cpc

Does anyone know how's RS in 3K on the 6K sensor? Same as 6K or faster?

Unnatural is part of the appeal. These days you don't shoot 24p for naturalism, on the contrary. Many TVs have a 24p mode, so do some graphics adapters for PCs. Of course, it is best to have choice. Dogmatic obsession with an acquisition framerate is as bad as any other obsession. Also, in Europe a lot of cinema is actually shot 25p because movies also go on air eventually, and pretty much all relatively recent DCP servers handle 25p fine (and 30, and 23.97, and 29.97 for that matter), so there is no need for speed canges and resampling audio.

Probably not a good idea to look at 5" for focusing. Most focus pullers will say 7" is small, let alone 5". 5" is more of a framing aid.

3D is terrible in traditional 24 fps. 2D and 3D are very different forms and translating into 3D any affinities cultivated from 2D makes no sense. While I mostly don't care about 3D, HFR at least makes it bareable and not just a mess of motion artifacts striving to look "real". I wrote this a few years ago, when The Hobbit's HFR debate happened: http://www.shutterangle.com/2012/why-48-fps-is-good-for-3d-movies/ In this context I can only applaud Ang Lee's perspective.

He says 4K. That makes the GH4 the only contender.

I think there are still some patents involved in Cineform for encoding outside of the Cineform GoPro installation codecs, even though it is SMPTE standardized. DNxHR in BM cams is probably a more likely development.

It certainly isn't Apple's fault. Users demand QT native codecs like ProRes, camera manufacturers listen to them, and software vendors follow up by integrating QT. DNxHD/DNxHR is essentially the VC-3 SMPTE standard and plays very well with another SMPTE standard -- the MXF container. Cineform is also a recent SMPTE standard (VC-5), and defined for MXF wrapping. It's not like if ProRes+QT is the only player in town.

Raw DR is more or less the same. Let's just say that FS7 raw is processed in such a way that it uses around 1 bit less than a FS700.

You don't need an interface unit with the fs700, only with the fs7. Fot whatever reason there is different processing on the FS7 raw. FS700 raw is denser. FS5 raw better be similar to the FS700.

Some cameras have power out from the camera: Bolex, Sony f5/f55/fs7+XDCA, etc. I imagine this will play nice with these. An HDMI option would be nice though. Also, Zacuto last gen EVFs have 1:1 setting, so seeing a pixel perfect image up to 1080p should be fine. They don't support a 4K input, so even if it was a 1080p display you wouldn't see the 4K image.

I got the Minolta TC-1 for a pocket camera a couple years ago. Probably the best camera purchase I ever made. Ultra pocketable, exceptional 28mm lens, spot metering, full frame (:P). A joy to use. Not cheap though. Between the TC-1 and the Fuji Klasse S I am seriously playing with the thought do I even need my Contax G2, Zeiss Ikon and Leica R when I almost always reach for one of the compacts.

1.7gbps means it is compressed internally while recording. Uncompressed would be starting at 2.4 gbps for 24p and going up with framerate. But I get your point - you are looking for lossy compression.

You can choose uncompressed DNG or losslessly compressed DNG when you export the footage. Their lossless implementation is far from optimal, but it's there; around 1.6:1.

Well, 30 minutes is almost 30% more than 23 minutes which is a significant difference when footage starts to add up. It is certainly another story for oversampled video (i.e. 2K from 4K or bigger sensors), but for full sensor resolution shooting ProRes444 is usually a handicap in terms of size. With the push for 4K that's happening now more and more cameras are recording (near) full sensor resolution in ProRes without any real benefit other than no need for raw processing.

ProRes4444 is significantly larger than Red raw 3:1 for the same resolution and framerate. In fact, ProRes4444 is larger than pretty much any lossy compressed raw codec out there, and this includes Red raw, BM lossy CinemaDNG raw, Cineform raw, etc.

5dm3's raw image is gorgeous. The fact that ML has tools missing in much much more expensive cameras (or any camera, for some of them) is the icing on the cake. The only substantial advantage for the A7sII would be its sensitivity, but for controlled shoots - 5dm3 all the way.

Picture Profile OFF is not a mode. This a setting that turns off picture profiles and uses whatever creative style is active at the moment.

I am sure the PIX-E is great, but this "explanation" is some hilarious BS.

A7s is actually really nice RS-wise in APS-C mode (similar to 5dm3, which does binning). And it is still full readout of the 1.5x crop area, so it's also pretty good in terms of aliasing. FF puts a lot of stress on the system cause there are infrastructure problems to resolve which depend on sensor area. FF will always have more RS than APS-C which will have more RS than smaller sizes (like 1") for the same number of pixels.

Yeah, but this is us and our wishful thinking. Sony people probably still see this as a photo camera with video added in. If they were positioning it as a hybrid (like A7s/A7s2) it would be a 12-16mp camera and wouldn't have these RS issues which the excessive pixel counts bring (and would likely cost a bit more). But it is still possible that they release such a hybrid. Not sure how Sony operates but my understanding is that photo and video divisions are separate. Hopefully they aren't THAT separate though so that features move in both directions. I'd certainly love an FS5 II with great AF and IBIS. Or an A7s III with electronic ND and great AF.

One of the problems is that this consumer camera has features that not some but all more expensive large sensor Sony cameras are lacking. Like decent AF. It will likely propagate up eventually, but it certainly can be a cause of frustration: putting it in a camera that fails spectacularly in other areas. And this does reinforce the image of Sony never being able to do everything right, there is always user frustration coupled with their technical inovations. The fact that this is a photo camera in the first place doesn't really help much on video related forums.

Both FS7 and FS700 output processed raw, but the FS700 raw image is denser ( = more color info in there). Also, FS7 bakes WB in the raw image, FS700 doesn't.

My dream camera has the electronic ND of the FS5 and the AF of the C300 mark II. Now only if Sony would catch up with Canon on the AF front.

The (relative) tint you see comes from the difference of native white color temperature between the displays. This can usually be matched fairly well to the 6500K standard even with cheap displays, although you will limit the brightness if the native white is way off (low brightness is rarely an issue anyway). The more challenging problem is getting the various colors in the ballpark. First, the display needs to be able to cover (almost) all of the color space you target, otherwise doing calibration is pointless. It is best to have a display with hardware calibration LUTs, this limits banding. The cheapest displays don't have any hardware LUT support. Moving up, you get 1D LUTs (I think some Dells are the cheapest you can get with 1D LUT support). Better displays will have hardware 3D LUTs. It should be noted that calibration doesn't guarantee correct colors, especially with 1D LUTs. In my experience with various low-mid range LCD displays, blues are usually the biggest problem even after calibration. Some calibrated displays oversaturate low saturation blues and push them to cyan, some undersaturate high saturation blues and push them to magenta. It seems that there is an inherent issue with blues and LCDs; I recall reading an explanation about this issue somewhere by an HP guy talking about the Dreamcolor (but it is related to all LCDs).

Dunno about Premiere but 2GB video RAM is not enough for 4K in Resolve. It should be fine for 1080p though.