Felix E Klee: What's with Berlin all the time? I thought you guys are in Seattle.
We (DPReview) are in Seattle but Andrew Reid (EOSHD.com) is in Berlin.
jesus_freak: What happened to the video mode Studio Test Scene comparison? I'd be interested in seeing the difference between native 1080p and downscaled 4K with this test scene.
Also, a low-light video mode test scene would be really helpful too!
The ones at [the bottom of page 3](http://www.dpreview.com/reviews/panasonic-lumix-dmc-gh4/3#sampling)?
Kevylloyd: What about focus speed in low light ?(better than any other camera I've used). What about focus speed with none dfd capable lenses? Still very fast from what I've read elsewhere. Ok so DFD might not be all that you expected per Pany's claims, but regardless I'd be surprised if many mirrorless cameras are as fast in aquisition.The AF performance section of this review seems to be unbalanced and lacking in breadth in my opinion. Perhaps you cover these points elsewhere in the review to some degree, but why not in the AF section?
Otherwise thanks for the review, I love mine for fast, silent, low light shooting especially combined with the stabilized F1.2 Noctiocron!
Those are fair points - I'll add a bit more detail (about things other than tracking).
RickPick: Just noticed a couple of small mistakes in the comparison chart above - you call the GH3 "GM3" - and the GH3 actually does support time code. Trivial, I know. Thanks for giving the camera a full review, I know you got some stick (Brits will understand) for your comparison with the A7S.
@RickPick - it was always my intention of including separate conclusions and scores. But either way, I'm hoping that people find things to enjoy or to interest them in this review. Sorry it took so long.
I'm not sure what more I could have said about DFD. The technology itself seems to work (the camera refocuses quickly and accurately as the subject moves in the Z axis), but, despite trying, we couldn't get the camera to lock onto subjects reliably enough for that capability to be the real-world advantage that the technology promises.
I'll go and look for those errors immediately. Thanks for pointing it out.
I must admit I still don't fully understand why publishing this separately from the a7S makes so much of a difference, but if separate reviews are what our readers want, then that's what we'll try to provide.
Terapixel: I'm just a still shooter, so that 85% is way lower when video is not taken into account, below 84%. So for just still shooters Fuji X-T1/ Samsung NX-1 is more interesting (not talking about Bokeh).
I'd put the GH4 on a par with the E-M1 for stills shooting (though I do slightly prefer the E-M1's 2x2 interface). It's less clear-cut against the X-T1, but there's still not a lot in it.
Our scoring system can't really cope with just how much better the GH4 is at video than the other cameras in its class, so it's not safe to assume that it would score lower than its peers, if video isn't important to you.
SulfurousBeast: Nice, some really good stuff from the likes of Pany, Sony and Fuji , Oly who are daring to innovate. I still wonder how long the big two will keep "milking" their customers with incremental updates and not truly innovating. i am a Canon user myself, but if it was not for the investments made already over 5K and I am a hobbyist, only thing preventing me to make the switch is my own inertia, sell the curent gear put together a new system etc. but always been amazed and lured to what Sony and Panasoic is doing in this place.
BTW, good review DPR though I almost did not care if GH4 was reviewed already or not.,.so late in getting this out.
Jorginho - we did test the AF fairly extensively, we've just illustrated it with this particular example. I tried using tracking AF in a number of situations, with different subjects, indoors and out, and found it struggled.
But, since I didn't shoot any birds in flight, I can't comment with complete certainty - it's possible that having just sky in the background is the necessary difference to make it more reliable.
However, even testing *extensively* doesn't mean testing *exhaustively* - there are too many variables and shooting situations to convincingly test every one of them. And for every one we did, someone could name another and say we haven't done it properly.
mpgxsvcd: Is it safe to assume that this review was written before firmware 2.0 was included? Will the review be updated with firmware 2.0 benefits?
Firmware 2.0 fixes some of the issues with AF and rolling shutter as well as adding multi-aspect ratios for video. No other camera offers the multi-aspect ratio video at any price. That is a huge benefit to get without paying any more money.
The review is based on firmware v1.1 but we'll be writing something about the 4K Photo Mode in the coming weeks.
mpgxsvcd: I can’t figure out what is more impressive. The fact that Dpreview placed this camera in the “Professional Interchangeable Lens Category” or the fact that it was actually competitive in that category.
I can’t believe that the GH4 is listed in the “Professional Interchangeable Lens Category” and the Canon 5D MKIII, Nikon D800E, and Sony A7R are not. Those are all listed as Semi-Professional.
mpgxsvcd - no, it was scored alongside its peers (E-M1, D7100 etc), it's just appearing with the wrong category title.
That's a database error, I'm afraid. I'm trying to find someone able to fix it.
kristian1: About AF performance I can say that GH4 has best AF from all mirrorless cameras , have tested samsung nx1, fuji xt1,olympus em-1, about c-sf fastest is samung but focus accuracy is very low not very useful , olympus is a bit faster, but focus accuracy not as good as gh4, xt1 is very difficult to lock first shot, its slower and accuracy a bit lower. about s-af GH4 is fastest and also very accurate.I have testes on leica 42.5mm f1.2 and oly. 75mm f1.8 at wides aperture also at difficult lighting conditions.
I am coming from 1dx , nikon d810 and gh4 is closing the gap very well on AF..Only about BIFs there is no lens with fast enough AF , and 100-300 is way too slow for BIFs , so I wouldn't spend time on that.
kristian1 - this isn't the only testing we did. We had the same problem when trying to use AF tracking in video mode and neither of the two people who tried it could get it to consistently lock onto subjects. We can only report our experiences.
Jay Turberville: Not sure what they mean by "No 1080 video from 4K crop region to minimize aliasing". If you shoot 1080 with Ex Tele Mode you get 1080 using a pixel for pixel center crop which certainly seems like the minimal aliasing mode they are looking for. Should be better than cropping the 4K image because the compression per pixel would be less.
I'll re-check that.
Menneisyys: The Scoring bar chart is missing at the bottom of Conclusions.
Sorry, we're having trouble with the database.
Tidewater: So is this camera able to focus follow birds or aircraft or motorsports better than existing models or not. There are many other scenarios encountered by action shooters other than bicycles coming directly at us. There must be plenty of erratic flying Gulls and Terns in the Seattle area to test with a moderate Telephoto.
There are an infinite number of things that could be AF tested, but testing all of them would take an infinite amount of time.
I experienced tremendous difficulty getting the GH4 to lock onto subjects, which makes me skeptical about its ability to track birds, but I'm not in a position to test it, so I couldn't be certain. The bike shots are simply meant to show how the camera behaves with a subject moving in two axes at a moderate speed and this is meant to give an idea about which aspects a camera does well and badly. It's not intended as the final word on AF, though.
falconeyes: Still largely useless tool. Ok for noise and corner performance. Nothing else.E.g., uses F5.6 70mm eq. for RX100m3 and F4 50mm eq. fir LX100. Can't be compared. E.g., F5.6 is F15.4 eq. and deep into softness by diffraction territory.
@rhlpetrus - we have a beta build of ACR that supports the LX100. We'll certainly be looking at the wide-open performance, since there's no point in having a fast lens if it doesn't perform well.
technomad: "it's most directly comparable to a Full Frame 70-200mm F5.6". What? Are they really not teaching basic arithmetic these days? There is only one variable where that statement is true: the DoF at a given f-stop. The light gathering capability is exactly the same for any lens at a given aperture and the focal length equivalent is whatever the crop factor translates it to - in this case 80-300mm. Which is not 70-200…
technomad - it has the effect of not only telling you the depth-of-field performance but gives an *indication* of what the low light performance could be like and how much diffraction it'll induce.
fastprime: Are these shots taken with pre-production firmware (0.2) or production firmware (1.0)? Panasonic has requested that shots taken with 0.2 should not be published.
We should be publishing our studio test shots from the LX100 in the next couple of days. Part of preparing for that involves shooting the camera at a range of focal lengths and apertures and we've seen nothing to suggest the camera is broken.
We'll certainly speak to Panasonic to see whether we can check it against another sample.
They're firmware 1.0 - we always indicate when we publish images from non-production cameras.
[This forum post](http://www.dpreview.com/forums/post/54525051) has a little more detail of some of the topics the editorial staff will be discussing. Please post any questions you'd like the panel to ask during the broadcast and we'll try to cover them.