ISO Comparison GH2 vs. G6 vs. GH3

As long as I still have the three beauties from Panasonic at home, I thought about doing a comparison to decide which one I’ll keep, and because some others may ask themselves a similar question: Here it is, the first excerpt of my comparison experiments:

All Shots are done with the Panasonic 20mm f1.7 at f3.2 with shutter speed set to1/25th at AVCHD 24p, Standard picture profile at 0 0 0 0 and manual WB 10000K.

The GH2 is hacked with Driftwood’s T7 Moon high bitrate hack. And because Nick Driftwood recommends it for 24p video, the GH2 uses i.Dynamic at LOW.
G6 and GH3 have intelligent settings disabled.

I found it very hard to match the cameras via their color profiles so I decided to use STD 0000 on all of them.

For post production in After Effects and Premiere, I decided to convert all shots to DNxHD first.
Sorry for the poor white balance, but I just had one LED spot that was dimmable and no proper video light. I then decided to leave it as it is and did not apply any color grading in post.

I’m going to publish a comparison at ISO 800 with some more individually optimized settings on each camera and – as far as my skills go – some optimization in post.

 

Panasonic G6 – noise and grain

Compared to the hacked GH2 I almost fell off my chair looking at the footage of my newly bought Panasonic G6 because there was heavy grain.

I figured out that setting i.Dynamic to “high”, like I knew it from the hacked GH2, caused most of the noise. Turning this off increased the image quality a lot.

Depending on how deep you’d like to digg in post production, you can set noise reduction to 0 or -1 and use some of the in camera noise reduction. I also encountered good results with NR set to -3 and optional post processing.

Currently using STD -3 -3 -1 -3 with Panasonic lenses. STD 0 0 0 0 works well, too.
You can see a comparison of the two settings here:

A more detailed video on the G6 vs. GH2 (hacked) vs. GH3 will follow.

What test setups do you use and how do you measure exposure while testing different settings?
I find the in-camera exposure metering a little uncomfortable. Maybe I’ll try Zebra as orientation so I get the most light without overexposing.