How do negative densities translate into positive?

[Gabriel Cortez]

I'll try to be as simple and clear as possible about my question, thus I'll give you an example:

 

So, the negative (say BW 7222, what I'm using now) is low contrast, gamma 0.65; I shoot a night interior with a man dressed in a dark jacket, for instance, then I go at the lab and inspect the neg and see the details in that dark jacket, clearly. Okay so far.

THEN, I go to watch the print, in the projection room, and the jacket in question is jet black! Why?

Normal processing on the positive, normal gamma there. Oh, and dead on exposure, no doubt there.

So these are the facts.

 

Now, getting down on the issue, I seek an explanation to the following:

 

- why does the positive have such a high gamma compared to the neg?

 

- what use is to me to have details on the neg if they are clipped on the final print, due to the print's contrast? how do you get around that?

 

- how is the density (tonal) range in a negative translated into the positive (maybe you could refer to the curves, and help with some sensitometric relations)? I can't see how there can be no loss of information in the process.

 

- everyone seems to overlook the issues regarding the positive, and concentrates on the neg (how to expose it, how to get details etc), and of course there's a good reason to that, there's the information in the first place!, but it's actually the final print that matters; I mean even Ansel Adams said that the negative is just the original score, the music comes in the print (now I know ansel darkroom practices don't apply to cinema, I was just trying to stress the importance of the positive).

 

 

I've never thought about this until now, taking it to the basics; and i've shot a couple of shorts on film, going the usual way (greaycard exposure tests, lighting tests and all), not approaching it sensitometrically-wise.

 

 

Please comment on these things.

 

Many thanks!

[David Mullen ASC]

Because a print is being projected onto a white surface (the screen) with a bright light blasting through it, the blacks have to be dense enough on the print to end up looking black on the screen. So the gamma of the print stock has to be high, otherwise the image will be washed out.

 

Negative is only half of a photographic process -- if you are shooting for print, you have to think of it as a neg/pos system.

 

If you had less information on the neg, you'd have even less information on the final print.

 

You can't really make a neg/pos system where the gamma of the final projected image is the same as on the original neg -- it would look horribly milky. For example, reversal film has the gamma of a projected image (neg + pos gamma) more or less, since it was designed for direct projection of the original -- which is why the contrast is so high compared to negative film.

 

So when you shoot, you have to take into account what information will be visible on the print, which is why you shoot a simple over and under exposure and lighting contrast ratio test and make a print before you begin a project for print, so you know when things fall off to black or burn out to white.

 

Another advantage to the fact that you have more exposure information on the negative is that you can print the image brighter or darker.

 

Most of what I shoot has been for print, so I always think of it as a two-step system, but most people these days shoot for transfer to video from the negative, or for scanning.

[Chayse Irvin ASC, CSC]

If he did want a low contrast print would it be as simple as adjusting the printer lights to as 20-20-20?

[David Mullen ASC]

If he did want a low contrast print would it be as simple as adjusting the printer lights to as 20-20-20?

 

That would make everything brighter overall, which is not quite the same thing.

 

You could process the b&w print to a lower gamma but then it might not look as rich; besides, print stocks are not really designed to have their contrast manipulated by processing variations, but it probably works better in b&w than with the color stocks.

 

Assuming you couldn't alter your style of shooting (by creating a lower-contrast negative) then you could do things like slightly flash the print, which would lower the contrast by darkening the whites (but now you have dingy whites) and then you could compensate by printing a little lighter for more shadow detail.

 

You really don't want to mess with the gamma of a print stock, not if you want correct whites and blacks on the screen -- you want to shoot your negative with the gamma of the print stock in mind. Just because there is more information on the negative doesn't mean that something is wrong because some of it drops out in printing -- it's designed to do that. You have to have a gamma of a certain level for a good screen image, so the gamma of the neg and pos are designed to add up to the correct gamma.

[Gabriel Cortez]

you want to shoot your negative with the gamma of the print stock in mind. Just because there is more information on the negative doesn't mean that something is wrong because some of it drops out in printing -- it's designed to do that.

 

 

That's the answer I was looking for, thank you very much Mr. Mullen!

 

I've never thought about the unused information on the negative until now, and your confirmation of that fact puts the "film stock latitude" issue in a "new light" for me now.

 

And let me say that "shoot your negative with the gamma of the print stock in mind" is one of the best advices I've ever been given. It's so true!

 

It's a shame they don't tell you that in film school these days (at least at my film school).

 

 

Many thanks indeed!

[Chayse Irvin ASC, CSC]

That's the answer I was looking for, thank you very much Mr. Mullen!

 

I've never thought about the unused information on the negative until now, and your confirmation of that fact puts the "film stock latitude" issue in a "new light" for me now.

 

And let me say that "shoot your negative with the gamma of the print stock in mind" is one of the best advices I've ever been given. It's so true!

 

It's a shame they don't tell you that in film school these days (at least at my film school).

Many thanks indeed!

 

 

Same. I dont remember learning this. But there is a learning curve that didnt hit me till I started shooting more and more. I doubt i would have remembered if it was even mentioned durring my school... I was really focusing on learning the terms. Thats whats so great about this forum... you can ask any question and get a great answer from so many great DP's.

[John Pytlak RIP]

Kodak has good tutorials on its website:

 

http://www.kodak.com/US/en/motion/support/h1/structure.shtml

 

http://www.kodak.com/US/en/motion/support/h1/

[David Mullen ASC]

You're back, John! I was worried about you after I didn't receive a reply to an email at Christmas... Hope you are feeling better.

[K Borowski]

John, glad you're back. There've been several technical questions pertaining to film that were over my head since you've been gone; you may want to see to the ones we all missed :-)

 

Sorry to hear you're off solid food. That must be tough. I wish you all the best and hope that the chemo continues to go well for you. Happy 2007 and Merry Christmas to you too.

 

So are you working out of the home den still or planning on going back into work soon?

 

Regards,

 

~Karl Borowski

[Dominic Case]

It's a shame they don't tell you that in film school these days

I've designed and taught quite lengthy courses just on sensitometry. However many hours there are in the program, there's never time to do more than skim over the basics. That's true of any course I've taken as well - and that's why reading lists are provided.

 

But I agree that this topic is prety fundamental, and should be explained thoroughly. It's not hard.

 

GAMMA:

 

To reproduce a scene on the screen, you need to see an overall gamma of 1.0 - that simply means that the contrast scale is neither increased nor decreased from the original - it's a lifelike reproduction.

 

Now due to (a) scattered light etc in the theatre which reduces image contrast and (b) a general subjective preference for things to be a bit more contrasty than real life, it turns out that a final print gamma of around 1.4-1.6 is the most satisfactory. That number is the product of the gamma of every stage along the way. B/W print stock is processed to a gamma of 2.5 - sometimes a little less, down to 2.1, depends on the lab. B/W neg is 0.65. No need to get out your calculator, the overall result is 0.65 x 2.5 = 1.6.

 

So that gives you the right contrast in the midtones.

 

BRIGHTNESS RANGE:

 

The human eye can respond to an astonishing range of brightnesses - at least a million to one - that's 20 stops. But not all at the same time (and that's important.). In fact, a given spot in the eye at a given moment only responds to about 100:1. Apart from iris changes, there are rapid chemical changes on the retina that alter sensitiivity.

 

Go back to your scene of a man in a dark jacket. There is probably a bright spot of light somewhere in the scene. If you look at that, you can't see the detail in the jacket. If you shade your eyes from the light, you can see the dark jacket details better. If the light is set quite apart from the dark area, you can probably see the jacket details anyway. But the point is, that as your eye "reads" the view in front of you, it is constantly making local adjustments to best examine the range of tones you are interested in.

 

The negative can't do that. It has to capture as wide a range of tones as it possibly can, so that you can "read" the image selectively later on.

 

The negative in fact can capture a range of about 2,000:1 quite comfortably, that's 11 stops (and rising, as Kodak bring out new emulsions!). That's a pretty handy range. It records that wide range by having a low gamma and extended toe and shoulder regions where it gets much lower still (so that the recorded image is only about 50:1, the tones are just shoved closer together).

 

Print film has to stretch that scale out again so the image doesn't look flat and muddy. Hence the higher gamma of print stock. But there's one final obstacle: the contrast range of the projected image.

 

Regardless of the gamma of the film image, or (in the digital world) of the claims of projector manufacturers, you can't have a scene with a brightness range of more than a few hundred to one. The light from the highlights of the screen image bounces around the auditorium and back onto the screen, where it pollutes the dark shadow areas.

 

Since the neg carries more tones than that, you have to make a selection from the range when making the print. That's where grading or timing comes in. It's much the same as the unconscious selection you make when viewing the original scene (see above), but it's locked into the print.

 

IN SHORT:

 

Psychology, physiology, optics, photochemistry and physics all combine to reduce the range of real-scene tones we can reproduce on the screen.

 

But - the bottom line - if your neg really IS well exposed, and you can see detail in the neg, then almost certainly you could grade the print lighter to reveal that detail while retaining good blacks in the really deep shadows. But if you have highlight areas, expect them to blow.

[Gabriel Cortez]

Thank you Mr. Pytlak for the links, and many thanks to you Mr. Case for your extensive reply.

 

"Since the neg carries more tones than that, you have to make a selection from the range when making the print. That's where grading or timing comes in."

 

That's what I was after, thanks.