Ask David Mullen ANYTHING

[David Mullen ASC]

Hi David, would it be a right approach if I do not have 80C/80D filter on a lens, just to put 1/4 or 1/8 CBT on 3200K lights?

CTB (Color Temperature Blue) not CBT... Yes, it's the same thing IF you are gelling every tungsten light source the same amount and there are no other sources of light in the scene. Obviously it can be hard to add gel to a lightbulb in a practical lamp, or to gel dozens of lights, or have a room of mixed color temperature sources, at which point a single filter over the lens would be a simpler way to increase the color temperature (i.e. shift it to blue).

Keep in mind though that the 80C and 80D filters convert more than a 1/8 or 1/4 CTB gel, they are more like a 1/2 CTB correction in terms of Mired Shift (in parenthesis):

80C (-81)
80D (-56)

1/2 CTB (-68)
1/4 CTB (-30)
1/8 CTB (-12)

[Jon MIchael Puntervold]

Dear mr. Mullen,

lately I've become quite interested in experimenting with the grammar and art form of stereoscopic cinema. The late Alain Derobe is often referred to as a brilliant resource on stereoscopy, but unfortunately I've not found any books by him. Neither good books or documentaries covering the subject of 3D-cinema. Do you have any suggestions for somebody who would like to go in depth on the subject?

Best regards,
Jon M. Puntervold

[David Mullen ASC]

Dear mr. Mullen,

lately I've become quite interested in experimenting with the grammar and art form of stereoscopic cinema. The late Alain Derobe is often referred to as a brilliant resource on stereoscopy, but unfortunately I've not found any books by him. Neither good books or documentaries covering the subject of 3D-cinema. Do you have any suggestions for somebody who would like to go in depth on the subject?

Best regards,
Jon M. Puntervold

Sorry, 3D isn't something I've studied in depth (no pun intended) yet...

[Brandon J. Garcia]

I am shooting a film coming up and we are going to shooting some macro shots of paper + typewriting.
I was wondering if you guys had any experience with this and could recommend paper stock and/or lighting techniques for these types of shots.

We will be shooting on 100mm Macro, I'm open to lens suggestions as well.

Of course, we will be testing soon but I thought I'd get some ideas about what we should test.

Thanks
BG

[scott william]

Hi David,

I’ve been reading quite a bit about noisy shadows and low light shooting here recently. Still a little confused about how the sensor and the noise floor work.

If I rate the camera at 800 ISO and light my subject at an f-stop of 2.8 but keep almost all light of the background as I want it dark, will that create more noise in the image than if I were to light my subject at 5.6, fill light the background and iris down until it is as dark as I want it? Or is it the same thing.

I guess what my question is, to achieve a cleaner image is it better to lift the shadows and iris down than to not light the shadows at all.

If you were shooting a dark scene on Epic and wanted to keep the image as clean as possible but shoot wide open how would you achieve it?

[Lorenzo Levrini]

If I rate the camera at 800 ISO and light my subject at an f-stop of 2.8 but keep almost all light of the background as I want it dark, will that create more noise in the image than if I were to light my subject at 5.6, fill light the background and iris down until it is as dark as I want it? Or is it the same thing.

I am obviously no authority on the subject so I'm ready to be corrected, but I would put to you that the sensor sees no difference between less footlambert on an open stop and more footlambert on a stop correctly closed to compensate, because the quantity of light actually getting through to the sensor is the same in both cases.

To use a film analogy, you are interested in the density of the shadows. More density printed down will make your shadows 'cleaner'. The density is the same whether you have less footlambert and open up or more footlambert and close down. So all other things being the same, shadow 'cleanliness' is not a matter of footlamberts or stop, but a matter of where the shadows sit on the density scale (the 'image value' bit scale in digital) vs where they are printed (or with digital, how far they are 'crushed' in post).

[David Mullen ASC]

Assuming that the lighting contrast ratio is the same, the noise level shouldn't change if you light to an f/2.8 and then shoot at an f/2.8 versus use more light and get up to an f/5.6 level and then shoot at f/5.6 -- in theory, the same amount of light is getting past the iris to the sensor, it doesn't know how much light there was before the iris. Same goes for using an ND filter, if you light to an f/5.6 and shoot at an f/5.6, the noise level should be the same as when you then add an ND.60 and then shoot at an f/2.8.

Now in the real world, there may be some minor differences due to the contrast of the lens if shot wide-open versus stopped-down, etc. And this is all assuming that the lenses are accurately marked.

But you said "fill light the background" which implies a change in contrast ratio.

I think you first need to separate some issues, one of which is the basic noise level. Forget the image for a moment and the mood lighting, etc. Just imagine shooting a grey card or pure black by shooting with the lens cap on. At lower ASA ratings, the noise will be lower and at high ASA ratings, the noise will be higher. So first you have to pick an ASA rating that gives you the noise level that you like or can live with. And obviously you are trading noise for highlight protection as you change the ASA rating.

But once you pick an ASA rating where you like or can accept the noise level, it shouldn't matter if one scene has a darker background than another anymore than it should matter if you panned through a black wall for a moment.

The problem really is that we sometimes end up making a correction, and some corrections will add noise if you are bringing up one or more of the channels, or boosting saturation in one of those channels. So how do you avoid making a correction? Well, you try to expose consistently and if not, you try to err on less underexposure.

You've already done a test let's say and picked an ASA rating where the noise level in midtones and blacks are what you want, so it shouldn't matter if one shot has a black background. But if you end up lighting that image, particularly lightening the shadows, you will be increasing the noise.

So when it comes to very moody and dark scenes, you really have to expose accurately, and if not, you have to expose conservatively knowing that you can make the scene darker in post without increasing noise.

Look at a moonlit scene, for example... you may do some tests and find that the correct look is when the key light is 2-stops underexposed. But as long as you do not lift that underexposed image in post, you haven't changed the ASA rating and you haven't changed the noise level. But let's say you accidentally underexposed by 2.5-stops instead of 2-stops and try to correct that 1/2-stop mistake. Now you may get some noise. So if you want the look of a 2-stop underexposure and you want some wiggle room for error, you may want to only underexpose by 1.5-stops, let's say, or only by 1-stop, and then further darken in post.

Same goes for shadow detail -- knowing that blacks will get more solid and less noisy if slightly pushed down (crushed), you may want to add a little bit more fill on the set knowing that some of that shadow detail will be darkened once you add that contrast to the blacks and low-end. You may even try crushing the blacks a bit on the monitor on set to preview the effect so that you tend to add more fill light to compensate.

But like I said, most of the time when people complain about noisy shadows, it's because they've picked an ASA rating that is borderline too noisy for their needs or tastes and then they accidentally underexposed a scene more than desired and have to bring it up a bit in post.

People also make the mistake in thinking that dark, shadowy scenes mean the lower levels of light have to be used, which is not true unless one of the light sources is naturally weak (like a candle). It's less about light levels and more about exposure levels and how that affects color-correction decisions.

To protect yourself, you need to give yourself a cushion (unless you have sophisticated noise reduction tools in post to save your underexposed footage.) When I shot 500 ASA film, I used to rate it at 320 ASA (a 2/3-stop overexposure) so that if I underexposed a scene by two stops for effect, I knew that I actually had 2/3's of a stop more exposure than what my meter told me. That was my safety net.

So if you really want cleaner images in dark scenes, use a slightly lower ASA rating and use a little bit more fill, and then plan on darkening the image overall, and also the shadows, in post. Or use the faster ratings and no fill, etc. but really be dead-on with your exposures and don't change your mind in post. The higher your ASA rating the less wiggle room you have to misexpose because you already are giving the sensor and the recording less information so you don't have room to mistakenly underexpose even more and then try and fix it later.

[scott william]

thanks david.

[David Mullen ASC]

I am shooting a film coming up and we are going to shooting some macro shots of paper + typewriting.
I was wondering if you guys had any experience with this and could recommend paper stock and/or lighting techniques for these types of shots.

We will be shooting on 100mm Macro, I'm open to lens suggestions as well.

Of course, we will be testing soon but I thought I'd get some ideas about what we should test.

Thanks
BG

Well, first thing is don't be afraid of lighting to a deep stop, there's no reason not to. Using a slightly off-white paper will help in a looser shot (like if you see someone's hands holding it) but once you get into a macro shot of paper you'd be exposing for the paper brightness anyway so it doesn't matter too much.

The real question is whether you can physically get the 100mm lens close enough to the paper as the keys are striking it -- at some point, you may have better luck with a telephoto lens with a diopter filter on it, like the long end of a 24-290mm Optimo, etc. It won't be as sharp as the 100mm macro lens however.

[dean merrill]

I am going to be shooting a car exterior ( no talent ), inside a warehouse type space mid next week. The idea is that there is a light up arrow on top of the car... and the shot starts black, and the arrow kicks on and we want to add fill light to look like its coming from the arrow illuminating the car and part of the room. A friend suggested using a very large (20x) silk, and I am wondering what you think. I was thinking a few open face 2k's to light the silk so we can spread the power load. We probably will have to bring in a generator for this. There is a budget, of course it isn't much.Any advice would be greatly appreciated.Thanks!