Sekonic and the FS7

Via: Wil Mathijs

Hello Group, a week ago i posted a message concerning the profile calculation issues i encountered for the Sekonic L758 Cine lightmeter for the FS7. Since SONY claims the sensor can go 14 stops in native ISO 2000 i was wondering if that would be true. Basically it means we could go up and down with 7T... For Sekonic gauging you need even +/- 8T though... Struggling with the data coming from the Sekonic Exposure Target II greycard I got in contact with the lovely Mr. Oda from Sekonic Japan. He's a wonderful creature I must say. We mailed daily and got the profiles (you can download them here for FREE GRATIS Emoticón wink No 20 euro needed! See below. Further we did all kind of testing, grey and color cards, china men and china girls, live situations, ultra low key set ups etc. The tests were run by me as Director/DOP at Elementrik Films and it's partners Ace Image Factory Brussels with Tom Mulder in the Pablo grading room and Eye Lite Brussels with DOP Louis-Philippe Capelle in the studio.

INFO ON SEKONIC set up:

Key 0T = F32 (F16 on a Canon 24-70 2.8 L with MB ultra SB +1/4ND)
+4T = F8
+8T= F2
then you change the lights so you go down to a key of F2 and add corrections in the Sekonic DTS software data : in Mr. Oda's words: "Please kindly set -8EV into "Light output Correction Value(EV)"of your -4.jpg and -8.jpg,because, you adjusted the light from F32 to F2"
then we underexposed to:
-4T= F8
-8T= F32 (F16 on a Canon 24-70 2.8 L with MB ultra SB +1/4ND)

The greycard jpegs are downloadable for free below.

Findings:

1. Although Sony claims that the camera does 14T DR, we found out in gradingtests at Ace that that might be the case for RAW (which were not able to test since the ridiculous high price for that plugbox) but certainly not for the XACV I codec. Without NR (like Neat via Nuke) we could say it's feasable to handle:

Slog 2: +4T and -4T = DR 8T (and not 14T!!!) (-5T is OK but denoising requested)
Slog 3: we skipped since nobody shoots on that Emoticón smile
Slog 3 Cine : +4T (in any case better looking then Slog 2 +4T more info in the highlights thus) and -4T (with some reasonable noise to be denoised in Neat via NUKE)

In rock 'n roll / run 'n gun situations like lowkeys in doc etc, you have 8T if you don't have lighting possibilities.

2. In the china girl/greycard situations we discovered some increasing Rose tone in the skintones in +3T in Slog 3 Cine and significant more noise in -3T then we had in the same aperture in Slog 2.

3. The new version of the FS7 firmware (V3) has improved NR. When we went up from low to mid to high we found that none of the settings gave us a plastic looking video image. Very OK.

Overall conclusion:
- If you have enough light and even high light situations (like skies in the back, sunny stuff, exterior daylight etc) use Slog3 Cine. Bear in mind that overexposing with 1T is still ok in grading and avoids noise in shadows etc if there is any...
- If you are in control of the light, like indoor interviews etc, over expose with 1T in Slog 3 Cine, shadow noise will vanish in grade. use NR high anyway to deal with the dark parts noise.
- If you don't have control over the lighting, or not enough and you have to film indoor or outdoor (dawn situations, night etc) use Slog 2 and forget about overexposing since you are already on open aperture probably.
-Don't use other ISO's then the native 2000 because others crack down you're DR
-Slog2 and Slog 3Cine can be easily mixed by a good grader.

here are the files:
Jpegs Sekonic Keylight tests
https://www.sendspace.com/pro/dl/iwbuu1

Sekonic Profiles for the FS7
https://www.sendspace.com/pro/dl/35sf43

Have fun!
Grtz Wil Mathijs