RED Komodo: Returning to ETTR!

After going back and forth, we’ve finally decided to stick with ETTR once and for all. No more false color overlays, just the histogram and traffic lights (increasing exposure until the lights turn on, then dialing back until they just turn off). As long as the traffic lights are off, there won’t be any clipping in the highlights. The advantages? For one thing, it means being able shoot at wider apertures for more pleasing shallow depth of field. It means faster operation. And perhaps most importantly, considerably less noise in the shadows (see below). In situations where the lighting is unpredictable, nothing prevents us from protecting highlights by stopping down the iris.

Below, you can see what the viewer looks like in the RED Control app. The screenshot to the left was exposed using false color video mode with skin tones placed between 61-70 IRE, while the one to the right was exposed to the right using the histogram and traffic lights.

One downside of ETTR is that it can be difficult if not impossible to see the image when giving a couple extra stops of exposure. In ‘Cinematography: Theory & Practice’ Blain Brown writes, “RAW, in this context, is an absolute scale based on the output of the sensor. It is not necessarily related to the viewing image in terms of brightness values. […] This is, of course, the basic concept of shooting RAW vs. shooting video that is more or less “ready to go.” Shooting RAW/Log is not about producing a final image, it’s about producing a “digital negative” that has great potential further down the line. The downside is that the images are not directly viewable and this makes using exposure tools like zebras and histograms pretty much useless—they can only be used as approximation.”

Below are the same two clips: false color video mode (L) and ETTR (R), showing false color exposure mode. Green represents middle gray, the absence of red indicates no clipped highlights.

Above, waveform monitor false color video mode (L), waveform ETTR (R).

Reducing exposure of the ETTR clip in post to match the clip exposed using false color video mode with the camera raw controls in DaVinci Resolve Studio 18.

The difference in noise between the two exposures is dramatic. Granted, this is a worst case scenario. Click on the screenshots below to enlarge (you may have to increase the brightness of your display to see the images).