Sensitometry and results from an artificial light source

[Nicholas Lindan]

Dear Rene Boehmer:

Ah, thanks for the context. Now we know what you are trying to do and maybe the advice will be more to the point and helpful.

The problem in forums is that often you don't know who you are talking to. It could be a teenager just starting out or a grizzled and seasoned photographer who has been there and done that.

I guess a solution to that problem would be if the name doesn't ring a bell then look at the poster's profile before replying. Won't happen, of course.

 

[ic-racer]

Dear IC Racer,
could you go into a bit more detail on how to create an ISO sensitometric daylight source? Is it a simple process? For B&W, I am sure it's not too critical, is it?

A xenon flash is probably the best bet for a homemade sensitometer. Otherwise, with a tungsten filament, a shutter may be required, like the Wejex sensitometer uses.

There are multiple choices for light source, all with advantages and disadvantages. Personally I'd just get a sensitometer off ebay for a few dollars. (which I have done and tested against standards like the EG&G and Wejex.)

 

[dkonigs]

A xenon flash is probably the best bet for a homemade sensitometer. Otherwise, with a tungsten filament, a shutter may be required, like the Wejex sensitometer uses.

There are multiple choices for light source, all with advantages and disadvantages. Personally I'd just get a sensitometer off ebay for a few dollars. (which I have done and tested against standards like the EG&G and Wejex.)

The problem with getting a sensitometer off eBay is that the only ones that are easy to find on eBay and in good condition seem to be those compact blue/green X-Rite units designed for X-Ray film. If you want to nitpick on light spectrum, these absolutely will not satisfy you.

Those EG&G and Wejex clunkers are much harder to find, and often look like something that was sitting in someone's dusty tool shed for 20 years, and may or may not work without a lot of fixing. Assuming you can find a decent listing in the first place.

 

[Rudeofus]

Thanks for this. I red through the threads, and it might actually be good to store the undeveloped negatives for some time. In order to approach a more reliable measurement.

Even if you get a precise way of handling your film after test exposure, you still don't know, what your customers will end up doing. How would you recommend an EI? Expose at box speed minus one stop for holiday snaps, but only minus half a stop for weekend shots?

 

[Milpool]

Thanks for this. I red through the threads, and it might actually be good to store the undeveloped negatives for some time. In order to approach a more reliable measurement.

Well, it is so many factors. I'll just try my best, do the work, see if it is good in a practical sense and if it is, ill leave it as is. After all, results are the best way to measure.

I suggest following the standard where possible for a baseline (for example see conditioning for hold times):

https://cdn.standards.iteh.ai/samples/3580/192a0f97667044f3ba51276b0ded056a/ISO-6-1993.pdf

 

[ic-racer]

I suggest following the standard where possible for a baseline (for example see conditioning for hold times):

https://cdn.standards.iteh.ai/samples/3580/192a0f97667044f3ba51276b0ded056a/ISO-6-1993.pdf

Light source specs from that already posted in j#27 above.

 

[ic-racer]

BLUE/GREEN senstiometers work just fine for comparing conventional photography film in a conventional darkroom. Green sensitometer peak indicated by the green line.

 

[reneboehmer]

Even if you get a precise way of handling your film after test exposure, you still don't know, what your customers will end up doing. How would you recommend an EI? Expose at box speed minus one stop for holiday snaps, but only minus half a stop for weekend shots?

This is intended to be an offer to more experienced clients. Clients measure scene brightness luminance and can look up the chart on our website, the chart will give them the appropriate sensitivity, based on my development. Therefore, giving people the opportunity to get very good negatives from a lab. (without making sesnitometry or densitometry measurements themselves.)
I am aware that this applies mostly to people who either shoot whole rolls in one scene, or who control their environment. But it's still a worthwhile thing to do, even if it only creates better negatives for one client.

 

[Romanko]

What are the tolerances of hand-processing film? Factors like mixing chemicals, their quality, measuring and maintaining temperature, time and agitation regime all affect the final result. Have you tried to estimate them? Likewise, shutter speed and aperture accuracy as well as measuring the exposure introduce further uncertainty. You might need to address these before attempting precise development to 0.01D or whatever is your target.

These small variations are integral parts of the analogue process and could be the reason why photographers choose to use film. In situations where precise control is required digital capture could be a better choice.

 

[reneboehmer]

What are the tolerances of hand-processing film? Factors like mixing chemicals, their quality, measuring and maintaining temperature, time and agitation regime all affect the final result. Have you tried to estimate them? Likewise, shutter speed and aperture accuracy as well as measuring the exposure introduce further uncertainty. You might need to address these before attempting precise development to 0.01D or whatever is your target.

These small variations are integral parts of the analogue process and could be the reason why photographers choose to use film. In situations where precise control is required digital capture could be a better choice.

It's not about changing the accuracy of things I can not influence. It's about a certain work moral, it's about trying your best for each film. (I'm going to ignore the digital part of the sentence)

 

[reneboehmer]

BLUE/GREEN senstiometers work just fine for comparing conventional photography film in a conventional darkroom. Green sensitometer peak indicated by the green line.

Dear Ic,
I saw this in the other threads and this is fine when comparing. But it should be an issue when doing speed tests, no?

 

[BCM]

Have you read Phil Davis' book Beyond the Zone System? Most of your answers are there. He addresses simple light sources, exposing film using your enlarger (including measuring the light and diffusion) and test methods. We taught this in workshops for years with great success using all B/W films.

 

[dokko]

I was thinking about a similar project but realistically won't have time for it in the next few months. a few thoughts:

Personally, I'd use a high CRI LED over tungsten bulb for three reasons:

- Tungsten light will change output as they get older (and when you have to replace it with a new one)
- LED can be pulsed very accurately, eliminating the need for a shutter (which again can develop variances and is hard to get accurate in the first place)
- you can arrange LEDs in an even array configuration, while tungsten light is usually coming from a point light source, which needs a larger box to get even lighting.
- the ISO might specify 3200K, but most photography these days is done around 5500-6500K light, so I'd be more interested in a films responsein this light.

the trickiest part is to build a box with very even lighting, but luckily with digital cameras is quite simple to improve with trial and error (take a shot, crank up the contrast like mad, modify until it's even).

 

[reneboehmer]

Have you read Phil Davis' book Beyond the Zone System? Most of your answers are there. He addresses simple light sources, exposing film using your enlarger (including measuring the light and diffusion) and test methods. We taught this in workshops for years with great success using all B/W films.

Hello, yes, I am very familiar with the btzs approach. As mentioned in this thread before, this is a possibility, especially when using a filter to change the color temperature. But my initial question, the reason this thread exists, was if a continued LED light source, that is very accurate to ECI D65 or ECI D50, is a suitable substitute to the tungsten sources mentioned in the ISO for sensitometry light sources. In my mind, as simple as it might be, I came up with the idea of such a LED being a good substitute since it will give off spectral properties that are very similar to a standard daylight source. I figured, as long as the output of the light somewhat represents a daylight source, it will be good enough to rate film speeds from. The reason this has not to be very accurate, again in my mind, is that the spectral products reaching the film, on a normal roll, will be completely random. The only goal for my setup is to reproduce an average of these random scenes. There is a good chance I am overthinking this, but well, where's the fun in something being a simple solution.

 

[reneboehmer]

I was thinking about a similar project but realistically won't have time for it in the next few months. a few thoughts:

Personally, I'd use a high CRI LED over tungsten bulb for three reasons:

- Tungsten light will change output as they get older (and when you have to replace it with a new one)
- LED can be pulsed very accurately, eliminating the need for a shutter (which again can develop variances and is hard to get accurate in the first place)
- you can arrange LEDs in an even array configuration, while tungsten light is usually coming from a point light source, which needs a larger box to get even lighting.
- the ISO might specify 3200K, but most photography these days is done around 5500-6500K light, so I'd be more interested in a films responsein this light.

the trickiest part is to build a box with very even lighting, but luckily with digital cameras is quite simple to improve with trial and error (take a shot, crank up the contrast like mad, modify until it's even).

Tungsten lights these days are quite ok when it comes to stability in brightness, I think. The lamp in a sensitometer will not really be on a lot, therefore this is a little concern.
Yes, I think it's a good point that LEDs can be switched on and off with very little warm up time.
The ISO norm, as far as I know, wants you to filter the tungsten halogen light with a chemical, in order to represent daylight color temperature.
I don't see any issue in building a box with even lighting, but as @koraks mentioned before, flare might become a concern.

 

[dkonigs]

I was thinking about a similar project but realistically won't have time for it in the next few months. a few thoughts:

Personally, I'd use a high CRI LED over tungsten bulb for three reasons:

- Tungsten light will change output as they get older (and when you have to replace it with a new one)
- LED can be pulsed very accurately, eliminating the need for a shutter (which again can develop variances and is hard to get accurate in the first place)
- you can arrange LEDs in an even array configuration, while tungsten light is usually coming from a point light source, which needs a larger box to get even lighting.
- the ISO might specify 3200K, but most photography these days is done around 5500-6500K light, so I'd be more interested in a films responsein this light.

the trickiest part is to build a box with very even lighting, but luckily with digital cameras is quite simple to improve with trial and error (take a shot, crank up the contrast like mad, modify until it's even).

Another tricky part is actually getting a high CRI LED with a decent spectrum. The vast majority of "white" LEDs have this huge peak in the blue, then a dip, followed by a gentle curve over the rest of the spectrum.

Having spent way too much time digging into this, your best bets are probably from YUJILEDS or Seoul Semiconductor. Both have white LEDs with a much nicer spectrum (still a bit lumpy, but much better than the norm). You just need to dig through their product pages to find the highest CRI models.

 

[reneboehmer]

The problem with getting a sensitometer off eBay is that the only ones that are easy to find on eBay and in good condition seem to be those compact blue/green X-Rite units designed for X-Ray film. If you want to nitpick on light spectrum, these absolutely will not satisfy you.

Those EG&G and Wejex clunkers are much harder to find, and often look like something that was sitting in someone's dusty tool shed for 20 years, and may or may not work without a lot of fixing. Assuming you can find a decent listing in the first place.

Yes, this is also my experience.
These LEDs I found seem very promising. Do you by any chance know if the CIE 50 or 65 norm is in any way similar or comparable to the standard light source for sensitometry?
Attached is a picture of spectral data from the before mentioned LED.
I am, no expert and half as smart as most people on this forum, my question stems from a lack of knowledge. I would just like to know if I can build a decent sensitometer with a modern white light LED source.
I'd be grateful for an answer from you Derek since I value your opinion on this! All the best! René

 

[reneboehmer]

Another tricky part is actually getting a high CRI LED with a decent spectrum. The vast majority of "white" LEDs have this huge peak in the blue, then a dip, followed by a gentle curve over the rest of the spectrum.

Having spent way too much time digging into this, your best bets are probably from YUJILEDS or Seoul Semiconductor. Both have white LEDs with a much nicer spectrum (still a bit lumpy, but much better than the norm). You just need to dig through their product pages to find the highest CRI models.

Well, just as you replied, I hit enter on my last entry. :-D These are the exact LEDs I am looking at. https://store.yujiintl.com/collections/cie-illuminant-technology/products/yujileds-cie-d50-sai-86-2w-led-smd-5050

 

[dokko]

Tungsten lights these days are quite ok when it comes to stability in brightness, I think. The lamp in a sensitometer will not really be on a lot, therefore this is a little concern.

yes, it's certainly not the biggest issue, but LEDs simply are much better in this regards.

I don't see any issue in building a box with even lighting

well, it always depends on how accurate it needs to be.
visually even is easy, but uniform to say 0.1% is very difficult.

the same is true for all engineering. if we only need 90% accuracy, a lot of approaches work well.
if we need 99%, a lot more planning is needed, and if we want 99.9% it becomes very hard (to the point that we probably would have to include temperature control or a feedback loop).

I generally tend to over engineer, since I like the challenge. for practical photography that's certainly overkill. but then again, one could argue that sensitometry is not needed at all for practical photography

 

[koraks]

I would just like to know if I can build a decent sensitometer with a modern white light LED source.

Whether you could do it, specifically, I couldn't say. But in principle, yes, it can be done, and I don't think it's necessarily extremely complicated.
The main 'trick' is to decide how you're going to do the different light intensities. One way would be to use a calibrated step tablet (Stouffer, Danes Picta etc). Another way would involve timed exposures on different parts of the film. This of course would introduce challenges w.r.t. reciprocity behavior, but within a reasonable margin, I'd expect this issue to be negligible.

 

[reneboehmer]

Whether you could do it, specifically, I couldn't say. But in principle, yes, it can be done, and I don't think it's necessarily extremely complicated.
The main 'trick' is to decide how you're going to do the different light intensities. One way would be to use a calibrated step tablet (Stouffer, Danes Picta etc). Another way would involve timed exposures on different parts of the film. This of course would introduce challenges w.r.t. reciprocity behavior, but within a reasonable margin, I'd expect this issue to be negligible.

Yes, I own a calibrated Stouffer 21 step strip. I figured that would be suitable. Now that you mention the approach of using the LED intensities to do the wedges, I am intrigued. Sounds like a cool idea. Although right now I might prefer the minimal effort route.

 

[koraks]

The step tablet approach will be easier, yes. You would need to determine the intensity of the exposure in lux*seconds at the film plane for ISO speed calculations though. Have you thought about this yet?

 

[reneboehmer]

The step tablet approach will be easier, yes. You would need to determine the intensity of the exposure in lux*seconds at the film plane for ISO speed calculations though. Have you thought about this yet?

I will probably rent a unit to measure it.

 

[Bill Burk]

I’d love to be able to tell a lab to develop my Panatomic-X to CI 0.57

And as a lab, you would want to be able to develop any film to any contrast.

It would take five test exposures for each film/developer combination to work up the data to create a usable time-CI curve.

As a lab you should probably seek a Wejex or EG&G. They can be had, and you could “afford” to buy two of them to ensure you get one working.

I think the electronic flash in a box is a good idea. You can always test output with a flash meter like a Sekonic L-758DR. The actual light output can be worked out by reverse engineering.

You want enough light to create a full scale on film, which you will know the first test you develop. You’re going to know if there are steps with under 0.10 density over base+fog or not, and if there aren’t… decrease exposure next time. If there are too many clear steps, count the clear steps and divide by two (for Stouffer scale 2115C having 0.15 density difference per step) - that tells how many more stops of exposure are required.

 

[koraks]

And as a lab, you would want to be able to develop any film to any contrast.

But what about curve shape? As a lab, wouldn't you want to be able to develop any given film with an upswept curve that kinks at 0.8logD, or 0.5, or 1.2, with a long toe or with a truncated toe, etc.
Where does it end? What's the practical utility of all this, given the fact that we can scan negatives and do curve-yoga all we want, or use VC paper in the darkroom combined with burning, dodging, flashing, bleaching etc.

I'd be very cautious in assuming you would "need" to offer app these options.