Hello everyone, thank you for reading.  In this installment I will be examining the similarities and differences of the F3 and RED Scarlet.  As an owner of a Sony F3 with sLog and a RED Scarlet, i am lucky in that i am in the positon to do a direct comparison.  These two cameras cost roughly the same, and both are targeted at the mid range feature, commercial and corporate video worlds.

After doing this test, i found a few interesting things.  First let me talk about what you will be looking at.  The same lenses were used for both tests.  The Skin tone test is with an 18mm Cooke iPanchro, and the CU is with a 50mm.  The Still life is with a 50mm iPanchro.

The first portion of the test is a skin tone test, and the two images were brought to “neutral” to my eye.  The F3’s Footage was graded in Apple Color, and the RED footage in RCXpro transcoded to 4k PRORES, then down scaled to 1080 for the export.  

The second portion is a Still life that has a wide dynamic range.  Roughly 12 stops spot metered.  The Shadow corner of the Playstation controller being the low point, and the upper left of the cell phones reflection of the light source being the high point.  The subject matter is specifically chosen.  The metallic greens, copper, brass and browns of the rams head and pocket knife are very difficult to reproduce.  Also the reflection on the cell phone is the real test of highlight retention.  The words “Motorola” should visible in the reflection.  White letter, under glass reflecting the source of light. A very subtle, but difficult highlight to render.

The F3 footage is “De-LOGed”.  Meaning, my metered grey point is brought to its proper place around 55IRE, from 40IRE, where it should be exposed in LOG, and the black portions of the signal hit 0IRE, and the highlights hit 100 IRE. When the slate reads “no clip” it means that i used false color to expose so that no portion of the brightest part of the scene is clipping. The same was done with the RED.  This is to see highlight rendition vs Shadow detail.

 The RED footage is exposed a bit more directly using Redgamma3 and Redcolor3.  I did do a simple grade in RCX, to make sure black hits 0IRE and the highlights hit 100 IRE.

So please, enjoy the Comparison Video.

The two cameras essentially pick up where the other leaves off.  The RED is sharper, but has more noise and less sensitivity.  However it records 16bit RAW 4k, even in its smallest configuration, a lens, a battery and a monitor.  The F3 despite its clean signal, infinite dynamic range, and 444 output, it needs an external recorder to take advantage of it.  Even when you do record to a cinedeck in 444, its still 1080p.  Less than the standard of 2k for digital projection.  

The usual workflow on F3 is as follows, Camera outputting sLog to Pix240 or Cinedeck.  I then feed the signal to the DIT who applies a LUT to the sLOG and then out puts the image to the video village, director and the back to a camera department monitor.   I set a base exposure, then the DIT applies the lut.  Based on each different scene, we then tweak the lighting to the LUT onset, and keep the LUT as a reference for the grade.

The workflow I’ve been using on the RED is a bit different.  Since the recorded image is RAW, on set LUT’s are slightly less important as it can be determined later on.  We simply feed the image to DIT and have him give a light base grade to output to the monitors on set.  I use the RED tools to measure exposure; false color and histogram.

Conclusion:

The biggest problem i have is deciding between the two cameras.  The RED is simply razor sharp, has a punchy slick look, and has the easiest onset ergonomics and workflow.  The footage is 16bit 4K RAW.  I mean, the camera is the size of a Hassleblad and gives you amazing images.  

The F3 on the other hand, is night vision, which makes for simpler indie shooting on lower budgets where you dont have access to bigger units.  It also has more useable dynamic rage, and its color is in my opinion the best in digital right now short of the Alexa. ( Though REDColor3 is REALLLLLY good!) Case in point, in the skin tone test you just watched, the color of the wall in the F3 footage is completely accurate.  Thats exactly what my wall looks like, but looks quite magenta in the RED footage. If I fight the magenta with a bit of green, the skin tone gets thrown out of whack.  This green/magenta balancing act is a definite pitfall of the RED, even with REDColor3.

This real issue is that it comes down to image quality.  I find myself, shooting films and documentary on the F3, but commercials, music videos, and corporate spots on the RED.  When it needs to look natural, i use the F3.  When it needs to look crisp and slick, RED.

Until next time,
-T

Feel free to follow me:
timurcivan @ twitter