GH2 driftwood hack in shootout with Canon C300

ehr

@kong@proaudio4
I don't see the D7000 skin blotchyness.
to my eyes the D700 had the least amount of banding (look below the lights) and blocking (look in the right corner), followed by the GH2, and all the rest after that.
The green hue of the D7000 is kind of nasty, but I think that can be fixed.

DrDave

Re: dynamic range--one thing on my list of things to do is shoot some vid with a legacy prime that is very sharp but has less contrast, say in the F2 to F2.8 range. In still photography, some people use this trick to get an extra stop of DR (which is then reconstituted in post). Have to try this on the Driftwood hack over the holidays, maybe with the Nikkor 35mm F1.4 or one of the Rokkor primes.

Elenion

Well an extra stop of dr only thanks to the softness of the lens seems to be a little too much...
Again: we're judging dr in an uncorrect way. An image can be very contrasty and have more dr of a flat one, because the details are still there in highlights and shadows until they're clipped.
It's just a matter of 'look': just flat the image if you like.

In these shots none of the highlights seems to be clipped in any cameras

proaudio4

ehr,
Yes, the green cast is too much on the D7000. I certainly see the difference due to compression, even more so viewing on a large display.

mozes

i just read that nr:4 is the gh2, wel that got me silent.
still dont have a gh2, i need more info on the canon 60d :'sorry.
But wow, that sets things different again.... lot off room to play with in post editing........

derek

were you read that?

DrDave

Extra stop would come not from the softness of the lens, but using a sharp, low contrast legacy lens and pushing the results in post. I haven't tried it in video, only still photography. So it may not work, just something to try out over the holidays.

mozes

*derek
on philips site he have updated the post

EDIT: Someone in the comments got in right!

Sony NEX5N, Flat PP

AF100 AB-RANGE PP

C300 C-LOG

GH2 Driftwood Hack nostalgia and dialed down

5Dmk2 Marvels Advance

7D Marvels Advance

FS100 G-LOG A

F3 AB-RANGE

F3 S-Log

Nikon D7000 Flat

derek

no no,is not like you say
1 5D
2 NEX5
3 AF100
4 C300
5 D7000
6 F3 AB
7 F3 S-Log
8 FS100
9 GH2
10 7D

kholi

@danyyyel

You have fair thoughts, my friend. Again, I wasn't trying to be a jerk, but people here and GH2 users out there should feel confident in what they're using. Not just GH users, but all camera owners. I know why the images are unsettling, it's very simple:

WE KNOW the camera is much better than what that still represented. It happens in nearly EVERY shootout, that's why they're pointless unless they're being done by tons of people with tons of time to commit to making sure every single detail has been accounted for.

The Digital Cinema Society shootouts were the most rich, detailed shootouts we could have gotten, and the results were handled professionally. If you want a serious opinion, then that's the kind of shootout you need to look out for.

Let me tell you this: I can grab an Epic and shoot whenever I want. Point blank. F3 S-Log? No problem. ALexa? I might have to pay a few bucks, but I know private owners. I could shoot with these any time.

As far as Epic vs GH2? GH2 can't truly compete with Epic at 5K or 4K, but at 3K and 2K? Not the case. The GH2 Hacked stands image to image with Epic 2K. And it's a freakin' 700.00 camera.

You can take that or leave it, but right now you shouldn't be worried about what the cam does in a shootout and, instead, showing moving footage of how insane it is.

kholi

@danyyyel

Oh and also on DR, I agree with you there. You can't base photo mode results on video mode results, it's two different animals. If there's a decrease in DR as you rise in ISO, I highly doubt it's more than negligible, and wouldn't be easily tamed by using NDs (Which you already understand, but most people do not) or exposing for flesh and highlights.

The rest is handled with apps like 5DtoRGB, which open up shadows and give you the options to go out to lossless formats like DPX for grading.

brianluce

@driftwood
If I was smart enough to hack a GH2, I'd never post footage because the expectations would be too high. Brilliant enough to turbo charge a DSLR? Then your footage undoubtedly would exceed that of Roger Deakins by orders of magnitude. Too much much pressure!

kong

@mozes

look in the comment itself. It is this:

1. 5D MKII
2. NEX5N
3. AF100
4. C300
5. D7000
6. F3 AB
7. F3 S-Log
8. FS100
9. GH2
10. 7D

mozes

That's what philips say. not me, is just a quote from there....

danyyyel

Are you sure, I see it but I can't relate to that. Can somebody ask for it because it would change a lot of thing. So from his post is it. Can't see the S-log F3 being one of the most contrasty image.

1 Sony NEX5N, Flat PP

2 AF100 AB-RANGE PP

3 C300 C-LOG

4. GH2 Driftwood Hack nostalgia and dialed down

5. 5Dmk2 Marvels Advance

6. 7D Marvels Advance

7. FS100 G-LOG A

8. F3 AB-RANGE

9. F3 S-Log

10. Nikon D7000 Flat

derek

he say ,Someone in the comments got in right!
and that person is spacemoose,
Here’s my Pink Panther:

1 5D
2 NEX5
3 AF100
4 C300
5 D7000
6 F3 AB
7 F3 S-Log
8 FS100
9 GH2
10 7D
philip say correct read

spacewig

I think half the board just committed Seppuku...

danyyyel

Yes it is a little confusing, his post on the main article and then him saying in the post below which one is correct. I think it is the one in the post that is correct with the gh2 being no 9.

DrDave

I would say 99 percent of the stuff I see online is not double blind, and therefore has no real "perceptual" significance. It's just random net stuff. There's no question the GH2 takes good vid. What else do we need to know? I couldn't even tell my Olly 45mm from my Vivitar 55mm once I had deliberately shuffled the deck. At least in this case they did not pre-announce the cam IDs.

The only reason there is a kerfuffle is because they didn't pre-annouce the IDs.

If they had, the ensuing discussion would have been completely different, guaranteed.
However, whether the camera is "actually" better or just "seems" to be better is not the whole story. If you take a musician with five violins, one of which is a genuine Stradivarius, and the other four are really good copies, more often than not, a highly trained panel in a blind test will pick the copy as sounding better. But that isn't a "measurement", it is a perception. And the musician would prefer to play the Strad, even if it sounded inferior. I will sometimes prefer to use an old Leica or Zeiss lens, even if my Panny 20mm trounces it on edge sharpness and distortion wide open (which it does). I would prefer to play the Strad, but I would still *really* want to know which one really sounded better. Time after time, double blind tests come up with radically different results than those that are not. What is interesting about this test, is that different people preferred different images, I don't see a consensus emerging--before the results were announced--as to which one was better, or even which one was which. I don't actually care which one "measures" better, I'm not going the measure the final product, just watch it on a big screen.

PhilipBloom

Relax chaps! The Driftwood hack did super well. Especially the resolution shots.

PLEASE read my post. Those are heavily compressed jpegs to represent the colour of skin tone only. Ignore compression artifacts.

With regard to PP. Nick Driftwood told me to use Nostalgia dialect down.

Whenever you do a shootout you will also have people say you should have done it this way or that way...

Remember none of these are graded. Once they are I can easily make them all match.

Stop obsessing over minute details are enjoy a supercharged camera.

PhilipBloom

Derek has it correct

danyyyel

Ok

Mimirsan

Okay guys sorry but lemme rant as ive nearly had enough I just cant take it anymore.
I think the only reason some GH2 owners get wound up over the bloomage shootout is that the outcome isnt going to be "OMG!!! The GH2 Is so awesome better than everything out there even the $3000 camera GH2 ROCKS MAN WOOO!"
Crying that its being shot wrong or overexposed or the shootout is deliberately uneven and biased just seems well...just typical and predictable.
Bloom is not the be all end all (Hes not blimmin Kurosawa...btw DONT get me started on his "critics" webisodes)...get over it and go and shoot stuff...y'know like what he does?
All the cameras are great you dont need a "shootout" by anyone to tell you what you use is good. Life is fucking short go do something with it!
So fucking insecure some GH2 owners are! Christ!
Sooooo sick of these fucking comparison videos! AAAAAARRRRrrrrrrgggghhhh!
*Goes for a pint* ;-)

danyyyel

I hope you post some of the uncompressed screengrabs. From what I have seen the gh2 exhibits oddly the less banding in the smooth background. It has very fine grain(noise) that breaks the pattern.

@kholi, I might seem a little DR obsessed but I live in a tropical island and contrast lighting ratio here is very high. As a working photographer you always have to fight it. But at least in photography with some flash I can manage quite well as a fill-in. I had a crash course with a DP on lighting for cinematography and I could overpower the sun with what would fill into my car... less wait than a 2k HMI. The same for continous light would need a lighting truck.

I have stated to read your blog and find it very valuable. The subject of banding was very interesting and I invite everybody to look at it here http://www.campcomet.com/archives/797 . I had experiment a bit with it and was getting to the same conclusion, will have to do more test with your findings to see what work best for me.

Roberto

Attached updated table with @PhilipBloom 's confirmed results.

Howdy, Stranger!

Categories

Tags in Topic

Top Posters