Ask David Mullen ANYTHING

[David Mullen ASC]

But it's the bounce off of the floor that fills in the faces. The point of the scene was how pure and white her carpet was before he tosses manure onto it to show off his vacuum (which doesn't work).

Anyway, often in the wide shots, I'll use an ND.6 grad at the bottom of frame to reduce clipping while keeping the ambient bounce in the shadows. In these examples, you are just looking at a Nikon JPEG snapshot.

As Mike Most can tell you, I love have small areas that are really hot in the frame. Caleb Deschanel once said that there should be something in the frame that is a bit out of control.

 

[Matt W.]

Caleb Deschanel once said that there should be something in the frame that is a bit out of control.

I agree it's a great look, though it's one technique for which dynamic range is a big deal so I've had trouble replicating it with the hvx, which only has +2 stops over 18% gray/45IRE. Virtually all of Robert Richardson's recent films have also employed this technique. It's tough because it forces you to compose well (make sure leading lines lead to faces instead of the over-bright floor/table, which must also be integrated into the composition) but those are both excellent compositions so it works even with the the blown-out highlights.

Just curious, how do you meter for this type of situation? My guess would be using an incident meter near the faces (one to one half stop under?) and then spot metering the table itself to make sure it's roughly within the range of what won't blow out (in this case, what won't blow out when using a grad filter, of course).

A question about source fours: beyond just focusing like stage lights and accepting cookies within the light, they can also redirect light so it creates more parallel beams, giving the impression that the source is very far away (like the sun) when it's actually close (within the room)? Is this correct or is something else at work to get those shadows? How does the inverse square law relate to this? Do lekos give roughly parallel beams but also behave like point sources w/r/t inverse square or do they also have inherently more "punch" at long distances through some crazy reflector physics? Really want to try one of these out. Seems like a great solution for day exteriors in small rooms.

Have you ever used, for night exteriors, a bright hmi for the backlight and a 4X'4 mirror for fill? I've heard this highly recommended by the steadicam op for the new Star Wars films; he says you can switch direction 180º and still have a set-up that's useable simply by rotating/focusing the light and moving the mirror--no need to move the big hmi for reverse shots.

And just out of curiosity, have you ever shot color reversal (either for still photography or film)? I'm very fond of it and it behaves quite a lot like digital, imo. I really enjoyed the look of Buffalo 66, which was shot on old ektachrome or something.

Even your 8mm stuff looks fantastic, btw. Helps explain why I'm getting rejected from every film school I apply to. Maybe in a few years...

 

[Bill Simone]

sopranos

sopranos

Hi David

Thanks for everything you contribute.
I've always admired the lighting in general on the sopranos.
Could you comment on how the attached screenshots may have been approached? In particular the quality of light on the faces noting the edge light on the lit side of the face.

Thanks,

Bill

 

[Matt W.]

Hi David

Thanks for everything you contribute.
I've always admired the lighting in general on the sopranos.
Could you comment on how the attached screenshots may have been approached? In particular the quality of light on the faces noting the edge light on the lit side of the face.

Thanks,

Bill

The lead DP for the Sopranos is an alumnus of the same film studies program I attended at undergrad and he did a roundtable discussion with us re: the lighting for the show.

Before I was more-or-less kicked out of the discussion, he gave us some insight into his approach:

They shoot on 35mm, obviously. The DP does not like lighting for high-def.

They never use lights behind the camera, instead using ambient light and (possibly?) reflectors for fill.

They shoot on set for a tungsten white balance (presumably 5218) and almost all light sources (except maybe kickers, hair lights, etc? I forget) are soft. The main instruments they use are enormous banks of 60w or 100w incandescent light bulbs in a grid pattern. These are installed on dimmers at the corners of the sets and only the ones facing the camera are used for a given shot. This way the DP can maintain a soft, offside key for all his CUs and get a dark "silhouette" look for straight-on wide shots. The soft light wraps nicely so it doesn't look so harsh and the considerable spill serves as fill light.

There's preference paid to getting offside keys and nice looking lighting over naturalism. You'll often see that the direction of the key and position of its apparently source (lamp in frame, etc.) are at odds.

The lead DP does not use a meter, but he estimates that most of the time he's shooting with the key (which is sometimes effectively a backlight in wider shots) at 1/3 stop or slightly more over and he lets the shadows fall where they may, often a stop or more under. I think. I'm sure David has more practical/useful info than this for achieving that look in broader terms, but maybe this will provide some supplemental insight.

 

[Steve Das]

on blown out highlights

on blown out highlights

...Virtually all of Robert Richardson's recent films have also employed this technique. It's tough because it forces you to compose well (make sure leading lines lead to faces instead of the over-bright floor/table, which must also be integrated into the composition) but those are both excellent compositions so it works even with the the blown-out highlights.

Blown highlights can look great. It's just how you use them. They need
to be justified and fit with story tone and subject matter.
Robert Richardson took something people avoided and made an
art out of it. Difference is he really pushed it intentionally..even used
large frame areas. Also check out Malik Sayeed work on Clockers.

Love the 8mm frames on p.318...great use of diagonals in the [] frame.

 

[Martin Stevens]

Hello again David,

At this point what do you see as the main aesthetic difference(s) between
images shot using 35mm negative and images shot with the Red MX (given that
in post you try to make the MX look as much like film as possible)?

 

[Pete Wallington]

Sorry Pete:

Instead of REPLY I hit EDIT, which I can do as a moderator, and I wrote over your original post, which was about dealing with filter reflection. Can you repost it? The photo you showed was quite lovely.

Generally the solution is to tilt the filter, but this requires a rectangular filter and a mattebox that allows you to tilt the filter. This is one reason I like Panavision's basic 4x5 mattebox, it has a large slot for holding two stacked trays, so you can also use a single wider tray that holds the filter at an angle (called an ARF: Anti-Reflection Filter tray). And you can squeeze two filters into the ARF tray. All the other tiltable matteboxes like ARRI's are too large and bulky.

M. David Mullen, ASC

Thanks David, here is the picture again, it's just about the end frame of a gentle track in, so you can clearly see the reflection moving around as the camera moves.

I suspected that tilting the filters might be a viable option. Getting hold of a specialist Panavision mattebox for every shoot (particularly those with only the budget for 5D) will be difficult though. I've spoken to my regular rental houses here in London and none of them have an ARF mattebox, I guess I'd have to go to Panavision direct, and I rarely have the budget for that.

How much and in what orientation are the filters tilted to get rid of this reflection effect? Ie are they tilted top to bottom or left to right? Does it matter? How much? 45 degrees like a teleprompter? Maybe I can rig up something to emulate this myself.

Also why does this "require a rectangular filter"? Do you just mean not circular filters, or do you mean specifically PV filters rather than square filters?

 

[David Mullen ASC]

Just curious, how do you meter for this type of situation? My guess would be using an incident meter near the faces (one to one half stop under?) and then spot metering the table itself to make sure it's roughly within the range of what won't blow out (in this case, what won't blow out when using a grad filter, of course).

More or less -- when I go for a nuclear hot spot, I tend to meter the area that the bounce is exposing and then decide how much under it should look, but I don't meter the hot spot, I just assume it's going to blow out. Now when shooting digital, I can see how much clipping there is and decide whether to favor more exposure towards the highlights, or use an ND grad to darken the hot area a bit. It's more of a problem with lower-end Rec.709 cameras than something shooting Log or RAW.

A question about source fours: beyond just focusing like stage lights and accepting cookies within the light, they can also redirect light so it creates more parallel beams, giving the impression that the source is very far away (like the sun) when it's actually close (within the room)? Is this correct or is something else at work to get those shadows? How does the inverse square law relate to this? Do lekos give roughly parallel beams but also behave like point sources w/r/t inverse square or do they also have inherently more "punch" at long distances through some crazy reflector physics? Really want to try one of these out. Seems like a great solution for day exteriors in small rooms.

They fall off over distance like any other light but the throw can be quite tight with a narrow lens without much spill. The beams are close to being parallel.

Have you ever used, for night exteriors, a bright hmi for the backlight and a 4X'4 mirror for fill? I've heard this highly recommended by the steadicam op for the new Star Wars films; he says you can switch direction 180º and still have a set-up that's useable simply by rotating/focusing the light and moving the mirror--no need to move the big hmi for reverse shots.

Haven't tried that. I can see how it would work but I also see that it's not a cure-all.

And just out of curiosity, have you ever shot color reversal (either for still photography or film)? I'm very fond of it and it behaves quite a lot like digital, imo. I really enjoyed the look of Buffalo 66, which was shot on old ektachrome or something.

Just a little, it's fun.

 

[David Mullen ASC]

How much and in what orientation are the filters tilted to get rid of this reflection effect? Ie are they tilted top to bottom or left to right? Does it matter? How much? 45 degrees like a teleprompter? Maybe I can rig up something to emulate this myself.

Also why does this "require a rectangular filter"? Do you just mean not circular filters, or do you mean specifically PV filters rather than square filters?

You basically have to try tilting them up or down on a shot by shot basis to see what works better. The tilt in the ARF is not extreme, sometimes I've had to take the filter out of the ARF tray and tilt it further in the slot and hold it in with some tape. Yes, I meant that round filters don't use a tray so they are hard to tilt.

 

[David Mullen ASC]

Hello again David,

At this point what do you see as the main aesthetic difference(s) between
images shot using 35mm negative and images shot with the Red MX (given that
in post you try to make the MX look as much like film as possible)?

The two main things I see are that the highlights still roll off to overexposure more gracefully with color negative film, and that film has grain. There are other minor things, some color reproduction is different, underexposed colors act different, the electronic shutter feels different than a mechanical one in subtle ways even at 180 degrees, etc.

My belief is that when you shoot any technology badly, it shows its inherent structure. So well-shot digital can look like film, but when things go wrong, digital goes wrong in uniquely digital ways.

 

[David Mullen ASC]

Hi David

Thanks for everything you contribute.
I've always admired the lighting in general on the sopranos.
Could you comment on how the attached screenshots may have been approached? In particular the quality of light on the faces noting the edge light on the lit side of the face.

Thanks,

Bill

You'll notice that there is a soft 3/4 back-edge from a higher angle, and then an eye-level soft side light coming from the same direction as the back edge (rather than the opposite side of the face, which is more traditional), so it all sort of blends and wraps around the face, but there is still a bit of a hotter edge and some soft edges on the shoulders, rounding them out and separating them from the background.

Try this trick: light someone from the side with two 4'x4' frames of diffusion side by side so it blends into an 8'x4' diffusion. Then make the first 4'x4' frame that is slightly behind the person hotter than the second 4'x4' frame that is coming around the front of the face more. The first 4'x4' frame that is hotter will create a little kick or glow on that lit side of the face but also blend into the soft key coming around the front.

 

[Jase]

Hi David..
I'm wondering what you rate the new Mysterium X Camera at. I know lots of people say they rate it at 800 outdoors to protect highlights... which to me sounds a bit funny. I've shot more digital than film, but have always approached lighting and exposure in a film like manner. So why rate it so high outdoors when you just have to add more glass in front of the lens to bring it back down?

Also, wondering what you would generally rate the MX at in outdoor and indoor situations.

I know the Mysterium sensor was rated at 320... however that seemed a bit generous. I think it was closer to 200. I think RED is still saying that the native ISO for the MX is 320.... have you found that to be the case?

Thanks in advance.

 

[Bill Simone]

Sopranos

Sopranos

Hi David

Thanks for your reply. Sounds right on target to me, including your tip.
What I'm curious about is I see the same quality of light on tight heads (as in the examples I attached), and also in much wider shots including shots with people moving through the frame.
Do you think the approach is simply scaled up and moved back?

Thanks,

Bill

 

[David Mullen ASC]

If you rate the MX at 400 ASA but stop down and/or use ND filters to darken the image by another stop to hold bright highlights, it's the same thing as rating it at 800 ASA, you end up at the same stop/exposure. The only difference is that by rating the camera at 800 ASA, the overall image will look brighter on the monitor so you don't have to look at a darkened image every time you are underexposing to hold another stop of highlights.

800 ASA is the sweet spot for the MX sensor, though if you look at RawView, you have to rate it at 400 ASA to get the brightness of the Raw image to match the brightness of a RedColor image. But considering the low noise of the sensor, you can rate the camera at 800 ASA without any negative consequences in terms of noise while gaining more highlight information. However, 800 ASA is a bit fast outdoors which is why I sometimes choose 500 ASA instead, though I get a bit more clippiness in the highlights that way.

 

[David Mullen ASC]

Hi David

Thanks for your reply. Sounds right on target to me, including your tip.
What I'm curious about is I see the same quality of light on tight heads (as in the examples I attached), and also in much wider shots including shots with people moving through the frame.
Do you think the approach is simply scaled up and moved back?

Probably. You usually light and shoot the wide shots, then bring in some softer lights closer for the close-ups.

 

[Matt W.]

Thanks as always!

Fwiw, on Sopranos they have huge (many feet by many feet) grids of bare incandescent bulbs installed permanently into the major sets and manipulated via dimmers for easy soft offside keys/backlights. For close ups and other locations, of course other lights are used (such as the set-up you described).

If the wide shots feel unnaturally soft, this may be why.

 

[Jase]

If you rate the MX at 400 ASA but stop down and/or use ND filters to darken the image by another stop to hold bright highlights, it's the same thing as rating it at 800 ASA, you end up at the same stop/exposure. The only difference is that by rating the camera at 800 ASA, the overall image will look brighter on the monitor so you don't have to look at a darkened image every time you are underexposing to hold another stop of highlights.

800 ASA is the sweet spot for the MX sensor, though if you look at RawView, you have to rate it at 400 ASA to get the brightness of the Raw image to match the brightness of a RedColor image. But considering the low noise of the sensor, you can rate the camera at 800 ASA without any negative consequences in terms of noise while gaining more highlight information. However, 800 ASA is a bit fast outdoors which is why I sometimes choose 500 ASA instead, though I get a bit more clippiness in the highlights that way.

Hi David... thanks for your reply. Sorry to ask this, but basically the reason why you shoot at 800ASA is so that you're technically underexposing the image to keep the highlights, then in post, you bring it all back and make it work with your colorist?

It seems backwards to me. because when shooting film, you want a slower ASA film like 250 so you don't have to ND down so much, etc. I know that if you over expose in film its fine as it looks great. But even in that case, I would normally try to keep the sky within range. Would you?

Thanks for you insight.

 

[David Mullen ASC]

Film has a lot more overexposure latitude than digital which is why you can get away with overexposing and printing down.

The primary downside to underexposing digital is noise, so it all depends on how noisy the image gets. At 800 ASA, the noise is low with the MX sensor. And since overexposure latitude is one of digital's weaknesses, then being able to rate it faster without much of an increase in noise is a big advantage.

You don't really "bring it up in post" manually by shooting the MX Red at 800 ASA, not if you process it at the 800 ASA metadata, it just comes up at the correct brightness. If you selected the output in post to be 400 ASA though you shot it at 800 ASA, then it would come out darker and you'd have to manually bring it up.

I think you have to get around the notion that the darker exposure at 800 ASA is a "wrong" exposure and rather think of it as an optimal exposure for going into post with a decent amount of highlight information and a minimal amount of noise.

On the other hand, it's also about the tonal range of the image, if all your subject matter was only dark-toned and you didn't have any hot areas to worry about, then you should give the image more exposure, whether by using a lower ASA value or just exposing more for the shadows.

Using a slower ASA film is a different issue altogether than how you expose for highlights or shadows. Yes, that's the only thing that is annoying about the new faster sensors, they need so much ND outdoors. You don't have the option of switching to a low-sensitivity sensor when outdoors. And what's tricky about digital is that in bright sunlight, the higher ASA ratings tend to be better because they hold bright detail better, which is why some people actually use lower ASA settings indoors and higher ones outside in sunlight.

 

[Martin Stevens]

The two main things I see are that the highlights still roll off to overexposure more gracefully with color negative film, and that film has grain. There are other minor things, some color reproduction is different, underexposed colors act different, the electronic shutter feels different than a mechanical one in subtle ways even at 180 degrees, etc.

My belief is that when you shoot any technology badly, it shows its inherent structure. So well-shot digital can look like film, but when things go wrong, digital goes wrong in uniquely digital ways.

THANK YOU David!!!

 

[Roberto Lequeux]

THANK YOU David!!!

+1! And I got a lot of backtracking to do. I see a lot of new posts to read.

I don't think I see the shutter difference, or at least can't put my finger on what is different about it, although it does already feel a bit different to me as I get more discerning. Maybe the acceleration of the shutter will need to be replicated in the electronic swipe. Or recreate the softness of the edge. Both...

I would believe that with Graeme still plugging away feverishly, RED's color science will get to a point where reproduction is hard to tell apart even for us who look for it. Perhaps not for a while longer if you have eyes as experienced and adept as David's, but soon enough for you to Mr. Mullen. Hopefully soon enough dynamic range will also be matched, hopefully Mostro does the trick.

One thing's for sure, all us will get our turn to say film is de...

...finitely replicated beyond reasonable concern.