Andrew Reid

  All the footage in this shootout is graded in Premiere with the fast colour corrector and digitally sharpened.   As I said in my review, the D5200 grades well for a DSLR. Of course it is a world away from the Blackmagic's raw or ProRes but then so are the C300 and 1D C and they are $15k and $12k.

  The D5100 wasn't usable because it lacked manual control of shutter and ISO in video mode. The image was poor too, with moire and aliasing issues. Not much better than a $200 GF1 really! Now the D5200 is a different league even though the body design and ergonomics are pretty similar.

  You're right. GH2 and GH3 still the resolution kings. Though the D5200 and 5D Mark III do sharpen up nicely in post. They need a bit of makeup.

The GH3 has some major advantages that the D5200 lacks.   Wait for my review...

http://vimeo.com/60135187 Let's pit the $700 Nikon D5200 against the $3000 Canon 5D Mark III and see which one comes out alive.  

The Nikon D7100 is here and appears to be based around the new sensor and image processor of the D5200 which is proving rather good at video so far. Whilst the D7100 gains a dedicated movie mode and headphone jack, it has a few things that may disturb video shooters.

  I don't read his forum. It sends me mental.   The ProRes is slightly better but the main issue is that it is 30p in live view, and drops to 720p if recording to the card. So no 1080/24/25p over HDMI from this camera until Nikon sorts it out with a firmware update. Might be possible to get 24p from the 1080i output but this needs more testing and possibly firmware updates from Blackmagic and Atomos as had to happen with the D800.

  One small problem with that test. Micro Four Thirds Speed Booster doesn't exist yet.

  Not tried JuiceLink.   It outputs the audio via HDMI so that is a solution.

  Great first post Sanka, welcome to the forum.   It is good to know some M42 lenses focus to infinity with the standard M42 to Nikon adapter. I'll get an adapter and try my range, as I have quite a lot of M42 glass including the Zeiss Flektagon 20mm F2.8 which is one of the best full frame wide angle lenses in my opinion.   I am a big fan of the Contax Zeiss stuff, it is just a shame it does need a modification of the mount to work on Nikon F.   The medium format stuff is very heavy and quite large and I once had the Zeiss Distagon 80mm F2.8 - it was pretty soft but as you say, good for organic looking portraits.

Have you tried setting HDMI to 1080p instead of auto? When I do this I get a true progressive feed but it is 29.97 frames (30p) even though the camera is set to PAL / 25p.   What about trying it in NTSC and trying 24p from 60i?

Nope. Nikon glass or Leica R fit.   Happily the Nikon AI/S stuff is not expensive and is very nice indeed.

Scaling and compression don't help either. I think the 8bit is only the main source of a banding problem when you grade heavily. (But don't mention that to Bloom!)   JPEGs are 8bit but usually have much smoother gradation.   The fix - well more of a bandaid really - is to add dithering or grain to the image.

  I can vouch for that. 1080/60p at 50Mbit is cleaner than the ALL-I 24p mode. It could be that the much higher bitrate on a per frame basis is allowing the codec to store more noise in the original output.

  Same issue for now. 29.97p progressive signal when set to 1080p. Not experimented much with it set to Auto or 1080i yet but the 24/25p seems elusive to me.   Could be a firmware thing on the HyperDeck / Ninja.   Find it strange that the camera, though set to PAL, would output a HDMI signal in NTSC. Maybe it is a firmware bug on the camera or some kind of compatibility issue.

My OCT 18 35mm anamorphic is actually sharper than my OCT 19.   It depends on the luck of the draw and what condition the lens is in. Olex is worth contacting if you want yours optically checking.

There's a further twist to this...   The first few frames of the D5200's H.264 files are very slightly muddier than the rest of the clip! So weird.   What you see above is one of the first few frames and I have only just realised this bug. But the HDMI is still giving me a better image for grading work and is uncompressed.   More to come on this...

Check out the Blackmagic HyperDeck Shuttle shots just in. Going to have to update the review. This cam via HDMI is absolutely killer.   http://www.eoshd.com/comments/topic/2125-exploring-nikon-d5200-hdmi-output-major-discovery/

It's uncompressed 4-2-2 guys.   I'm finally getting it to work with the Blackmagic HyperDeck Shuttle and it grades amazingly. Noise grain is finer, there's obviously far less compression going on via HDMI. There's also about a stop more dynamic range and a more gentle roll off to the highlights. Amazing stuff.   If you get this camera, I recommend you shoot via an external recorder.   Now here's the proof... (blown up 200% in PS)...   http://www.eoshd.com/uploads/nikon-d5200-prores-vs-h264.png   More detail too - look at the thread on the outer ring of the Iscorama... Thread lines visible in the ProRes and smudged in the H.264 internal recording at 24Mbit.

  H.264 MOV ALL-I 72Mbit 1080/25p.

  Not really since the aperture ring is on the lens with a lot of the Nikon glass. I use the AI/S stuff.   With the modern AF stuff you need to flick it into optical viewfinder mode and change the aperture there, then switch back to live view mode.

  With different sensor sizes, the lens performance is always in the equation. It is just a matter of trying to match them as closely as possible.   Given the same lens, a 2x crop sensor will have sharper corners than a full frame or 1.5x crop sensor because it uses more of the centre where the resolution is highest.

The banding noise pattern is only visible when you boost the lows very high using the built in picture profile. There's a workaround in my review here: http://www.eoshd.com/comments/topic/2120-nikon-d5200-review/   Discussion moved to the review thread.

Julien - This is the D5200 review not a commentary on the GH3 and I should clarify the GH3's strong points which aren't mentioned in the article...   The mirrorless mount means you can use a wider range of glass. Have no Nikon stuff? Then it won't be a $700 purchase, to say the least.   The GH3 is weather proof and has an OLED EVF which you can use during video recording. No mirror to get in the way.   Rolling shutter is less severe.   You don't need to sharpen in post as the resolution is superb out of the box.   Battery life is miles better on the GH3.   It has more physical control, more programmable buttons and 5 dials in total despite the small body.   The moire problems are pretty minor compared to the Canon 650D.

  Well I touched on it in the review but let's summarise this...   I'd take the articulated screen and fixed moire or aliasing on the D5200 any day over a full frame sensor.   The D5200 is also smaller and lighter, and a third of the price!   On Super 35mm there's very few situations where you cannot just use a fast aperture and add a slight vignette in post to mimic the look of a Nikon lens on full frame.   The D600 meanwhile has even worse moire & aliasing than the D800, but with the Mosaic Engineering filter it might be more appealing. I haven't tried that filter yet but the one in the 5D Mark II did soften the image and introduce issues with wide angle lenses. I liked it - but it is a bandaid not a fix.