

I have done all this test when it was 5M on the head but will do a new test tonight

 

[Kino]

It's a common occurrence for bulbs to get brighter or have inconsistent output before failing so it is hardly "far-fetched." The filament deteriorates providing less resistance and therefore more current flows through it. The heat produced further affects the lamp's output. You might not notice the difference but the paper will.

Google it if you don't believe me. There are plenty of references for this. And, you might want to look at some references for good manners as well before calling someone's well-intentioned suggestions "far-fetched," especially when it turns out the bulbs appear to be the issue here as I suggested.

Our motion picture printers use quartz halogen bulbs and, not frequently but on occasion, the bulb will suddenly gain in brightness when filaments eventually "bridge" with redeposited metal particles and jump in brightness. This is a byproduct of long term use and the random redeposit of tungsten on close-spaced gaps in the filament which eventually bridge.

While halogen gas does extend the life of the bulb with this redeposit of cast-off filament material, it also can cause this problem.

So, yes, it can happen and I have seen this happen a handful of times. It usually happens shortly before the lamp blows...

 

[mshchem]

The Lamp for the DeVere 504 Dichromat head uses conventional quartz-halogen lamps, which are a type of tungsten filament lamp. These attain full output within a few milliseconds, assuming, that the power supply and any voltage stabilizer (if one is used here) work properly.

It might be a good idea to check the contacts of the lamps. About once every year or two, I remove and insert the lamps several times in succession. The idea is that this lightly abrades the contacts of the lamps and sockets to ensure good conductivity.

My 504 and 5108 Dichromat heads appear to reach full brightness immediately when switched on. Yours should too.

Yes, I agree. I would suspect that something is going on with the head. I've seen something like this in old Beseler heads.

 

[Nicholas Lindan]

I believe the DeVere 504 uses a 'Trans Stab' power supply. The problem is probably here.

 

[MattKing]

I wouldn't discount the bulbs as being the source of the problem.
But in my experience, this sort of problem with bulbs doesn't repeat itself. By that I mean if you have a fluctuation like this between two exposures that follow each other promptly, if you try to repeat the test shortly thereafter, you will see variation in the result.
I would be more likely to attribute this problem to bulb behavior if it was a "one-off".
Which light source is this? Do the bulb connectors or the wiring need work?

 

[Doremus Scudder]

What I need is a continues light meter, which will output its results into an excel so I can understand how long exactly it takes to reach full output. Any such a device known?

Faith,

The test strips you are making tells that tale as well as observing the light intensity change over time with your meter.

Here's a test to do: Set your timer to beep for each second and set it to a high number (mine's set to 99 seconds - as high as it will go). Then, hold a card under the lens, start the timer and let the bulb warm up for 5 seconds or so. Then remove the card and start counting. Make an f/5.6 at 8 seconds exposure, an f/8 at 16 seconds exposure and an f/11 at 32 seconds exposure (or whatever times you need, doubling time for each stop closing down). I'll bet they will all be very close. (Note: how you count seconds is important, so count like a musician: "1-2-3-4-5-6-7-8-off" for the 8-sec. exposure, "1-2-...-15-16-off" for the 16-sec. exposure, etc.

FWIW, this is how I print everything.

Best,

Doremus

 

[Fatih Ayoglu]

Ok, some news. I have swapped the bulbs with completely new ones, opened a fresh box of Ilford MGRCV, cut 0.5x5” strips and exposed 4 strips as f/4 at 4s, f/5.6 at 8s, f/8 at 16s and f/11 at 32s. I have developed all of them with a single Nova holder to make sure development is consistent as well. Results below;

I’m not going to tell you which one is which but clearly they are equal to me. All have some density at 13, and I can’t see below 6 at any of them. They all read 2.3 at density meter at their darkest edges.

We can see how dense the strip is for f/4 at 4s. Nearly over 2 stop of exposure!

I’m so glad it was the bulb issue but not some very expensive part.

Thanks all for your help, much much appreciated!

 

[pentaxuser]

Fatih said: "We can see how dense the strip is for f/4 at 4s. Nearly over 2 stop of exposure!

I’m so glad it was the bulb issue but not some very expensive part.

Thanks all for your help, much much appreciated!"

I am glad that you went down that road rather than down what appeared to me a very large rabbit hole of bulb warm-up as the issue

pentaxuser

 

[koraks]

rather than down what appeared to me a very large rabbit hole of bulb warm-up as the issue

But that's what it was, though - to my surprise. Of course, if it turns out to be the bulb and the problem stays away, that's great! It's a simple and cheap fix, after all. Fingers crossed that this does indeed turn out to be the solution!

 

[snusmumriken]

I’m so glad it was the bulb issue but not some very expensive part.

Excellent! But you changed two things: bulbs and paper. So it’s still possible it was the paper. I don’t see how, but one can always be surprised.

 

[Fatih Ayoglu]

Also when I have looked to the results from the fresh bulbs, I can see f/11 + 32s is 2 stop overexposure. So realistically, I should have exposed the paper at f/11 + 8s to achieve deepest black in the shortest time. This correlates with the old bulbs, I could achieve deepest blacks at f/11 8s, as I have mentioned before. That means this problem occurs only when the exposure was too short and cut before the bulbs have reached full output and given the correct amount of exposure. Anyways, I have kept the old bulbs in a box, labelled as old, for backup and will order a new set of bulbs to have a fresh backup.

 

[MattKing]

The electrical specialists here will know whether this makes sense, but I wonder if the lamp sockets and the connections on them are the source of your problem.

 

[Fatih Ayoglu]

But then I wouldn't get correct results after changing to new bulbs?

 

[MattKing]

Removing and replacing the bulbs can help remove corrosion or other build-up from the connections.