Rating Film Stock

[Chris Durham]

Please forgive me if this question is a hair dense. Gearing up to make the leap to film, but I want to understand as much as possible first.

 

I read a lot about people rating film stock, as far as the ASA; and I'm not sure I fully understand. But I think I have an idea. I thought the film was already rated, just like ISO for still film. Right? and I understand that the more sensitive the film, the more grainy it's likely to be. Is that right?

 

So then, when you talk about rating a stock, are you talking about when you're making calculations with a light meter? Having not used a light meter before I'm not sure, but as I understand it you would enter the shutter and film speeds and it would give you a reading of what you should set your aperture at. Is rating a film stock then a way of manipulating this output?

[Michael Collier]

Rating is a method of giving a standard level of overexposure to the film, usually to take advantage of the smaller grain in the upper end of the films curve. It will also give more detail in the shadows if you were to overexpose slightly and print down.

 

If you have 200asa film, and you rate it at 160, you will overexpose everything by 1/3 of a stop. As it relates to metering, when you enter the asa value of your film, you would enter the value you are rating it as, not what the manufacterer rates it at. then when you meter, every reading you get will be 1/3 of a stop higher than the actual setting, if that makes sense.

[Chris Durham]

Excellent. That's pretty much what I'd surmised and was asking.

 

Now when you talk about grain in the upper end of the curve and printing down I assume you mean that because darker exposure is grainier if you overexpose you'll get less grain on the negative which means a finer image; however when you print from the negative you can elect to "darken" the image by effectively reducing the exposure on the print while maintaining the tighter grain of the negative. Is that right?

 

Can the same trick be pulled on reversal stock? I would think not as easily because if I understand correctly then to print from reversal you have to go from positive to negative to finally be able to print positive; but if you're viewing the reversal stock directly you would be viewing the overexposed image (isn't one of the main purposes of reversal to view without having to print?) Am I wrong in thinking that the better use for reversal stock might be if you knew you were going to telecine; in which case if you were overexposing you would simply bring the image down in post?

 

There's so much I don't understand yet, but talking about film just feels cooler than talking about video (not to spawn the aged debate, just a visceral response - like listening to vinyl).

[Jonathan Benny]

Rating is a method of giving a standard level of overexposure to the film, usually to take advantage of the smaller grain in the upper end of the films curve.

 

Rating a particular film stock is usually based on other decisions and factors that are made beforehand (not just a desire to overexpose). Do you want a generally denser neg? Or do you want a generally thinner neg? Are you bleach-bypassing the neg (ie: are you going to "underexpose"?). Are you going for a milky look? A contrasty look? grain etc? Am I going to print or transfer? What filters am I using on the camera? What are we going to show the audience and how do we want them to see it?

 

Rating a stock is the final step after a number of other decisions.

 

My last two projects on film I rated 5218 at 1000. That is, I underexposed the film a stop (a choice that falls outside of your definition of "rating").

 

 

AJB

[K Borowski]

I read a lot about people rating film stock, as far as the ASA; and I'm not sure I fully understand. But I think I have an idea. I thought the film was already rated, just like ISO for still film. Right? and I understand that the more sensitive the film, the more grainy it's likely to be. Is that right?

 

So then, when you talk about rating a stock, are you talking about when you're making calculations with a light meter? Having not used a light meter before I'm not sure, but as I understand it you would enter the shutter and film speeds and it would give you a reading of what you should set your aperture at. Is rating a film stock then a way of manipulating this output?

 

Don't worry about it. We all had the same trouble with the concept of film speed and exposure indeces that you're having now.

 

ASA and ISO are essentially the same thing. ASA stands for American Standards Association, and ISO is a hybridization of DIN (the German Deutsche Industrial Norm or somethign like that) with ASA in a hybridized form. While the ASA number is often used interchangably as ISO value now, the true ASA 400 corresponds to I believe DIN 20, and the correct full form in ISO is ISO 400/20 with a degree symbol after the 20. There were other systems, GOST, and earlier, but essentially they were all variants of the ASA's multiple growth as opposed to DIN's logarithmic system. With ASA, every time a number is doubled (excusing the rounding off to make it fit in base 10) you gain a stop, whereas DIN gains a stop when you add three to a value.

 

EI is exposure index. A film can be given an ISO of 400/, but be given an EI of 320 or 500 or 800 or 1000 when it is exposed. That's the speed at which the film is treated when it is shot. You treat the assigned EI as if it were the film's rated box speed. In essence that become's the film's speed for all questions concerning exposure. You adjust f-stop and shutter angle/exposure and your light meter corresponding to the EI not the ISO.

 

Generally, when a film is rated at an EI slightly less (1/3 to 1 stop) of it's "box speed", that film is developed normally. This has the effect of giving the film slightly more density. There are usually slow and fast grains on a film. When you give the film more exposure it is akin to "recording in the red" with analog tape. You don't get necessarily as natural of color or contrast as you would with properly exposed negative film (necessarily being the key word here; films used to be rated a full stop faster than their "true" speed), but it reduces graininess, especially in the midtones, your "greys". So not only does this make for a finer grain in the negative, but it also gives you more underexposure latitude. Generally you can't underexpose more than a stop from the box speed without losing important details in the finished print or transfer. Overexposing a full stop gives you about two full stops either way of latitude. You can probably get more than two stops of overexposure, but I digress. Anyway, a lot of actual scientific data tends to agree that films are about a third of a stop slower than Kodak and Fuji rate them with color negative and B&W. With reversal films, this information is totally incorrect, because slides have so little latitude, you can only underexpose by about 1/3 of a stop (which has a similar effect to OVEREXPOSING negative film 1/3-2 stops) before you lose detail.

 

Does this help?

 

Generally, you want to do tests before shooting a whole movie to determine with your lab's chemistry what the "true" contrast of 18% gray is and what exposure you need to give the film to get maximal detail with the least amount of overexposure over box speed, unless you're going for an unusual effect.

 

Now, when you alter processing, this adds more variables. There seems to be disagreement as to whether push processing film (extending development to increase aparant and/or the film's true speed), but I'd say that for every stop you push a film (extended development by the lab over the standard development time), you gain less than a stop of true speed, maybe 1/3 to 1/2 of a stop. So if you need true speed increase, not just retention of highlight detail, you'd probably only get a true speed of 640 from a 500-speed film you push a stop to "1000", and a speed of only 800-1200 for a 500-speed film you push two stops, to "2000".

 

I hope you aren't confused more.

 

Regards,

 

~Karl

[Chris Durham]

Rating a particular film stock is usually based on other decisions and factors that are made beforehand (not just a desire to overexpose).

Thanks John. Yeah, I kind of figured that there's a lot more to it than that; however putting it in terms of exposure is simple enough for me to grasp at this point. My follow-up question was kind of academic, but it all helps me understand a bit better.

[K Borowski]

Can the same trick be pulled on reversal stock? I would think not as easily because if I understand correctly then to print from reversal you have to go from positive to negative to finally be able to print positive; but if you're viewing the reversal stock directly you would be viewing the overexposed image (isn't one of the main purposes of reversal to view without having to print?) Am I wrong in thinking that the better use for reversal stock might be if you knew you were going to telecine; in which case if you were overexposing you would simply bring the image down in post?

 

There's so much I don't understand yet, but talking about film just feels cooler than talking about video (not to spawn the aged debate, just a visceral response - like listening to vinyl).

 

If you alter any more than 1/3-1/2 of a stop for your Exposure Index from the box speed, you must alter development (either push or pull) to retain the same detail in the highlights. NEVER overexpose reversal a full stop without a -1 pull at the lab. While there is some small degree of adjustment possible with telecine and/or analog printing, you aren't "burning in" more information with increased exposure. It's designed to render a viewable image from the fiml that comes out of the camera, so increasing exposure a stop over box speed, certainly without pull processing and probably even with it, will have the consequence of lost image detail. As far as optimizing for telecine, in the still photography world, slides were earlier on optimized for scanning due in part to the ability for you to gauage the image off of the original directly, and also because slides show less grain for a given speed with films of the same emulsion technology. However, negative films have been similarly optimized for scanning now, certainly with movie film and with the majority of still films, so there's no real advantage to shooting reversal for scanning anymore, unless its an approach you're comfortable with because you've been shooting the same reversal emulsion for the same purpose for twenty years and you've become totally familiar with its characteristics. That's the reason why so many "old school" E-6 emulsions are still made by Kodak (don't know if Fuji has any older stocks left in production), because catalogs demand the same film emulsion, with minimal changes to make shooting as easy and repetitive in matching quality as possible.

 

Vinyl has the same sort of "recorded in the red" sound as saturated tape does. It's a compression, even distortion, that a lot of people find pleasing, just as a lot of people find overexposing neg. film to be an improvement, though it's not a perfect analogy.

 

~Karl

[Chris Durham]

If you alter any more than 1/3-1/2 of a stop for your Exposure Index from the box speed, you must alter development (either push or pull) to retain the same detail in the highlights. NEVER overexpose reversal a full stop without a -1 pull at the lab. While there is some small degree of adjustment possible with telecine and/or analog printing, you aren't "burning in" more information with increased exposure.

 

Pull and Push. Thanks for the right nomenclature - I knew I knew it, just couldn't get it out. and yeah, burning in still film was kind of the analogy I had in mind with my question.

 

So, how do you know what numbers to use if you want to say overexpose 1/3 stop? Is there a formula or a table? Is it an experience thing? Is it something you test and eyeball?

[K Borowski]

My last edit of the day: (The rest of this didn't want to go into my prior post, just the first sentence made it): Back when our only option for a color negative film of any kind was Kodak's 100-speed stock, cinematographers would cheat it for two stocks by pushing the stock one stop, underexposing it two (essentially using an EI of 400, maybe 250-320) and adjusting for as much of the remaining underexposure when printing the negative onto positive film. Kodak never recommended this, and I don't think modern T-grain emulsions are as flexible to this type of process, but a lot of pushing in the past was out of necessity, whereas now, with film speeds as fast as 500, pushing tends to be done more for "look" or "aesthetic" rather than out of absolute necessity. If you're starting out, I'd recommend not doing any push- or pull-processing except out of the same necessity.

-------

To know what numbers to use? It's the reading for an exposure meter (calibrated of course, and not a $10 light meter, a good one) when you set it at the same EI as the one at which you're rating your film, and point it at an 18% gray card illuminated evenly, and without direct glare, by your "key" or main light, or evenly illuminated by the sun. It can be experience how to interpret a given meter. Some are too accurate, and f-stops are often only adjustable to half stops with certain lenses, or even out of alignment by as much as a half-stop with beat up old 16mm student lenses. My Angenieux 12-120 lens is off by about 3 stops, although it has a compensation mark that I (hope) assume is accurate. Let me answer a question you're probably going to ask later: With zoom lenses, the T-stop calibration is the one that you want to set your lens to to correspond with your meter reading. F-stops show the depth of field a lens will get when you look through the viewfinder, but they aren't as accurate for exposure purposes as T-stops. If the meter reads"F/2.8" set your lens to "T 2.8". If you look through the lens and you don't like how little depth of field "T 2.8 gives you", then adjust your lighting to compensate, as there's no set conversion factor between F/stops and T stops on all lenses, because different lenses have varying amounts of inefficiency in how much light that enters through the lens doesn't make it to the film at the other end. Just set your lens using the T-stop ring to match what the reading on the light meter is, and make sure you have your meter set in the right mode, at the right speed to match your EI, and pointed at something that is 18% grey and illuminated by your primary light source.

[Chris Durham]

I think you've just enlightened me a bit Karl. I was still a bit in the dark as to how a gray card should be used. I've been playing with digital video in which you point the camera at a white card, but I couldn't figure out the gray. So you point your light meter at the gray card, is that right?

 

If so I still don't fully understand the numbers. I get how you would set your T-stops and all based on the light meter readings (and adjust for any variables like filters); but how do I know that 320 is 1/3 stop down on 400? What is 1/3 stop down on 500? That's kind of what I meant by formula or chart. I'd guess there's a curve for that sort of thing instead of a simple calculation - and that it might vary on certain stocks. That may make it more of an experience thing.

 

Thanks for all the help.

[K Borowski]

I think you've just enlightened me a bit Karl. I was still a bit in the dark as to how a gray card should be used. I've been playing with digital video in which you point the camera at a white card, but I couldn't figure out the gray. So you point your light meter at the gray card, is that right?

 

If so I still don't fully understand the numbers. I get how you would set your T-stops and all based on the light meter readings (and adjust for any variables like filters); but how do I know that 320 is 1/3 stop down on 400? What is 1/3 stop down on 500? That's kind of what I meant by formula or chart. I'd guess there's a curve for that sort of thing instead of a simple calculation - and that it might vary on certain stocks. That may make it more of an experience thing.

 

Thanks for all the help.

 

This really is my last post; my car died and if I don't get a jump soon I won't be able to go home tonight :ph34r:

 

Sorry you've been living under a rock all that time with DV ;-) There are two different modes for your "average" light meters: incident and reflected. IIRC, incident you point the meter at your light source, with a whitish diffusion globe over the sensor, and reflected you take off of a gray card or something like caucasian skin that is "close enough" in a pinch. Most movie productions I believe use a combination of both. Your meter will probably be designed for still photography too, so don't get confused and use EV numbers or shutter speeds instead of frame rates and shutter angles you'll need for the correct reading.

 

How do you know? I got really into the numbering system behind ISO ratings, so I memorized them back when I first got interested in photography. Basically (although I'm not entirely sure below 10 if my numbers are right; rules probably weren't standardized back when films were that slow; the list tends to fall apart past 3200 too in terms of consistency) this is the list, with every three numbers being a full stop, and the two in between being 1/3 and 2/3 of a stop, respectively. IDK about half-stops. Just making up a number between the thirds on either side is the technique most of us "professionals" use ;-) :

 

0.1 0.2 0.3 0.4 0.5 0.6 0.8 1 1.2 1.5( or1.6) 2(or 2.5 or 2.2) 3 (or 3.2) 4 5 6 (or 6.4 or 6.5 or 6 1/2) 8 10 12(12.5) 16 20 25(occasionally 24) 32 40 50 64 125 160 200 250 320 400 500 640 800 1000 1250 1600 2000 2500 3200 4000 5000 6400 8000 10,000 12,500(occasionally called 12,800 or 12,000 by people who don't know better) 16,000 20,000 25,000 32,000 40,000 50,000 64,000 80,000 100,000 125,000 160,000 200,000 250,000 320,000 (or 300,000) 400,000 500,000 640,000 (or 600,000) 800,000 1,000,000 1.2(5) million 1.6 million 2 million.

 

See the pattern of 1 1.25 1.6 2.5 3.2 4 5 6.4 8 being repeated over and over again? Just don't get confused into thinking that 1 million as ten times as fast as 100,000. It's only 3 1/3 stops faster than 100 grand.

[Chris Durham]

Wow Karl. Thanks a lot. I think that clears it up for me. I mean, as much as can be in theory. Thanks for taking the time with such verbose posts. I really appreciate it. Good luck with your car.