ilkka_nissila
Offline
• • • •
Upload & Sell: Off
|
 p.2 #9 · Zr - the review I was waiting for | |
bernardl wrote:
I am of course not saying that Nikon shouldn't improve their H.265 output and I agree that the appeal of the ZR would be further extended if they did.
But the logic is quite obvious isn't it? You get a much better quality and flexibility to grade and correct issues in post. Even if the scale of the benefits may depend on the application.
It's the exact same logic photographers have always been using when shooting raw vs jpg.
There is indeed a cost associated to this and the ratio of file size is larger than for stills. But even with this amount spent on media, you are still at the cost of a C50 or FX3 without anything, right?
Considering the cost of the events being captured with a proper video set up and a videographer, I am a bit puzzled that temporary file size during capture and edit phases would take such precedence over quality.
Cheers,
Bernard
...Show more →
(Still) photography and video are not that similar and for video, the demands for storage in raw video are obscene. Because of this, it is not that common to use raw for video and there are ways that add editing flexibility and provide high quality which require less storage than raw, this includes shooting 10-bit log, there are also intermediate codecs where the storage requirements are intermediate between raw and h.265. In video, most cameras which are able to shoot raw only have 12-bit raw (unless we're talking about big cinema cameras), and since it's linear it's not that different from 10-bit log in terms of information content. For photography, the common standards are just jpg (8-bit) and raw which is typically 14-bit, and the difference between the two (in information available for editing) is more obvious (there is HEIF but support is not that widespread). Since photos are individual frames, it's much more practical to edit them sometimes significantly whereas for video such editing with masks on a frame by frame basis would be prohibitatively costly in terms of work time required, unless making a large production. Photo editing software support lens profiles for most modern lenses that provide identification information, aperture and focus distance are stored in the EXIF, and thus RAW files are easily corrected for vignetting and distortion within reason. For video such automatic correction is only available for in-camera processed formats such as h.254, h.265, Prores 422 etc., not raw. So raw for photos is much more established and much better supported by software than raw video, because there are so many people using raw for photos but not for video. Perhaps this will change in the future; I would not hold my breath though.
For photos I always keep the original raw files since I may want to do a different edit later on. For video, let's say I'd shoot 3 hours of r3d 25p per week for 15 years and make two backup copies of the raw footage for safekeeping. That's 1.4 petabytes of storage, just for the video, and then add the edited versions, the photos and so on. It's a significant amount of storage and since with it comes the extra work of bookkeeping to know which lens was used for each clip and what focal length, distance to subject and aperture was used (since distortion and vignetting are focal length, aperture, and distance dependent) creating one's own lens profiles, distortion and vignetting correction algorithms and running them through, there is a lot of additional work in creating something that can actually be presented. And because of the extreme limitations on bandwidth for online sharing of video, the result will look crappy compared to when viewing the original or even the edited file on a modern OLED TV. It just doesn't make any sense except for a large multimillion-dollar production where the stylized vision with strongly edited "moody" color is prioritised vs. realistic rendering that is more typically needed for documentary video and which is pretty much as the footage comes out of the camera as it is.
Sony is a lot more experienced in video than Nikon and it makes sense that they'd prioritize high-quality h.265 over raw video in their mid-range video products, since it's a lot more practical. Nikon is trying to find a niche for itself that would emphasize its strong areas (optical design, reliability etc.) in a crowded field. I'm just glad that Prores 422 HQ is supported probably I would not touch video if I had to shoot raw in order to get good video image quality. I'm able to avoid the lens correction hassle by using Prores 422 HQ and the quality is outstanding. I do delete the original files since I can't spend that kind of money on storage to keep them but it's a loss. I do not do much in terms of editing, slight cropping, cutting, apply a LUT, and adjust curves, insert audio and synchronize it and adjust it for consistency in volume, and such things. I don't think I would benefit from RAW video since I do such minimal edits to the image and the final result is typically a h.254 FullHD video, in some cases 4K but then people in most cases will never go through the trouble of viewing it in 4K. The footage looks so much worse in the online version than when the original exported file is viewed on an OLED 4K TV that fussing about RAW vs. other capture formats seems silly. Will things change as displays get better? No, because most people will still look at the footage on their mobile phones and the human vision in those cases greatly limits the perceived quality. What's funny is that even when BBC use 8K RED cameras to capture their wildlife documentaries, they're still distributed in FullHD, not 4K, so even in those cases there is so much information lost before finally presented. There is some quality improvement over the decades; the BBC documentaries produced in the last 10 years are technically visibly superior than what they had 25 years ago, but ironically it still shows through in FullHD delivered image when you have access to the right channels of distribution and produce it correctly. Free distribution channels that are available to most of us don't seem to provide such quality. So, does it really make sense to put into the cost of capturing raw? Already video streaming is a significant cost to the planet in terms of carbon emissions and if future distribution is improved in quality, the CO2 emissions will likely grow, as old men grab onto power and try to reverse time. IMO it would be much better to try to optimize practical formats rather than capture raw video. I'm guessing Nikon's cameras might need new hardware to support better h.265 (as Sony offers h.265 4:2:2 10-bit) and so it might come in the next generation if Nikon see the light, but they might be so in love with R3D that their focus is in that. Problem is it shuts out a lot of users with more modest means.
|
