Breaking down Lighting Ratios through mapping stop values to False Color

[Ryan Sherwood]

Hi everyone.

 

I'm very interested in being able to breakdown lighting ratios from inspiration using the same set of standards that I use when shooting (mapped IRE false color values on a production monitor). However, I'm having a bit of trouble grasping the best way to accomplish this.

 

It's clear to me that images read best on camera between 2 stops over/under key. I'm shooting in slog3 on FS7, but monitoring in rec709 LUT on my monitor.

 

I've brought up a chart from the app "LutCalc" that displays relative IRE/stop values for a certain monitor LUT- in this case going from slog3->rec709(800%). (Attached)

 

So from what I understand here are the following values for +- 3 stops from key.

 

-3 = 14%

-2 = 20%

-1 = 30%

0 = 44%

+1 = 63%

+2 = 84%

+3 = 97%

 

Does this mean that if I assign each relative stop to the corresponding IRE value in False Color using the same monitor I shoot with that I will be able to accurately breakdown lighting ratios and seamlessly recreate said ratios on set by matching colors?

 

 

 

 

[Bruce Greene]

In the "old days", I did something similar for learning exposure and lighting ratios for shooting color negative film. Using movie film in a still camera I exposed frames of a white towel from - 5 stops to + 6 stops using a light meter and printed the results on color print film. Placing the frames, in order from dark to light, on a light table I was able to visualize the effect effect of the exposure (assuming "standard" printing lights) on the resulting film print. I found this very helpful to understand what my spot meter was reading.

 

But, film processing and printing were standardized by the film manufacturer and we could only print to the gamma curve designed by the manufacturer. Basically, a zone system, pre-viz for movie film.

 

Today, there is no standard gamma for our final products. All of this can, and is changed and adjusted during computer color correction. There is no standard.

 

You are using your in-camera REC709 LUT as a standard. But, it is important to realize that there is no "Standard" REC709 LUT. Each camera is different, and many cinematographers use non-standard or customized viewing LUTs as well.

 

So, I think you might be throwing yourself down a rabbit hole here. You will do a lot of work learning this one particular camera and LUT. And the final product may be changed to look nothing like the preview (if you choose). And your work will not apply to any other camera or viewing LUT. There are just too many variables here.

 

My best advice, from my experience, is to not use "false colors" in the viewfinder and to just make preliminary judgements about lighting from a good on-set image display. If you can toggle the LUT on/off to see the full LOG image exposure, that can keep you out of trouble, but basically judge the lighting by the REC709 LUT from your camera on the viewing monitor.

 

To really learn your camera capabilities, I think it's far more useful to view your LOG exposures in a color correction software and do basic color correction on them to see how the exposure is working to give you the final result that you like. So, instead of making "false color" graphs, play more with color correction to learn your camera and exposure.

 

When I'm shooting, everyday I go home with a copy of the day's work on a little USB3 hard drive. I take the LOG footage and select shots that I'm interested in and color correct them to something that approaches what I had in mind on set. Doing this, I learn how far I can stress the camera exposure and still get what I want out of the image. It's not very time consuming as I'm not watching and correcting all the dalies, just selected takes.

 

You can even do this with selected frames in Photoshop if your more familiar with that software than something like Davinci or Speed Grade. On my MacBook Pro, with calibrated (ish) REC709 screen I usually use Speed Grade for this task as it's less demanding on the computer than Resolve. I also watch the video scopes carefully as well.

 

So, because the REC709 viewing color space is standardized, but the viewing LUT's are not, I think your attempt at a "false color" zone system is admirable, but, unfortunately, will not be very helpful in the end.

 

But, if you really want to learn a lot of stuff, keep at it until you see why it's not so easy to visualize your results this way. A lot of the best education here is by trial and error. And, maybe, who knows, you'll stumble upon a system that works really well for you. Good luck!

[Mark Kenfield]

False Colour (or the detailed versions of it at least) is a brilliant exposure tool. But it's really not ideal for ratios. I'd highly recommend picking up a lightmeter instead, I'm practically blind without mine.

 

It'll allow you to set the ratios you like regardless of the camera. As well as explore ones you wouldn't try otherwise.

 

Here's a grab from a recent project I shot, where I pushed things much further than usual:

 

 

The key light here is four stops over, but nothing's clipped and I'm quite pleased with the results. Understanding what different ratios can bring to a scene is a huge part of the job, and having a meter is really the only reliable way to know what you're lighting to.

[Ryan Emanuel]

Ryan I did exactly what you did with my first cinema camera. There is value in the the process and your thinking but Bruce is right. After diving to deep into the rabbit hole myself, I can attest that for a DP there is marginal utility to learning the deepest knowledge of gamma curves. My advice is focus on the slope of what ever curve you are using specifically in the mid tones. Thats whats most important. The higher the steepness of the slope the more contrast in the mid tones, which gives you a more poppy saturated image, but it also increases how crushed 3-4 stops under will get, for this you need more fill light. I find 2:1 to be good for really high con looks where you still want info in the blacks. If your gamma curve is less steep and closer to log in the mid tones, you'll get a flatter milkier look, which makes the difference between stops in the mid tones sometimes unnoticeable. If you have a 2:1 contrast ratio with a low contrast gamma curve, it might look like a 1:1 so I usually almost always go at least 3:1 or 4:1 for low contrast curve lighting ratios. I guess what I'm trying to explain is that for DPing, I would recommend focusing more macro for your gamma curves. Does the project require a low con, medium con, or high con look, and then have lighting ratios that you find pleasing prepared for the three cases.

[David Mullen ASC]

2:1 would be very flat, low-contrast lighting... that's having the shadows be only one-stop darker than the key!

[Jae Solina]

False Colour (or the detailed versions of it at least) is a brilliant exposure tool. But it's really not ideal for ratios. I'd highly recommend picking up a lightmeter instead, I'm practically blind without mine.

 

It'll allow you to set the ratios you like regardless of the camera. As well as explore ones you wouldn't try otherwise.

 

Here's a grab from a recent project I shot, where I pushed things much further than usual:

 

 

The key light here is four stops over, but nothing's clipped and I'm quite pleased with the results. Understanding what different ratios can bring to a scene is a huge part of the job, and having a meter is really the only reliable way to know what you're lighting to.

Mark how do you deal with ND filters and light meters? Thanks.

[Mark Kenfield]

Mark how do you deal with ND filters and light meters? Thanks.

 

 

Well I use a Sekonic L758 Cine at the moment. I'm not a big fan of the meter, but it allows you to (labouriously) dial in exposure compensation in 10% of a stop increments. Most modern meters offer similar control.

 

With an meter without any kind of exposure compensation, I'd simply calculate my 'revised' ISO by rating the camera a stop slower with every added stop of ND. So if I were shooting at 800 ISO with a 6-stop ND (ND 1.8), in my head I'd count down: 800-400-200-100-50-25-12 and so set the meter for 12 ISO. Or, if the meter doesn't go quite that low - 25 ISO with a 90 degree shutter (moving from 180 degrees to 90 degrees is equivalent to a stop of light).

Edited August 3, 2016 by Mark Kenfield

[Ryan Sherwood]

Hey Mark- I have been thinking of getting a light meter for a while...and the Sekonic L758 Cine was the one I was planning to get. Why don't you like it? Is there another, cheaper meter you'd recommend for my first one? Thank you.

[Mark Kenfield]

Hey Mark- I have been thinking of getting a light meter for a while...and the Sekonic L758 Cine was the one I was planning to get. Why don't you like it? Is there another, cheaper meter you'd recommend for my first one? Thank you.

 

 

Well I got the L-758 Cine because I wanted a combined spot and incident meter, and it was the "latest/greatest" meter that the internet was raving about when I went looking. It had all these fancy extra features listed, so I assumed that could only mean good things... unfortunately, operationally the thing is a nightmare. It must have the single most unintuitive user interface I've ever encountered on any device.

 

What that means, is that in regular use, you have to perform a literal ballet with your fingers to access and modify almost any setting. It makes operating the meter slower than it should be, and just painful in general.

 

Now it's not all-bad. Performance-wise, it's great. Both the incident and spot meters always ring true, and it works well in very low-light conditions. Certainly it's a serviceable meter, I use mine every day on set and it does it's job. I just wish it was faster and simpler to operate. The complexity of the operation means that I can only use it's most basic features though (so basically I spent a lot of extra money for features that I'll never use.

 

I started out at uni with a trusty old Sekonic L-398A analog incident meter, and some kind of ancient Gossen spot meter that the school provided. And they worked fine, though carrying two meters is a bit of a pain. In an earlier thread where we were talking meters, Satsuki recommended the Sekonic L508, which is a digital combo meter in the same vain as the L758, just with less of the fancy-features-that-are-too-impossible-to-access-anyway-so-they-don't-really-count, and should be a lot cheaper to source.

 

If you don't need a combo meter, there are tonnes of cheap and cheerful options, but I find having the spot meter on hand, very handy at times.

[Dennis Hingsberg]

For ratios I use the Sekonic Zoom Master L-508 which has great features yet isn't overly "electronic" like the L-758 which Mark has stated.