Don Kotlos

Since you are using premiere your best bet is adobe media encoder . You can even automate it with watch folders. In the next premiere update there should be also a proxy editing function that does this automatically.

DNxHD/DNxHR no question about it. Cineform YUV 10bit also works great. Prores is currently unsupported and should be abandoned as soon as possible.

Hard to compare the two videos. They were not shot at the same time, so vibration & movement patterns can be very different. Stabilization looks good though. I guess having the second video helps illustrating the importance of stabilization with lightweight handheld cameras.

My opinion is that you shouldn't expect too much improvement with ProRes if Premiere is using GPU acceleration for mercury playback. The reason is that your GPU is your bottleneck and not the CPU. If Premiere was not using your GPU then yes, CPU friendly codecs such as Prores would help. Lowering the resolution helps with real-time playback of the effects (here more details), and truth be told you should only have it at 100% resolution only if you have a full screen preview on a second monitor. The advantages of the Prores codec will be smoother scrubbing (see previous link), and a 10 bit environment that can help with some effects.

Quantum efficiency is very good indeed but in only affects low light sensitivity and not dynamic range. Dynamic range depends on full well capacity & read noise, which on this camera gives ~93db (~15 stops) of dynamic range. Pretty good but typical with scientific cameras. The biggest problem (other than the excessive prices of most scientific cameras & the need for a PC to run them) is that it doesn't offer global shutter that you can find in other scientific cameras.

I agree with Phil. Surface book might give you the best performance for porxy editing because of the d GPU but you will need a desktop even to create the proxies and render the final clip, otherwise you will end up slowly cooking your tablet to death.

Yeah I wouldn't try this on a >$2000 camera. Maybe when they implement RAW video I will be more motivated More coverage:http://www.newsshooter.com/2016/04/25/hacker-claims-to-have-removed-the-30-minute-record-limit-from-alpha-cameras-including-the-a7s/

And global shutter

12 bits should be enough for 14 stops even when spread linearly. Most photography cameras in the past had 12bit files and even my a7rii shoots 12bits in continuous mode. From what I read the problem with FS7 was putting a 10 bit signal in a 12 bit container. I guess that's why they could fix it in the first place.

It depends on the camera. .. For EM1 the 25 0.95 SLRMagic almost never left the camera. For the A7rii the CV 35 nokton gives me everything I need, range (35-105), macro, low light & shallow DoF.

4K is overrated. Especially since you started taking videos last week I wouldn't jump into a new camera just for that. For example I found the stabilization that EM1 offered was far more important in creating nice looking videos since my style is handheld with manual lenses. Others might have a need for good AF during video, or ability for a shallow DoF or lowlight performance or large dynamic range or .... All these things take time to find out and in the meantime 6D is a very capable camera that you can use to learn a lot. Shooting technique, framing, audio, editing etc can make a much much bigger difference than 4K. New cameras are always going to come and get cheaper & cheaper so no need to rush for something that you want but might not need... Anyways G7 is a great camera.

Try another color mode, it will look even better

A big part of the look depends on the lenses and their coatings. With D16 it is highly more likely to find footage with vintage s16 lenses that have this distinct look. Then there is also the sensor that is slightly more bloomy and that appears more with very bright scenes.

Banding depends on multiple factors such as compression, bit depth and color sampling. Most critically it depends on how much of the tonalities you want to recover from an 8 bit codec. That is why less aggressive profiles will always give you less banding, no matter what the LUT is. Now I agree that correct interpretation of color in s-gammut is indeed critical in order to reduce the visibility of banding (which is apparent in both of the examples and just less visible in the second one). Since banding can also depend on the codec, if you have an external recorder that you have tested in order recomend it: It could be great to show these results as well. Cheers

RED issues a firmware update to unlock DNxHQ/HD recording in their cameras. More companies should follow. That should take care the prores bias in the industry and relying on crappy programs, such as QT, to decode proprietary formats. Another critical step is that Davinci resolve has to be patched to get rid of QT dependencies for most codecs. Here is an article from cinema5D.

Most often video shot with high gain looks bad not because of the noise at the pixel level but because of the shitty noise reduction that is applied in camera. So while internal recording might have less noise because of heavy noise reduction, the RAW files will still look better because the noise is at the pixel level and "organic" (random) and not the ugly macroblocks that internal codecs with low bitrates produce. Moreover, the noise reduction that you can apply on the RAW file will be far far better than any internal noise reduction any video camera has. In short, external recording will indeed help make the footage look better. I don't know with FS5 but with the BMPCC there was at least 1 stop advantage relative to the prores footage. Now, if you want 1080p at the end you have 2 options: shoot at 4K recorded externaly and downsample after you apply noise reduction in post. shoot at 1080p and use a cheaper external recorder such as the atomos ninja star. Additionally, for if large DoF is not necessary, you can get a speedbooster to gain another stop right there. Also I don't know what footage you saw, but after the recent firmware update there are reports of better quality with high gain (https://vimeo.com/149383051) I bet Andrew will soon have more information on the low light abilities of FS5 with the external recorder

As far as I know, 6D does not have continuous AF in video. And even if it did it would be contrast only and wouldn't work that well. Manually focusing is your best bet, but don't worry you get used to it and it becomes easier (and more efficient than most AF systems) after a while. Shooting with a smaller aperture (large f#) also helps! I don't know what premiere defaults to when you first use it but it looks it has the wrong aspect ration and a low resolution. Try h.264 format , 1080p resolution (1920x1080) and a ~20 bitrate for better results.

How great would be a 4K ninja star, or even a 4K ninja blade for run&gun type of shooting. >=5" monitors/recorders are just too big.

Sorry I meant the micro Studio camera. So confused with all these names...

Wait is the Micro Cinema Camera out of the question?

Yeah it's funny that Cinema5D has an article on the Craft Camera that is just a design concept but not for the Terra that actually works. Good thing that newsshooter had a somewhat extensive coverage.

I am glad you worked hard on this and I am eager to try out your LUTs. If you have done tests that show the importance of an external recording, it would be great to share the results. There has been conflicting reports on their advantages with the A7rII/A7sII.

I talked about the whole video/NLE market not the just the filmmaking industry. And that is general percentage but it is an educated guess based on actual data. Funny thing I am the only one providing actual data. To both of you, you missed my point that's fine but I am done participating in this silly discussion.

From the data that I linked before ~30% of operating systems is windows 8,8.1 &10 all of which are unsupported.

I see you present these percentages with great confidence. Do you have a source for these numbers? The only percentages that I could find is the ones about Mac having less than 10% of the operating system market here. Also as far as the NLE goes only ~20% is the mac specific FCP and ~14% is the PC specific SVP from here. So you are telling me that >80% of premiere users use Mac. Yeah right. (Don't forget that resolve only recently became really a usable NLE system). But whatever it seems that you missed the point that I wanted to make about lack of prores support for many users and instead decided to be condescending. Good for you. For video acquisition and post production. It is from the Society of Motion Picture & Television Engineers.