Jump to content

M Joel W

Basic Member
  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by M Joel W

  1. A lens's focal length is the distance between the optical center of the lens and the film back when the lens is focused on infinity. (I think?) When you change focus to focus on closer distances, you're moving the lens forward and effectively "zooming in." At least with most standard unit focusing lenses... with my old Nikon still lenses, I notice they zoom in as I move toward "macro" mode. Very slightly. The effective f-stop also decreases but only slightly. Which is why there's exposure compensation for macro and for large format. A lens designed to minimize breathing minimizes zooming in or zooming out when you change focus. Perhaps the lens zooms out a bit as it focuses close.
  2. I usually consider a focal length equivalent to the diagonal of the sensor/film back size to be the "normal" regardless of aspect ratio. 36X24 would be: 43mm Super35: 25mm Canon APS-C: 27mm 6x6: 80mm I shoot most often with 50mm on 36X24 and 28mm or 35mm on Super35/APS-C, though, and consider those the "normal." So I guess I round up. If 50mm is "normal" for you on Super35, then the full frame equivalent would be 85mm.
  3. I once graded film to match digital and paid special attention to vectorscope peaks. Even when the chroma and saturation matched, the digital vectorscope had a more diffuse point cloud, as if blur had been applied to the image. (It hadn't.) Subjectively, the film looked more saturated to me. This is just one camera and one film stock and one person's experience. But I noticed... something. 2383 is a print stock, for which there are emulation LUTs. Never had the fortune to finish on film (well, except when I shot reversal) so I can't speak to how saturated it is or isn't compared with scanned film, an emulation LUT, and digital projection. Guessing the real thing looks better still, but I do post digitally so I may never know. There are other differences, too. Reversal/digital vs color negative film. Look up the "Linny LUT" for an interesting take on that. The spectral acceptance curves for some very saturated films (Velvia, etc.) are quite narrow compared with what I believe a Bayer pattern filter achieves. The "trichromatic" digital medium format back is similar in this regard, though the marketing is misleading and there is still a lot of overlap between filter acceptance curves despite what they imply. Regardless, look at images from it and compare with the standard back. The "colorspace" etc. is all the same. Just different dyes in the Bayer pattern and a slightly slower (ISO) sensor as a result. Could you just grade one to match the other with a LUT? I couldn't say. I've read a lot of information indicating a lot of contradictory things about this topic. I think you just need to trust your eyes. I think another part of it is that everyone is drawn to video or film for a different reason. If you like shooting in one format or another, embrace it, I guess. Or maybe not. I thought Michael Mann's style or Fincher's in Zodiac would be the future of digital cinematography, but instead the Alexa and a "film look" seems more popular now..
  4. I don't think they're interchangeable. Every lens is different the same as every type of diffusion is different. The artifacts you're getting from a Speed Panchro, for instance, include edge softness and different types of chromatic aberration and flares that are far more complex than a simple promist filter could ever simulate. Even the MTF characteristics of modern lenses are distinct from older classic designs and those design choices filter down into bokeh and CA and coma, too. On the other hand, I agree diffusion is given too little consideration and lens choice too much, particularly on Internet forums. A lot of DPs are already doing what you mention, I've liked the look of modern Angenieux zooms with a light Promist filter, for instance, on von Trier's films. If that look works for you, it certainly seems easier. It can look great. It just won't replicate a Cooke Panchro or Super Baltar imo. I have no idea if it's true, but I've heard Game of Thrones is shot with a Hollywood Black Magic filter except during vfx sequences. Could be totally wrong. So the approach you're mentioning is a really good one, it just results in a different kind of softening imo. I personally miss over the top diffusion like Kaminski and Richardson used in the 90s. The way Kaminski used nets and classic softs is just as (if not more) interesting to me than the way contemporary DPs use vintage lenses... I really love the look of the non-HD classic softs at higher strengths.
  5. Look for responses from others, too. I'm an amateur DP (work in post), but I know a bit about HDR/SDR workflows as a result, too. So take the above with a grain of salt.
  6. Personally, I aim to keep everything (except point sources and specular highlights) within the camera's dynamic range. You can always burn highlights out in post or lift shadows (to an extent). However, I suppose if I were after a really punchy, rich, vibrant look I might aim to shoot really flat and add a high contrast grade. I'd never considered it, but using an ND grad filter would be an interesting thing to try instead of a power window, however I sold mine once I got a camera with dynamic range similar to what you mentioned above. I wonder if it offers any advantages. I think color negative film looks different from digital in a way where a lower contrast scene can appear more vibrant, and there's an interesting LUT I saw that emulates this but it occurred to me most while watching 90s blockbusters. Regardless, those films were exposed for the film then telecine'd for video without throwing away half the dynamic range, so I wouldn't do that with digital, either. For me, I would worry about exposing properly at key or where I want to be relative to key first (not the case with those who strictly ETTR with digital), then I worry about not losing highlight and shadow detail, and adding fill or an ND grad or streaks and tips or a flag or ND gel or whatever as necessary. But aesthetics are a separate concern. If I want a very rich image, I might worry about my contrast ratios and keep them conservative then add contrast in post. So I wouldn't say shooting for a smaller dynamic range is strictly crazy, but I would say it's a stylistic choice, not a technical one. I do find that these cameras encourage lazy lighting, and if using a lot of light gets you an image you like, go for it.
  7. Thanks. The original plan was shooting 4:3 16mm... I might still... The best vfx I saw this year (the Thor fat suit) were mostly practical. But if you do it wrong practically it's much harder to do it right in comp. Vfx deaging imo requires the cleanest source possible and that's not just grain, but practical effects and makeup that muddy it. Smaller things like wigs present good questions... I just need to shoot tests, which I don't want to do since 16mm is expensive, but whatever. I think shooting coverage a little tighter will help sell the effect because I'll have a little more meat to work with and/or shooting additional plates digitally in close up for the wide shots (to composite in over film). Figuring out how to stack the lenses to obviate parallax might be an issue, though. I lack resources to Timecode sync my digital camera and a film camera, I think. Or maybe I don't. Not sure. I'm very very curious about the deaging workflow on Irishman and how they used multi-cam (for tighter coverage to cut in with, for motion capture, etc.) but it has little to do with what I'm doing personally, so I won't pretend it does. They're working on a much higher level. I'm working entirely in 2D for a quick scene. Still, could be an inspiration for how to tackle a tricky problem. But just from this thread I've decided to light flat and user a slower stock, so it has been useful, thanks everyone. And I'm sure I'll be asking about timecode syncing a digital and film camera... or might just roll 120p for digital and hope to get some overlap. Very curious to see the breakdowns here, too.
  8. Does Pfister mean he's lighting at +6 stops over key (through the incident meter) or +6 stops over 18% gray through the spot meter? I think CML found the Monstro to hold to about +2.5 stops over key in their test scene, which would translate in their test scene to +5.5 over 18% gray or just under I suspect. Alexa held to +5 I believe, which would correlate with the rated +7.8 over 18% gray. I believe... I like the look of the Vimeo link btw.
  9. Yes, that makes a lot of sense. Well maybe I will shoot 7213 rather than 7219 for this sequence at the very least. Of course, film grain is also the best thing there is to help sell a composite. 😉
  10. Any idea why they had to shoot digital to do de-aging? For the multi-cam stuff or to get a cleaner image to start with? I was planning to do a project with simple 2D de-aging and shoot on super 16. Now I'm contemplating shooting those shots digitally–in fact, I had been before–but I can't imagine it would matter much since I'm just doing 2D de-aging. (Yes, I have done this in the past.) My needs are modest, I think I can get by shooting super 16, but I'm very curious about this if anyone has more information about the de-aging process they've used here. It's funny, I can see some tells here (signs of frequency separation: too smooth base, too much fine texture), but this blows me away: Whereas the CGI-based de-aging largely doesn't. The Tony Stark work was very very excellent, though, and I believe combined 2D and 3D.
  11. And do I know it. I might just settle for ultra 16 but that's a discussion for another thread.
  12. I missed the boat! A friend told me the same thing when I asked about getting back into shooting 35mm. Well, it's a good sign so far as I'm concerned. I far prefer the look of film. 16mm is more my budget anyway.
  13. A somewhat random thing to point out, but 135mm is just about where I can't handhold a lens acceptably steadily anymore (on Super 35, at 24fps) and when I begin to need IS. 85mm is already tentative. I don't have the most stable hands, nor do I work with larger cameras. For music videos, I have seen longer and wider lenses used well and creatively, as wide as 9mm equivalent and as long as 200mm+. I can imagine a 70-200mm f2.8 II IS L is enormously helpful for events. But I think the 100mm f2.8 IS L macro is pretty good, too, so tastes vary. But frankly the rendering on the 70-200mm is much closer to the 100mm f2.8 IS L and to other L zooms (very limited green/magenta bokeh fringing, more blue/red and more color neutral but less smooth bokeh) than it is to some of the primes you're shooting with, so I'm not sure it would be the best complement aesthetically, but it's wonderful if you like sharper more clinical imagery. Not really my preference, but it's an amazing lens. Cuts okay with the Sigma 18-35mm, also incredible, also not really my thing. I love, for instance, the 135mm f2 Nikkor but on full frame. It's not so sharp at all. Sharp at infinity then smoother as you focus closer with really smooth bokeh with more soft fringing. But I don't find the focal length terribly useful on Super 35, to answer your original question, but for events and if it had IS I might. For even work with enough light, don't discount the 55mm-250mm STM. Unremarkably rendering, slow, but tack sharp and $99 refurbished from time to time and easy to handhold. For daylight work it's quite useful.
  14. Thank you, this might be the ticket. This project is for fun/on spec with no crew. Would shoot properly for a proper project. Will also test to see if the result is acceptable.
  15. I'm shooting some content just for fun and I have currently have a tungsten kit and a daylight kit. I get it, tungsten looks way better. But for non-serous work, how terrible would it be to instead put CTO on my daylight LEDs? They're mostly consumer-grade, but 95+ CRI. Will they match? Will this be a total disaster or will it sort of work? Again, the goal isn't perfection but adequacy.... moving target, I know. Shooting with C100 and EVA1. Thanks.
  16. Thanks. That's good advice about looking to specific references and considering more than lighting. I sort of assumed the sources I have wouldn't be bright enough to put out a window, but hadn't considered that light coming through the window would be something I could deal with just by waiting for the right time of day. I assume backlit afternoon light is usually best? Looking to specific references is an even better call because I'm trying to mimic something pretty specific, but also just curious about how to approach things generally (if I'm in a rush I'll shoot a tighter frame so I can fill brighter/softer, etc. but want to have the right principles). Anyhow, thanks again.
  17. Thanks so much, that's really really helpful. "Bland" is the wrong word, I just meant whatever the conventional high key look is. I really like that look, actually. Good point about natural light, that will be something I look more for when choosing a location. I'll also look for off-white walls, which generally I avoid. I could probably get my hands on a couple 575w HMI pars. Is the fastest "book light" I could make with them bouncing into a 4x4 bead board and then hanging my own 8X8 of 250 from a c-stand? Or would I do better just handing heavier diffusion and punching straight through with the widest lens? These are basically "sketches" sort of like SNL or something (but a little more meta) where it's meant to mimic the look in an identifiable without requiring the actual level of polish you'd get on a "real ad." But obviously aspiring to get close. Thanks again! I'm kind of impressed how good the production value is on some of the new fake SNL ads. And makes sense about the cyc. I assume the space lights are also working as fill on the talent, though?
  18. This stuff is so interesting. I was about to mention that I tried their upscaling software and it inadvertently took some out of focus areas and brought them into focus. Really bizarre. There's a tree in the foreground that machine learning somehow brought partially into focus. Not really a desirable result, but pretty interesting. I've had to do this in post from time to time and there really is no good solution for video. But the best solution I've found is using (carefully) a good Neat Video profile to denoise the footage, then a wide radius unsharp mask (possibly masked over just the areas you want to target), and then regraining. I've actually gotten pretty far with this approach, but mostly on focus pulls that dip out because the motion masks that it's just added contrast, not added detail. There are also deconvolution plug ins, but I'm not sure they work that well or are compatible with video software. But yeah, no good solution available yet. With machine learning I expect there will be soon... that's going to change the vfx industry in a huge way I suspect...
  19. How is the lighting style achieved that I see in most sort of blandly high key (tastefully so, not too flat) ads that take place in say a living room or kitchen. Maybe like a prescription drug ad? I imagine it's a huge soft source as the key and then filled at a high key ratio but are there any other tricks to getting this look. Now the tricky party... any ways to get this look on a budget? If I just have an astra or two and a 200w HMI through a soft box, let's say, am I hopeless to even get close? I don't mind if the windows blow out a bit. Would the best approach then to be bouncing the astra or heavily diffusing it? Is this easier than I think? Also, any tricks for lighting on cycs? Just turn the space lights on to light the cyc and add fill and then add a soft key?
  20. I'm relatively inexperienced compared with many here, but I believe the Alexa changes its ISO such that the gain isn't really changing, it's more like it's baking in a different curve. Hence the over/under changes (unlike Canon and Sony) between 800 ISO and 3200 ISO to increase the highlight detail at the expense of shadow detail: http://1.bp.blogspot.com/-S2PefWOPeWo/TzuvhYONf9I/AAAAAAAAAhU/-cqjtNT-3hs/s1600/427243_10150575080764065_592699064_8940257_1181583253_n.jpg So I don't understand why someone shooting Alexa/Amira would rate at 1250 ISO and underexpose and monitor with an on-set LUT pushing it further rather than just rating at 3200 ISO and previewing with a normal LUT. With a different camera it would make sense to me, but the Alexa already pushes the image in software rather than with gain, I believe. (Or something similar.) There's probably a reason that's over my head or I'm misinterpreting what you've written, but I suspect you can achieve the same thing simply by rating the camera at 3200 ISO and metering at 3200 ISO with no custom LUTs. I worked on an Alexa show shot on 3200 ISO and part of the reason was an attempt to get more "film grain" and it worked beautifully. The over/under was closer to film, too. Just be sure to shoot like film–you don't want to underexpose further or it gets muddy fast. I didn't notice a "softening" effect so much as the over/under changing and the shadows getting grainy (with great texture; the Alexa's noise is quite nice). The Panchros and K35s have significantly different rendering; the Cookes imo have a stronger vintage feel with nisen bokeh whereas the Canons are better-controlled with the only weird quality being onion-ring bokeh from the aspherical elements. Manchester by the Sea I think is K35. I love Super Speeds and B Speeds, which also have a vintage look, but that look is a little colder imo in comparison. Standard Speeds are pretty nice, too, and seem more affordable and easily available. I haven't worked with the Kowas so I have no idea what they're closest to. Just a disclaimer, I'm not very experienced in this area, but have worked on shows where DPs were making similar choices. The feature I worked on where the DP rated at 1600 or 3200 ISO on the Alexa was shot on rehoused Leica primes from Panavision. I'd run your own tests because the Alexa does get noisy fast and it might not be desirable. Its over/under even at 800 ISO favors highlights more than most cameras do. In my experience, tungsten white balance also makes it noisier.
  21. I can strongly relate to this. I'm selling a lot of equipment now and losing a lot of money. But I find the more I sell, the less obligated I feel to pursue whatever project it is for which I bought the gear and that I wasn't really committed to or have moved on from or changed my mind about since. And it's liberating. I sometimes find myself spending more and more time or money in order to justify a purchase, or inventing a project so I can justify a purchase. And it's really stupid, it's just the sunk cost fallacy snowballing. I envy friends who only rent; it seems they live in the moment creatively, but I like to have a camera around, whether or not it does me much good to have one. Perhaps a very cheap one. On the other hand, I used to shoot 4x5 slide film and sold my view camera and picked up a digital camera to replace it. It wasn't the same. The colors weren't close and tilt/shift lenses perform poorly in my experience compared with view camera lenses and now I'm selling the digital camera, too, and I miss the view camera. So don't pick up a digital replacement just to have something if what you truly want to do is to shoot film. But film cameras are affordable to rent now anyway, so maybe sell and rent whatever it is you want at the time. I can't say, I like shooting digitally. (Perhaps what I really want is a 645 digital tech camera for stills, I just can't afford one.) But you're making me wonder if I should sell my Lite Panel that I have yet to use, and which I bought simply because it was very cheap. I think you're mourning the dreams you conflated with the gear through the promise you saw in its purchase. And selling it is letting go of the person you were when you bought it. Hmm do I sell my Lite Panel...
  • Create New...