M Joel W

Tri-X is notoriously hard to expose and a third stop slower under tungsten than under daylight. In theory the above (1/50) will work (although some 16mm cameras have 1/60 shutter speeds or finders that eat 1/3 stop of light), but I agree--just borrow a meter. Edit: Re: the above answer, there are lots of us in the dSLR crowd who use meters, I'd like to think most serious dSLR shooters use meters.

You might be looking at lead acid batteries because of their high c rating, but those are huge. Alkaline batteries have a terrible c rating (very low drain). I discovered this the expensive way. I would use CFLs rather than tungsten bulbs for their efficiency and lack of heat. There are 75w 12v tungsten lights but even then... 6 amps. You could get a few of these and switch them out: http://www.bhphotovideo.com/c/product/815655-REG/Bebob_Engineering_A90RM_3_Stud_Snap_on_High_Draw.html Normal Anton Bauer camera batteries might offer similar performance. You could rent them to save money. My guess is running 12v lights at 14.4v won't make them explode. I've been wrong before. You could run 12 9v batteries in sequence to a 110w fluorescent light. With lithium batteries you probably can get some acceptable amperage, too. Warning, you could get shocked at a potentially lethal voltage? I'm not an electrician. Probably don't do this. There's a video where someone runs a 220V fluorescent off a single AA alkaline battery by cannibalizing a fuji disposable camera circuit board. I recently ran a tungsten balanced triphosphor 15w 12v CLF off 8 lithium batteries. I used the battery pack for a marantz sound recorder but similar battery packs are a few dollars, $20 for the batteries. Very lightweight and pretty small. 60w equivalent and about 1.25 amps each so with a lithium camera battery that can output 12 amps you'd have a very bright prop indeed. NihM batteries are rechargeable but I'm not sure they have a high enough c rating. Alkaline batteries do not! The bulb flickers a lot. Maybe I'll build a portable fire rig out of a bunch of these. Getting a spherical diffuser that doesn't eat too much stop will be a challenge for your art department. There are spherical hard plastic soft boxes for some strobe lights. Look around! Is this the only source in your entire movie? You might want to put a chinese lantern on a fishing pole at the very least and bounce some very soft fill when the light is on.

Super 35. Well, t2i. Same thing. It's amazing to me how many directors and DPs (particularly capable ones) pick that as their go-to lens. Sometimes the 40mm. I was just thinking Brazil as an example of wide lenses used well when you brought up Gilliam. But he goes pretty extreme and makes the distortion visible. I've heard Spielberg's favorite lens is the 27mm, but that's not too wide and I don't know the source of that information. I didn't realize until I looked it up that there was a 10mm Primo (still t1.9!). I had the chance to try the 14.5 but didn't take it. Partially I'm just wondering how to approach 2.35:1. I rarely go beyond 17mm at 1.85:1. Do people who shoot wide screen find themselves going to 14mm or wider?

I want to shoot a scene, city night futuristic slums interior with most lights off or dimmed way down, primarily lit through windows--soft 1/2CTB 1/2 green for mercury vapor ambient light and 1/2 CTO 1/2 CTS for sodium vapor key (street light outside window). The look is Darius Khondji, the shots are Spielberg. Hopefully! I want it to look sort of like the talent is crying because it's raining outside and the street light is projecting rain shadows onto faces...would I need a leko and how would I go about focusing it to make this read? Planning to use a hose to spray rain on a window. Thanks!

I might try directing something (for fun and practice) and need some advice... For narrative video I don't know if I've used a lens wider than 17mm or the format equivalent. Space starts to feel very distorted beyond that point and I hate the "ultra wide HDR" look in landscape photography. But I shot a music video with a very wide lens and it was a lot of fun, very dynamic... For the purposes of this short video, I'm a big fan of Spielberg and Michael Bay. Maybe Ridley Scott (Alien and Blade Runner). Also enjoyed Star Trek and Hellboy 2/Pan's Labyrinth and even Avatar a whole lot. I figure I'm going to have to shoot 2.35:1. Going for an anamorphic feel but shooting spherical and cropping... I was wondering what kind of focal lengths I'm looking at. I've heard Spielberg is big on the 20-30mm range, I assume that's in terms of cropped super35. Does that sound about right? So would I be all set with 17mm as my widest lens or do I have an excuse to buy this 11-16mm zoom? Narrative DPs, how wide do you go and how do you get away with it? I feel like an ultra wide could be really fun for some action scenes.

That's the idea, or at least my interpretation of it. As for light sources in scene, that's one time I personally use a spot meter--when a light, the sun or moon or clouds, an overcast sky, a lampshade, some 1000H, etc. is in frame. I guess in those cases the scene dynamic range is usually pretty extreme and the source just blows out, but I feel cool using a spot meter. I'm sure photo printers would love to get their hands on some of that black hole emulsion, btw. The best people I know work primarily intuitively, and I often feel embarrassed talking technically around them since their work far outstrips mine and with less pretense, but I'm a verbal thinker so I go through these cognitive hoops with any creative endeavor. You should read the (somewhat brilliant) twenty page neoformalist manifesto I wrote to pitch the horrible short film I never finished. Some day, hopefully, my work will speak for itself--or at least speak the same language I do.

At least in terms of how I approach it, contrast ratio is the key to fill ratio as read through the incident meter. So 1000:1 is a ten stop contrast ratio. Dynamic range is the range from brightest value to darkest value in a given scene. So, assuming a 1000:1 contrast ratio, and assuming a subject with a roughly five stop range under even lighting (a big assumption, but one I'm guessing the author made), you'd end up with 15 stops of dynamic range in the scene through the spot meter. It still doesn't totally make sense--each scene has a different color palette and five stops is a lot of contrast for flat art. But I think this is what the book is getting at. Either that or it's a typo! Anyway, what you and David have said makes perfect sense; I'm just not sure the author is outright wrong about this, though. I'm sure every capable DP or gaffer approaches this differently, just as some people light primarily with an incident meter (which makes a lot more sense to me, personally) and some with a spot meter. I'm self-taught, too, excluding one intro course in school, so I could be wrong.

No, Chris. Of course you're right but you misunderstand what I wrote. Under even illumination flat art is at most six stops (in reality closer to five stops) apart (in terms of diffuse, not reflective/specular light). So if you have a scene that's 1000:1 (ten stop contrast ratio), chances are the dynamic range of the scene is about 15 stops. So it's sloppy writing (and obviously the exact number varies based on subject), but since the book references the subject and not the lights themselves, a 1000:1 contrast ratio correlates generally with a roughly 15 stop subject dynamic range. Six stops on the print itself.

The difference between a very bright and very dark object (through a spot meter) is generally around 4-6 stops. In my experience, black fabric and white fabric are about four stops apart and I believe that the best black and white photographs approach six stops of contrast. Obviously reflective objects, street signs and reflective tape, specular highlights, and anything with UV dye pushes that a bit, but who meters for specular highlights? My guess is the 15 stop figure refers to the approximate dynamic range of an average scene when lit with a 1000:1 (ten stop) contrast ratio. Especially since the book mentions the subject, not the ratio itself.

I see Sekonic dual meters on set more often than any other meter, and yet everyone I talk with says to get a Spectra, since they have the reputation for being more accurate. Odd. Fwiw, I don't use the spot meter much, but I do use it. If I'm trying to blow out a window or dim lights up to the same level or check and see what something in the distance looks like it's useful. For shooting slides it's great. It all depends on your priorities. If you want a general purpose light meter, the 758 cine seems good and it's pretty accurate, apparently more so after calibration. But for general use an incident meter is fine. Stopping and checking the spot meter to see how bright everything is can be very interesting, but for actually shooting it's too slow. The incident meter is much more useful (and the surplus of cine modes kind of pointless when 99% of the time it's 1/48 24fps).

From what I remember it looked good, though I saw it a while ago and don't remember specifics. It's hard to protect yourself as a DP, especially on smaller projects; directors almost always take performance over lighting/focus/composition, which I guess makes sense but I've suffered from that in the past, thankfully on movies no one will watch. I suppose that makes sense, lighting for a consistent style in day exteriors rather than based on source, but it seems like a ridiculous amount of work to me. But I guess that's what differentiates the huge movies from smaller ones. Thanks for the insight!

Thanks, Jon, that makes sense. I'm surprised you can get away with doing 180º flip in terms of backlight direction for daytime shot/reverse, not due to continuity between shots, but because you'd assume the background would be lit totally differently from the foreground in terms of the direction of the sun for at least one side of shot/reverse. Did you ever find yourself lighting the background, too, when the sun went behind the clouds? And hiding the cut off point between real sun light and what's artificially lit under the grid cloth seems mighty tricky, too... I suppose shallow focus and careful composition helps, but big day interiors seem crazy to me. I remember I worked on a short you shot in 2010, btw--didn't know you were on Indy 4 or I would have pestered you then since I was really obsessed with Kaminski at that time. Also re-reading my post I'm endlessly frustrated by Apple's spell check changing kinoflo to "kinfolk" and beadboard to "breadboard."

Does the 5D display RGB histograms or just one for brightness? If it's the former, that would be helpful. For the sake of consistency and to get my ratios right I prefer to meter, but since each color will likely start clipping a given channel at a different point from the next, it's probably futile to just guess "one stop under" as I've been doing--so I can see the histogram being much more helpful for this particular video. I'll look into it. Thankfully I'm not doing post, but that might be great in this case.

Any advice on metering light with an extreme color cast (for a dSLR)? I'm talking effects filters, totally outlandish music video stuff. My worry is one channel will get saturated (not to mention the 4:2:0 color space or however bad it is) and this is particularly an issue with a very deep color on the key light. In general I try to expose at least a stop under with a deep blue green or red key light? Maybe more? Good idea? bad? Will my meter (a Sekonic) read different colors differently or does it respond to all light equally? Thanks!

For the most part that's how I think of it and why I think "objective" (proscenium or on the sidelines) "indirectly subjective" (not a POV shot, but motivated by a character's feelings and closer to the action and/or somehow stylized) and "directly subjective" (POV or dream sequence) makes more sense in terms of classifying the same three shot types. No shot is totally objective and it's ultimately a matter of what's appropriate for the scene, so an OTS shot should be more subjective (especially if you're following with the character's eyeline or doing a steadicam follow with them) than a really wide shot because you're closer to the action and following a character's eyeline, but on the other hand dirty shot/reverse is a predominantly objective editing pattern (especially with longer lenses) and that's OTS, too. So maybe I misspoke--I was mostly referring to OTS follows like in Black Swan being very subjective shots (maybe not a great example, also not my favorite movie). None of this matters at all except to the extent that it helps directors and dps classify the shots they choose and articulate their motivations. If it's all inherent to how you're thinking you don't even need to articulate it. I doubt most directors approach this really academically while they work, but if you want to stand back and emulate your favorite director it's nice to have a vocabulary to discuss things or if you're talking with your dp it can help, too. And chances are some of them do approach it academically. I suspect the Coens do. The one place where I disagree with Brian's statement is that I don't think the director is usually putting his or her spin on the shot (though they can be, especially in the case of shots like the aforementioned Barry Lyndon zooms) so much as articulating a diegetic character's emotions through an indirectly subjective shot, especially in the cases of transparent directors like Spielberg. This isn't something to get hung up on, just something to consider while storyboarding. And helpful for picking functional equivalents when one shot doesn't work.

I completely agree (and really like Barry Lyndon). I think the same thing might be true about the zoom out at the end of Alice Doesn't Live Here Anymore and maybe the zoom as a director's POV has something to do with the popularity of zooms in the 1970s in general. But I need to watch more movies before making that generalization.

Agreed. Physical comedy plays best wide. Comedy is tricky since you need access to each character's subjectivity (so you know their goals), but you can't be too closely aligned with their experience or you feel the pain. You need access to intention but distance from pain. I guess comedy is primarily objective/very omniscient, but modern character-based comedy has more subjective access. I agree about the wider frame for comedy in that it's a) more objective and B) offers more distance. In an odd way, the Jackass movies seem to be the most direct descendants of older physical comedy--a lot of tableaux staging, wide shots, clearly articulated set up and pay off, etc. except pushing the boundaries of how much pain the audience can take. "Awkward" comedy like the British Office is interesting in that it's subjective and "dangerous" to start, but scenes resolve safely so it kind of becomes funny retrospectively. Comedy is a tricky genre. I guess you're right about Kubrick. His movies are all so different, though. I suppose he's mostly objective, but his use of the zoom and careful control of the frame puts him in a very different category from Preminger. Kubrick's movies feel more authorial and less formally transparent to me. I'm not sure what to think about Coppola. The Conversation is an interesting film; probably it's about the failures of subjectivity so it's subjective/restricted (in terms of range of narration), but I need to rematch it in the context of its forebears (Blow Up, etc.) to really say. The Godfather movies do seem objective but I haven't really considered them in this context before. I think dramatic stories that are more objective/omniscient than traditional dramas (in terms of storytelling) are generally tragedies and the Godfather seems to fit that description in terms of both story and form, so maybe The Godfather is a modern tragedy. Lars von Trier is one of the trickier directors. He'll give direct subjective access (Dancer in the Dark), he'll make a completely Brechtian movie (Dogville), and he'll swear he hates animation and artifice (Five Obstructions), but then make one of the more over-the-top operatic movies I've seen (Melancholia, which is excellent). I've always assumed the final image in Breaking the Waves was meant to be disingenuous, but I don't think the ending of Melancholia is disingenuous at all, and that makes it very hard for me to figure out what's going on in either movie.

How do you light day exteriors in these huge movies? For wide shots I imagine all you can do is "paint" from behind the camera or the side. For close ups are we talking enormous overheads of grid cloth (or solids) and then diffused light from the front, direct from the back, the idea being the background is far enough out of focus that you don't notice the difference in light's directionality relative to natural daylight? Or are these musco and bee bee lights big enough that they don't need to be diffused? Are some used for background lights when the sun falls behind a cloud? The day exteriors in War Horse trailers that are clearly lit seem to follow two patterns: overcast days have an additional offside key meant to look like direct sunlight and sunny CUs are heavily backlit with an unnaturally bright fill that, based on reflections in eyes is quite soft. But for that latter look you could backlight with a medium/big HMI and use a breadboard for bounce--no need for bee bee lights (or whatever they are called, I have never seen one). And how do you move fast enough with this huge gear? To me the trailer looks a bit over lit but in an intentional and controlled way (far more pleasing than Indy 4) and the photography overall looks stunning. I'll reserve judgement until I've seen the movie itself.

I agree, but think of it a bit more broadly: The screenwriter starts out with a story (a series of events), from which they have to make a plot (the employment and arrangement of those events). When we first pick up a camera or write a story, we try to present everything that happens over the course of a story and it's usually very boring and there's no suspense or surprise, just a straight sequence of all the pertinent events. Then as we progress as storytellers we omit what's superfluous or implicit between scenes, omit events wholesale to allow for surprise or ambiguity later on, choose from which perspective scenes happen and how much we know relative to one character or another (some films will have the main character present in every scene, most won't), choose to show things in a linear fashion or non-linear fashion and/or with varying degrees of objective truth. You can motivate non-linear storytelling plenty of ways: Tarantino motivates it through authorial presence (his movies are obviously crafted by an outside voice, his own); in Memento, Nolan motivates the non-linear story through the subjective experience of a crazy person. So objective/subjective/authorial storytelling starts with the script. How much do we know relative to each character? How much is the writer guiding us or do we have a lot of freedom as audience members? No shot is entirely objective. Just by selecting a moment time and space that choice favors whatever is present there. But an objective shot is one that lets the viewer see everything without a lot of stylization. A long take tableaux or a big crane move following everything somewhat transparently would be relatively objective. Then there's indirectly subjective storytelling. This is where the camera evokes what a character thinks or feels. You could have a shaky cam, a Vertigo zoom, a zoom at all (zooms are complicated), a push in, a pull out, a very wide lens, a very long lens, edits on eyeline, a close steadicam follow, inserts motivated by a character's thought, a canted angle, whatever. Indirectly subjective shots are shots that are stylized or motivated explicitly by one character's subjectivity. POV shots are directly subjective. You literally see what a character sees. (An OTS shot is somewhere in between; it's very subjective but not literally a POV so it's strongly indirectly subjective. Eye lines are HUGE in film and are overlooked, especially by naifs life me.) Authorial shots (frequently inserts--or camera motion NOT motivated by a diegetic character's subjectivity) are indirectly subjective from the director's perspective. Maybe even a "director's POV" shot. This is complicated territory. So that's objective/subjective/POV (though I prefer objective/indirectly subjective/subjective since then you can discuss sound and dream sequences, too). But it starts at the script stage. And while I agree with the examples above, you can have a movie with multiple subjectivities (a horror movie where different characters die and you identify with each one prior to that, a screwball comedy or thriller, a story told from multiple perspectives, etc.) and that's very different from something that's straight objective. Zooms in POV shots are directly subjective. This is the basis of a zoom: when you look at a detail in a scene, your eyes and brain discount the surroundings. The cinematic equivalent of this is a black frame or blur or something engulfing the entire frame except the detail. But now blow up that area of detail to the full size of the screen. You get a zoom. The eyes can't zoom and yet the zoom is arguably the most subjective camera move (if you can call it a move at all). It's also the most "reflexive" because it's used in news footage and because the eyes can't zoom but we can move in all the other ways a camera can move. Zooms not in POV shots may be indirectly subjective or authorial (Kubrick). Fincher, Kubrick, etc. are authorial directors, but Fincher is more transparent which is why Kubrick uses more zooms. Just to give some context: Welles is a tremendously authorial director with a somewhat distanced/objective/authorial camera. Lots of stylization, but a broader focus than one character's experience. Influenced by theater and radio. Hitchcock is a subjective/authorial director. You usually have a range of narration (how much you know relative to each character, a choice made at the script stage) similar to the protagonist or protagonists, but you might learn about a threat to the protagonist(s) before he/she/they do…the infamous bomb under the table (or whatever, I forget the exact quote). So Hitchcock creates suspense with authorial inserts, character identification with indirectly and directly subjective cues. He loves POV shots and uses them better than any other director. Spielberg is the master of indirect subjectivity. He is not a very authorial director, with Munich being his most authorial film. How you feel about a character is largely predicated on proximity--the closer the more empathetic. Spielberg is great at blocking so that character relationships are revealed through figure movement/relative proximity, but he's also great at placing the camera and moving the camera to add a visual emotional trajectory to a scene. Push ins, pulls outs, aperture framing and mirrors, etc. He's amazingly transparent for such a formalist. Underrated. The Coens rely on multiple subjectivities. Essentially every Coen film is a screwball comedy/thriller hybrid (the two genres are very similar to start with) and that requires having emotional and narrative access to a set of characters with interrelated/conflicting goals. They are great with POV shots, too, but they follow more characters than Hitchcock (or Raimi). The Wachowskis are all about transcendent experience and unity. So for them time/space/subjective/objective/etc. all comes together following the protagonist's enlightenment. See the Matrix or, better yet, their flawed/bizarre opus Speed Racer. Genres are important in terms of placement of the audience: Comedy requires safety parameters (establishing up front who can and cannot get hurt--you can't laugh if you play the pain!) and it's a predominantly omniscient genre with subjective access but still more of a focus on empathy than vicarious experience. The director must be subtle and transparently authorial. Inserts and reaction shots are the soul of contemporary film comedy. Superbad is one of the best comedies in recent years in that it provides subjective access but then cuts to authorial and objective shots to articulate the misunderstandings inherent to comedy and make the "dangerous" safe. It's superbly directed. Horror modulates between subjective (POV) and indirectly subjective (shots motivated by suspicion or feeling, frequently pain, but not seen through a POV) during scare sequences with more conventional storytelling during story-driven and expository sequences. Like Hitchcock's bomb, shots of the killer outside the victim's subjective realm can be used for suspense. Better but more difficult is to use potential threats to build suspense, reveal the danger for surprise. Screwball comedies and thrillers modulate between semi-subjective semi-omniscient storytelling, following a few parties toward a common/interrelated goal. Tragedy is all about distance. But comic distance and tragic distance are quite different (and different from voyeuristic distance and Brechtian distance), though the same cinematic techniques can achieve both. Brechtian critique is about a LOT of distance and distance that is not transparent. You could call Brechtian cinema objective/authorial. Maybe. Musicals…are complicated.

A combination of it being overcast and the bleach bypass process used in the film adding contrast and reducing saturation. The scene also looks pretty bright whereas sometimes dps will underexpose overcast footage a tiny bit. In general, blowing out the sky is easy; keeping detail in an overcast sky is hard!

I watched most of this movie with a friend and thought it looked pretty bad overall, some of it unusually bad. There was some nice magic hour photography and I've seen much worse from the era, but this was not a good looking movie and the music video effects didn't age well, either. The lighting in particular is quite bad; there's a lot of poorly done hard HMI fill during the day and the interior (sound stage) lighting can be reduced to two looks: hard HMIs through each window (at totally illogical angles) during the day and the same lights bluer and a stop darker for "night." Really inelegant stuff, but fast to work with, I assume. That alone isn't remarkable, but this movie was shot by Janusz Kaminski with Mauro Fiore as gaffer. Anyone know the story behind this movie? I assume the two met at Columbia College Chicago, but how did they land this gig and how did the relatively poor photography on it launch two very impressive careers? Or am I missing something? Just curious if anyone knows, not trying to disrespect either as both went on to shoot some gorgeous stuff.

It's like a book light except instead of the bounce coming from the side it's coming from under the board. I think. I just use soft boxes.

Being future-proof is a terrible argument for shooting red because 95% of these movies are finished at 1080p or 2k. So how is that future-proof? Oh, I'll just export ALL my footage again from raw, redo all my effects, recolor the whole thing. If you happen to make the next Star Wars, maybe 20 years down the road someone will do this for it. For now, shooting Red does more to date your footage than it does to future-proof it; like it or not Red has a pretty distinct "look" and for reasons that have nothing to do with the quality of the look (for better or worse) the camera has been adopted for indie and low budget movies and so the red look has become associated with "cheap." Kind of a victim of its own success. Lately, the 7d look has become associated with "even cheaper," which is too bad since both cameras have their merits. The red camera's constant state of flux doesn't help either; original Red footage (old redcine; old sensor) looks bad and The Informant, for instance, is already visually dated. There's a solution to this, though, and that's shooting and grading really well and the red footage is super flexible to grade and the mx sensor is not bad at all. The red workflow remains horrible but its a pretty awesome camera for the price if you can coax a good image out of it in post. For stock footage, I can see 4k being totally useful, though. Way more flexibility and a longer life in this case. For vfx there is also a good case for red; it keys nicely and can be blown up in post. 2k vs 1080p is almost semantics. It's a matter of one format being 18 pixels wider. 3D 4k is not part of the DCP spec. If you look at the mtf of a 4k image vs a 2k image and the integral of the curve (which equates roughly with perceptual sharpness) a 2k red down convert has like 75% of the useful resolution of the original file and the Alexa has almost as much, too. From a marketing perspective, 4k is a big deal, though.

My recommendations are as follows: If you're shooting video ONLY: get 28mm (and/or 35mm), 50mm, and 85mm f2.8 or faster manual focus nikon primes (used from KEH or eBay) with a canon mount adapter and then complement those with the kit lens (which is only a half stop slower than f2.8 at 18mm!). Old Nikon glass is as sharp (or sharper) than modern zooms, but has a prettier look, nicer bokeh, etc. BUT no image stabilization so put that 85mm lens on a tripod! Or, for $550 get the Tamron 17-50mm zoom. I have mixed feelings about this lens. It seems to perform right on par with the twice-as-expensive Canon version and have better IS but the bokeh is inferior. For a gorgeous silky look...not my top pick, but still okay. For the money, though...it's a great lone lens. That said, f2 is useful and 85mm is a nice focal length (so is 105mm) so you may want to branch out.

For what it's worth, we've never had QC complain about images being too dark or grainy even when they've been really dark and grainy. So it's probably more of a subjective judgement there. 10 IRE sounds a bit extreme to me, though. If you watch Rob Zombie's Halloween it gets pretty dark and maybe it approaches that, but apparently that was a producer overruling the DP and director and it's incredibly hard to follow what's going on, which doesn't matter much since the movie is awful. 10 IRE might read, but not on all screens. I'm sorry if this seems really obvious but one technique to try is darkening and desaturating red and warm tones in general in your image almost completely and adding a subtle blue tint (very subtle, unless it's motivated by moonlight or mercury vapor in which case you can make it more pronounced)... I don't know how to use Resolve, but the luma curve and saturation curve in the secondaries in Color are very good for this (except that the luma curve breaks up the image in a bad way, bad algorithm I guess). The eyes see red as near black in the dark so you can darken something very little and make it look much darker subjectively. This might be grading 101, but I skipped that class...