zlfan

from a business point of view, Nikon's hybrids may have the best raw in the future, as Nikon acquired red with its decades of knowledge of one of the industry's best raw codec tuning under the hood. also, Nikon does not have pro video camera department for protection. however, from a historical point of view, although Nikon introduced the true DSLR revolution with its d90, earlier than 5d2, it is not willing to add more buffer to Nikon v2 and v3, convinces me that Nikon is still controlled tightly by Japanese industrial conglomerates. anyways, I still have high hopes on nikon, as its z mount is also the shortest flange distance. sony, sigh, except its accidental error on fs700/r for excellent sdi out raw, its later low to mid tier cameras have crippled sci out raw to protect its high end Venice/2 cameras. on the other hand, Sony spent decades on codec implementation in hardware and firmware, its internal dsp is actually good enough. I will not bother with Sony's low end cameras' sdi out raw. canon is similar to Sony, except there is pressure from the free open source magic lantern. so canon has to compete with magic lantern on feature providing. however, seems to me, canon is trying to use sensor as the differentiator, r5/r5c sensors are behind c70/c300 mk3, etc. I guess the logic behind is that even magic lantern can hack the latest canon mirrorless, it cannot do much with the lower quality sensors. panasonic is the only one really aggressive, possibly due to its lacking of still photography camera department before. Panasonic seems also like canon, using sensor to differentiator its high end cinema cameras from its Lumix offerings. olympus, its board controlled by Sony investors, is not willing to step on Sony's toes, even willing to go belly up. its ibis was supposed to be the most disreputable technology 10 years ago. yet it did nothing to exploit this.

if a camera is famous for its log shooting, better to follow the crowd. its raw may not be that juicy.

seems to me that it is better to choose the cameras with the tradition or reputation of raw if wanting to shoot raw. at least to test the camera even if the spec says that it can sdi out/hdmi out to raw or can record raw internally. seems to me that "raw" is a glowing hat now, not necessarily true raw, which is actually very dumb, but fresh, juicy, real meat. it is funny that a real good raw, tmho, is the free offering, magic lantern raw.

In the comment section, this video's author agrees that Lumix is far better than Sony. Maybe the wrong setup on gh7 causing more jittering. @KenCoCreates 2 weeks ago Boost I.S. Is meant for tripod static shots when handholding. Be careful when walking around and panning on it because it will cause jitters and I think it hurts the sensor a bit. Other than that E-stabilization on high is so good. I use the LUMIX G9ii Microfourthirds all the way for me. Show less7 Reply · 1 reply @JasonMorrisphotocinema 2 weeks ago That makes sense. Realistically the lumix was far superior to the Sony in stabilisation anyway

https://www.youtube.com/watch?v=1GokSOqok1k at 10:10, seems zve1 is pretty close to gh7, maybe equal to s5ii/x.

on crop sensor, what is the advantage of anamorphic lenses over ff lenses plus a metabones reducer? the reducer can give you almost ff look, one stop more light, all the current available ff spherical lenses. Philip bloom used to shoot his cat a lot. sometimes, there are hidden talents in cat shooters. lol.

to be more specific, 5d3ml has a 1920x1920 sub crop mode in 3.5k or 3k crop modes. I used it several years ago. the footage was ok.

this I totally agree.

i totally agree with your analysis. i think the bottleneck is the end user's monitor/tv/screen+projector now. if these can reach to the level of very high end production monitor or even better, like 12 to 14 bits, ips, oled, 120 inches, dr of 15 USEABLE stops, then a35 and raptor etc are really essential.

I know you are an experienced dp, actually I am honored that you are interested to see my footage as you only want to see very high quality product as I can tell. just I want to keep anonymous on this forum, so that I can speak freely. I truly appreciate your interest, bow to you.

anamorphic lens originally was to increase the resolution of the final footage, although this notion of resolution increases is debatable itself. nowadays, for 4k, 6k, 8k cameras, resolution is not an issue. shooting for TikTok and traditional aspect ratios together seems the current reason of open gate and anamorphic modes. actually, for TikTok, there is no need for over 1080p, as people watch on their phones. at least I don't see TikTok like big screen tvs. on 5d3ml. there is a 1080px1080p crop mode. if the true purpose of open gate and anamorphic is not as a differentiator using that cross flare artifact to show that I am using expensive anamorphic lenses and you cannot afford them, I think 5d3ml 1080p 1x1 square mode is the best, as it is 1:1, raw, 14 bit, 1080p enough res for TikTok. also, rolling shutter is smaller in this mode as the sensor size is about 1 inch.

depends on codec, the behavior is different. for mpeg2 50 mbps, which is a standard for broadcast news, the panning artifact is blurring. once panning or tilting starts, the image is not clear. once the camera movement stops, it is clear. for xavc (Sony's mp4 implementation), the artifact is jumpy or choppy. for prores, similar to xavc. actually, theoretic discussion of the reasons of panning artifacts is not useful to me. as I am a practical man, the only way I find out that can solve this problem is using uncompressed or true lossless raw cameras if want to stay at 24p, or go 60p if the camera can do 60p and does not have a thick codec. I have seen blurred or choppy panning or tilting all the time, even on famous camera reviewers' footage. there is no deny that this kind of artifacts exist. I doubt that there is a practical way to prevent this happening on a camera that does thin codec 24p. most of time, experienced dps avoid camera movements at 24p if they use thin codec cameras, as you see from their demo reels.

from my use on Nikon v1, I doubt the notion that bigger sensor has better dr, as v1's dr is better than many apsc sensors. smaller sensor will have worse low light, that is for sure. bokur wise, I think m43/43 is good enough for most cases, maybe not enough for portrait and fashion. maybe a long lens at 200mm on m43/43 is fine. I use 43 55-200mm hg at 200 mm f3.5, out of focus is actually very nice, and you don't need to worry about half face is out of focus. I think 43/m43 for shallow depth of field seems an optimal format if background still needs to keep some information.

on my YouTube channel, I did a rough statistical analysis on the upvotes, 80% from go pro 12, 15% from magic lantern cams, 5% from f3 and red m-x cams. when I use go pro 12, my style is very different, as I know the camera movements will not reduce the image quality. in almost every shot, I do panning, tilting, walking, etc. also the camera is so small, I can go closer as people accept this as a toy cam. many times I attend events that I will not bring bigger cameras at all, I bring go pro 12 in my pocket, and can get some shots. the end results actually are very compelling. magic lantern cameras are very good for color. f3 is also very good for color, just ergonomically not easy to handle. red m-x cameras may have film like rendering so that please people's eyes. in general, I think a small action camera with m43/43 sensor, so that the low light is borderline usable, with gh7's ibis and af, with a pancake lens at f2.8, I doubt people will have other cameras. at least this camera will be used much more frequently than other cameras.

the same. either 29.97p or 59.94i. there is a difference of the loss of quality, though. 59.94i footage rendered in 29.97p has horizontal line skipping artifacts when moving the cam. 29.97p has jumpy movements. just did a 23.98p prores hq 4k from c500 og sdi out. jumpy or choppy panning. from my experience, only uncompressed or low compression raw holds up at 24p.

for a scene to look really good, actually the lighting needs to be controlled to be within 10 stops. otherwise, the results will be like in this video, even the camera is the legendary 15-stop usable Alexa 35.

for big budget production, camera is not a concern and a small part of the budget. for indie film makers, camera still is a big part of the budget. it is safe to say that big budget production still want to differentiate itself from indie shorts. this includes camera and the style only high end cinema cameras can provide. shooting a long take of the main character walking though a narrow street packed with group actors lighted with hundreds of meters of led tubes and reflectors is another differentiator. I have seen so many such scenes that do not matter much to the story line. also, large scale of green screen and heavy time consuming vfx, like dune 1/2. on the other hand, for the last ten years, tv series are getting so good, that its cinematography is better than that of most of the movies in the 1970s and earlier times. indie short films are also very good now. Denis Villeneuve spent his earlier and mid career in indie film making, then breaks through into the block bluster making. this is very rare case almost impossible 30 years ago.

it seems to me that the camera is not a differentiator like it was 20 years before, no matter what kind of style. intensive camera movements also can be mimicked by indie makers using s1h with good ibis, or gh7, or gimbals, or digital reframing and power window if shot in 8k or 12k. I think right now that hiring many actors and applying large scale green screen are the key differentiators, although these will make dps less essential to the whole product process.

totally agree. "It's high level art dear, you don't understand". lol I guess the true reason behind this trend is that low budget cameras are so good, like black magic bmpcc ff, Lumix s1h, etc, that Alexa users, i.e., big production dps, have to use dr as the differentiator, just like bokeh, etc, to separate themselves from low end indie makers. the problem is that they can not control the viewers' environments, like the black levels on viewers' monitors, the ambient light in the viewing environment, etc. nowadays, I like news, sports, and commercial shots more, as they just look normal, the colors are rich, the contrast is fine, and the exposure just looks all right.

the new tv series warrior has so many dark scenes. it is understandable that this series is about ganster, crime, etc. but still too dark for the whole series. the John wick series is the same. cannot find the camera info for warrior. John wick was shot in Alexa cameras. so Alexa cameras also want to explore deep Into the darkness to maintain its highlight rolling off legacy. everyone is doing this dr rat race. the new landmark of the current era cinematography style. lol.

seems to me, in theory, this mid grey shifting approach makes many cameras' useful dr very close to alexa. very clever detour. the achillie's heel is that the end viewers' monitors have to have very good black levels. in other words, the end monitors need to have dr distribution similar to these cameras that have more dr stops under the 18% mid grey. if this kind of monitors get popular, the advantage of alexa cameras will diminish.

possibly this workflow is trying to take advantage of the more dr stops under the 18% mid grey in most cameras. basically, the "new mid grey" shifts to the darker side, to maintain the high dr and keep the highlight rolling off, and to keep certain levels of contrast and saturation to please ordinary human eyes.

i know now why so many latest movies have so many dark scenes. my current understanding is that the camera is operated at the highest level of dr possible, mostly using log, so most dps use 60% to 70% ire for high light protection. then colorists may increase black level in post too to add the contrast. the final results are contrasty, color-rich, and, have high dr (high dr is normally not associated with high contrast high saturation style), at the expense of looking dark.

just did another shoot out with c300 og at 29.97 fps with the camera set at 59.94i using the in camera codec 50 mbps mpeg2. in resolve it recognizes as 29.97p. when the scene is static, it is great. when there is camera movement like panning and tilting, or even when there is subject walking in the static scene, on 4k monitor, the loss of image quality is very apparent. on laptop, it is not as apparent as on the 4k large monitor, but still visible. so this confirms again my previous observation. when the codec is weak, at 24p or 30p, either the camera movement is to be avoided, or use 60p. at 24p uncompressed raw or low compression raw, slow camera movements seem ok. faster movements may have issue. but I normally don't do whip panning. also, rolling shutter may kick in, so it is hard to define the effects of the 24p, the codec, the rolling shutter.

Lumix GH7 + Sigma 18-35 | Kraków Cinematic - YouTube seems working on gh7