CINE D. All Settings at "0". CC and Grading in Premiere Pro CC 2014

@maddog15 Damn - there's some really nice shots there. Thanks for posting. I'm curious to see how this camera does with narrative drama scenes of close-ups and mediums with actors in frame, but these shots here definitely pique my interest. Very interesting.

@Matthew Great video. Can you share with us your settings, equip (lenses, etc), and post workflow software(s)? Very interested!

@Matthew Love the video and the music. Great job!

@maddog15 That's interesting re 25mm Lumix as that lens has kinda been on my mind and I wondered how it played with GH2 and GH4. I'd love to see some of those close-up shots if you've got time to post. Also like the idea of more latitude for push and pull in post. Great looking stuff you're doing - thanks for your thoughts on GH4.

Thank you all for the nice comments - really appreciate it! @revitdazio, as far as all of that is concerned, I shot the video with only one GH4 and two lenses (SLR Magic CINE 12mm T/1.6 and SLR Magic CINE 25mm T/0.95). I used the updated Light Craft Workshop Variable ND Mark II filter (way better than the older version) to help control the harsh outdoor lighting (client wanted shallow close-ups), and I shot the entire video on a rented Zacuto Stinger + Z-Finder Pro rig (looking into buying the Tilta III rig for my personal equipment, however). The GH4 settings were constant throughout the entire shoot: Cine-D with -5 contrast, -5 sharpness, -5 noise reduction, -5 saturation, and 0 Hue. The highlight and shadow curve was set to 0/0, because (for me) boosting the shadow curve any further with an already -5 Cine-D contrast introduced way too much noise (nice to hear that @shian is getting better results, though). I had my zebras set at 100% instead of 95% because I noticed that with the GH4 there is already a little extra room before highlights are fully blown out. The footage was cut and stabilized (when needed) before being exported to a 1920x804 ProRes 444 file (perfect for going out to DaVinci Resolve LITE 11, as Cinema 4K is only in the full version) and from there I had quite a lot of room to work in post. I utilized a lot of power windows on the balloons and his red jacket to help retain the deep/bright reds without them turning orange from some of the LUT's I used. My Resolve workflow consisted of adding a Kodak 2393 LUT on the final node (read up on node structure from Juan Melara here: http://juanmelara.com.au/basic-resolve-node-structure-and-order-of-operations/) and adding a Rec.709 Kodak Elite Chrome 200 LUT from VisionColor's new Impulz LUT's on the node prior. From there, the node structure varied with exposure/color adjustments (I had my wheels set to log and my highs at 0.333 to better target the highlights before adjusting - again, great suggestion from Juan Melara). That was about it - I didn't take any time to de-noise the footage, but for the most part it was okay (a little noisy considering it was primarily shot at a base ISO of 200, but I was pushing the footage quite a bit). Any slow-mo footage was shot at 1080p in 60 FPS - I didn't need 96 FPS and it wasn't worth the tradeoff in quality loss - and the final shot in the video was 4K 30 FPS, conformed to 24.000, just to add some subtle slow motion (the entire video was shot at a 180 degree shutter, by the way). Hope this helps you/anyone else who is interested. Sorry for the lengthy post - just trying to be detailed!

So I put together another round of test. This time comparing the Ninja Star 10Bit Recording to the GH4 8Bit recording. I shot the three profiles I liked from my last test, (Cine-D Regular, Dragon, and Film Convert Low Contrast), and made a vid for each one. There's no doubt that the GH4 internal recording at 4K looks great, but the Ninja Star can record a better range of color and light (though you have to be prepared as it may not match your histogram and be over exposed).

It was a foggy overcast day, so keep that in mind as you watch.

The GH4 internal footage just looks a little murky in comparison to be honest. There is a little more write up about the footage in the descriptions.

Cine-D (which ended up a little over exposed)

Dragon

Film Convert Low Contrast

And I also shot a Slow Motion Test with the three profiles. This one was hard to judge myself, at times 48fps over crank looks worse than 96fps does, but I do believe that 60p slowed in post came out looking the best, especially with retaining the color information.

@zachasurp you listed many different profiles. In what cases are you using each? Just curious if you've found some of them particularly helpful in certain situations.

Dragon settings and no matter what the light conditions are using just ISo 1600 and a fixed white balance? how can this make sense? anyone experienced with these settings and other ISO and kelvin?

Last weekend, I shot this live performance with of a band with the GH4 and Olympus 12mm. For the first time I used my Came 7000 3-Axis brushless gimbal (still some fine tuning needed, visible movment). In this condition I tried some low light settings instead of flat cine like D. I used Cinelike V, Contrast +0, Sharpness -5, NR -5, Saturation -5 and ISO 1000. The original 4K file was way crispier and better looking than I thought

When people make blanket statements along the lines of "high ISO is not good" or this camera "cannot be used in low light", shouldn't these opinions at least be qualifgied with what type of lens you're shooting with?

Telling us that ISO 1600 is noisy, is useless. Instead, tell us ISO 1600 is noisy with Pan 12-35 f2.8 or whatever. At least give us some aperture values. Cinema lenses and speedboosters should yield better results.

@Mistas @AdamT is right. Lenses should not determine noise patterns. What they allow is for you to shoot at lower ISO due to better aperture values. But I can tell you that I had the same noisy results with the Voigtlanders at f 0.95 and the Panasonic 12-35mm. Also with Zeiss Master Primes the result was the same. :-) Even in well lit scenes and stopping down the lenses keeping the 1600 the sensor is really noisy. So I can say that it is a rather noisy sensor. There is a compromise between ISO/Noise that can be achieved but the lower the ISO the less noise I get.

My post presumed the same exposure.