Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Ask David Mullen ANYTHING

Hi David- I have a color timing/filmstock question for you. I recently shot a scene with a woman in a golden field on a sunny day. Her wardrobe was bright, saturated red and fuchsia. We were shooting 16mm with 7201. The director wanted the end result to have very saturated colors, including the golden field. The problem is that in the telecine, when we tried to saturate the sky and the field-- saturating the image overall-- the colors in her wardrobe became overly saturated and the colors started bleeding and also seemed to be clipping. If I had it to do over, I would make sure that the saturation of the wardrobe was normal, so that we could increase saturation overall with no problem. My question is-- is this an issue with the stock being extra sensitive to the saturated red and fuchsia?
 
No, there isn't a technical problem, it's simply that your frame has pastels and saturated objects and when you boost chroma saturation in post, you will hit the ceiling sooner on the saturated objects. Of course, there are tricks to deal with this, like isolating parts of the frame or select colors using keys and windows in order to color-correct them separately.

But I'm not sure dressing everything pastel and boosting it overall in post would necessarily be better because skintones look odd when they are overly saturated. Better to dress as much of the frame in strong colors as you can.

You may also be running into the limits of Rec.709 if this was for an HD monitor.

There is always an issue with pastels in a frame taking longer to get saturated, or falling to b&w faster when desaturated. You can sometimes use that fact for creative effect though.
 
Thanks, David! One more color timing question for you-- I'm about to color time my first Red feature. We'll be doing both the dvd and the print. Any tips on this process? When I work on film prints for festivals, I sometimes err on the bright side due to older projectors with older bulbs. When you're timing a Red feature do you think about that or do you just make it look it's best on the calibrated monitor you're working with?

Thanks again for your time!
 
I would time it for the film-out (final master likely to be 2K or 4K RGB, 10-bit Log) but make it look the way you want it to, assuming you are viewing everything in a proper D.I. environment using the right LUT's.

After it has been filmed-out, you can always make the print off of the digital negative a few points lighter overall if you really feel it is necessary for dim projection.

Then you'd just do a second pass based on the color-corrected master for Rec.709 broadcast video, for any home video / broadcast version (and for HD projection at a film festival). Ideally, even that submaster would be 10-bit 4:4:4 1080P and you'd make 8-bit 4:2:2 versions from that at the resolutions needed. For example, once you are accepted, Sundance for years has asked for a 1080i HDCAM tape copy for digital projection if you are not sending a 35mm print.
 
Hi David, I have a question regarding lighting with a more sensitive film stock or in my case the new MX sensor.

Is there a school of thought that says lighting with the new sensor is easier as you don’t need as much light and therefore easier to achieve softer lighting?

When shooting film do you prefer to choose the most sensitive stock you can get away with?
 
I'm a bit old-school in that I tend to pick the slowest stock that is practical for the scene, though I limit the overall number of stocks for efficiency's sake, I don't want my loaders having to keep inventory on several stocks. Usually I stick to two or three stocks total for the whole shot, a slow, medium, and fast stock. But it varies, I've done some shows that were all shot on 250D and 500T stocks, for example (I just read that Wally Pfister did that on "Inception") or maybe I've tossed in a little 50D (or 64D if Fuji) as well, but other shows I've used only tungsten stocks like 100T, 200T, and 500T.

To some extent, it also depends on the current generations... often a new 500T stock comes out that is nearly as fine-grained as the previous generation's 200T stock, making it less necessary to carry the 200T. But then the 200T stock gets replaced by a new stock that is more clearly finer-grained than the 500T, etc.

Faster film certainly makes soft light techniques easier since they tend to be power-inefficient (i.e. you lose a lot of intensity in the process of softening the light.) However, there are balancing issues that fast film doesn't solve, like soft-lighting someone against a bright background.

Though it's a minor problem, when you starting working at extremely low levels of light, you do have an issue with contrast because there is so much ambience being captured... which is normally great, you get that natural ambient detail all-around, but sometimes, it's not so great when you want deep blacks or have problems with the ambient light spilling the wrong colors into the shadows, like greenish or reddish light in the shadows that you don't want. Then an overall higher light level will help overpower some of that natural ambience.

For me, it's only been in the last couple of generations of 500T stocks that I've become less bothered by the grain; I used to be a lot more motivated to shoot as much of the interiors on 200-to-250 ASA stocks, if not slower (I did a 16mm short film on 50D and 100T stocks that looked great, back in 1989.) My favorite tungsten film back in the early 1990's was Fuji 250T, I shot a lot of features on that, including most of "Twin Falls Idaho".
 
Hi There,

I was wondering if anybody can help. I need to know if there is some sort of multiple EDL support for RM.

I render a lot of DPX. sequences from EDL's made in AVID, and they are often different versions of the same edit.
This means that in each EDL some shots remain the same, and a couple change each time.

How do i avoid rendering duplicate shots for each EDL, without manually working out which shots havn't changed, and then unticking them?

ANy help will be appreciated

Thanks,

Shaun
 
I started a similar thread last month on Cinematography.Com after the first thread appeared on this site (this is the second time this thread has started on RedUser about that youtube clip.)

It's very lovely and shows why even if Kodak stopped making film someday, they have a tremendous value as an institution preserving our film history.
 
Hi David- Thanks for your color timing help!
I have another Red timing question for you. I'm in the middle of timing my feature right now. We shot with 2 different Red cameras on our shoot. The bulk of the film on one, and 4 days on a second body. The RAW files from the 2nd body look very different from the first body. The greens and yellows are very, very saturated-- which lead me to believe that these are not actually the RAW files that we're working with. The colorist assures me that they are the RAW files. I don't understand why the 2 bodies would have such different RAW files, since a camera setting wouldn't affect the RAW saturation or color. Any thoughts on what might be going on? Thanks!!
 
Hi David-
We've even gone back to the proxies and looked at them-- the greens are supersaturated there too. It's not the overall green/yellow shift you get when importing the new Red sensor files incorrectly. It's bizarre-- everything else looks RAW- skintones, everything else desaturated, but green grass has never looked so green!
Julie
 
Julie K said:
Hi David-
We've even gone back to the proxies and looked at them-- the greens are supersaturated there too. It's not the overall green/yellow shift you get when importing the new Red sensor files incorrectly. It's bizarre-- everything else looks RAW- skintones, everything else desaturated, but green grass has never looked so green!
Julie
Click to expand...

Are you sure that both cameras were set to default? Can you reset the strange clips back to default? I had an issue with two cameras last week on a shoot where one camera had the Saturation set to 1.2 while the other was at 1. Big difference!
 
Remember that you can't actually "look" at a RAW file, it has to be converted to RGB to be viewable, so there is always a conversion process going on. So often major mismatches between Red cameras (assuming all else is equal, same lenses, filters, etc.) are due to metadata settings or conversion settings.

For example, in the camera you can play with RGB gains and overall Color levels, which will cause shifts to the converted files. When I did "Stay Cool", I had some conversion problems because I had pushed the Chroma in the camera metadata from "1" to "1.5" I believe, causing some artifacts to pop up. But I also had some dailies come back mismatched in saturation -- for some reason, the metadata off of one camera was not being read correctly in the conversion software.

But I've also made mistakes, like lowering the green gain to correct some fluorescents and left that settings change in the next day, only to end up with a magenta tint to my footage (again, that's metadata only, it's not in the actual shot).

But if you are sure that all camera settings match and all post software settings match, like in RedCine-X, then send the R3D files to Red and see if they can tell what is going on. Perhaps you have a bug in one of the two cameras that is causing the metadata to be recorded improperly. But odds are high that the actual original files are fine.

Now keep in mind that RAW files have to be converted correctly to RGB or else you often have green casts and whatnot due to the fact that 50% of the photosites are green, 25% are blue and 25% are red in a Bayer filter pattern. The earlier color science behind conversion to Rec.709 color space was a bit off in the Red, often resulting in exaggerated greens, causing the most minor green spike in fluorescents to be very visible and exaggerating green objects in the frame. I don't know what color space you chose for shooting and outputting (I recommend RedColor in the camera and RedGamma for the output if for TV.)

The other issue is black shading, whether this was done properly; if not, your blacks can have a color cast like green, but that doesn't sound like the issue here.

It terms of timing a feature, either the D.I. facility is working directly from the R3D files and doing real-time conversion, or they first converted all the footage to 10-bit DPX RGB Log files. If the second, perhaps the conversion was done wrong, but again, they would just be following the camera's metadata likely so that could also be the culprit.
 
Last edited:
Thanks so much for your help, guys!
We did some more investigating and it turned out to be the way the files were brought into the system-- wrong gamma, wrong color space. All fixed now. Thanks again!
 
Hi David,
So, last weekend I was shooting a short. It was intended to be practice for a 48 hour film contest, so we were shooting very quickly and I tried to use as few angles as I needed. When I went through the rough cut with my writer, he remarked that I kept using "the same shot". I noticed, that I had, in fact, taken many shots of my lead actress using similar angles: profile CUs and MCUs with medium telephoto (examples below).

In general, how do you feel about using similar framing of the same character in several scenes? I feel like if you do it too much it will start to get boring, like the DP can't think of anything better to do. On the other hand, it does establish a certain consistency for the characters, and I feel like using a million angles can be jarring.

Also, how do you think the mood conveyed by profile shots of talent compares to that conveyed by more straight-on shots. My tendency would be to think that they put more distance between the viewer and the character, because they aren't looking onto the eyeline, but I don't think that this is always true.

14-0010503408T.jpg
 
This is all a judgment call based on the needs of the individual project. There can be artistic/symbolic/thematic reasons for visual repetition in how a character is shot, you just have to weigh that against the fact that repetition can be boring for the audience if they don't "get" why it's being employed and just think the director is unimaginative or lazy.

As for the eyeline issue, generally when an actor's gaze is more direct into the camera, it connects the viewer to the actor in a stronger way. "Silence of the Lambs" is a great example, the interchanges between Hopkins and Foster are often clean singles (close-ups) where each actor gazes almost directly into the lens, or actually into the lens (it's hard to tell which.)

However, I also feel that semi-profile angles can be emotionally rewarding in a different way. They give the viewer the effect of observing the actor from a slight distance, but that doesn't mean necessarily that the emotions of the actor aren't being communicated to the audience. Truth is that sometimes the back of an actor's head can be very powerful emotionally.

One of my favorite shots is from "My Darling Clementine" -- when the three Earp brothers come back to their camp and find their youngest brother murdered and all their cattle stolen... rather than show a reaction shot of the brothers, John Ford shot a highish angle looking down at the dead brother in the rain, and in the foreground you saw three cowboy hats looking down, with the rain pouring off of them. A very sad and powerful image, these three big men just looking defeated with the rain beating down on them, as they gaze on their brother's body. Much more effective than seeing their faces.

Anyway, the other reason to use less frontal angles now and then is simply that when you do use the frontal close-up with the tight eyeline, it will have a greater impact due to its rarity and novelty, whereas if every close-up was that way, it would start to lose its power.
 
Super 35 DI vs Anamorphic

Super 35 DI vs Anamorphic

Still a noob, and maybe I'm just seeing things but I've been watching some recent Super 35 scope movies and for some odd reason they seem to be sharper than some other anamorphic movies. Obviously the opposite should be true but No Country For Old Men is razor sharp -- almost jump off the screen sharp. Same with parts of Kingdom of Heaven, Quantum of Solace and The International (which I know had some 65 aerials). Watching the Nolan Batman films (intentionally knowing they were photochemically timed) they didn't seem have the same bite (though some of the contrastier stuff in Batman Begins does and also taking into account Wally seems to have shot TDK at a shallower stop, the focus falls off pretty fast -- maybe that's what I'm seeing).

Watched Indy 4 (different style, diffusion, deeper stop) and the same thing. It's sharp, in focus, and parts of it are pretty contrasty but rarely is it that 'make you bleed' sharp look and nor does Star Trek (noting the frost used on the bridge scenes). I guess Generations is pretty crisp.

My question is am I seeing the difference between anamorphic with a 2K DI, versus Spherical with a 2K DI, is it lenses, film stock, everyone these days on commercial films seems to stick to a handful of stocks, lighting, or over sharpening in the DI, or as I suspect a combination of all the above with unique variables for each film. Maybe coming from a video background I'm noticing the Super 35 stuff seems almost HD-ish these days.


There are a number of shots in Quantum of Solace (the first shot after the titles with the flags for example) where I felt like I could put my hand through the screen. As an aside I really like Roberto's work in QoS. How would you describe that look? Seems very sharp, deep focus with funky compositions to capture the production design, dolly moves across glass from weird angles, funky blue-green in 'colder' scenes, soft toplight in others but always naturalistic. His ASC interview didn't really give me a whole lot of insight other than basically framing for 'scope and working with Dennis Gassner and Marc Forster (who has a very unique almost Bauhaus aesthetic in his own right) to develop the look.
 
Most of the perception of sharpness in some Super-35 films is coming from the use of harder lighting or higher contrast lighting because that creates more visual edges to things. "No Country For Old Men" was shot on Zeiss Master Primes, which are razor sharp, but the really crisp scenes in the movie also tend to be shot in harsh desert sunlight on slower-speed film stock as well. "Quantum of Solace" used harder lighting for the contrasty, dramatic, noir-ish feeling; Roberto in general is not someone who lights every movie with big soft lights, he likes more control and directionality, more drama. Also Roberto has a tremendous sense of composition, inspired a lot by Antonioni I think -- that always stands out to me about his work, this architectural sensibility.

The biggest advantage to anamorphic isn't so much edge sharpness as it is the lower grain from using a larger negative, and that you can capture subtle detail better, even in soft light. But a lot of those subtle advantages are only visible on very large movie theater screens.

However, the use of D.I.'s lately has made both Super-35 and anamorphic movies look much closer in quality than they used to, so a Super-35 movie shot in crisper light with more contrast may look snappier than an anamorphic film shot in softer light or with less contrast. Certainly Roger Deakins' Super-35 movies look as good as anything shot in anamorphic in terms of sharpness. But you can also find a number of mediocre Super-35 movies where you don't feel the same depth and detail as with anamorphic photography.

I have noticed that on Blu-Ray, Super-35 movies look better than they did in the theaters in film prints. They also look better when digitally projected. I think it's partly because the slightly grainier look of Super-35 (compared to anamorphic using the same film stock) is less noticeable in digital presentations, particular for home HDTV viewing.
 
Thanks so much for your response. Once I got past the clunky editing in Quantum and wasn't trying to figure out what was going on the visual style jumped out at me as very unique for any genre.

Along those lines, as both a DP and a film fan, are there any franchises that you would love to be a part of like Star Trek or Batman or Bond (some of those DP's seem to almost moonwalk into those roles seemingly due to sometimes idiosyncratic director -- Star Trek is a culprit -- choices which gives them unique opportunities to show off their talent I suppose)? You seem very adept at managing both soft and warm (Akeelah and Astronaut Farmer) with sharp and edgy, almost graphic novel-like (certainly parts of Jennifer's Body and Stay Cool). Apart from obvious story demands and input from the director how do you think you might go about lighting a Bond or a Star Trek should you be asked since those franchises tend to have characteristic or at least signature looks?
 
My favorite Bond movies for cinematography have a strong sense of classic film noir + glamour to them, particularly "Goldeneye", "Quantum of Solace", "The Living Daylights", and "From Russia With Love". So harder, noirish lighting for some scenes, traditional glamour lighting for women for others, and great widescreen travelogue photography for day exteriors is all part of the Bond look. I think it would be a mistake to light a Bond film too softly or naturally, there has to be some gloss to it when necessary, and some grit and edginess when necessary.

The "Star Trek" movies, on the other hand, well it depends on the script but I like it when I can believe the lighting inside the spaceships, that it comes from the sources on the set. I'd lean more towards a "2001" and "Alien" approach if I had a choice rather than mimic too closely the 60's TV show lighting, other than the use of colored light, which I would do IF I could motivate it logically into the set design (like the use of colored light in "2001"). I liked the look of the new "Star Trek" movie, the darkness of the Kelvin, the clean Apple Store look of the Enterprise, the bright lights flaring the lens. I have a few quibbles with the production design when it mixed real locations (like the Anheiser-Bush brewery) with sets because sometimes the technology seemed mismatched, though that happens in real life, not all spaces and equipment are built at the same time, you can have a new computer in an old building, etc.

I'm probably one of the few fans of "Star Trek: The Motion Picture" (though some of that pristine look is in the new "Star Trek" movie, so maybe J.J. Abrams liked parts of that movie too...) -- there are elements in most of the Trek movies I like photographically. The opening Vulcan interior scene in "Star Trek 4: The Voyage Home" is quite lovely. The camera moves in the ones directed by Nicholas Meyer are great. Some of the darker interior scenes in "Star Trek 3" are nice and moody. The look of Enterprise bridge in "Star Trek: Nemesis" was good, as were the scenes on the Romulan ship.
 
You must log in or register to reply here.
Back
Top