cantsin

No, the Nikon V1 shoots 30 full-resolution raw frames per second. See here: https://vimeo.com/groups/nikon1raw

The secret, in most cases, is using Neat Video Pro (especially its temporal noise reduction). It works wonders on any uncompressed/raw footage, especially if you use it as an OFX plugin in Resolve directly on CinemaDNG footage, with no previous transcoding.

It's an eight year old sensor. Although it was one of the best APS-C sensors back then, its ISO sensitivity and dynamic range aren't up to the standards that we take for granted today. Compared to today's best APS-C sensors, the 50D gives you about half the ISO sensitivity, two stops less dynamic range and 2 1/2 bits less color depth. (The raw codec compensates for these shortcomings a bit, but still can't give you any image information that the sensor hasn't been able to record.) https://www.dxomark.com/Cameras/Compare/Side-by-side/Sony-A6300-versus-Samsung-NX1-versus-Canon-EOS-50D___1072_976_272 Even the 1" sensor of the Sony RX100, with 25% the surface of APS-C, outperforms the 50D in most areas: https://www.dxomark.com/Cameras/Compare/Side-by-side/Sony-Cyber-shot-DSC-RX100-IV-versus-Canon-EOS-50D___1033_272

Sorry, but that's nonsense. Panasonic collaborates with Olympus, both companies jointly created the Micro Four Thirds system. And both have divided the M4/3rds market with Panasonic focusing on video hybrid cameras and Olympus on traditional stills cameras. Sony has more market share in mirrorless cameras than Panasonic and Olympus combined. Olympus even makes losses with its cameras that it only compensates with its medical imaging revenue.

The camera industry is currently fighting its crisis with steep price increases - in a dramatically shrinking market, profits can no longer be made through sheer volume, but through profit margins per sold unit. Some indicators: The official price for a Sony a6000 body official price was $650 when it was released in 2014, the price for a Sony a6300 body climbed to $1000 in 2016. The Sony A7s was released in 2014 for $2500, the A7s II in 2015 for $3000. Nikon has increased the prices for its lenses by 18% this year. So: the GH5 promises be an as substantial update to the GH4 as the a6300 was to the a6000; the price increase from the a6000 to the a6300 was 54%. comparing Panasonic to Sony makes sense since both consumer electronics giants generally compete in the same markets, and are major competitors in the mirrorless camera market like Nikon and Canon compete in the DSLR market; the GH4 body cost $1700 when it was released in 2014, an initial price of up to $2600 (53% increase) for the GH5 body could be possible, and wouldn't be excessive.

Oh, and I forgot built-in ND, a killer feature which the GH5 won't have...

10bit won't be enough for the GH5 to obsolete cameras like the FS5. Some limitations will remain: the MFT sensor will have about one step less light sensitivity/twice as much noise as a S35 sensor, the body won't be as robust (and not as optimized for video shooting), audio interfaces will be less professional, and the image will be more geared to amateur demand (with no possibility to completely dial out artificial sharpening).

Another point in favor of Canon: robustness.

You're aware of the fact that on APD-C, 35mm is normal focal length while 50mm and 85mm are tele lenses? For vintage lenses, the standard recommendations are Contax/Zeiss, Leica-R or Nikon lenses, alternatively any other quality Japanese camera brand (Pentax, Olympus, Fuji, Yashica, Minolta, Ricoh, Konica, Canon...). I would stick to only one brand for optical consistency. Alternatively, you could buy modern Samyang lenses or the excellent Sigma 18-35 and 50-100mm APS-C zooms.

I think that this is unlikely to happen since the specs only mention conventional SD and CF card slots. Like with the 5D MkII and MkIII before, the storage controller (along with the uncompressed raw image signal from the camera) creates the bottleneck for higher frame rates and/or pixel resolution when recording raw video.

After seeing this... ...I really wonder whether you need any more than two lenses - the Sigma Art 18-35mm/1.8 + Sigma Art 50-100/1.8 - when shooting video on S35/APS-C sensor or speedboosted MFT and S16 sensors. The 18-35 already obsoleted my 24mm, 28mm and 35mm Ai-s Nikkors + DX zooms. It's optically much better than any of those. Now it seems that the 50-100 will also obsolete my 50mm, 85mm and 105mm Nikkors. Get two bodies of whatever camera you use, mount the 18-35 on the first body and the 50-100 on the second, and you should be able to cover almost any shooting situation. And get peace of mind because you'll no longer need to hunt lenses.

The question is: What's still in focus at 105mm f1.4? The tip of your talent's nose with eyes blurred out?

Better stay away - the video quality of the A7 is terrible, with tons of softness, aliasing and moirés. It's not comparable to the A7s and A7R. It's a great stills camera though.

Well, the problem is that all digital moving image cameras - even those shooting raw - have less color information than film (whether Vision negative film or Kodakchrome reversal film). Since film emulation further degrades the video image, you at best end up with the equivalent of a low-end digital film scan. It should work better the other way around, emulating video colors on film stock using color filters.

I suspect the opposite - and it's already happening in cellphone photography, compact cameras and mirrorless systems like MFT: Fully corrected, high-resolution images are and will continue to be expensive to manufacture. These are simply the laws of optics and physics. The workaround is to design cheaper optics with known optical aberrations, and correct them in software (i.e. through in-firmware photoshopping). More powerful CPUs/SoCs will also mean that more powerful image reconstruction algorithms (like superresolution) can be used. The downside of this is that images look overengineered, artificial, oversharpened etc. So for true optical quality, and high-quality images, you will still hold on to your expensive classic photo/cinema lenses.

What VisionColor does - unlike FilmConvert and more simple film emulations - is to emulate a full analog process, i.e. (a) camera film stock + (b) print film stock In analog film, you normally shoot on negative film and, in the end, cut your negative and print it on special print film stock (very slow [+/- 10 ASA], very low-grain film material). If you use a "FPE" ("film print emulation") LUT, then Vision color allows you to choose a camera film emulation (such as Kodak Vision3 50) while applying a standard print film emulation in one run. If you choose a "CIN" LUT, then Vision color allows you to choose a camera film emulation + choose one of the 5/6 print film emulations in the "Cineon Conversion" tab.

@funkyou86 - here's my shot. Graded/corrected your PNGs in Resolve. In the first shot, I also applied Neat Video's artifact removal filter. You will be able to get rid of the image noise if you apply Neat's temporal filter to the material. (Which I couldn't do with the PNG stills.) I also included the drx files so that you will be able to use these grades in Resolve (if you work with the program). 1_1.1.1_1.1.1.drx 3_1.3.1.drx 3_1.3.1_1.2.1.drx

My own 2 cents: Don't obsess about cameras, learn to master the one you have, unless you're deeply frustrated, and keep working with it. Have been a happy user of the Blackmagic Pocket for now three years, and won't ever go back to an 8bit camera. (Even my A7s is only being used as a stills camera.) Here's one of my latest concert videos.

Can't really comment on the Blackmagic vs. Samsung decision since the two cameras are not in the same category. It's like deciding between a compact, bare-bones offroad car (let's say a Lada Niva) versus a family sedan that is more comfortable yet not fit for offroad driving (let's say a Toyota Avensis). If you go Blackmagic, the Video Assist isn't optimal because you won't need its built-in recorder. The camera has a built-in recorder that is even better since it records Raw. Better go for a pure external monitor that is lighter and has a better display, for example a SmallHD DP4 for the same price. As the focal reducer, better buy a Metabones Pocket Speed Booster. You won't have electronic focus control, but 0.58x instead of 0.75x focal reduction (i.e. the possibility to use a 30mm lens instead of a 24mm lens as a normal lens) and, even more importantly, 1.67 stops instead of 0.9 stops gain of light sensivity. (I.e. you will be able to shoot scenes on ISO800 with the Metabones for which you would need ISO1440 with the Aputure).- Regarding pulling focus on the gimbal, I don't think that you will be able to do this unless you have someone else as a focus puller and a more remote display. So I don't see a genuine practical advantage of the Aputure's electronic focus control unless you've really tested the whole set-up and are sure that it will work for you. Regarding the Canon EF 24mm, this will be a normal (not a wide) lens with the LensRegain (50mm full frame equivalent), and even still a normal lens with the Pocket Speed Booster (40mm full frame equivalent). To go wide, you will need at least 18, better 16mm. For gimbal shots, I'd just use a plain, cheapo vanilla Canon stabilized 18-55mm kit lens - it's lightweight, the optical quality is decent enough, and you won't need fast apertures for gimbal shots anyway - and then get a fast standard prime, for example a Sigma 30mm/1.4 or Canon 28mm/1.8, for everything else.

Here's a thought experiment: A Full Frame camera and an APS-C camera, both using a sensor from the same die, thus with the same dynamic range, same noise and same ISO per pixel. If this sensor had 12 megapixels in APS-C (24mm*16mm), it would have about 27 megapixels at full frame size (36mm*24mm, 225% the surface of APS-C). Now, light sensitivity in ISO would still be the same for both sensors/cameras, but: if you downsample the picture to video resolution, the full frame chip would yield about 50% less image noise, therefore about twice the usable ISO. [EDIT: If the full frame-camera line-skips instead of doing a full pixel readout and downsampling, then indeed, it will simple throw away the extra light/image information, and the bigger sensor size will yield no advantage in terms of image noise/usable ISO/light sensitivity.] It's the same when you would create same-size photographic prints from both pictures: The picture from the APS-C sensor would be noisier, and hence have less effectively usable light sensitivity/ISO. The experiment is even easier when both sensors have the same pixel count (let's say, both 12MP/4K). Then the pixels on the full frame camera would have a higher full-well capacity, thus more dynamic range and less noise, and hence a higher ISO than the APS-C sensor with the same pixel count. - One has to take into account that a bigger sensor captures more light/photons simply because of its bigger size. This is also true for lenses. A full frame f1.8 lens transmits twice the light that an APS-C f1.8 lens transmits.

None of the two models is weather-sealed. Here's more on the OIS jitter (bad stabilization) issue of the newer model: http://***URL removed***/forums/post/54459813

My advice: Keep the old model because of its better optical stabilization.

Of course it is clipping, just look at the highlights in the RGB parade (there's a hard clip in the peaks) and look at the roof of the booth on the left where all detail is blown out. And no, exposing to the right (ETTR) is not what you should do with an 8bit Rec709 image because you will get less than 8bit in your grade. An 8bit Rec709 image should always be exposed as the final image. It's a different story with 10bit, Log and Raw.

Your footage is overexposed (clearly visible in the scopes as lifted shadows and clipped highlights), and the white balance is off (too warm and too green). If you record internally, I'd really take care of good exposure and white balance since 8bit material can't be corrected that much. Here's an attempt to normalize the picture, correcting lift + gain and color offset, followed by attempt of a more stylized look (still based on the previous normalization.)

Sorry to go (literally...) against the grain, but all videos posted here don't really look like projected Super 8, rather like Super 8 scanned with 1990s low-end SD video transfer. This is what well-scanned Super 8 reversal film really looks like: And this is modern Super 8 with negative (Kodak Vision) film stock: