Adventures in film characteristic analysis

[grain elevator]

So maybe my exposure is about 2 stops too low. What do you think?

Yes, in comparison to the Stouffer step wedge, your test strip needs more density=exposure (if your development is locked down).

 

[radiant]

Yes, in comparison to the Stouffer step wedge, your test strip needs more density=exposure (if your development is locked down).

Pretty standard development, Rodinal 1+25 for 6 minutes.

Does anyone have approximation in which density 18% tone for HP5 400 box speed + std. dev. should end up in Stouffer T2115 ?

 

[Jonno85uk]

This is interesting stuff. One thing I can't find is the details of the LEDs. What's the part# on them? Just wondering if there is a spectrum sensitivity issue at play.

 

[koraks]

WS2812. They're common as muck - and they would not be my first choice for this to be honest, due to the inherent limitation of not having control over the integrated drive electronics and the speed limitations of the communication protocol. I'd suggest simply using discrete R, G and B smd leds (any 0805 type for instance would do quite well) which at low power levels can be driven directly from a microcontroller pin. Then plot R, G and B measurements separately and you get a sense of spectral sensitivity for almost free on top of normal measurements.

 

[radiant]

WS2812. They're common as muck - and they would not be my first choice for this to be honest, due to the inherent limitation of not having control over the integrated drive electronics and the speed limitations of the communication protocol. I'd suggest simply using discrete R, G and B smd leds (any 0805 type for instance would do quite well) which at low power levels can be driven directly from a microcontroller pin. Then plot R, G and B measurements separately and you get a sense of spectral sensitivity for almost free on top of normal measurements.

I'm planning to do spectral sensitivity too, which can be done with WS2812 as well. Out of curiosity, what makes WS2812 not your first choice? Edit: more accurately in optical way.

If you are driving the LEDs seaparately, you need to control 3 * 16 leds = 48 leds. Quite a wiring and would need dedicated PCB for keeping sanity

 

[RalphLambrecht]

But in fact, film photography and modern semiconductor electronics are an absolutely fantastic combination.

looksto me more like optimizing wagon wheels.

 

[ic-racer]

There's no doubt in my mind that most engineers today would do something quite similar to what @radiant.

Did you study sensitometer design in photography school or engineering? I did and have degrees in both. A modern semiconductor design is detailed in this thread. https://www.photrio.com/forum/threads/x-rite-334-sensitometer-inspection-and-analysis.180579/
radiants project is a lot of fun and clever, but not a plactical sensitometer.

 

[radiant]

Did you study sensitometer design in photography school or engineering? I did and have degrees in both. A modern semiconductor design is detailed in this thread. https://www.photrio.com/forum/threads/x-rite-334-sensitometer-inspection-and-analysis.180579/
radiants project is a lot of fun and clever, but not a plactical sensitometer.

I've seen that device somewhere. Nice design!

Yes this project is fun for sure, however I need to take care it doesn't eat my time from photography itself..

 

[Jonno85uk]

WS2812. They're common as muck - and they would not be my first choice for this to be honest, due to the inherent limitation of not having control over the integrated drive electronics and the speed limitations of the communication protocol. I'd suggest simply using discrete R, G and B smd leds (any 0805 type for instance would do quite well) which at low power levels can be driven directly from a microcontroller pin. Then plot R, G and B measurements separately and you get a sense of spectral sensitivity for almost free on top of normal measurements.

Thanks. I'm looking at the datasheet, more specifically the LED characteristic parameter section. I'm curious if those values would end up with a very spiky looking spectrum (low CRI)? Would that matter?

 

[koraks]

what makes WS2812 not your first choice? Edit: more accurately in optical way.

Optically I'm not too worried about them; perhaps the red is a bit short in wavelength; I'd prefer to work with 660nm leds, but for your specific application it wouldn't matter all that much.

If you are driving the LEDs seaparately, you need to control 3 * 16 leds = 48 leds.

That's the downside indeed, although it's manageable. The alternative of course is to do the strip exposures serially instead of in parallel, but that would make the mechanical design much more complicated.

radiants project is a lot of fun and clever, but not a plactical sensitometer.

If by 'sensitometer' you mean the combined system of making the strip exposure and doing the measurement, it's quite clear to me it's not intended as the full system; it only manages the first half. I'll leave the discussion to you if a (flatbed) scanner is appropriate for the measurement part. For making the actual exposure, it's crystal clear that semiconductor light sources are very suitable indeed. If you believe otherwise, I'd love to hear your arguments. The argument "my education is top notch" quite frankly doesn't contribute much to the debate. It only suggests you have relevant information to share but are unwilling to do so. If that's the case, why post at all?

looksto me more like optimizing wagon wheels.

Yeah, well, if you look at it that way, any fairly modern camera made from 1980s onwards would fit that bill. It's been discussed ad nauseam here and elsewhere that some people just like to tinker.

I'm curious if those values would end up with a very spiky looking spectrum (low CRI)? Would that matter?

CRI would be abysmal, evidently. If and how much it would matter depends on what you want to get out of it - i.e. how important is it to have good insight into a particular part of the spectral response of the film? In this sense, these leds would not necessarily do much worse than the old-fashioned approach of filtered continuous light. The profile would evidently be different, but not necessarily better or worse.

 

[radiant]

That's the downside indeed, although it's manageable. The alternative of course is to do the strip exposures serially instead of in parallel, but that would make the mechanical design much more complicated.

For sure. Also very fast speeds are possible, with my setup I can update the leds at 680 us interval. I checked with oscilloscope and the optical pulses are really that short so the speed is limited by the communication protocol alone. This speed is fine for my needs and for ISO 400 films.

 

[Craig]

Does anyone have approximation in which density 18% tone for HP5 400 box speed + std. dev. should end up in Stouffer T2115 ?

If we are assuming step 11 is the mid point that approximates 18% grey (and I don't know if that is a valid assumption) the density is approximately 0.70.

On my tests I contacted printed the 21 step wedge onto a sheet of 4x5 and I can see a distinct tonal seperation between all 21 steps. It's a very slight difference between steps 21 and 20 that is difficult to see, but there is a easily visible distinction between steps 18 and 19.

 

[radiant]

If we are assuming step 11 is the mid point that approximates 18% grey (and I don't know if that is a valid assumption) the density is approximately 0.70.

On my tests I contacted printed the 21 step wedge onto a sheet of 4x5 and I can see a distinct tonal seperation between all 21 steps. It's a very slight difference between steps 21 and 20 that is difficult to see, but there is a easily visible distinction between steps 18 and 19.

I assume you contact printed on negative film? 10.5 stops for a paper is a bit far stretch

But yeah, there is a big underexposure going on. I'm a bit puzzled because I measured the lux amount and compared to lux seconds for the ISO. I cannot understand why there is such a big underexposure, many stops actually. Could it be the spiky CRI, I don't know. But it is much more easier to increase the exposure than lower it ..

 

[Craig]

Yes, I contact printed the Stouffer scale onto HP5. I adjusted my enlarger to give EV4 when set to focus, then expose for 0.4 seconds. It's the procedure that Phil Davis lays out in his Beyond the Zone System book.

 

[radiant]

Yes, I contact printed the Stouffer scale onto HP5. I adjusted my enlarger to give EV4 when set to focus, then expose for 0.4 seconds. It's the procedure that Phil Davis lays out in his Beyond the Zone System book.

Wow! Interesting! Did the 18% gray end up in position 11 on Stouffer? I assume so.

Anyways that is nice experiment. I think I can show such sample from my device too with some adjustments.

 

[Craig]

I don't know where 18% gray ends up. The procedure is to expose 5 sheets, then develop them at different times. Once developed, then read the density of each step on all the sheets of film and put that into a piece of software he's developed. That then generates the curves for contrast index, effective film speed, developing time vs contrast etc. The idea is it them gives you a custom film speed and developing time table for your chosen film/developer combo and method of developing.

Here are the charts it gave me in replentished Xtol and HP5. Ignore the 70F temperature, I did it at 20C ( 68F)

 

[radiant]

That is pretty cool software. But I see it; if you know how things work you can extrapolate such data from those 5 sheets.

I think I need to shoot one frame on white wall to get 18% gray on that certain development method. I've been trying few things out so then I could atleast check what is the density of 18% with that process.

The density of 18% is about 1.0 based on this: https://kinasevych.ca/2014/04/26/densitometry-for-ilford-hp5-plus-developed-in-rodinal/ and on T2115 it should be somewhere between steps 7-8.

 

[Craig]

The nice thing is all I need to do is meter the brightness range of the scene I want to photograph and then from the charts it can tell me the effective film speed for that particular situation, and the contrast index and developing time.

I can't remember what section it was in, but there was discussion here very recently how there isn't a standard for 18% grey, and it can vary from one grey card to another. That was the joy of Phil Davis's approach, it measured density of the exposed film. You can buy a calibrated step wedge from Stouffer, so you can calibrate your densitometer to the step wedge. Then read the density of the film you exposed and you can generate the charts.

 

[radiant]

The nice thing is all I need to do is meter the brightness range of the scene I want to photograph and then from the charts it can tell me the effective film speed for that particular situation, and the contrast index and developing time.

That is very pro. Perfect negs every time, at least technically

I can't remember what section it was in, but there was discussion here very recently how there isn't a standard for 18% grey, and it can vary from one grey card to another. That was the joy of Phil Davis's approach, it measured density of the exposed film. You can buy a calibrated step wedge from Stouffer, so you can calibrate your densitometer to the step wedge. Then read the density of the film you exposed and you can generate the charts.

That is true. I've been using the 10/ISO for lux-seconds of 18%. I'm not doing any real science with my device so it doesn't have to be in any standards, just in the ballbark. But I think I will get there on next tests I'm exposing. If not, I will probably just expose my Stouffer directly on the film like normal people do.

I believe all Stouffers are pretty much really dead on right and the difference is that you get certificate + calibration report on those calibrated ones. But that is totally not for me, way too accurate for my needs.

 

[Craig]

That is very pro. Perfect negs every time, at least technically

The practical result is my negs are much more consistent and easier to print since I started doing this. I also found for my methods anyway, that the Ilford recommended HP5+ developing time in Xtol was nearly a stop less than it should have been. I was losing shadow detail due to underdevelopment.

I gained a lot of useful information for the modest cost of 5 sheets of 4x5 film and some time.

 

[radiant]

that the Ilford recommended HP5+ developing time in Xtol was nearly a stop less than it should have been.

This is what I read somewhere that when people talk about the film characteristics, they in reality talk about development methods. But really? Currently Xtol 1+1 box speed is 12 minutes. One stop is 50% more time, so it would be more closer to 18 minutes? Both Ilford datasheet and Massive Dev chart notes it is 12 minutes.

It would be awesome if we had some actually measured database somewhere, instead of some list that everyone refers and nobody knows where those times are coming from. Some kind of database that has the original 5 sheet data and then the analysis method could be changed based on needs as the "original" data is available.

 

[Craig]

I wasn't diluting Xtol. Recommended time at ISO 400 is 8 min, at ISO 800 it's 11 min. For the contrast index that is considered normal, I needed about 10 min.

The film data sheets always say the times given are starting point recommendations, they may need to be varied according to personal working practices and desired contrast index. People can agitate in different ways, they might have a thermometer that isn't calibrated etc, so there isn't a "universal" developing time. That's why I did the testing, it gives results for my agitation style (in this case a Jobo) and however the thermostat in it is calibrated.

 

[radiant]

I wasn't diluting Xtol. Recommended time at ISO 400 is 8 min, at ISO 800 it's 11 min. For the contrast index that is considered normal, I needed about 10 min.

The film data sheets always say the times given are starting point recommendations, they may need to be varied according to personal working practices and desired contrast index. People can agitate in different ways, they might have a thermometer that isn't calibrated etc, so there isn't a "universal" developing time. That's why I did the testing, it gives results for my agitation style (in this case a Jobo) and however the thermostat in it is calibrated.

Actually I wasn't referring to your times (I noticed you use replenishing + const.agtation). I was just mumbling in loud about the times.

I might have underdevelopment issues. I haven't calibrated my temperature meter and I'm using my self built tank agitation device. So it is actually a good idea to do calibration at least once. However as I don't know any facts if my exposure device is working, it makes this a bit more challenging I would guess.

 

[otto.f]

Hmm, I don't quite get this. I was talking about maximum density of negative = highlights in print. The exposure of the film was not enough to reach the maximum density of the film, that I was just pointing out. The exposure of the last slot is +4 stops above 18%.

ok!

 

[radiant]

Ok, new test done. Film drying atm. Now I got full range of density. The negatives look very well exposed.

The first example strips 18% slot is exposed at 12.5 milliseconds as it should be 125 milliseconds. There is your problem! On these tests that are currently fine I have 100, 200 and 400 millisecond 18% slot base times and 8 stops over and 7 stops under exposed around that. So plenty of dynamics available on next measurements.

One strip has all 16 stops visible, but I think I have reached dMax on HP5 as the highlight exposures are about 10 stops over 18%

BTW: I cross-checked my thermometer and it is fine.