jgharding

Perhaps try ticking the maximum bit depth option at export, it processes the video in float point. If you untick it it will be processed in 8 bit space.

When editing, timeline video tends to play back at draft banding and scaling unless you tick max quality and max bit depth in sequence settings, which could be the cause.

You may find that such things aren't on your exports if you use maximum settings, even if they appear in the timeline draft mode. Certainly I have no such problems.

Many After Effects users complain of banding in their animations, until they realise they need to render in something other than 8-bit for the engine! It's often missed by users.

Talking about shooting yourself in the foot...

There's no reason to remove Magic Lantern, it actually sells camera bodies. Those who don't want it ignore it.

Those who would install it go elsewhere if they can't install it. For Canon, blocking it would be lose/lose.

C100/300/500 do a downscale from 4K in the body. If you record that from the HDMI ot's pretty damn sharp.

C100 MKii is the first to take all three 4K channels, then combine then downscale. So it appears it's even sharper 1080, and will upscale to 4K pretty well!

C series are very sharp, sometimes it feels like too much. I'm not yet delivering anything 4K so it doesn't bother me for now! I'd say if it's a huge effort for little reward don't do it, and if you can see the difference and feel it's worth it, do.

Lovely job! The first few shots in the forest are epic, and I do love the film look.

A7S seems to lend itself to a very filmic tone I think.

The musician reminds me of Daughter...

It's a shame the Shogun is so big! It makes the package look quite unwieldy

You always use post, but the material you take into post affects where you can go, and how far.

Canon starts off very over warm, but there's a lot of wiggle room to remove that, so I like it. With others I find it hard to convincingly add that warmth if I want it, so I like Canon quite a lot.

Arri just nailed it for natural "do anything go anywhere" skin, amazing. So really the Alexa and Amira are the winners, at a price. That's one reason why so much film, TV, advertising, drama, so much everything are shot on them.

Red is good too, though there's sometimes a funny green/magenta tint that I find hard to knock out in a consistent way in post. I find it less forgiving.

Sony is very cool feeling but quite Kodak'y so not at all bad I don't think, just different. The rest I don't really have the experience to call.

That's my two pence from my experience, but YMMV of course!

I saw one of the earliest big 4K monitors at NAB one year.

The shot was of a Samurai walking down a gravel path. I quickly realised I was staring at the detail of the gravel rather than the actor's face!

I also saw plenty of Sony 50p and 60p that year, and found it plastic and dead. Too much like real life.

Since then I never really obsessed much over resolution or HFR, in fact I often blur footage slightly on purpose! It feels dreamier, and the best movies are kind of like waking dreams I find...

Wildlife, sport, reporting and some kinds of documentary suit both well I'm sure.

Even at £10,000 I wouldn't buy a C500. It needs an external recorder to do anything impressive.

Canon camcorder division just need to stop playing games and actually answer the FS7. That's the only piece of kit around now that says "buy me and your work will look better and I will be worth more to you than I cost", at the business end of the market.

As an addendum, and slightly off topic (apologies) the debate about the quality increase when increasing sample rate never ends. When mixing many tracks together 96khz allows frequency interactions above hearing range to be blended in a non-truncated manner, resulting in a final mixdown that's less congested, though it's quite subtle really. It is worth doing though.

interactions between high-frequency elements can create peaks of frequencies beyond the capability of the original capture rate, and slicing these off by working a mix at 48khz can create a bit of a tight feeling -- a lack of air -- so basically record at 48 and mix 96 will do you good.

quad rates like 192khz are pretty excessive unless you're altering audio speed. Some argue this point. Dan Lavry wrote a pretty deep paper on it. If I'm honest, the DSP maths in that paper is lost on me. but a 96khz environment is beneficial even with 48khz sources, audibly so with good monitoring.

Bitdepth wise, 16 bit dynamic range is fine for delivery. For recording, do 24bit. Your software will probably mix in 32 or 64 float by default. If not, set it to 32bit yourself.

​in order to get audio at higher speed you need to multiply the sample rate, the audio equivalent of frame rate, in order to get smooth slow audio.

if 25p is 48khz for example, 50p needs to be 96khz, and 100p needs to be 192khz. This requires a capable AD chip and similar code etc.

most camera companies don't bother.

Looks beautiful! It's the first time a company other than Canon has produced colour response that nice in a consumer camera. I think it'd cut with Cxxx footage well

Im sure h265 will be in Premiere in not too long! The only thing missing here is some kinda stabilisation in body (I'm being greedy)

id love to see some 120fps examples. In fact a slow motion shootout with a ton of cameras would be a great article

I must say Hitfilm looks great though, I'll trial it, thanks for the tip!

Premeiere + Speedgrade, Filmconvert, Looks, Colorista II, LUTS and grain loops all in different blends.

no cages here. Cheap shoulder rig and tripod will help you keep a small body still, I use Canon C series so no need for cage really.

Red files don't require transcoding in Premiere. Drag them in and play.

i nearly went for this too. The only issues were size complexity and PL mount.

lenses are expensive, the whole thing weighs loads Especially with big Anton Bauer batteries or similar, but the image is brilliant and gives you an edge. I like to work quicker and lighter than Red allows so I went against it, but do consider it. Hire one and try it out!

I guess it's just much harder to stabilise full frame compared to four thirds.

It's a much larger area and probably much more difficult to eliminate rolling artefacts across it, without greatly increasing scan speed, that is.

You'd probably need some clever software stabilisation too, analysing the movement sensor and trying to counteract rolling shutter too.

Surely it must work with legacy lenses? That's most of the point... if not I am disappoints

I have no faith, i have wandered from the path. I use C100s for nothing but convenience (plus Ninja for lovely quality), and unless Canon rise again and roll away the rock an FS7 will come next.

I loved my 550D, but all things must die...

The 5D MK iv will not be an inspiring product for video makers. I'm happy to state it outright and wait to proven wrong. I'm sure I won't be.

Hmm it always takes quite a while for a standard to settle... they probably jumped the gun a bit with H265, I'll bet there's plenty of optimisation work to be done at all ends of the process.

The sensor sounds cutting edge though. Is there an uncompressed out?

The original was sadly completely serious.

I feel that's the funniest outcome and thus should be true and so believe it!

I can't find whether the original was a parody or not... Did it parody the idea of parody, by being realistic enough not to appear to be a parody?

So... many... irony... layers...

How can the sensor output 6k video (or even 4k video) when it only has 4.44M active pixels?  That's more like 1080p-class resolution. The only way I can see it outputting 4k video is if it upscales its 2166 vertical lines to 3180, which can't help but look nasty.

assuming Sony use their RGBW system, creating a red, green, blue and white pass using an electronic filter:

4.44M * 4 = 17.76M = 18MP

18 megapixels is about 5760 * 3240, roughly 6K

So the maths points to a very high colour depth indeed, combining full colour and a monochrome channel. Of course, they're then having to stretch this out to 6K final spacial resolution, so like Bayer, there's trickery involved.

The only way to get true 6K (as with any resolution) is either user three 18MP sensors stacked, split by prisms, or use one Bayer sensor with four times the pixels. Still, it could be the best picture ever... ;)

My guess is that where Foveon uses stacked sensors, this effectively uses  RGB pixels side by side, hence the pixels being wide.

The filter will work at an extremely high speed and the three images will be blended together to create full RGB image. 

Yes unfortunately it is horrible. The LCD has a lot of options for brightness and so on though, and can be used in bright light.

Interesting! Great to see new non-Bayer sensors.

I love the image of Sigma Foveon, glad someone else is going for a similar full RGB technology.

It was well shot, but the original was so pretentious and narcissistic, many thought it was a parody. This parody though, manages to parody it:

I've been fooling around with using DNxHD 220x with Ninja 2 attached to C100 MK 1, and for C-log it's pretty incredible how much of a difference it makes. You gain a ton of detail (yeah even more!) and colour separation, secondaries work incredibly well... you basically gain a stop or so of usable dynamic range as you don't lose all the detail in the darks. it's quite brilliant.

The price of C100 MK 1 and Ninja 2 are both dropping so in combination it's gonna be really good for indie film making, just jack the audio into the camera too! I don't mind using the AVCHD for reporting and that kind of thing, but the DNxHD has ten-times the data and whips it into touch if you want to do post. IT ain't 4K, but it'll look damn good.

For promos and slow motion stuff though the FS7 is a much better bet, but it'll cost you twice as much. OR you can hire it for 150 a day. It seems it maintains the Sony cool, Kodak-y look, which is certainly a look. It's one i like, though it's not as friendly as Canon.