Episode One of the Cinema Camera Mega Test gave us the opportunity to see how the #RedCine Epic, #SonyF5, #SonyFs7, #SonyFs5, #URSAMINI4.6K, and Canon #C300MK1 performed in our over and under exposure stress test. If you haven’t seen it yet, go watch the results in 4K here.
In this episode, colorist Eric McClain, of Digital Pix Post, details the color grading process that was used for this test and discusses the strengths and weaknesses of each camera.
RED EPIC DRAGON
Of all the cameras tested, the RED Epic proved to be the easiest to work with. The image right out of the camera rendered a well-balanced image. Hardly any work needed to be done to produce a neutral image from which to start the grade. The Epic performed very well in our overexposure test. For the most part, recovering the image to neutral from an over exposed capture was simply a matter of readjusting the exposure compensation in RED RAW tools. There were no adjustments required to the color or gamma settings.
Underexposure was a bit more of a challenge. As with most cameras, there was a fair about of noise in the underexposed images. At -2 to -3 stops, it could be argued that the noise was acceptable, maybe even pleasing. Starting at about 4 stops, we found the noise to be far less acceptable. It suffered from harsh coarseness and poor color response.
The F5 was captured in S-GAMUT/SLOG-3. In Divinci Resolve, a REC 709 Input transfer was applied to render a neutral starting image. Much like Epic, The F5 produced a very well balanced image right out of the camera. In our over exposure test, the F5 seem to thrive from being over exposed and pulled back down. It easily retained highlight detail, even at +4 stops. When the overexposed image was pulled back into proper exposure, much the shadow noise was eliminated. This produced a cleaner image than the properly exposed shots. It was necessary to make a few color adjustments to bring each over exposed shot back to neutral, but it was very light work.
Underexposure also proved to be an area where the F5 performed at the top of the group. At as much as -4 stops, the F5 continued to render a fairly usable image. With a little noise reduction, -4 stop pulled up to proper exposure could be quite acceptable. That said, the underexposed F5 had much more difficulty reproducing the color and gamma of the properly exposed image.
If you’re looking for a good companion or alternative to the F5, the Fs7 proves to be an excellent choice. In our stress test, it performed almost identical to the F5. The differences presented themselves at the extremes. At +3 and +4 stops, the Fs7 isn’t able to retain the highlight detail and roll-off was well as the F5. In our under exposure test, the Fs7 at -4 stops produced far more noise and unacceptable color artifacts.
URSA Mini 4.6k
We chose to shoot RAW with the URSA Mini. Using the RAW tools to transform the image to REC 709, produced an image with a strong blue bias in the shadows and mid-tones. (It should be noted that this test was conducted before the Black Shading firmware update. Maybe that would have help to produce a more neutral image) With some light color and Gamma adjustments, the URSA Mini produced very pleasing images. Of all the cameras tested, the URSA Mini rendered the sharpest image, without any sense of harshness. Much like the Sony F5 and Fs7, it seemed to thrive when overexposed, with any visible noise being eliminated when pulled back down to proper exposure. Both the F5 and the URSA Mini were strong performers at -4 stops.
Canon C300 MK1
The Canon aesthetic has long been a favorite. This stress test proved that Canons also produce a very a robust signal. Our test footage was captured at 1920x1080p 23.98 (all other cameras were 4K) with REC709 baked in. The put the C300 MK1 at would should have been a significant disadvantage. In reality, the upscaled C300 footage looked just as good as the native 4k from the Sony cameras. In general, the C300 had a slight green bias, but not nearly as pronounced as the blue bias of the URSA Mini or even the slight yellow bias of the Epic. At -4 stops the upscaled C300 performed about as well as the F5, in terms of noise and color artifacts. While this test is not the most comprehensive look at the image scaling capabilities of the C300, It does show that the canon workhorse can hold it’s own in a 4k workflow.
The big disappointment of this test was the overall performance of the Fs5 internal 4K. In HD, this camera shows great results. In 4k, It would seem that the combination of the 4:2:0 color sampling, 8-bit data, and log-GOP compression put it at too much of a disadvantage. The Fs5 required the most manipulation is Resolve to render a neutral and balanced image. Like all the cameras in this test, the Fs5 benefitted from over exposure in terms of noise reduction when pulled back down to proper exposure. Of all the cameras, it required the most re-balancing of color and gamma to return to a neutral image. At -2 stops it was possible to recover a fairly clean image, but the challenge was returning to neutral color balance. More the 3 stops of underexposure proved not to be recoverable.
In the end, the choice is up to you and the needs of your production. This test shows that not only are all of these cameras capable performers at overexposure and a fair amount of underexposure but with just a little effort in Resolve, you could easily mix most of these rigs in a multi-camera production.