Characteristic curve

[Damien Bhatti]

I have been looking at characteristic curve diagrams and I am a little confused as to how to apply the information in a practical shooting sense. Could anyone please enlighten me as to the fundamentals of it? Thanks,

[Krishna Chandar]

I have been looking at characteristic curve diagrams and I am a little confused as to how to apply the information in a practical shooting sense. Could anyone please enlighten me as to the fundamentals of it? Thanks,

 

Hi Damien:

 

Charecteristic curves depict the way emulsion responds to light intensities. The best way to compare this is to relate a scene to your eye. our eyes can distinguish and still record brightness variations with details to great ranges. Let's say if eyes can do a variation of 1-100 film may not be able to do the same. C.Curves tell us its efficiency range called latitude (a property within which it can accomodate the brightness variations or intensities with best comparable results) and how emulsions can respond to lighting variations. In order to get the best results close to what our eyes can see, film and lighting levels should act in cohesion. As a photographer, understanding what a film could do and controling lighting levels accordingly is one of the great techno aesthetic factors. In short it is a scale showing how much of response a film has to the lighting levels. I hope this helps and I am sure other experienced peers have more to say.

 

Krishna Arunachalam

[Damien Bhatti]

Hi Damien:

 

Charecteristic curves depict the way emulsion responds to light intensities. The best way to compare this is to relate a scene to your eye. our eyes can distinguish and still record brightness variations with details to great ranges. Let's say if eyes can do a variation of 1-100 film may not be able to do the same. C.Curves tell us its efficiency range called latitude (a property within which it can accomodate the brightness variations or intensities with best comparable results) and how emulsions can respond to lighting variations. In order to get the best results close to what our eyes can see, film and lighting levels should act in cohesion. As a photographer, understanding what a film could do and controling lighting levels accordingly is one of the great techno aesthetic factors. In short it is a scale showing how much of response a film has to the lighting levels. I hope this helps and I am sure other experienced peers have more to say.

 

Krishna Arunachalam

 

thanks Krishna, so would I be right in assuming that if you wanted to know how a specific filmstock would react in a given situation you would consult it's relevant curve diagram. Do all films come with an individually worked out scale? Or do we talk about 'latitude,' in a general sense for any given stock, or rather is the characteristic curve something that is quite definite in making a decision about what film to use.

 

Thanks again,

[Chris Keth]

It's like any other graph. The scale is given on the side and bottom. On the side is always density where the zero point is a density referred to as base+fog. It's exactly what it sounds like, the amount of light that doesn't make it through the film base and the film fog caused by ambient radiation. Along the bottom is LOG exposure. The logarithm is just a convenient way to make an increase of one log exposure equivalent to double the amount of light. Yep, one log exposure equates to one stop.

 

With this information you can visually size up gamma (slope of the straight line portion of the curve), knee (top of the curve), toe (bottom of the curve), and lattitude (length of the curve from lowest x value to highest x value) information. For the knee and toe portions, the severity of the bend will tell you how quickly or gradually the film will go to black or white.

[Damien Bhatti]

It's like any other graph. The scale is given on the side and bottom. On the side is always density where the zero point is a density referred to as base+fog. It's exactly what it sounds like, the amount of light that doesn't make it through the film base and the film fog caused by ambient radiation. Along the bottom is LOG exposure. The logarithm is just a convenient way to make an increase of one log exposure equivalent to double the amount of light. Yep, one log exposure equates to one stop.

 

With this information you can visually size up gamma (slope of the straight line portion of the curve), knee (top of the curve), toe (bottom of the curve), and lattitude (length of the curve from lowest x value to highest x value) information. For the knee and toe portions, the severity of the bend will tell you how quickly or gradually the film will go to black or white.

 

again thanks for the input, I am starting to get hold of it now,

[Patrick Neary]

I have been looking at characteristic curve diagrams and I am a little confused as to how to apply the information in a practical shooting sense. Could anyone please enlighten me as to the fundamentals of it? Thanks,

 

Hi- I don't think it's just me, but I find that curves have little to no practical value when it comes to actual shooting. Load up a camera, shoot some stock, have your lab process and print it, then look at it, that's practical.

[Chris Keth]

Hi- I don't think it's just me, but I find that curves have little to no practical value when it comes to actual shooting. Load up a camera, shoot some stock, have your lab process and print it, then look at it, that's practical.

 

Curves don't replace testing. They give you somewhere to start when you test so you're not trying every possibility under the sun.

 

They're plenty useful when you know exactly how to read them. I could read a curve and go out and soot something with that stock, with that EI and be plenty comfortable with the outcome. The trick is that we rarely shoot a stock under the exact conditions of the test.

 

For that reason I did a few tests or a class where I make a curve for one stock exposed and processed a variety of different ways. That was extremely helpful. It lets me see a concrete numbers representation of how the gamme, toe, knee, and lattitude change under different conditions.

Edited May 13, 2007 by Chris Keth

[Patrick Neary]

I hope I didn't give the impression that I'm advocating ignorance of characteristic curves. I think it's a given that a working dp should know what they are looking at. Often times it's the first point of reference for a new stock, especially now that both Kodak and Fuji offer "camera-stop" curves, which are a far more useful reference (at least for the math-impaired, like me) than the older ones.

 

Basic testing of any filmstock isn't about "trying every possibility under the sun." I think that's obvious enough I don't really need to elaborate.

 

What I was trying to point out, and what for me is important to understand about (camera negative) curves, is that unless your finished exhibition is a strip of original camera negative held up to the window, the characteristic curves aren't all that useful (again, only speaking for myself) in making artistic or creative decisions, any more than the RMS granularity or spectral density graphs. And quite honestly, lining up the curves of our currently available stocks, I'm hard pressed to see glaring differences among any of them. If one suddenly showed up with a curve that looked like the slopes of the Matterhorn, or the floor of the Gobi Desert, I might sit up take notice.

 

The problem is that if a shooter relies on curves alone, that person is ignoring the rest of the cinematographic process, namely printing or telecine/scanning and whatever other post-production variables await the camera neg. Not to mention the variations in processing time and temperature, shooting, storage, etc. that might throw the manufacturer's curves off even a little bit. or a lot.

 

Especially when you start playing around in the nether regions of the stock's toe, you can't expose by curves alone, you have to know (again, by testing, or have a pretty good guess based on experience) where things are going to fall off in the final print or at telecine/scanning, because that's what we're shooting for, what Steven Poster (ripped from the Emulsion Testing chapter of the ASC manual) calls the "chain of events that result in the presentation of images that we create during production."

 

Hope that's clear, I realize my first post was pretty abrupt.

[Brian Pritchard]

It's like any other graph. The scale is given on the side and bottom. On the side is always density where the zero point is a density referred to as base+fog. It's exactly what it sounds like, the amount of light that doesn't make it through the film base and the film fog caused by ambient radiation. Along the bottom is LOG exposure. The logarithm is just a convenient way to make an increase of one log exposure equivalent to double the amount of light. Yep, one log exposure equates to one stop.

 

With this information you can visually size up gamma (slope of the straight line portion of the curve), knee (top of the curve), toe (bottom of the curve), and lattitude (length of the curve from lowest x value to highest x value) information. For the knee and toe portions, the severity of the bend will tell you how quickly or gradually the film will go to black or white.

 

Actually 0.3 Log Exposure = doubling or halving of exposure which = 1 stop. (The log of 2 is 0.3010)

 

Brian

[Damien Bhatti]

Hope that's clear, I realize my first post was pretty abrupt.

 

 

thanks for that, I really wondering whether these graphs/stats/diagrams are of paramount importance when thinking about stock. But now I guess that its important to know of these things but work with a particular stock in a situation and having the results back is the basic way of knowing how it looks, right?

[Patrick Neary]

They're of paramount importance to some and not to others. I'm sure there are plenty of cinematographers and photographers who produce stunning work without ever having glanced at a characteristic curve. But it doesn't hurt to know what it is and understand what it can and can't tell you.

 

Shooting tests is more fun than looking at graphs anyway... :)

 

It would be interesting to hear from anybody else about how they use (or ignore) camera negative data...