BasiliskFilm

I am not worried about hundreds of lenses. I am happy mounting adapted lenses for special purposes, and obviously all my manual focus lenses will be fine. If the F mount lenses focus at native speeds, there is not even any need for long wildlife lenses; a Z mount version of those is hardly going to be much more compact.
What I really need though is an affordable, practical, compact, lightweight f2.8 run and gun standard lens like the Tamron 28-75. We know it can be done.

At least Canon video shooters can go mirrorless happy in the thought that they can keep their beloved 1.7 crop...

If the evidence continues to grow that adapted lenses on the Z series work well, then the mirrorless option looks increasingly attractive for video shooters. Sadly many of my D series lenses with lovely aperture rings will become MF only if I go that way, by the looks of things, unless they bring out a screw drive adaptor. For video use I doubt they would ever be smooth enough to use in AF mode anyway.

At least Canon video shooters can go mirrorless happy in the thought that they can keep their beloved 1.7 crop...

If the evidence continues to grow that adapted lenses on the Z series work well, then the mirrorless option looks increasingly attractive for video shooters. Sadly many of my D series lenses with lovely aperture rings will become MF only if I go that way, by the looks of things, unless they bring out a screw drive adaptor. For video use I doubt they would ever be smooth enough to use in AF mode anyway.

The Z6 seems to be staying under wraps for the time being. I guess if it gets launched with fully stable and optimised firmware then that might be a good thing.

So we unfold the optics with mirrorless cameras, and end up folding them in smartphones. I can't keep up
 

You could make a smaller sensor with an effectively longer lens (in equivalence terms). Or you could just make the sensor denser, for improved digital zoom; certainly for 1080p shooting. This certainly won't help with low light shooting, but maybe multiple cameras can work together to improve noise reduction. At the longer end parallax differences are less significant.

I had an Apple Newton and I'm not sure what was the bigger death knell for it fashion wise - me buying it or Steven Segal using it as a recipe book in Under Siege 2.
My patronage also saw off Windows CE, the PalmPilot and the Compaq Ipaq so Light will be encouraged by the fact that I haven't bought an L16 
 

The Light L16 is probably the Apple Newton (https://en.wikipedia.org/wiki/Apple_Newton) of photography. Too many cameras/lenses and not enough processing. But it is probably showing the way ahead, when good small sensors cost a tiny fraction of the price of good large sensors.

I am never cutting edge with phones, but have just got a iPhone 7 plus and have a couple of observations. I have been using it to film barn owls in flight. OK that sounds a bit insane, but the direction of travel with this tech is clear. 1. Higher density sensor with improved digital zoom - at least for decent light conditions and HD resolution, and better processing for noise reduction. 2. Extra lenses - we already have 28mm, 56mm equiv. (I think) so why not 110? Blending the output of more than one camera opens up all sorts of possiblities. 3. Combined optical and digital stabilisation - filming at 6x zoom on my iPhone (is that around 180mm equiv?), one handed, and it is still remarkably smooth. 4. Usable, and suprisingly directional audio from a tiny microphone. 5. Very adequate, intelligent and responsive video AF. 
It isn't going to challenge the top of the range, but it will certainly push out a lot of the mid market competition.
 

I am never cutting edge with phones, but have just got a iPhone 7 plus and have a couple of observations. I have been using it to film barn owls in flight. OK that sounds a bit insane, but the direction of travel with this tech is clear. 1. Higher density sensor with improved digital zoom - at least for decent light conditions and HD resolution, and better processing for noise reduction. 2. Extra lenses - we already have 28mm, 56mm equiv. (I think) so why not 110? Blending the output of more than one camera opens up all sorts of possiblities. 3. Combined optical and digital stabilisation - filming at 6x zoom on my iPhone (is that around 180mm equiv?), one handed, and it is still remarkably smooth. 4. Usable, and suprisingly directional audio from a tiny microphone. 5. Very adequate, intelligent and responsive video AF. 
It isn't going to challenge the top of the range, but it will certainly push out a lot of the mid market competition.
 

I am never cutting edge with phones, but have just got a iPhone 7 plus and have a couple of observations. I have been using it to film barn owls in flight. OK that sounds a bit insane, but the direction of travel with this tech is clear. 1. Higher density sensor with improved digital zoom - at least for decent light conditions and HD resolution, and better processing for noise reduction. 2. Extra lenses - we already have 28mm, 56mm equiv. (I think) so why not 110? Blending the output of more than one camera opens up all sorts of possiblities. 3. Combined optical and digital stabilisation - filming at 6x zoom on my iPhone (is that around 180mm equiv?), one handed, and it is still remarkably smooth. 4. Usable, and suprisingly directional audio from a tiny microphone. 5. Very adequate, intelligent and responsive video AF. 
It isn't going to challenge the top of the range, but it will certainly push out a lot of the mid market competition.
 

Apparently we are only 15 years away from commercially viable nuclear fusion reactors. Funny, but when I went on a tour of the Joint European Torus 35 years ago they said we were about 15-20 years ago from commercially viable nuclear fusion reactors then too. 

Apparently we are only 15 years away from commercially viable nuclear fusion reactors. Funny, but when I went on a tour of the Joint European Torus 35 years ago they said we were about 15-20 years ago from commercially viable nuclear fusion reactors then too. 

Apparently we are only 15 years away from commercially viable nuclear fusion reactors. Funny, but when I went on a tour of the Joint European Torus 35 years ago they said we were about 15-20 years ago from commercially viable nuclear fusion reactors then too. 

I think you are probably right that it is a bit of a myth that cameras have an intrinsic "motion cadence". There are a range of problems with shooting motion: incorrect frame rate conversion, wrong shutter speed, rolling shutter artefacts, jittery lens/sensor stabilisation; some of these come in the class of user error, or can be avoided by the skilled operator. Once these are ruled out, most cameras actually record motion in a very similar way, and I would be surprised if many could spot the difference at 24 fps.

Most folk who shoot 30 minute+ takes are doing it in quite specialised circumstances. If you are leaving a camera on a tripod pointing at a speaker on a platform and pressing record there are 100 other video cameras that can do a decent job. Who shoots more than half and hour shallow DOF video? I am trying to think of the last time I shot more than a minute of anything.

Need to see figures, but assuming that S35 mode on the A7III will have 1:1 pixel readout for 4K, which could be significantly faster than the 6K downsample from the whole sensor. So if you want minimum rolling shutter, and extra reach for action, then crop mode might be a good option.

The 105 f1.8 might be a great lens. If you stop it down you will get heptagonal highlights, with only 7 blades. If you want to shoot at f2.5 then the classic 105 f2.5 AI/AIS might be a better, cheaper and lighter option, with the same 52mm filter thread that the 50mm f1.2 has. It has the advantage, as a stills lens, of being usable with the OVF focus screen at full aperture. Anything faster than that and you probably need to switch to live view to be sure of focus as the focus screen is optimised for f2.8 zooms (unless they have improved in current models)

The end of King Kong would have been epic in vertical widescreen - the whole of the Empire State, plus monkey!

The end of King Kong would have been epic in vertical widescreen - the whole of the Empire State, plus monkey!

If it matches the video of the A99, also has decent AF, and lets folk use their Nikon glass natively, that would be an decent option for those invested in Nikon glass.
On the other hand if they brought out a video oriented D750 upgrade, 24 MP full frame sensor down sampled to 4K (A9 style) for half the money, then you can keep the 40+ MP camera.

If it matches the video of the A99, also has decent AF, and lets folk use their Nikon glass natively, that would be an decent option for those invested in Nikon glass.
On the other hand if they brought out a video oriented D750 upgrade, 24 MP full frame sensor down sampled to 4K (A9 style) for half the money, then you can keep the 40+ MP camera.

I think another model with a 12MP sensor only makes sense if they upgrade the autofocus to match the A9 etc, otherwise what would be the justification for an (inevitably) more expensive model. There is no one snapping at their heels to keep prices down.
If they can read 20MP @ 25fps (as in the A9) then I guess maybe they could read 8MP @ 60 fps, and maybe (lineskipped) 2MP/HD @ 240 fps. The challenge though is writing this much data to the card.

I wouldn't be surprised if they put the A9 sensor in an A7S body for a video oriented release. Given the advances in sensor technology, it is possible that it would be just as good in low light as the A7Sii as both have full pixel readout. The advantages of a 24 MP sensor are clear for photography and video, and give you digital zoom/full 4K at super 35. Maybe even 4K/60p at super 35 if the readout is fast enough, and hopefully better quality for 120 fps. The new longer life battery would be a bonus, but dual cards are not essential as video users are used to dual recording. Improved AF and a more affordable price would make for a very attractive option.

If they could do 8K downsample to 4K they wouldn't use line-skipping in the A7Rii, but as we know from the A6500 they have the processing to perform 6K > 4K.
The size of the photosites is less important than the total area covered by photosites - the A7S has better coverage than the A7RII because it reads from the whole sensor without skipping like the A7Rii. In which case I would expect the A9 to do a similarly good job in low light.