Next feature, sci-fi, 2.39:1 delivery. The rental house offers me a 2x set and a 1.33x set for roughly the same weekly price. Camera will be a 16:9 sensor mode because that is what the post budget can handle. Which one do I actually want, and what do I lose with each? I like the oval bokeh, I don't like throwing away half my sensor.
Do the math first. 1.78 x 2 = 3.56:1 after unsqueezing, so to reach 2.39 you crop about a third of the width. You paid for those pixels and you throw them away. 1.78 x 1.33 = 2.37, which is effectively 2.39 with a hair of crop. ARRI's own guidance on its 1.3x de-squeeze mode says the same: 2.39 divided by 1.3 is about 1.84, close to the 16:9 shape of the sensor. What the 2x set gives you is a stronger anamorphic character: more oval bokeh and more pronounced flares. If you want that look and stay in 16:9, plan the frame lines for the crop from day one. Source: ARRI, ALEXA LF and anamorphic lenses white paper, https://100-assets.arri.de/arri_assets/lf_assets/downloads/en/ALEXA_LF_&_Anamorphic_Lenses_-_White_Paper.pdf
Editor note so this doesn't bite you later. Whatever squeeze you pick, put it in the camera report and in the clip metadata, and set the de-squeeze on the dailies so editorial cuts the image the audience will see. I've had a 2x show arrive as squeezed proxies with no frame lines. We spent a week cutting skinny people. Also agree with the DIT on the exact crop for 2x on 16:9 before the first shoot day, and burn the frame lines into the dailies.
Went with the 1.33x set. The lens test sold it: the bokeh is still oval, just less extreme, and I keep the full sensor for the VFX plates, which our supervisor was very happy about. Frame lines are in the dailies LUT now, the squeeze is in every camera report, and editorial confirmed the de-squeeze on the first day of rushes. Thanks both, this saved me a week of second-guessing.