Ask David Mullen ANYTHING

[Stefan Christou]

Is there a set list of filters you will always have at hand? What is each used for? (Sorry if this is a bit long and tedious to list.)

 

[David Mullen ASC]

Is there a set list of filters you will always have at hand? What is each used for? (Sorry if this is a bit long and tedious to list.)

Besides ND's and Polas?

Well, if I am shooting tungsten-balanced stock, I'll also carry 85/ND and 85/Pola combos.

ND grads, usually a ND.6 and ND.9 Soft Edge Grad and a ND.6 Attentuator. An ND.30 Soft Edge Grad is too subtle to be worth it, and hard edge grads only work on longer lenses. Ultimately, you end up with an ND.60 grad 90% of the time because an ND.9 grad is pretty obvious.

I usually carry a +1, +2, +3 Diopter for close-up work.

I sometimes carry Split-Diopters in those strengths as well, but if I sense a director doesn't like that effect, I usually send them back after awhile.

Optical flats for anytime the lens may get sprayed with something, like during a blood squib, dust explosion, etc.

That sort of leaves diffusion / contrast filters and moderate color-correction filters for warming or cooling off of the image. That tends to be more show-dependent. If I'm doing a show with a specific diffused look, then I'll pick the filter that gives me that look, in different strengths. However, if I'm shooting a "sharp" show, then I might carry a mild diffusion that is pretty subtle, just in case an actress needs it one day, or there is a special scene that needs it. I might also carry something for dream sequences or flashbacks. In emergencies, I have some homemade old pantyhose net filters I carry everywhere with me, even if I've hardly ever used them.

And as I said, if I'm doing a smoked show, I'll try to carry a light ProMist and a #1 and #2 Smoque filter for matching shots, inserts, etc. If I couldn't get a Smoque, I'd probably try some heavier UltraCons, or some lighter LowCons or Fog. I recently had a sequence in my last show with a lot of smoke on one set, and then they asked me to shoot a wide shot of a house in the daytime as an establishing shot before we cut into these smoked interiors. So I used a #1/4 Fog on the wide shot so that it would have some of the softness and milkiness of the interior shots, though a #1/8 Fog would have been better (I was carrying the #1/4 Fog for another reason, for a flashback memory scene we had just done.) A #1 Smoque would have been even better. But it's a hard filter to find at rental houses -- I probably should just buy one from Tiffen before they stop making them due to lack of interest...

I don't use light warming or cooling filters much because it's easy to time that effect in post, or to fool the digital camera by playing with the color temp. Of course, with the RED camera, some cooling filters would help to reduce some noise in tungsten-lit scenes. On film shows, I do carry some warming and cooling filters, but not for the scenes, but to shoot the grey scale at the head of the roll, this way when they time-out the filter on the grey scale, the unfiltered scene that follows has the opposite color cast. This is a quick and dirty way of getting dailies timed with a deliberate color bias. I don't bother actually shooting the scene with the filter because it's just another piece of glass on the lens and the effect is so easily cancelled or created in post that half the time, it disappears at some point in the post chain and you end up recreating it in the final timing anyway, so why mess with filters on the camera lens if you don't need to? All you really need is for the dailies to have the effect so everyone, through editing, knows that the intention is a warm or cold color bias.

 

[David Mullen ASC]

Hi Mr. Mullen,

I really enjoy your erudite visuals! In a shot such as this, maybe after reviewing dailies or watching the DI in progress, would you ever suggest to add a smoke/haze effect in post just to keep the continuity of the look?

One often plays with the blacks and gamma in a color-correction session to match smoked shots in a sequence. Beyond that, you can easily add a hazy grad effect to simulate a shaft of light. But anything more complex requires creating masks around foreground objects and whatnot to simulate smoke in depth, which gets time-consuming in an expensive D.I. session.

"Northfork" was finished photochemically, but when I did the home video transfer, I was able to match some smoke levels better. I was filming in some very drafty sets built out on the real prairies of Montana, so it was really hard to get the smoke to stay inside. In this scene, for example, I did the over-the-shoulder but the close-up that followed did not have any smokey haze in it because the smoke got too light. So in the home video version, I lifted the blacks on the left side of the frame of the close-up to give the impression of haze to match the over-the-shoulder better:

 

[scott william]

Hi David, i watched Assassination of a High School President a few weeks ago not realising it was your film. It must have been fun to shoot, great job. Sad to hear it didn't go to cinemas and probably wont even make it to Australia on dvd.

 

[Kenneth Elkington]

Daivd, how do you feel about the use of telephoto diopters and wide-angle converters? Are they difficult to match to prime lenses? What are the real barriers to using them? Would you say that they can be used to replace the wider and longer ends of the focal length range, or just used as a narrow supplement?

And given all that, who, in your opinion, makes the best ones?

Thanks.

 

[Sarah C.]

@David, re: Northfork images

Beautiful! Thanks for posting them.

~s

 

[David Mullen ASC]

Daivd, how do you feel about the use of telephoto diopters and wide-angle converters? Are they difficult to match to prime lenses? What are the real barriers to using them? Would you say that they can be used to replace the wider and longer ends of the focal length range, or just used as a narrow supplement?

And given all that, who, in your opinion, makes the best ones?

Thanks.

I don't use converters too often so I can't say who makes the best ones, I guess Century Precision Optics is a big name in that arena. It's extra glass, so I'd rather use a prime lens of the focal length I want than add a converter to get there, unless there is no other choice. Plus you have to figure a light loss with some converters. But for special shots where there are no other options, sure, I'd use a tele or wide-angle converter. For example, ARRI tilt-focus lenses only come in three focal lengths so most rental houses supply converters to achieve focal lengths in between. And I've used the extender on ENG zooms before to get super telephoto shots.

And I've used wide-angle adaptors on DV cameras to get security camera POV shots and whatnot.

I just don't see them as a general substitute for getting prime lenses in the correct focal length; I see them more as tools to use in a pinch to get select shots now and then.

But other people use them more freely than I do, especially tele converters. A lot of action scenes in "A Bridge Too Far" were shot with anamorphic zooms with tele converters to get very telephoto shots during battle. Not the most optically sharp-looking shots, but I think Unsworth pulled his customary Fog #2 filter on some of those super telephoto shots to compensate for the loss of sharpness from the tele converter.

 

[Stefan Christou]

Thanks David that was really informative. I might have to get a #1 Smoque before they go too.

S

 

[albert rudnicki]

David
Are "effect" filters better then the ones made in post production?
Specifically the smoque.

My background is post and I am reluctant on using effect filters on the lens.
I always imagined, that you can have more options in post, as long I you have well lit and exposed image to start with.

Thanks

 

[David Mullen ASC]

Well, I'm not a post person so it depends on who is doing the post -- the problem is that if it's done in post, it goes out of my hands. I listed the pros and cons for doing filtration in-camera versus in-post in a different thread. Here's what I wrote:

But if you are talking about diffusion filters, there is no consensus on that, a lot of people subscribe to the "save it for post" theory of not baking-in a look like that into the original recording.

However, reasons for doing it in-camera may be:

1) It's more fun.

2) Everyone on set and in editorial later sees the intended effect immediately, which may be particularly important if the diffusion is being used to make your lead actress look better... and she's checking playback or looking at dailies.

3) People get so used to how dailies look in editing over months that when it's time to do the final color-correction, they don't want to create new looks.

4) You may not be there for the final color-correction or the project may be taken over by someone else who won't added the intended diffusion effect.

5) Adding diffusion to an image in post does involve a certain amount of extra processing time, and some people end up dropping the idea rather than deal with that extra step.

6) Most forms of post diffusion have the side-effect of softening grain or noise, which may or may not be desirable.

Having done one feature where we planned on doing the diffusion in post, I now regret I didn't do it in-camera. For one thing, the producer took the project away from the director and myself and decided to not add the diffusion. Second, even while we were working on the post, adding diffusion to every shot was slowing down the D.I. process, which was costing more money.

The truth is that if I'm not there to actually supervise the effect in post, I'm not sure I want to leave it to others.

Also, not all color-correction suites in Los Angeles have decent diffusion effects software available. So the post house charges you more to move the session into a room with the proper software (too many basic color-correctors simply have defocusing tools, not diffusion tools.) I don't want to shoot a whole movie clean if the intended look is diffused, only to enter the post phase and find that the producer has picked a post house that can't do the diffusion effect I want for the budget that we have.

Also, it's not risky to use a diffusion filter if you know what you're doing -- movies have been using diffusion filters for 100 years now, still photography even longer. It hasn't suddenly become too risky to do it in-camera.

I was just watching the new "Star Trek" movie on Blu-Ray and noticed in the behind-the-scenes footage that the Panaflex camera mattebox has a #1/8 Schneider Black Frost diffusion filter label on it during the Enterprise interior scenes. So if a movie that costs over 100 million dollars can risk using a light diffusion filter on the lens, certainly a small-budget movie can. Even the efx-heavy "Lord of the Rings" movies used both digital diffusion and regular filters like ProMists and nets to diffuse the image.

But the main reason I do it in-camera is that I'm a cinematographer, so most of my work is done using photographic tools -- once it goes into post, I may or may not be asked to be involved, and I'm not the one operating the equipment, I have to rely on post houses, colorists, etc. and I'm at the mercy of the post budget. So I don't want to save an intended visual look for post IF it's very fast, simple and easy to do it when shooting... only to have that look dropped in post for various reasons out of my control.

I'm still of the old-fashioned opinion is that you get close to the look you want in the original photography so that the post process is mostly about enhancement rather than creating the look out of whole cloth from scratch. But on the other hand, I'm a cinematographer who mostly shoots dramas, dialogue scenes with actors, etc. not effects-driven CGI-heavy movies, not commercials, not music videos, etc.

I also think optical diffusion has a certain organic quality to it, it "feels" optical in nature, not electronic in nature -- it feels like light hitting glass and reacting, which is exactly what is happening. And you get nice happy visual accidents now and then from the interaction of the filter and the light. I like optical accidents and textures to happen live rather than to be created artificially in post. But it all depends on the project.

Now with actual smoke on the set, that's very dimensional so it works a lot better than doing it in post.

 

[albert rudnicki]

In ideal world where you have the control from the start to finish.
Would you go for post effect over the "on set" one ?
Is there a quality advantage?

Thank you so much for sharing your knowledge, experience on this forum!

 

[David Mullen ASC]

Well, in this imaginary ideal world, I'd probably use a mix of in-camera and post techniques to create the diffused look, like "Lord of the Rings" did. It's not so much a gross issue of quality so much as a difference in look and the question of time and effort, on set and in post. If it's easier to do it on the set and it looks good, then you've saved yourself some time in the final color-correction to concentrate on other, more important things. Unless this ideal world also includes an unlimited budget for post so you can take as much time as you want.

But honestly, if all you want is a lampshade to halate a little in a shot and it's easy to do it in-camera, then I don't see why someone would save it for post unless they're the type of filmmaker who can't commit to a creative impulse and wants the opportunity to endlessly change their minds. I tend to believe that one grows as an artist by making bold choices and living with the consequences, good or bad, and learning from the experience and then moving on. So it's partly a philosophical issue, I like the creative process of making movies during the photography phase -- that's why I became a cinematographer! If I believed one should save every important creative decision that affected the image until post... I would have gone into post.

Not to say that post isn't creative, because it is -- it's just that I subscribe to Kurosawa's notion that shooting and post, or picture, editing, and music in his words, is a multiplicative process, not an additive one. You build on the previous stage to create something stronger than merely the sum of the parts. He also said that if directing could be simplified into a few rules, it would be that you put the best elements you could in front of the camera, you shot it in the best manner you could, and then you edited it the best way you could, etc. So production is not merely fodder for the post phase -every step has to be done as well as possible. Hence why I believe you have to go "all out" and shoot the movie in the best manner possible, figuring that post can only make it better -- rather than see production as merely gathering raw material to then create the art in post.

But coming down from my high horse, obviously imagemaking and cinematography involves an understanding of the entire chain of events and techniques leading to the final presentation, and you have to know what is better or more efficient to do in post versus doing it in camera, so you use your time and budget efficiently.

Practically speaking, I've mentioned this before, but I'm of the "half-and-half" or "50-50" philosophy, that you get some of the look baked into the original and then you finish it off in post, rather than create it wholesale in post. From merely a technical standpoint, this often looks better anyway because you aren't pushing an image signal around to extreme levels trying to create a look from footage that isn't close to being correct, you are instead putting the final touches, the final polish, onto the work. Garbage-in, garbage-out still applies to digital photography and post. The best images in a movie, that go through the D.I. process the best, often were the best images going into the process, and the worst shots often still are the worst shots at the end of the process.

But as for whether a Classic Soft or a ProMist diffusion effect looks better done in-camera or in post, it just depends on your taste -- to my eyes, I can generally tell when a filter effect on the large screen (not so much on TV) is done optically or through image processing. And philosophically, I feel that diffusion should feel like an optical phenomenon or artifact in most cases -- however, I also admit there are exceptions, when the diffusion effect is more of an added visual texture, like printing a photograph on parchment paper would be a texture and not an optical artifact. So it partly depends on the visual style and intent of the project.

As I also said, one problem with a lot of post digital diffusion tricks is that they have the side effect of defocusing noise and grain because most processes are based, like optical diffusion, on the notion of superimposing an out-of-focus layer over a sharp layer (except that when it's done with an optical filter, it doesn't soften the grain). Sometimes that's a good thing, when you want a cleaner image, but other times, it can make the image look too plasticy and sterile, too clean, especially if it's a film-shot project and you're intercutting shots that aren't processed with ones that are, so the grain level keeps jumping.

 

[Antonio Forjaz]

Quick question David,

I start tomorrow shooting a 26 min film that will only be on TV (PAL 16:9) it will never end up in the cinema.
I have the Zeiss Prime lenses, f1.9 18mm,25mm,35mm,50mm and 85mm.
Should I shoot it on 4.5K or 4K?
What would your choice be?
I have enough storage.
Thanks

 

[David Mullen ASC]

I'd probably shoot "4K HD" mode, aka "Quad HD", which is actually 3.8K (1920 x 1080 doubled in all directions, so 3840 x 2160). Apparently this downconverts to 1920 x 1080 HD a little better because it's a simpler whole integer process. I haven't tested it myself yet, that's just what I've heard here.

 

[albert rudnicki]

Well, in this imaginary ideal world, I'd probably use a mix of in-camera and post techniques to create the diffused look, like "Lord of the Rings" did. It's not so much a gross issue of quality so much as a difference in look and the question of time and effort, on set and in post. If it's easier to do it on the set and it looks good, then you've saved yourself some time in the final color-correction to concentrate on other, more important things. Unless this ideal world also includes an unlimited budget for post so you can take as much time as you want.

But honestly, if all you want is a lampshade to halate a little in a shot and it's easy to do it in-camera, then I don't see why someone would save it for post unless they're the type of filmmaker who can't commit to a creative impulse and wants the opportunity to endlessly change their minds. I tend to believe that one grows as an artist by making bold choices and living with the consequences, good or bad, and learning from the experience and then moving on. So it's partly a philosophical issue, I like the creative process of making movies during the photography phase -- that's why I became a cinematographer! If I believed one should save every important creative decision that affected the image until post... I would have gone into post.

Not to say that post isn't creative, because it is -- it's just that I subscribe to Kurosawa's notion that shooting and post, or picture, editing, and music in his words, is a multiplicative process, not an additive one. You build on the previous stage to create something stronger than merely the sum of the parts. He also said that if directing could be simplified into a few rules, it would be that you put the best elements you could in front of the camera, you shot it in the best manner you could, and then you edited it the best way you could, etc. So production is not merely fodder for the post phase -every step has to be done as well as possible. Hence why I believe you have to go "all out" and shoot the movie in the best manner possible, figuring that post can only make it better -- rather than see production as merely gathering raw material to then create the art in post.

But coming down from my high horse, obviously imagemaking and cinematography involves an understanding of the entire chain of events and techniques leading to the final presentation, and you have to know what is better or more efficient to do in post versus doing it in camera, so you use your time and budget efficiently.

Practically speaking, I've mentioned this before, but I'm of the "half-and-half" or "50-50" philosophy, that you get some of the look baked into the original and then you finish it off in post, rather than create it wholesale in post. From merely a technical standpoint, this often looks better anyway because you aren't pushing an image signal around to extreme levels trying to create a look from footage that isn't close to being correct, you are instead putting the final touches, the final polish, onto the work. Garbage-in, garbage-out still applies to digital photography and post. The best images in a movie, that go through the D.I. process the best, often were the best images going into the process, and the worst shots often still are the worst shots at the end of the process.

But as for whether a Classic Soft or a ProMist diffusion effect looks better done in-camera or in post, it just depends on your taste -- to my eyes, I can generally tell when a filter effect on the large screen (not so much on TV) is done optically or through image processing. And philosophically, I feel that diffusion should feel like an optical phenomenon or artifact in most cases -- however, I also admit there are exceptions, when the diffusion effect is more of an added visual texture, like printing a photograph on parchment paper would be a texture and not an optical artifact. So it partly depends on the visual style and intent of the project.

As I also said, one problem with a lot of post digital diffusion tricks is that they have the side effect of defocusing noise and grain because most processes are based, like optical diffusion, on the notion of superimposing an out-of-focus layer over a sharp layer (except that when it's done with an optical filter, it doesn't soften the grain). Sometimes that's a good thing, when you want a cleaner image, but other times, it can make the image look too plasticy and sterile, too clean, especially if it's a film-shot project and you're intercutting shots that aren't processed with ones that are, so the grain level keeps jumping.

David,

I guess even on big shows the same rules apply: once shot, it ain't yours....
In ideal world :001_smile:you would grade your shots your self, or at least assist colorist in establishing the look.
Do you tend to use the same colorist ?

You have mentioned Classic Soft and ProMist as the two you would use on camera as a better alternative to the one in post.
Are they any others you would recommend?


Thank you once again David !

 

[David Mullen ASC]

It's a little like camera operating: I always want to work with someone more talented at it than I am, or else what's the point. Same goes for coloring -- most colorists are better at it than I am, because they do it all day long, day after day, year after year. The same reason why I'm better at lighting than a colorist would be, because I do it for a living, many days a year.

So unless a colorist is incompetent or obstructionist, someone who for some reason, colors to please himself and not the client, then I'd rather work with a colorist than do it myself. I need that experience, talent, and a second pair of eyes. I also like having someone to bounce ideas off of.

However, the more experience I have with post coloring, the better and more accurately I can communicate with the colorist, or know what their problems are going to be.

Now I suppose a little personal short film, it may be fun to try doing all the coloring work myself in post, though I may not be any good at that...

On rare occasion, you do end up with a colorist whose own visual taste is so far off from yours, or someone with their own agenda, that doing the color-correcting is unnecessarily painful and confrontational. But that happens rarely (in fact, it's only happened once in my career.)

I'm not touting the superiority of Classic Softs or ProMist, I only used those names because they are commonly recognized. There are a lot of diffusion filters on the market and they all have their quirks, their looks, etc. and it's just a matter of taste which your prefer. Some create artifacts that would be harder to replicate in post than others.

The thing to keep in mind is that I'm speaking as someone who shoots features and dramatic TV shows, not commercials and music videos or effects shots. By that I mean that I shoot a lot of footage every day and the final product will be a 2-hour movie or maybe 12 hours of TV episodes, etc. That's a lot of footage to process every day if you wanted to add a post diffusion effect to it. And even if you save it for the end, a typical movie may have a 1000 cuts in it, which again, is a lot of shots to individually add a diffusion effect to. So whether you do it in-camera or save it for post may partly be a practical matter dependent on the volume of shots you are talking about.

 

[Antonio Forjaz]

I'd probably shoot "4K HD" mode, aka "Quad HD", which is actually 3.8K (1920 x 1080 doubled in all directions, so 3840 x 2160). Apparently this downconverts to 1920 x 1080 HD a little better because it's a simpler whole integer process. I haven't tested it myself yet, that's just what I've heard here.

Thanks for the advice again

 

[Miguel "Macgregor" De Olaso]

Could you explain a little bit the lights involved in this shot?

I belive you are bouncing some HMIs inside above the windows. Would it look different if you had used kinoflo banks? How do you balance the amount of light between the outisde in inside?
Any filling? thanks

 

[David Mullen ASC]

Actually, in this room, the windows were so large and tall that I seem to recall that I shot it mostly with available light. I had one Kino off camera left to side-light the guy at the window just because I didn't want him to become a silhouette. I may have bounced a little light off of the ceiling too, in which case it probably was the HMI Source-4, but knocked down. I really wanted to use mostly natural light in this room because of the large windows. In the scene, someone pops up big in the foreground and for that person, I had a little Kino off camera.

The hard part was picking an exposure between the exterior and the interior because I wanted to hold some texture outside, though it was a hazy day looking out at the harbor, so there wasn't much to see.

 

[Ryan E. Walters]

David,

A follow up question to the previously linked image- how do you manage coverage for a scene lit like that- when you go in for singles are you adding fill / lighting each of them? And if so, does it cause much of a problem for continuity / editing since they are not lit in the wide? Or does that get "hidden" in the compositional choices of the singles?