Ok, Natural -2, -5, 0, -5, 50p MOV

natural, -5, -5, -2, -5, 25p, iso1600, ss25, f1.7, a touch contrast-saturation in post, neatvideo. 20mm pancake probably needs coated filter, it is not ca round the light, the lens is spherical.

@starios That's a good question my friend. The wedding I recently shot got me thinking the same thing. I shot most of the footage at 60fps so I could get beautiful slow motion in post by slowing the footage to 50% speed. The 60fps and the 24fps raw footage is.... well.... identical in quality. I'm sure there's some math to explain this right or wrong but it boils down to this... My eyes (I am a pixel peeper) would swear on the life of my GH3 that there's 0% noticeable difference. Maybe I'm and idiot and I'll learn that in the future but for now ignorance is bliss? lol X^D

i have huge complex avoiding the incamera settings that I destroy everything trying not to use them with the hope of a flat image. Conclusion: In very low light forget flat settings. Denoiser, sharpness, saturation and HIGH iso must be used to boost image before compression. As I discovered it is better in extremely low light conditions to go at even 6400 -5 +5 +2 +5 and press down mids in post than trying to shoot a flat image and boost in post everything. I hope for a hack that at least let us shoot in 25 and 30p and use all the color information for those few frames, not like now, the issue with same image quality in both 25 and 50 (30-60) is like the camera tells me that it just records and then deletes half frames

Found something today and I think it is very interesting. When underexposing by 2/3 of a stop (in sufficient light), skin tones seems less chalky when recovered in post and shadows fall off more natural. Overall better video quality.

I'll be posting clips later this week.

@deang001 You said it brother. That would be some fantastic meta data to have with each shot.

I shoot always underexposed, except in daylight cause I don't have nd filter, but I feel that using in-camera contrast changes all the time skintones and generally the way that colors act cause in every different exposure the point that the camera starts pushing mids to black also changes. The camera will select as "black area" a different area in exposure -2 than +2. I am not sure but I still hate in-camera contrast :P

I dont know that it does exactly what you need, but I've found using a free program called 'MediaInfo' extremely helpful in recovering certain metadata from video files.

@larrysanders and @wgtwo When we've been mentioning "metadata" that's probably the wrong term we're using. We're saying how convenient it would be to be able to look at all the information in a clip as it pertains to the GH3 specifically. So not just your typical "metadata", but.....

eg: Clip 1080 24p, 50Mb/s, Color Profile: Standard, Contrast -5, Saturation -3, Sharpness -1, Noise Reduction -2, White Balance: 3700K, Lens: Summilux Leica 25mm, f1.4, SS: 60, OIS: On....etc. etc.

That kind of data.

"@larrysanders and @wgtwo When we've been mentioning "metadata" that's probably the wrong term we're using. We're saying how convenient it would be to be able to look at all the information in a clip as it pertains to the GH3 specifically. So not just your typical "metadata", but.....

eg: Clip 1080 24p, 50Mb/s, Color Profile: Standard, Contrast -5, Saturation -3, Sharpness -1, Noise Reduction -2, White Balance: 3700K, Lens: Summilux Leica 25mm, f1.4, SS: 60, OIS: On....etc. etc.

That kind of data."

Yes, exactly. That sort of info on a text file with the same file name as the clip would be very handy. I'm sure it would not be that difficult for them to do.

My footage was poor, we managed to shoot in 2 hours one main shot that I used crops and about one shot closeups, in a very small room, pancake 20mm to its limits. The sound is live from static shot, one mic to amplifier and gh3's microphones. About 50watt light :P natural -5, -5, -1, -5, iso 800, ss25, f/1.7 This is my first music clip, next time it will be better because we will do ONLY what I say. I wish I had one more shot closeups :(

@maddog15 The sound for me is the easiest part in reality, I was able to get it organised with a motu soundcard and record flat channels remaking the loops then digitally and correct anything, but as I wrote, next time will be only the way I say in both sound and video :)

Hi all - I've been doing extensive tests with the GH2, GH3 and the DSC OneShot chart. It has Rec.709 colors, skin tones, black, gray, and white on the front. Nice chart. Been trying to figure out the sweet spot on the GH3.

I've also been using a firmware-updated Ikan D5w with built-in waveform monitor and vectorscope. I've been very interested to see exactly what the changes are in the different profiles. It's been quite a learning experience.

I brought all the footage into FCP X to double-check the scopes in there. I also have a calibrated Videotek TM-621 It can zoom in on the colors 2x, which is necessary to balance to "perfect" Rec.709 colors. Reason for this is that the chart prints the primary and secondaries at 50% of full saturation.

I was testing with both the 12-35 and the 35-100 Panny lens. There is some additional vignetting on the 35-100, but I would balance colors toward the center of the frame on both lenses.

My method was to shoot the chart with both cams, then shoot some footage with a face in it. My own face seemed to work just fine for this test, although it looks like I'm not yet "professionally calibrated." Perhaps a trip to the dermatologist will help with that ;-).

In addition, I would then bring footage into Resolve 10 and do some basic Hue vs Hue fixes to see how much correction needed to be made. I don't know Resolve 10 very well but this seemed like the way to go about it based on Art Adams' original explanation on Pro Video Coalition when he was using Color.

The goals here were two-fold:

A. Figure out a best-case mode for skintones, since I find them to be generally flat on the GH3. B. Try to find the best modes between the GH2 and GH3 for matching.

Sorry, no pictures of scopes and such. This has taken a bit of time and I was focused on the work and not so much on documenting it with screengrabs.

Here are my initial findings - one of which was a big "duh" moment for me ;-P.

1. Vivid -4, 0, -2, 0 seems to be the most accurate in terms of how the colors are lined up with their vectors on the vectorscope. It also lined up the contrast on the chart most accurately. White at 90%, Gray at 50%, and Black at 7.5%. In addition, it also gives an extra boost to Green, Magenta, and Cyan, which are generally lacking in the other modes.

2. Accuracy does not necessarily mean "most favorable." And all the testing in the world doesn't reproduce real world conditions where lots of happy accidents occur. Shooting on vivid yielded more yellow skin tones. Standard yielded more magenta/red. Both could be balanced out to match in Resolve.

3. Currently, I prefer the skin tones the best on Standard 0,0,0,0 (not the "duh" moment!). This was a bit of a surprise since I thought Vivid would be better. My reason for this is that there seems to be more contrast in the mid-tones without taking too much into black, when set to standard. Faces were a little more plastic-y in Vivid.

4. While I was able to manually balance to my flolights (rated for 5400k, but actually closer to 5200k according to the camera), I found that auto balancing the GH3 off of a gray card (not white) produced more favorable and accurate results.

5. If you want to balance a GH2 and a GH3 together, again the gray card is the way to go. Line them up and balance off the same light source if at all possible. I was able to get the cameras pretty close. Yeah, that was the "duh" moment that occurred after WAY too much testing :-D. I'm just using both cameras on Standard, currently. More testing needs to be done here to see how close I can get them before balancing in post.

6. The GH3 on it's own looks great. But whenever I compare the image to the GH2, I almost always prefer it. More testing necessary here though to see if I can really get the GH3 closer to the look of the GH2.

7. There is very little room, in my opinion, for color balancing or correction on the GH3. You really need to nail the white balance on set if skin tones are involve. Highly recommended for interviews.

8. On the GH3, the Green, Cyan, and Magenta were a bit shallow on the vectorscope. Adding saturation to them did not noticeably or dramatically affect skin tones, although some slight changes could be perceived. Outdoor testing might prove otherwise.

9. On the GH2, the midtones seem to be a bit more contrasty and bit more blue. This is an unscientific assessment. More testing here.

I still plan on doing some more testing to try out a few more odds and ends, but I feel like I'm armed with what I need to know for some field shooting.