Film Emulsion test

Ken Willinger · January 14, 2008

Can anyone give me a step by step guide on doing a film emulsion test in a controlled setting?

Thanks.

Saul Rodgar · January 14, 2008

Can anyone give me a step by step guide on doing a film emulsion test in a controlled setting?
Thanks.

This subject has been dealt with in the past quite extensively, I suggest looking it up in the archives.

Ken Willinger · January 14, 2008

This subject has been dealt with in the past quite extensively, I suggest looking it up in the archives.

Thanks for your kind help Saulie. I did a search on all forums for Film Emulsion Test and Exposure Latitude test and got nothing that was close to giving me a step by step description of the process. Can you direct me to the archives? I see no link to it though I did check the resource links. Your help has been immeasurable however. Thanks again.

David Regan · January 15, 2008

It's not a topic I understand well enough to give you really definite directions, but if you get ahold of a copy of the ASC manual, it has articles in there about finding printer lights, and processing. Also the Reflections book starts with Stephen Burum discussing the film tests he runs, basically putting two subjects in front of a light and dark background, with a greyscale in shot, and then exposing the scene basically at key, over and under, and then using a one light transfer and a timed print to analyze how the stock holds up under different exposures. I know your probably more interested in a step by step guide as opposed to me just referencing you to other sources, but they are good places to start I think, and with my limited knowledge on the subject I wouldn't want to get you off on the wrong foot by giving misleading information.

Good luck

Ken Willinger · January 16, 2008

the Reflections book starts with Stephen Burum discussing the film tests he runs, basically putting two subjects in front of a light and dark background, with a greyscale in shot, and then exposing the scene basically at key, over and under, and then using a one light transfer and a timed print to analyze how the stock holds up under different exposures.

Thanks David, I just picked up this book. There appears to be some really great info in it!

Satsuki Murashige · January 16, 2008

What is it specifically about the film stock that you want to test? Do you want to see how much latitude (i.e. ability to under/overexpose and print back to normal with acceptable results) a particular stock has?

In this case, one method is to set up a scene in a dark controlled setting like a soundstage, a windowless classroom, etc.

You would ideally include in the scene:

1. 18% gray card for neutral gray reference

2. black reference tone like Duvetyne or black felt

3. white reference tone like foamcore or gatorboard (styrofoam)

4. an actor for fleshtone reference, preferably two actors with different skintones for comparison

5. Macbeth ColorChecker Chart for color reference (optional)

Light the scene so that everything is lit evenly and in focus. Tungsten balanced film should be lit with 3200K tungsten light. Daylight balanced film should be lit with 5500K light. What you will do is shoot the scene several times: exposed normal, 1 stops under, 2 stops under, 3 stops under. Then in the other direction, normal, 1 stop over, 2 stops over, 3 stops over. You can go further and/or add intermediate steps if you like. This will allow you to see what the effect of under/overexposure in various increments will do to the look of your image. You can then decide which exposure looks best for that particular filmstock and where the image becomes unacceptably "wrong" to your eyes.

When you take your film to the lab, you will instruct them to make two prints:

1. A one light workprint which is timed to the 1st gray card on the roll (the normally exposed one).

2. A timed workprint in which every under or overexposed shot is timed back to normal.

The one light print will allow you to see at which point you will lose detail in the shadows and highlights, when the fleshtones begin to look unnatural, what a face two stops underexposed will look like, etc. The timed workprint will allow you to see how the overall image degrades or improves when that exposure is printed back to normal. For example, underexposure combined with printing up will generally lead to grayish blacks, more grain, less color saturation, and possibly a color shift. Overexposure combined with printing down will tend in the other direction up to a certain point. You will readily be able to see all this for yourself when you view the timed print.

Several pointers:

- To make your test as efficient as possible, light the scene to the highest light level you will need. For example, if you are testing a 200 ASA filmstock and you want to see what 3 stops overexposed looks like, then you would have to light the scene 3 stops brighter than normal. You would then light the scene so that the scene would exposed normally at 25 ASA. You can then add three double scrims (red wire) to your lights, each of which reduces the amount of light by 1 stop, for a total loss of 3 stops which gets you back to normal at 200 ASA. Or you could add 2 double scrims to each light and an ND.3 filter on the lens which would also reduce the light output by 3 stops. The point of all this is that it is much faster to add or remove scrims from a light or a filter from the camera than it is to relight the scene by moving the lights around.

To go in the other direction, you also want to see what three stops underexposed looks like. That means you need to light so that the scene would expose normally at 1600 ASA. Now you have a problem because you can't fit any more scrims into your lights. However, it is much easier to take away light than it is to add it. You could add ND gel to the lights, but that could cause a subtle color shift if the gel is not pure. You could also turn the lights away from the scene and bounce them off of foamcore, effectively lowering the output. Or you could simply walk them back (inverse square law).

- Ideally, you want to make all the changes in lighting level with your lights and not the lens aperture or with ND filters, as these variables can change the contrast, color cast, and sharpness of your final image. Then you won't know for certain whether a certain effect was due to the emulsion or the glass. Practically, you may have to change the lens aperture or use NDs but be aware that they can alter image.

- Light with two sources at 45 degrees to the subject, one on each side of the camera. This will help you keep the left and right side of the subject evenly lit and help to eliminate glare into the lens (angle of incidence = angle of reflection).

- If you have a spot meter, measure the black and white reference tones from the camera position. This will tell you exactly how far over/under the tone is from neutral gray. Then you will easily be able to see in the one light print when the white and black tones lose all detail.

- Keep the lab numbers (RGB numbers, ie. 25, 25, 25) from the timed workprint and see how the numbers corollate to your various exposures. The numbers should go up by about 6-8 points when you overexpose every stop, and they should go down about the same amount when you underexpose. So the next time you pick up a workprint, you can immediately look at the RGB numbers and tell if you were under or over before even watching the film.

I hope this was helpful. :)

Ken Willinger · January 16, 2008

Thanks Satsuki,

This is exactly what I was looking for. Your help is greatly appreciated.

Cheers,

Ken

Patrick Neary · January 16, 2008

- Ideally, you want to make all the changes in lighting level with your lights and not the lens aperture or with ND filters, as these variables can change the contrast, color cast, and sharpness of your final image. Then you won't know for certain whether a certain effect was due to the emulsion or the glass. Practically, you may have to change the lens aperture or use NDs but be aware that they can alter image.

Hi-

First, everyone has their own way of doing these things, but...

For practical purposes it's a lot easier to just light your set-up to 5.6, shoot 10-15 feet or so, open up to 4, shoot 15 feet, open up to 2.8, shoot 15 feet, then 2.0, then do the same thing at 8, 11, 16 and 22 (if you want to go that far.)

Lens/aperture color, contrast and sharpness changes aren't really a factor when you're talking about over/underexposing the neg that much (it brings its own effects), and a timed print will make small color shifts (from a lens iris change) a moot point anyhow, and the one light will be so dark or light at the extreme ends that those shifts disappear into the tortures that the neg is experiencing just in the exposures. At 3 stops under or over you'll see color shifts inherent in the neg more than anything caused by an iris change.

And unless you're lighting your reference scene with 10Ks (and a LOT of wire!), I can't imagine it being a terribly practical or time-efficient way to work.

Sam Wells · January 16, 2008

I think the controlled setting test is a good one for learning.

But to be honest the most meaningful tests I've done have been to: shoot the particular emulsion under conditions and with subject similar to what I want to use the stock (or stock/processing combination) for.

Don't deprive yourself of doing that, it's good to get a feel for these things, not jusyt the numbers.

Sorry if I've editorialized instead of answered !

-Sam

ps a good lab timer is your best friend.