Personal View site logo
GH2 driftwood hack in shootout with Canon C300
  • 471 Replies sorted by
  • I'm not disappointed in the GH2 low light test either. That was a very extreme torture test, any shoot of mine that matters will have adequate light. Of course, journalists and documentary shooters may have different low light requirements. Wonder if Smooth -2 -2 -2 -2 would have been less "baked in" than Nostalgia for the skin tones footage?

  • Does anyone know why the Canon 5D's footage during the skin tone test was banding more than the others? It seemed to be the case on my screen.

  • gh2 is not perfect camera,4/3 sensor not help in low light,red epic is also bad in low light

  • This is an incomplete low light test, we don't have enough info yet.
    Did he shoot a low-light test at F2.8? If so, why?

    Yes, when sticking the SAME LENS at F2.8 in front of each camera, the GH2 is POSSIBLY a poorer performer at any ISO.

    There is a FLAW in the low light test so far- no camera looks good WITHOUT a lens. They all look bad. You must factor in APPROPRIATE LENS, not same lens when doing extreme test. I don't see any lens detail yet- hopefully that will change as he updates his post

    Now to REALITY, for $1000 you get the 25mm F0.95 Voightlander and now which is worst at same ISO setting? Why not use a Canon 85mm F1.2 for the low light test? With both lenses, then you better low light performance and resolution at the ISO setting on GH2.

    LENS speed is critical in a low light test, so not discussing it up front in detail is incomplete in my opinion.

    If this is a real-world camera body test (and lens), then tell me a combo that outperforms the GH2 with F0.95 Voightlander combination for under $2000?

  • @NickBen

    A NEX5 with a Voightlander 50/1.1 Nokton?

  • It's true, even the Epic is not as good as Alexa and F3 at low light. If the same lens was used in all tests, I don't care, you can stick a fast lens to any camera with interchangeable lenses.

    But point is: for any kind of planned shooting, we would take care of lighting. A doc shooter might need a FS100 with a fast lens, but for indie features the value for money of a hacked GH2 with a good lens is incredible. I'm quite relieved that the skin tones were not really so hard to grade. Maybe VK can pull some tricks to access better profiles to even improve on that. I like how the skin tones are not so blotchy after correction with the help of higher bitrates.

  • @bitcrsuher I agree with your point, we NEED details on lens settings for low light test. Until then, we are guessing in the DARK!

  • @NickBen it looks like a 85mm ZF F1.4

    The GH2 kills in the resolution but losing in the picture profiles. I really wish we had a flater setting.

    As far as low light the GH2 is not the king but still ok. Its not like I am filming vampires, if I was then I would get a FS100.

  • I have the zeiss 35mm ZF F2 and for some reason,the lumix Lents are more sharper than the zeiss one, in my GH2.

  • I was disappointed in Bloom's failure to include the Nikon D7000 in the low-light tests. While I wouldn't expect it or any DSLR to seriously compete with the FS100, F3, or C300, the D7000's specs would lead you to expect it to show its best performance on this test. In my experience with the Nikon D5100, I can rely on it to deliver comparable noise and dynamic range at ISO 3200 as the GH2 at ISO 1600.

  • LPowell Nikon D700 is not good camera D5100 also.you say D5100 is better camera?

  • I was fairly pleased with the GH in the skin tone test. It wasn't perfect, but not unusable either. I wonder how hard it will be to get Pany to come up with a flatter profile. it seems like a really important thing for both the AF100 and GH line. Gotta make that a priority along with more fast glass.

    I liked the look of the F3 S-Log the best. The FS100 2nd and then the C300 in terms of low light. I will say that I thought it was remarkable how fine the grain was when the C300 was cranked all the way up to ISO 20K. Unbelievable how these cameras do in low light.

    I'm not too put off by the GH performance in low light compared to those monsters. It was an extreme test and hopefully we can augment the GH with a Nikon or Sony DSLR. I think those are acceptable low cost tools for very low light situations.

  • @derek Yes, I did say Nikon, sorry if you happen to disagree...

  • I wonder if 'smooth' might have been a better choice for the skin tone test?

  • Glad 5n shows its benefits. Its all in the sensor size . Remember why camcorders/compacts are worse in lowlight? Smaller sensor. None of this is surprising and the GH2 is a great camera and good enough in lowlight and great with fast glass. I will say it is nice not to have to rely on expensive fast lenses in lowlight when using my 5n.

    @Lpowell The d5100 & d7000 are great cameras. You know it would pretty darn great in lowlight anyway. ;-)

  • The low light test is nonsense He has filmed with f2.8? Besides - use light sources, when filming

    and the Winner is: the GH2, the camera with the best resolution

  • Thx to Philip for part 2 ! It was great.

    Concerning the extreme low light test, hmm we can obviously call it THE NOISE TEST. Very dark (dracula conditions of light with no candles) and here we don’t have the NASA lens of Kubrick . No much micro contrasts to see, flat still surfaces except hair of the guy. And of course small sensors are bad here as they need both super-primes and denoising. Note that we don’t know nothing about internal denoisers of all the cams here while compressing…

    So yes, NOISE for GH2, what else? Look at image at 6400 iso for gh2 : details are still here. Noisy of course but still details inside.

    Gh2 should be used under 1600 iso with lens aperture under 2 and (neat) denoiser in post process. 3 conditions and details I think are here : that is what matters. The only drawback is shallow DOF if not needed…

    We don’t need a NOISE test Mr Bloom. These tests has been done already. We need something real : a low light test for details (like in the zacuto test). But PB chose the worst clinical test for small sensors. Not low light but…dark. It is not England it is Romania 

  • i don't mind noise to a level, as long as it's decently 'filmic', like 70's stock film. or even 16mm noise for that matter. and the high driftwood patches DO make it looks so, to a degree. so- what are the differences between this 'nice' noise and 'ugly' totally unusable noise?

  • il is a test here, no artistic feeling about noise allowed :-)

  • IMO there's no problem with seeing how far cameras can be pushed. Now when it comes to whatever camera you have, then you learn that camera and work within it's limits. I have always believed that's how you work with any tool. Stay in it's sweet spot and even cheap things can perform at a high level. I've had to engineer sound with some awful equipment, but I still manage to get a good sound working within the limits of the gear. Unless you're working in extreme environment, that would require only the best.

    All of the cameras had acceptable image when not under stress, which is a win for low budget film guys everywhere. I will say tho that Pany needs to upgrade the AF100. No reason that camera should be using that sensor. Everything else about it is great, but that sensor.

  • Hmmm.....wondering what's the deal with new picture profiles with the GH-2. Anyone know if that's still being worked on?

    I think for the skin tone test it would have been wiser to use smooth. Nostalgic is too baked in.....still loked great though.

  • I'll be trading up to the GH3 if it has the sharpness of the GH2 and the low noise of the GH1 ( or better!) and an awesome hack to up the bitrate indoors.

  • @DrDave the low noise of gh1? IMO it's at least as noisy as gh2, if not noisier, and the grain is not as fine and absolutely not uniform.


    The one problem I have with the vid about the test is that Bloom makes it sound like the GH2 fares worse than the other DSLR, when, really - it cannot match F3, C300 and FS100 for signal to noise output. The other DSLR, whilst some can be cranked highly - macroblocking and other problems will ensue.. Some of the surfaces literally come alive on the other dslrs, and the noise is not uniform - it will be a bitch to get rid of. Anyway, who cares?

    I thought the FS100 looked good in that test, obviously my concerns for the camera would be to the functionality of it's operation - on paper it always looked like a very interesting prospect. A bit of a shame that it only has a HDMI output.. (would it be wortwhile to exchange your EX3 for a FS100?)

  • GH3. Another year, another cam. Why can't we have it all? I expected by now the Canon 5D Mk III would be a big step up. Maybe Fuji's organic rainforest brew will open the gates. In any case, time for a new sensor with lower noise and better everything.

  • We're spoiled.