Don Kotlos

I put my money on Samsung. Nothing to lose.

I don't expect more from an extreme grade of an 8bit log footage. Here is the ungraded version:

This is where Sony has an advantage, sensor development. Making their models obsolete might seem like a bad strategy, but what they achieve in the meantime is market domination. I think it is great to have companies innovate as fast as Sony. I was really tired of waiting for even simple things from companies such as Nikon, Canon, etc. I also don't understand why people buy cameras as an "investment", if the camera is not worth ( to you ) the asking price don't buy it. Maybe that will force companies to stop using this aggressive pricing model.

Yep I agree. If you look in the footage from my previous post around 3:18 you can see that motion is captured very well.

"One thing I discovered about half way into the shoot was that the camera has a tendency to overheat when recording in 4K for extended periods" I wonder what an "extensive period" might be at 90F outside temps.

The video is shaky like hell. I can't tell from his words whether he used IBIS or not.

From what I got, he used the S35 mode with speedbooster for that test.

I would have never thought that A7rii can have better dynamic range than the A7s. Not bad at all Sony.

Mattias, I don't disagree that you can push the image quite a bit with an 8 bit log file. What I am saying (and seeing) is that it is hard to reproduce tonalities since color information is just not there. Especially with skin color that we have been trained to discriminate tiny differences i.e. in redness (blushing), blueness (blood oxygenation-sickness), etc. The test that Andrew proposed will be quite telling. Get a frame from a log 8 bit video and try to match it to a perfectly developed raw still from the same camera. From my experience with 8 bit video in general, That will be hard. A log file will just make it harder.

Very informative link. Thanks

There is also a A7rii. But sensors do generate heat.

Lack of detail Terrible bokeh Aliasing WTF Panasonic?

Of course. I hope I didn't imply such a thing. My point was that an 8 bit codec can be very limiting when grading a LOG file. There are plenty of other things that limit the grading too. When the V-Log gets released it will be interesting to redo that test.

Neuroscientist, studying how we (our brain) processes visual input. Could be true too. In the following video (hope Andrew does not block it ) they used two cameras FS7 and A7s. Can you guess which images come from which camera? Hint: compare 1:43 to 1:44. https://vimeo.com/134440445

If they make an internal 10bit camera with less rolling shutter they will dominate the low budget film industry (if there is such a thing...) .

Technical articles will help too. For example we should know that Bayer interpolation results in slightly blurred edges. So some digital sharpening is necessary. The problem is that 1DC and NX1 target a very different market. Professionals will most probably post-process their footage, while enthusiast most probably will not. So Canon made a gradable file, whereas Samsung made a "put me in a Samsung 4K TV" kind of file...

Unsharp (hate these icons).

Exactly.

By soft he means "leave the signal alone so I can apply as much sharpening as I feel like". When sharpness is applied in camera, it cannot be removed. Here some information: "For digital cameras, resolution is limited by your digital sensor, whereas acutance depends on both the quality of your lens and the type of post-processing. Acutance is the only aspect of sharpness which is still under your control after the shot has been taken, so acutance is what is enhanced when you digitally sharpen an image (see Sharpening using an "Unsharp Mask")." http://www.cambridgeincolour.com/tutorials/sharpness.htm

I believe most of you know this but for the ones that are not that keen with the idea let me explain. The last few months I see a great deal of enthusiasm for LOG profiles. I also see a lot of people complaining about colors when trying to grade these profiles whereas specific cameras such as the Canon 1Dc can give great results. As Andrew has mentioned many times there is a great deal of effort put into customizing not only the colors (how signals from the RGB sensors are combined) but also the LOG profile to a specific camera. It is similar to the effort that camera companies put into their JPEG algorithms for stills. For example you have companies such as Fuji that have great experience with color and can provide very usable JPEG profiles. That is not always the case with other companies. Now when shooting RAW (>12bits/channel) you can most probably get any color you want. Yes it will require quite a bit of effort but you can do it given infinite amount of time. Of course time is limited, as is space - recording space. Here comes the 8bits/channel (and reduced color sampling). Everyone knows that when shooting with a LOG curve you map a greater range of input values to the image. In such cases skin tones are mapped into a tiny fracture of the color spectrum. To give you an example the variation of skin typical pixel intensities in RGB space (to make things easier) is usually +/-10 for each channel: 20x20x20 --> 8000 colors. In this space you have to accurately map the tones of the skin. Yeah right. **PERSONAL OPINION** The most grades that I see with A7s & GH4 log profiles end up with plastic skin colors. I wonder why... Kudos to Canon. Do yourself a favor when you are using any log profile and record with a 10 bit codec. Here is a great article that dives a bit more into it: http://www.newsshooter.com/2015/07/27/looks-picture-profiles-luts-and-log-why-when-and-how-you-should-use-them/ P.S. I am not an engineer and neither I am a professional videographer so I could very well be totally wrong about all this. But I am just fed up with the boxing match threads between different cameras. Let's move on.

I have played extensively with different ways of getting a 1080p 10bit 4:4:4 from a compressed 4K 8bit 4:2:0 file and from my results, working with the original 4K file just until the final render always got me the same quality. Maybe with a relatively uncompressed 4K file the results might be different. There is no way around it, camera companies need to get that 10bit on consumer products.

I believe Ebrahim covered that pretty well. I only disagree on the analog noise which I find more random with less color.

An important thing to test is given the overheating problems of the RX100IV whether external recording will help.

Indeed. ​