Andrew Reid

Yes but once you start getting down to such tiny resolutions, can you properly measure the results?

What camera does real acquired 4:4:4 though? Not many, especially not at $1999.   Let's see the end results and compare to 10bit 1080p 4:2:2 on other cameras and then judge it.   The theory holds tight mathematically.

Be sure to check out the GH4's new trick - internally recorded '1080p' 4:4:4 with 10bit luma, at EOSHD here One of the first to receive a Blackmagic Production Camera in the UK (hoping to get mine next week for a review, shipments allowing) is James Miller, Philip Bloom's friend and frequent shooting partner. As this impressive footage shows even though the camera is designed for general production rather than cinema, I am not missing that BMCC dynamic range here. The footage looks wonderfully cinematic and the camera appears to be a powerhouse of image quality, given the right handling, good light and material. Read the full article here

Danyyyel, it's no joke. Name one Nikon innovation for video since the D90's shoddy 720p heralded others to do better....   As for the things you listed, they're not so unique on their own, but yes somehow the D5300 has accidentally ended up with quite a lot of them.   No moire / aliasing - see 5D Mark III, which came out before the moire & aliasing Nikons. Nikons were a veritable moire and aliasing PARTY before just around 9 months ago when they finally got wise. Shame D800 did not come out this year instead, maybe they could have done a better job.   Frame rates 24p to 60p... numerous other cameras had this a long time before Nikon (2-3 years). Sony were first to 1080/60p. Panasonic I believe offered 1080/24p first with the GH1 before the Canon 5D Mark II firmware update.   Very good DR... I'm not seeing it myself. Way more on others, such as Blackmagic cameras and even the tiny Panasonic GM1 does better in the highlights.   Very good low light... It's not hugely different to others. 5D Mark III is better and GM1 is near enough identical but with more fine detail.   Proper S35 sensor... yeah, but crappy lens mount which I can't use 80% of my glass on! 1.4x crop with Speed Booster on GH3 = larger than S35.   Clean HDMI out... been around since GH2 times. For external recorder this output on the D5200 and D5300 like on other cameras brings little-to-no real image quality increase but for a lot of hassle. Wait until it is 10bit 4:2:2 or 4K then it will actually be useful.

The Panasonic 4K capable sensor... rumour has it that this is present in the GX7, GM1 and Olympus OM-D E-M1 not just the 4K GH4. But the key thing is giving it enough processing power, faster SD card slot, cooler readout circuitry and avoiding other heat issues, plus a few more things I don't know of because I'm not a camera engineer!   So firmware updates for 4K are probably out on the other cameras despite the powerful sensor. Well done Panasonic for making such a great follow up to their GH2 sensor and avoiding a Sony chip.

Previous discussion thread: '?do=embed' frameborder='0' data-embedContent>>   Discussion continues here, thank you to all contributors so far (and please let's stay on topic as I feel there's a lot to be gained from this info...)   1D C - yes same should apply to that. As for the actual workflow technique, perhaps others can chip in on how they would go about doing this... I have never tried it myself.

Another thread merging moment. Please continue here as before, very welcome discussion   '?do=embed' frameborder='0' data-embedContent>>

Cinema quality is a very broad term, and since GH2 and 5D shot films have been shown in cinemas, let's not even go there ;)

The end results do matter more than specs but the D800 is old news guys!   Here we have some pointless stats on raw stills on a video forum, no real new info, no understanding of how the video quality truly rates on the D800 (below D5300 and 5D Mark III) and therefore I'm afraid it's time to close the thread to free up room in the EOSHD zoo for a better giraffe. Bye marius.

Hazy and soft at F1.2, you will notice that the bokeh is nicer than a F1.4 50 (wide open) but the overall image quality much worse. Main problem is purple fringing. And stopped down they perform just like a much cheaper FD 50mm F1.4. Consider the Olympus 55mm F1.2 instead.   The exception is the old FD 50mm F1.2L which maintains good colour and contrast even wide open, and it's very sharp. But pricey.

Have you Birk aside from DXO mark stats got anything more, erm, useful, to add? Otherwise thread closed.

Why would I need to see Dexter? I know enough about the D800's image from first hand experience to know that it falls massively short of the Alexa.   Whatever they are doing in post, whatever they are doing in the cinematography and lighting, to cover up for the D800's inadequacies has nothing to do with the argument here.   Your argument Birk is that the sensor does 14.4 stops of dynamic range in stills. That's what DXO measures. Raw sensor data.   Line skipped compressed 8bit video is another matter.

I think Nikon lobbied for the D800 to be used on Dexter so don't read too much into those claims. It's still a very cut down video feed from the sensor.   You can compress those 14 stops from the sensor into an 8bit codec with a flat picture profile, but then when you grade it you get major stepping, poor tonality and colour, the exact opposite of what 14 stops in a 10bit 4:4:4 codec would look like.   There's just not enough headroom in the codec to do it justice. That's why most flat profile stuff has odd colour, weird looking highlights, crappy low contrast, grey shadows which should be dark, etc.

Seems possible guys   http://www.eoshd.com/content/12140/discovery-4k-8bit-420-panasonic-gh4-converts-1080p-10bit-444

Pros are wondering what the benefit of 4K is to them in terms of overall image quality, when mastered and delivered for 2K / 1080p. A lot of work is still shot in 1080p and cameras like the Canon C300 are the workhorses of the moment. In the case of the GH4 it may appear from the specs that it's just an 8bit 4:2:0 camera internally. Actually the theory is 8bit 4:2:0 4K material from this camera can be taken through a workflow in post that converts it to 10bit 4:4:4 1080p - with all the smoother tonality, better colour and workflow advantages that format brings. This is a big leap for 1080p based on the much more expensive C300 which only does 8bit 4:2:2. I asked Go Pro's David Newman (Sr. Dir. Software) whether this theory was correct... Read the full article here

So bit depth stays around 8bit but sampling is indeed 4:4:4 after conversion?

Thanks for the K-3 info. I'll think I'll pass on that one then!

  This is exactly what I suggested to Panasonic in our meeting :) "Super 4K'. A near 5K camera in a sea of 4K cameras this year, would be good for sales as well as for us.

  Only if you want 4K 4:2:2. HDMI on the GH2 does 4:2:2 in 10bit for 1080p to the Ninja.

The idea is to oversample from the 4K, so you build better quality pixel data from 4x the data available in a normal 1080p stream.   I think it needs special software to do, and yes some transcoding to ProRes 10bit 4444.   But even with a basic downscale in Premiere or FCPX you are smoothing out any 4:2:0 artefacts in the 4K like aliasing on brightly coloured red, blue or green edges for example.   4K is like a big pipe from the sensor to the card.

Oh and anti-aliasing simulator, though not sure what affect if any this has in movie mode. There's no evidence of moire and aliasing on the K-3 videos I've seen so far, so you probably won't need to use it.   Check out the 4K time-lapse footage here - vimeo.com/search?q=pentax+k-3   And download some original 1080p from this stabilisation test (not inspiring but look at the detail... looks sharper than D5300 and more organic?) - >click through to download don't watch it below

Speaking of Pentax the K-3 has the same if not identical video quality to the D5300 from what I've seen of it so far. I haven't tried it first hand yet so don't take it as confirmed 100% but it looks to have the same 24MP sensor - BUT with two very neat features. First is 4K time-lapse mode in MJPEG format (3840 x 2160) and second is in-body stabilisation. Don't think the stabilisation is effective as on the E-M1 but the camera itself seems pretty nice. Proper pro-DSLR body, weather sealed, looks better than the Nikon D7100 for video thanks to those features mentioned above but you do lose 1080/60p and the tilty flippy screen from the D5300.

  Still looking at cheap cameras objectively, it's mainly about the year we're in, 2014. Indeed if the D5300 had come out when the 5D Mark III had come out of course I'd have been more positive about it. Such little innovation will no longer cut it. Look at it another way... if I had glowingly recommended the D5300 this month, in 2 months there would be a lot of upset people with buyer's remorse come NAB 2014. Wait to see what's coming.

  Yes please, vs 5D Mark III stock ALL-I codec with flat profile too.

Let's be clear...   You can flatten out the image... iron it right down to such a low contrast that you basically have LOG or CineStyle and 12 stops dynamic range on a DSLR.   And you then have no choice but to add contrast back to the image with the grade and that loses you almost all the dynamic range you just gained because you're doing what the image processor did instead automatically, crushing the blacks so they are black and making the highlights bright again. Unless you keep the flat image as it is, which looks... dreadful.   So let's be clear by what we mean by dynamic range... I'm talking about the final end result. There's a big difference between flattening out an image, removing all the nice tonality and colour in the process, and getting those extra 2 stops, and to having 13 stops in your FINAL GRADED IMAGE with 14bit raw.   To have 12 or 13 stops as your starting point with LOG or a flat image is one thing, to have the final end result look like it has great contrast, colour AND a wide dynamic range is quite another...   With the necessary hard grade of flat DSLR footage, the codec just falls apart.