Determining ISO of your emulsion

[Nodda Duma]

I'm hoping I can begin a discussion about the best methods to determine the ISO rating of a newly-made emulsion. When mixing a new emulsion, we always assume it is "about ISO 1, give or take". As we all know, this is a very rough approximation.

Let's say that we know how to determine ISO / ASA rating of film. Many of the examples I've seen involve devoting a full roll or most of a roll of 35mm film to the task. Determining ISO rating of home-made emulsion, I think, is a little trickier because a) the plate cameras we use have no meters, b) the ISO is so low that normal charts don't help us much, and c) the labor and economy of DIY emulsion motivates one to search for a method that uses perhaps one or two plates. That is, some method a bit more scientific than the trial and error method of evaluating the scene of a developed plate and adjusting exposure if it's under- or over-exposed by some unknown number of stops.

I made an attempt today to measure the ISO of Batch 2 -- my second emulsion mix after the batch I mixed back in the fall. It was sunny and pleasant day outside, I had the day off, and I could thus take advantage of the Sunny 16 rule. I also have a gray card (18% gray) and a white poster. I set those up in sunlight, so that the gray card and white poster each filled half the frame. I then exposed sections of the plate similar to what you do in a darkroom under the enlarger when determining the proper exposure for a good print.

Developing this plate, I evaluated the densities and determined this emulsion batch is running at about ISO 0.5. The deduction involved calculating EV based on the shutter time corresponding to best density, and then using the EV equations to back out S in EV(S) = EV(100) + log2(S/100). A couple of test shots indicate I might actually be close to the actual ISO rating.


So this method seemed to work. However, I'm not even going to get into how "...I evaluated the densities and determined...". Suffice to say it involved a scanner and leaning heavily on my image processing background as an optical engineer to remove the effect of the scanner itself (I assume "scanner" is not off-topic for this subject as it was used as a evaluation tool and not used as part of a photographic process per se). Very susceptible to GIGO.

Please feel free to point out issues with my method. I'd be curious to know how emulsion sensitivity was evaluated originally -- i.e. at the turn of the 20th century, before densitometers were available but yet they would have used the scientific method -- or at least a better method which is less reliant on suspect methods such as scanning and compensating for the gamma curve. Any reference texts I can review?

For reference, here is an example image from today's testing. The top half is the white sheet, and bottom is the gray scale card. f/# was set to f/45, and each strip is increasingly exposed in 1 second increments (1 second on far right, through to 11 seconds on the left).

 

[Photo Engineer]

Ok, determining ISO is a problem due to spectral sensitivity.

The easiest way with blue or ortho emulsions is to coat on paper and expose along with a blue or ortho emulsion and test. However, if it is on paper, then you must subtract 1/2 of the speed due to back surface reflection. But this does not compensate for UV radiation.

The easiest way is to just expose your material and some other material on the same substrate (film or paper) and then compare. That is what I do.

It seems to work.

PE

 

[Nodda Duma]

Thanks Ron, yes the spectral sensitivity bugs me. It's bugged me ever since I first started using a meter to estimate exposure. I don't meter off vegetation for this very reason and just let plants act as a dark background. So far I've been able to "dodge the bullet" with careful composition and subject matter (people, rocks, and buildings).

For my testing, one underlying assumption is that the spectral reflectivity of the gray card is flat. I'm trying to fight the urge to bring my grey card into work and measure the reflectivity across UV / VIS spectrum on a spectrophotometer. I also have the equipment (light and filters) to set up a source to mimic the spectrum that the blue emulsion would see, and could see how my meter responds.

Ok I think I understand how your approach can get me there. I just have to assume the emulsions have linear response...valid if I test correctly. I'll give it a try (despite what Yoda would say).

-Jason

 

[Bill Burk]

Ah you should bring that card in to work...

Have you got any UV meter? I found a cheap one that acts like an incident meter and indicates mW/square meter. I have no idea how to correlate that to exposure, but it may be helpful information to record alongside your exposure tests and photographic exposures.

One way to "deal" with UV is to filter it out. That way the meter and camera will agree. But that would waste an awful lot of the very light you need to expose with so I'm sure that approach is impractical.

I think comparing to "Sunny 16" is a good sanity check of your estimated speed.

I'd be inclined to draw a D-LogE graph and to the best of my ability estimate the LogE. Usually I estimate LogE by comparing graphs of unknown film to graphs of a known film as PE suggested.

Stouffer sells calibrated transmission gray scales for a very reasonable price. You can scan one of those with your samples and compare the scanner output against the marked step wedge patches. I haven't been able to get good density estimates from a scanner but you should be able to come close by "comparing".

 

[Nodda Duma]

Thanks for the insight. I like the idea of plotting a D - LogE graph. Also of scanning the gray card itself. Tying back to a meter is sort of secondary...if I tie it back to an EV chart then I can just evaluate based on the scene (EV 15 is sunny, EV 12 is open shade, etc). I actually generated an Excel chart which gives exposure times for f/# vs EV and the chart is compensated for emulsion speed. I can enter my ISO speed which updates exposure values, print this out and then stick it in my camera bag.

I've also been tempted to use or acquire something to measure density - I have a Beseler color meter (how to use that and convert to density?) or could get something like an old Xrite 820 off eBay. I'd have to figure out how to properly use the densitometer.

My goal here is two-fold: figure it out for my own use (done that I think), then share an accessible procedure to advance the knowledge of our fledgling modern-era emulsion-making community.

 

[RalphLambrecht]

Ah you should bring that card in to work...

Have you got any UV meter? I found a cheap one that acts like an incident meter and indicates mW/square meter. I have no idea how to correlate that to exposure, but it may be helpful information to record alongside your exposure tests and photographic exposures.

One way to "deal" with UV is to filter it out. That way the meter and camera will agree. But that would waste an awful lot of the very light you need to expose with so I'm sure that approach is impractical.

I think comparing to "Sunny 16" is a good sanity check of your estimated speed.

I'd be inclined to draw a D-LogE graph and to the best of my ability estimate the LogE. Usually I estimate LogE by comparing graphs of unknown film to graphs of a known film as PE suggested.

Stouffer sells calibrated transmission gray scales for a very reasonable price. You can scan one of those with your samples and compare the scanner output against the marked step wedge patches. I haven't been able to get good density estimates from a scanner but you should be able to come close by "comparing".

keep in mind that the lens glass itself is a UV filter.

 

[dwross]

Determining the ISO of a particular emulsion is both far easier and a bit more complex than I think I read here. An emulsion has a general baseline ISO, but that is determined by an average of conditions over the course of an entire year – at a given latitude (and to a lesser extent, elevation) and with a given lens (as Ralph alludes to). A single test, no matter how rigorous, will only give you a snapshot of the emulsion. ISO varies enormously over the course of the year, especially with colorblind and ortho emulsions. The higher the latitude, the greater the variation throughout the year. The developer used also influences the speed.

If a reliable, actual, working speed determination is the goal, one or two plates from each batch of emulsion should be used for speed determination. Make at least one batch a month of the exact same emulsion for one year. At the end of the year, you’ll know your emulsion backwards and forwards, and have developed your very own technical data sheet for the emulsion. Your own technical sheet is the only one that matters.

The kind of meter isn’t as important as it is to have a consistent strategy for using it throughout the year. I like my inexpensive incident meter and the palm of my hand. I meter my hand in the shadow of my body with the sun directly behind me.

A good starting rule of thumb for beginning testing is that an unsensitized plain silver emulsion will be ISO 1-3 at high noon at winter solstice and 6-12 at summer solstice. An ammoniacal silver emulsion is twice to four times faster.

My technique: Take one of your film (or plate) holders and draw lines, about ¾ inches apart, parallel to the pull bar. (I like a silver calligraphy pen.) Set up a test shot in a place that is convenient to replicate and repeat at different times of the day throughout the year. Set the ISO on the meter to one step below the low end of the expected range. Meter and get the shutter speed for your preferred f-stop. Pull the darkslide out just far enough so that you see the first line. Expose. Pull the darkslide out until you see the second line. Expose. Repeat until you have exposed with the darkslide fully removed. Process carefully in a way that you will repeat for each and every test. You’ll have created a step wedge for that time and place. Correlate the steps with ISO. The last step (the lightest after processing) is your beginning ISO. Each step denser is one ISO jump greater. The best density is the best ISO for the given time and time of year on a sunny day. For a more complete picture, test in the shade and/or early and late in the day, or on a cloudy day. Needless to say, keep a good notebook. It’s also is a great help to keep all your test plates with a complete description of the exposure conditions and the processing.

Of course, you don't need to test for a year before you can start using your plates for "real" photography. Ballpark the speed and bracket as much as you feel you can spare the plates. My opinion: Use them freely and fast. That way, you'll be needing to make more emulsion more often and the more you make emulsion, the better you get!

Good luck and fun no matter your technique to learn.
d

 

[Bill Burk]

dwross,

Is it the emulsion which changes speed over the year?

Or is it the amount of UV which changes?

If it's the UV... and glass cuts UVA as Ralph pointed out, so it would be the UVB that's making meter readings less than reliable.

I took my little UV meter out and read 143 mW/sq.m and behind glass it falls to about 40 mW/sq.m

When I held up a UV or Skylight filter the reading dropped to 0.

This makes me think you could use a regular exposure meter and put a UV filter on the camera. You could also put a UV filter on the light meter. I found it makes about a 1/12 stop difference (lower reading with filter) "whoopee!" with a UV filter on a Pentax Spotmeter V.

Or is it nothing. Is this a solution looking for a problem to solve?

 

[Photo Engineer]

The UV transmission of any given system is constant, but the light varies throughout the year. Kodak established a baseline of Noon in Washington DC on a given summer day (I have forgotten, but it might have been the first day of summer) and we used that for our Sensitometers.

I use a MacBeth color checker for exposures and monitor the density of the gray scale and the colors (as I vary spectral sensitivity) and I use Stouffer step wedges on-easel to get curves. You can get curves from the MacBeth but it is more involved and includes system optics.

I've even exposed commercial films on-easel and in-camera for comparison exposures and have found that with an old lens I can get ISO 25 from an Azo emulsion on paper due to the UV in the light and the old optics. This is with an emulsion over 5 stops slower than an enlarging paper. So it can vary.

What I have found is that you can find a speed simply by using a meter and avoiding greenery in the measurement.

Nick and I actually reviewed his procedure last week and found that it works. It is pretty much as I describe. They even used it in the motion picture workshops.

PE

 

[Bill Burk]

My technique: Take one of your film (or plate) holders and draw lines, about ¾ inches apart, parallel to the pull bar. (I like a silver calligraphy pen.) Set up a test shot in a place that is convenient to replicate and repeat at different times of the day throughout the year. Set the ISO on the meter to one step below the low end of the expected range. Meter and get the shutter speed for your preferred f-stop. Pull the darkslide out just far enough so that you see the first line. Expose. Pull the darkslide out until you see the second line. Expose. Repeat until you have exposed with the darkslide fully removed.

Minor White explains it as pulling out the darkslide all the way except the last 3/8 inch (that last 3/8 inch receives no exposure and is Zone 0)

Expose as for Zone I (this gives the whole sheet except that last 3/8 inch a Zone I exposure). Then push in the darkslide 3/8 inch and expose same again.

(This gives the whole sheet double the Zone I exposure except for the last 3/4 inch which got Zone I and Zone 0) That's the Zone II exposure.

Continue doubling the exposure and pushing in the darkslide... It's the doubling that makes every step equivalent to the next Zone.

This test ignores the intermittency effect. I assume you are starting with the maximum exposure and cutting each exposure in half when you start with a closed darkslide and pull out to the next line for each test exposure. There is probably no difference in the result.

This method does rely on a very even lighting on the test target, and you must have no vignetting on the camera lens. I once did such a test in a way that I had serous vignetting and it led me to conclude the EI of 400 TMAX was 64.

 

[Photo Engineer]

Multiple exposures on a home made emulsion will emphasize the problems inherent with multiple re-exposures. A factory made emulsion is designed to minimize this. Modern Kodak and Ilford emulsions have practically none.

PE

 

[Bill Burk]

The UV transmission of any given system is constant, but the light varies throughout the year.

Nick and I actually reviewed his procedure last week and found that it works. It is pretty much as I describe. They even used it in the motion picture workshops.

PE

So when working with home-made emulsions, do you want to take advantage of the UV to get all the speed you can get?

Then it might make sense to "evaluate" the UV transmission of your "system" and use a UV meter (such as a Blak-Ray J-221 that varies in price from 35 to 200 dollars on eBay or a more modern "UV Checker" for about same price, example this is the one I used today http://www.ebay.com/itm/UV-Checker-/252441120892?hash=item3ac6a9d07c:g:yygAAOSwHaBWirjz ).

There must be a correlation between incident UV light and the bump it gives your exposures, such that you might meter with a normal exposure meter and then apply an additional factor of increase due to contributory UV

 

[Bill Burk]

Multiple exposures on a home made emulsion will emphasize the problems inherent with multiple re-exposures. A factory made emulsion is designed to minimize this. Modern Kodak and Ilford emulsions have practically none.

PE

So to rule out intermittency effect...

Maybe make a set of three or four darkslides, each with a rectangular opening that covers all but a sixth or an eighth of the plate (on one side)... Then use them in sequence to make independent exposures that are in the double or half exposure series... after running through the set with the openings on the left side, flip them around and run through the set on the right side.

 

[Bill Burk]

f/# was set to f/45, and each strip is increasingly exposed in 1 second increments (1 second on far right, through to 11 seconds on the left).

Do you understand the idea of doubling (or halving) exposure for each strip? It will give you a fuller range of samples and they would be logarithmic.

 

[Photo Engineer]

Bill, I would NOT use UV for exposure of any emulsion unless it was intended solely for that purpose. I used it with the AZO type emulsion because it has little visible sensitivity unless treated. I've actually done that as well and exposed it with tungsten.

I do not suggest using anything that causes multiple exposures for measurement. It is too chancy.

PE

 

[Nodda Duma]

Do you understand the idea of doubling (or halving) exposure for each strip? It will give you a fuller range of samples and they would be logarithmic.

Yes, I did that today. One of my light shields (light blocking sleeve in the plate holder?) has a strip-sized hole cut out to avoid measurement error from overlapping exposures due to the plate holder shifting around in the 120 yr old camera. Learned the hard way what PE says above.

Did some strip tests and then bracketing shots today...looks like I'm running within a half stop or so of ISO 0.5 for these plates. Doubling the exposure from there gave negatives that seemed overexposed, and halving the exposure from that of 0.5 gave underexposed negatives. Had to let the plates dry before checking.

 

[Bill Burk]

OK so we

Bill, I would NOT use UV for exposure of any emulsion unless it was intended solely for that purpose. I used it with the AZO type emulsion because it has little visible sensitivity unless treated. I've actually done that as well and exposed it with tungsten.

I do not suggest using anything that causes multiple exposures for measurement. It is too chancy.

PE

Sure. While half of each exposure is a single exposure, the other half of the exposure for each strip is a hodgepodge.

 

[Bill Burk]

Yes, I did that today. One of my light shields (light blocking sleeve in the plate holder?) has a strip-sized hole cut out to avoid measurement error from overlapping exposures due to the plate holder shifting around in the 120 yr old camera. Learned the hard way what PE says above.

Did some strip tests and then bracketing shots today...looks like I'm running within a half stop or so of ISO 0.5 for these plates. Doubling the exposure from there gave negatives that seemed overexposed, and halving the exposure from that of 0.5 gave underexposed negatives. Had to let the plates dry before checking.

What does your light shield look like? I figure it would have to be a roller blind to allow just one strip to be exposed at a time.

 

[Photo Engineer]

Regarding the speed, are you sure you coated enough silver and developed long enough? A fully exposed plate or film should achieve a Dmax of about 3.0 if you coated enough silver halide and developed long enough. This maximizes or rather optimizes speed.

PE

 

[Prof_Pixel]

Just to try and keep things accurate. What is being discussed here is NOT ISO speed which is well defined! See for example https://en.wikipedia.org/wiki/Film_speed#Current_system:_ISO



What is being discussed here is effective film speed or perhaps (if you prefer) equivalent film speed. Since the early emulsions used by most here are just blue and UV sensitive, and since the blue and UV portion of daylight will change as a function of time of day and time of year, the effective film speed will change (although the actual sensitivity of the film doesn't), so the best an effective film speed setting can do is give you a suggested exposure starting point.

 

[Bill Burk]

Since the early emulsions used by most here are just blue and UV sensitive, and since the blue and UV portion of daylight will change as a function of time of day and time of year, the effective film speed will change (although the actual sensitivity of the film doesn't), so the best an effective film speed setting can do is give you a suggested exposure starting point.

OK I've been focusing on the problem of UV... You can negate that problem of the UV portion of daylight changes over time of day and year... with a UV filter.

But that doesn't account for the variability in the blue portion of daylight. That still varies.

It might be easy to meter this.

 

[Prof_Pixel]

OK I've been focusing on the problem of UV... You can negate that problem of the UV portion of daylight changes over time of day and year... with a UV filter.

But that doesn't account for the variability in the blue portion of daylight. That still varies.

It might be easy to meter this.

With early emulsions, I'm not sure you really want to throw away the UV, since in general, the emulsions are slow enough that every little bit of exposing light helps.

 

[Nodda Duma]

What does your light shield look like? I figure it would have to be a roller blind to allow just one strip to be exposed at a time.

Dark slide is the term I was looking for (sorry had a brain fart). It looks like a dark slide with a 3.5" X 0.5" sized rectangular hole cut into it.

Regarding the speed, are you sure you coated enough silver and developed long enough? A fully exposed plate or film should achieve a Dmax of about 3.0 if you coated enough silver halide and developed long enough. This maximizes or rather optimizes speed.

PE

I am fairly certain that the emulsion is thin (as opposed to not having the proper concentration of silver halide). This is the main reason I am going through this exercise. Holding one up in the darkroom, I can (barely) see the safelight through the dried emulsion. This tells me that not all the photons will be absorbed as they pass through the emulsion layer. For 4" X 5" I guesstimate I used about 3 ml of emulsion to coat vs the (from what I understand) ~ 5 ml I should be coating it with.

I've sourced a Kodak No. 21 Photographic Step Tablet so I can at least estimate D values until I find a suitable densitometer.

I do have some 5" X 7" plates that I believe are correctly coated (about 10 ml).

Just to try and keep things accurate. What is being discussed here is NOT ISO speed which is well defined! See for example https://en.wikipedia.org/wiki/Film_speed#Current_system:_ISO



What is being discussed here is effective film speed or perhaps (if you prefer) equivalent film speed. Since the early emulsions used by most here are just blue and UV sensitive, and since the blue and UV portion of daylight will change as a function of time of day and time of year, the effective film speed will change (although the actual sensitivity of the film doesn't), so the best an effective film speed setting can do is give you a suggested exposure starting point.

A subtle but important point. Thanks.

 

[Nodda Duma]

Here's a strip sample with associated exposure times. After all my talk about cutting holes in dark slides, I couldn't find the dang thing when I took this shot.
I've also included exposure times for each strip as well as EV readings at various regions of the scene.

Shot at f/16. Developed 1 minute in Dektol, 1:2 dilution, constant agitation, 74 degrees.

I selected the 6s strip as the most balanced, understanding that vegetation should be slightly darker.

Here is proof that kids can't sit still for 3 seconds when they are busy with water balloons. 3 seconds at f/11 (overexposed .. should have been 1.5 second exposure):

After this we all piled in the car to get ice cream and drive around so I could take a couple of photos. However my wife started talking chores and upcoming itineraries for weekend activities and work -- which isn't very inspiring for doing any type of creative photography. She once said that marriage is finding that special someone that you can annoy for the rest of your life. We've both successfully done so! So it'll be tomorrow probably before I get to play again. I plan to set up down on main street for the Independence Day parade. That should be fun -- our fire department has fire trucks dating back to the 1920s which they usually bring out for the parade.

 

[Photo Engineer]

Your development time may be too short. IDK with such a low silver layer, but we use 12' with D76.

PE