GH3 GH3 v Gh2 with color check and star chart

maarek said:
I always find it odd how people blame 8bit or 420 for getting a bad key. Getting a good key out of a green screen depends on a multitude of factors with 444 being at the bottom. Lighting is extremely important and no amount of 444 will help you key it if you mess up the lighting. The software you use and your proficiency in it. Also your skills and amount of time you put in in (especially if it's not just a talking head but something way more complex). And planning the shots.
Click to expand...

+1

Chroma detail just means the amount of pixels that make up the edge of your key. The act of pulling a decent key is still subject to the same level of skill and finesse regardless if you have 4:4:4, 4:2:2 or 4:2:0. With proper chroma filtering such as that in Adobe CS6 4:2:0 quickly becomes 4:4:4 albeit with slightly softened edges. If you use chroma filtering then the only thing 4:4:4 gains you is sharper edges with more raw detail. Keep in mind your luma channel is still full detail so not everything is softer.

I work with 4:4:4 from our RED One and RED Epic at work and trust me it isn't any easier or better really. If it is shot or lit like crap I have to spend just as much time cleaning it up with multiple masked layers.
 
Stab said:
http://www.youtube.com/watch?v=WNrctjyPqxs

As you can see there, there is a huge increase in dynamic range for the GH3 ... Like 2 stops. Or, the camera is by default more light sensitive which seems strange... Or the tester is a retard.

What do you guys think?
Click to expand...

This test didn't do much. The Gh3 seemed to put out a brighter image. With that there was more detail in the shadows and less highlight detail. Not a good way to compare DR. Still waiting for someone to pick up the test i proposed in another thread so we can get some real comparisons going.
 
1. I've done plenty of keying with GH2 (hack) footage, it's fine for small things. I would not use it for anything very big, and keying a 5K image definitely does turn out a better product, but you can get by with GH2 and I'm sure GH3.

Zephyrnoid said:
Meanwhile, Andrew Reid is having a very tough time coming up with BMCC raw footage actually edited in raw (vs converted to ProRes), so what's the point of that camera unless you have the horsepower to edit it in raw?
Click to expand...

2. Some people don't know what they're doing. If you can't work with the footage then that's why. Also, you don't "edit" RAW.
 
Fresno Bob said:
How can you tell anything about the DR from that?
Click to expand...

Just look at the ships. It's barely visible in the GH2 footage, and very clearly there in the GH3 footage. This is exactly the kind of test I've been waiting for, and a real test of DR. How well can the two cameras see into the shadows? This test shows that perfectly.
 
dusty, any new test mate? You've been MIA in the thread for a while. Still waiting on that exposure climb on both gh2 and gh3 to compare detail in highs and lows while ramping exposure.
 
PaPa said:
dusty, any new test mate? You've been MIA in the thread for a while. Still waiting on that exposure climb on both gh2 and gh3 to compare detail in highs and lows while ramping exposure.
Click to expand...
I've been kinda busy buddy. I'll see what I can do over the weekend.
 
:) Sorry to be such a nag, but it will be great to see a real test comparing the two cams. Everyone throws around dynamic range numbers, but very few know what it actually means. This test will go a long way in explaining just that.
 
You must log in or register to reply here.
Back
Top