ETTR: A Cinematographer’s Perspective

We still ETTR with the Komodo, but not nearly as much as we used to do in our early days with the Sony a7s III. For years, there’s never been a clip that we didn’t expose to avoid important shadow detail falling in the noise floor, and then some! Interestingly enough, some shots with the a7s III where skin tones had what we believed were unrecoverable highlights a couple years back turned out to be just fine once we began to understand color management a little better. But ETTR is not all that common on high end productions. Here’s John Brawley explaining why:

“ETTR is not the best technique if you want consistent results. It’s really a stills technique…. It works differently from stills than it does in moving images.

ETTR is really way of trying to maximise the DR you have and capture as much of it as you can….BUT…

You have to be able to know what’s clipping in your scene…It’s a top down approach…you’re basically exposing for the highlights, trying not to let them clip….and letting everything fall underneath that….

If done correctly, it optimises the DR for each given shot by keeping the recorded information from the noise floor, but there’s a catch…

What if the DR changes in the middle of you shot? What if the sun pops out? What if you start looking out a window on an interior and then pan the camera off the window…..Moving pictures are dynamic…the lighting CHANGES as the shot changes by virtue of the weather or camera moving or character dynamics…

These kinds of scenarios where you ETTR based on a given shot, doesn’t really take into account what can happen when shoot MOVING pictures that then have to cut and grade together seamlessly…

It can lead to very uneven and inconsistently exposed pictures shot to shot… Yes, you may have optimised the DR for each shot but you now have a hot mess of differently exposed shots to try and grade and match…

Consider a day wide interior of a person at a desk with a window just off to the side behind them.

In the wide shot with a window in the shot, ETTR would have you try to hold the highlight detail out the window, and your person inside may be a few stops under that exposure…

Time for a close up. Without changing lighting*, you frame up the shot and there’s now no window in the back of the shot anymore. ETTR would have you open this shot up a few stops because you’re no longer holding the highlights of what you see out the window, you want to optimise your DR by ETTR.

If you try to grade those two shots to intercut you’ll likely really struggle to make them match.

Think about it….you’re changing where the noise floor is on each shot by using ETTR…now each shot has a DIFFERENT noise pattern…each shot will have a potentially different noise structure to it….so each shot will draw attention to itself as the cutting pattern flows…some shots are noisier, some are perfectly clean….

ETTR has limited use in video, and you have to really understand what its good for and when to use it….keeping the noise floor as low as possible and thus maximising your DR for whatever your record file can contain…

But as I mentioned this typically means your exposures will rollercoaster and you’re going to really struggle to pull the shots together to actually LOOK consistent if you’re doing any kind of drama pattern editing / delivery. This is then further compounded by image compression and especially bit depth limitations of the sensor itself and the recording format….

Often we try to hold onto information that’s NEARLY clipped, and I’ve found even the DR king Alexa can show information in highlights that isn’t clipped, but if you expose someone’s skin tone that normally looks great at say 35% but you place it up at 95 % instead…I mean yeah it’s not clipped and you’ve ETTR’d like a beast, but if you try to bring that 95% skin tone back to 35% in the grade it looks thin and brittle and not very pleasant at all compared to if you’d just left it at 35% in the first place…

It’s flexible, but it’s not magic….

It’s truly best to use the native ISO of the camera were the DR is typically optimised anyway, and then try to adjust your scene contrast through lighting and other control techniques to fit within that, rather than the gymnastics of changing your exposure for each shot. I try to find an optimum exposure for the whole scene and stick to the same exposure for all the shots…it gets me a very consistent noise floor and look and the shots grade much more easily.

JB

*Of course you can try to change the lighting to reduce and to better balance the difference but again, you’re still all over the place with exposure and you’re CHANGING the lighting so it will also be harder to match shot to shot.”