The Sophomore

[David Mullen ASC]

We worked a sixth week plus two days of a seventh week.

 

Week Six was mostly spent at a small upscale house in Clifton, NJ. We picked it because it was something of a time-warp, sort of that late 1960's modernism you see in the Brady Bunch house. It was quite beautiful actually, with nice (not cheap) wood-paneled walls alternated with pale green walls, a bathroom in sort of green & gold tile, fluted glass panels around the doors, losts of brass everywhere, etc.

 

Unfortunately at the beginning of the second day, one of the crew members cracked the fluted glass panels on each side of the front door on the day we started shooting all of the scenes at the front door. So there was some panic trying to decide how to deal with the problem, since we couldn't hide the crack (the glass window was next to an important side door leading to our other set area, so a tall plant wouldn't fit, we couldn't get someone out there to replace it, etc.) We ended up using a mix of methods to distract your eye from that window but it was still annoying to deal with, not ideal.

 

The interiors were lit with tungsten lamps, either a Soure-4 bounced off of the ceiling or Chimeras, or the actual track lighting of the house, which gave me enough exposure with the practicals. The house had a lot of big windows overlooking the backyard below (it was a split-level house so the first floor living room was a floor above the backyard) and outside there were a lot of bamboo trees. I used an HMI lighting balloon in the backyard (since it was sloping downhill into a bunch of trees, there was no placement for a condor or scaffolding) which put a pretty soft blue light on the bamboo and treetops. I kept a little brighter than was realistic, to give the impression that even at night, it was darker inside than outside, so people would be framed against the greenery outside the windows.

 

The front of the house was lit from a condor with a Maxibrute with 1/2 Blue, so I was mixing somewhat full blue from the HMI lighting balloon and 1/2 Blue from the Condor, but on green grass and trees, it all somewhat blended anyway, colorwise.

 

After four days at the house, shooting nights (5PM to 6AM everyday), we moved to an old mansion in Newark, NJ that was playing for a frat house. Again, 1/2 Blue from the Condor for the front yard (half blocked by trees though, so a lot of shadows on the house), tungsten coming from the inside of the mansion. The interior mansion was all wood-panelled with wall sconces everywhere that gave me a T/2 just in available light, so I just augmented the closer shots with a tungsten Chimera. We had a dark office night scene in one room, in which there were five LCD video screens in the center. I kept the room dark, just two small spots on the walls from Dedolights to keep the background from going pitch black, then a side light from a Kinoflo on the actor. We worked 15 hours that day, which meant blacking out the house at sunrise. After a week of night shooting, we were all exhausted.

 

I pushed the film one-stop for the night exteriors of the frat house but otherwise shot the 500T normally inside.

 

The final week had two shooting days scheduled, the first in an office space playing for the school's newspaper office, then we had to skip a day of work in order to shift from an 11AM to Midnight schedule (which was the earliest we could begin the week after ending the previous week of night shooting) because we needed a full day of sunlight for the final day -- so we shot on Tuesday, skipped Wednesday, and had an dawn calltime on Thursday. Thursday was a light day of exterior MOS shooting with a skeleton crew, no sound department, etc. We ended the day with a nice dusk shot on the main street of Bayonne, as the main character rides his bike. I got a set of Zeiss Master Primes for this shot since we had no lights other than a single 575w HMI PAR plugged into a portable generator. Even so, when twilight began, I had an T/5.6-8 split at 400 ASA, which I underexposed by a stop and a half for a dusk look -- so we began the first take at T/11. But by the last take, I was at a T/1.3. So we had something like five takes done over a half-hour period at different degrees of dusk light, the last being nearly night.

 

On Wednesday, which I was planning on using to pack-up for my flight early today (Friday), I got a call that the first scene shot on Tueday in the office had flicker problems. I knew immediately that the problem was the old overhead flourescents in the office because I spent time swapping out tubes trying to get them to stop flickering. Luckily after the first scene, I decided I preferred the look of the room only being lit by the windows and photolight box tables in the newsroom, so I turned off the overheads. But still, it was a big first scene with flickering light. I looked at the footage at Postworks in an HD suite and the problem was subtle, luckily. The wall behind one character in one direction was flickering, and in the reverse angle, the face of one person under the overheads was flickering -- and since she was lit by window light and fill from the lightboxes (which had Kinos in them) the overhead flickering was just on the top of her head and tip of her nose.

 

Still, it wasn't acceptable but we discovered that simply by using some noise reduction, which averages grain over six frames or so, 80% of the flickering disappeared. We determined that later in post, we could get rid of most of the final flickering in the small parts of the frame where it still persisted using more advanced tools. So in the end, a reshoot was not required. Again, I am thankful that I decided not to use the overheads after the first scene, because it was a huge seven page day.

[Travis Cline]

After you changed the bulbs out and decided to shoot the scene with the overhead flousescents on, was the flicker noticeable to the eye? If not, how would one go about avoiding this problem? I ask because I finished shooting a feature a few weeks ago and we had flicker show up on the film even though we had a flicker meter telling us the lights were in phase. Any advice?

 

 

Travis

[David Mullen ASC]

The gaffer said he metered them and they were OK, not great. We could see some very subtle flicker to the eye, we could see some flicker on the videotap image, but when you're talking about fixtures built into the room, in the shot, you don't have much choice sometimes but to take a chance. You stare too long any any fluorescent and you are going to see flickering.

 

I think in the future, though, anytime I shoot in an old building, I will insist, if feasible, for an advance electric crew to go in and replace the ballasts in the overhead fixtures, assuming there aren't too many of them. But all that costs money and planning. I'll also just assume that if I can see the flicker, if the tap can see it, there's going to be flicker.

 

This was in an old Catholic school in Jersey City; at the old high school in Bayonne, we also had flicker problems from some of the industrial lights on the building at night. Maybe because I'm from LA, I'm not as used to filming in places with such old wiring and fixtures still in use...

[Evan Winter]

Hi David,

 

Quick question - I always thought that the rule of thumb with flicker was that if you see it through the eyepiece or on a tap then it may not be on film; since the 'frames' we see in the eyepiece/tap are not being recorded on film.

 

Again, this very loose rule of thumb is only helpful as a guide (depending on shutter angle, film speed, etc) when one has no other way of determining if there is flicker... but, a trick I've heard but never used suggests that when shooting TVs without phase/shutter control devices that one should shift the camera's shutter/film-speed until the roll bars on the TV screen are the most evident in the eyepiece/tap.

 

This, in theory, will minimize or even remove the roll bars from the footage that is actually being shot.

 

Any thoughts on this? Urban legend? :)

Edited September 22, 2007 by Evan Winter

[Kieran Scannell]

David,

 

I'd just like to thank you for you're posts on this feature, it's such an invaluable insight into the cinematographers magic box, and quite special that you take the time to share. I for one have been waiting for every post, as I'm sure have many others.

 

Just one question, you mentioned pushing one stop for the night exteriors is that for more detail on wider shots? And is it something you always do, or does it depend on what your seeing at the time?

 

Kieran.

Edited September 22, 2007 by Kieran Scannell

[David Mullen ASC]

You don't get more detail when you push -- you're just increasing the density of the negative in processing to compensate for a lack of density from underexposing (in this case, because I was working at very low light levels.) So whatever got exposed onto the negative gets more density, but you can't create information that was never recorded in the first place, hence why pushing also increases contrast (your highlights get more dense but whatever was black stays black.) You also get more grain because your image exists mostly on the larger grains that got exposed, whereas the smallest grains never got enough exposure.

 

Now it's a good question if you're doing a D.I. whether there is any advantage to pushing versus just brightening the shot in post, since as I said, there should be no increase in information from pushing, just density. I don't know but I suppose it's the difference between grain versus noise in terms of which method you use -- increasing the brightness from pushing gets you more grain, increasing the brightness from boosting the image electronically gets you more noise.

[Kieran Scannell]

So you would always go for film grain rather than post noise?

[David Mullen ASC]

So you would always go for film grain rather than post noise?

 

Ideally, I'd go for neither, but at least film grain in a movie shot on film is more accepted by viewers who notice such things -- we've had over a hundred years of grainy movies but (for the most part) less than a decade of electronically noisy ones in the cinema... At least, that's my theory and I'm sticking to it for this week at least.

[James Steven Beverly]

Dave, Let me ask you a question, This whole push thing you've been discussing here made me begin to think (which is ALWAYS a dangerous thing). I'm. planning on using Lomo anamorphics on my next project as you know, and as you also know, I'm planning a lot of night work in the desert, Here's the question, what about lighting brightly, (over exposed for what should be night) to compensate for the F stop needed to shoot anamorphic properly (3.3 or better) then using a pull process to make the image look more like it was shot at the proper exposer HAD I been using sphericals at night, rather than shooting slightly underexposed due to night conditions and then forcing the film by a stop or 2. Are there inherent problems with a pull process? It seem to me that LOGICALLY you would end up with less grain by over exposing and pulling the film rather than under exposing and pushing the film but sense push ALWAYS seems to be the choice rather than pull, there MUST be a reason to go with underexposed/push rather than overexposed/pull. Can you elaborate on this? :huh:

[David Mullen ASC]

There isn't much reason to overexpose and pull-process these days unless you simply want a pastel, lower-contrast look - modern stocks like Fuji Eterna 500T or Kodak Expression 500T are already lower in contrast if you want that look.

 

As for pulling to reduce grain, if you have enough light to overexpose and pull one stop, you probably have enough light to switch from a 500 ASA stock to a 250 ASA stock and accomplish the same thing with less hassle, less cost.

 

Plus, the anamorphic format is already less grainy than the spherical one, and modern 500T stocks are fine-grained rated normally, more so in 35mm anamorphic, so what's the point of trying to reduce the grain even more?

 

Besides, few people -- including me -- ever have enough light to shoot at T/3.3 outdoors at night PLUS overexpose and pull-process one stop on top of that.

 

If you have the light for it, just rate a 500T stock at night normally at 500 ASA or 400 ASA, process normally, because with anamorphic, you'll be fine, grain-wise.

 

The main reason people underexpose and push for night work is that they have no choice -- they don't have enough light to be able to shoot at 500 ASA and light a large area. Or they are balancing with existing low-level practical illumination and don't want to overpower it. For example, in "Solstice", I shot the night stuff in the swamps and woods on Eterna 500T rated at 320 ASA - I had enough light for that because I was underexposing for a moonlit look by two stops, so if I lit the woods to a T/1.4 at 320 ASA, that meant I could shoot at T/2.8 for a dim moonlit look. But I had one scene at a lake with string lights over the dock and I wanted those to expose brightly, so I lit the scene for a one-stop push at 640 ASA so that I could use less light and let the string lights play fairly brightly. Same goes for the carnival scenes in "The Astronaut Farmer" -- I could have lit them to a higher level so that I didn't need to push the stock, but I wanted most of the exposure to come from the existing rides and carnival booths and add as little extra light as I could so as to not overpower them.

 

So it just depends on the place you are lighting and the look you want. Sometimes you accept more graininess from a push in order to capture more of the natural light sources in the location; sometimes that doesn't matter because you are providing the only light in the scene.

 

My advice is simple: test, test, test. Learn how your stock looks, printed and projected, when you process normally, and when you underexpose and push, or overexpose and pull, and decide for yourself.

[James Steven Beverly]

Good advice of coarse and thank you for the reply. One other question, given the nature of modern stocks, 5218 in particular, when does the push process become a problem on 35mm? I mean how far can you push it (forgive the pun but I couldn't resist :D )?

[David Mullen ASC]

That's a matter of personal taste, just how much underexposure and grain can you tolerate. Generally, most people only push by one stop, and a two-stop push is sort of the most anyone pushes.

 

Gordon Willis once said that there was no such thing as a three-stop push, in that, you've recorded so little information on the negative to be worth pushing (remember, pushing does not add information that was never recorded, it only increases the density, brightness, of what DID get recorded.) Although some of "The Deer Hunter" was pushed three-stops by Vilmos Zsigmond. But that doesn't necessarily mean that he underexposed by three stops.

 

It's an important distinction -- how much you underexpose versus how much you push-process. Because pushing is not an exact science, and because it increases graininess, most people underexpose less than they push. For example, I usually rate at 500 ASA stock at 640 ASA when I push it by one stop, so I have a denser-than-normal negative. Sometimes I rate it at 800 ASA.

 

One of the more interesting looks is what Dante Spinotti did for "Red Dragon", shot in anamorphic -- he push-processed 500T by two stops but rated it at 800 ASA. So the final negative was 1 1/3 of a stop denser than normal, which he then printed down. This gave him a somewhat contrasty look.

 

"Eyes Wide Shut" was shot in standard spherical on the older EXR 500T pushed two stops and rated at 1600 ASA instead of 2000 ASA, so there was a slight leeway in case the pushing didn't add enough density.

[timHealy]

Hi David,

 

Quick question - I always thought that the rule of thumb with flicker was that if you see it through the eyepiece or on a tap then it may not be on film; since the 'frames' we see in the eyepiece/tap are not being recorded on film.

 

That typically works with things like gunshot flashes say using a strobe. Fluorescent flicker is a different animal.

 

Best

 

Tim

[timHealy]

I think in the future, though, anytime I shoot in an old building, I will insist, if feasible, for an advance electric crew to go in and replace the ballasts in the overhead fixtures, assuming there aren't too many of them. But all that costs money and planning. I'll also just assume that if I can see the flicker, if the tap can see it, there's going to be flicker.

 

Replacing the ballasts in a old building is the only way to be sure and even then you still may get caught with one. I have a friend who replaced practically every fluorescent ballast in the Eero Saarinen TWA terminal ( http://www.greatbuildings.com/buildings/TW..._New_York.html) at Kennedy Airport including the long corridor for "Catch Me If You Can". But they did have a little bit more money than the average low budget film.

 

http://www.greatbuildings.com/cgi-bin/gbi...._mce_121_20.gbi

 

Best

 

Tim

Edited September 23, 2007 by timHealy

[Michael Nash]

Hi David,

 

Quick question - I always thought that the rule of thumb with flicker was that if you see it through the eyepiece or on a tap then it may not be on film; since the 'frames' we see in the eyepiece/tap are not being recorded on film.

 

Again, this very loose rule of thumb is only helpful as a guide (depending on shutter angle, film speed, etc) when one has no other way of determining if there is flicker... but, a trick I've heard but never used suggests that when shooting TVs without phase/shutter control devices that one should shift the camera's shutter/film-speed until the roll bars on the TV screen are the most evident in the eyepiece/tap.

 

This, in theory, will minimize or even remove the roll bars from the footage that is actually being shot.

 

Any thoughts on this? Urban legend? :)

 

That rule of thumb is for individual flashes, and even then it's not completely reliable. Flashes visible in the viewfinder might also be partly exposed on the film if the flash duration overlaps the moment when the shutter opens or closes.

 

As for TV shooting, you're talking about two different variables; synch and phase. Synch is getting two items to share the same frequency, in this case getting the camera to run at 29.97fps (eliminating flicker). Phase is aligning the start point of the cycle between two items that already share the same frequency, in this case getting the roll bar to appear in the mirror viewfinder and therefore not on the film.

 

In any case, you can't avoid flicker from a faulty fluorescent ballast this way.

[David Mullen ASC]

Here are some crew photos taken by Walter Thomson, our unit photographer. I was lucky that someone was taking photos of me at work because I end up with so few sometimes.

 

Here I am in the gym, standing on an apple box to get a meter reading of a basketball net:

 

 

My 1st AC, Frank Rinato, in the library with some stand-ins:

 

 

Key Grip Anthony Gamiello (left) and Gaffer Kevin Janicelli (right) at the swimming pool location, lit with hanging flos and one HMI Source-4 bounced off of the ceiling:

 

 

The camera crew in the streets of Bayonne, NJ -- you'll recognize the place from "War of the Worlds". L>R: Me, Leigh (2nd AC), Jeff (Operator), Frank (1st AC), John (Dolly Grip), Brian (Loader):

 

 

1st AD Tom Fatone standing on a spot where the actor will be later, in a frat house scene shot in Newark at an African-American Association of some sort in Newark. It's so similar to the final set-up that I cropped it to scope. He's lit with a 5K bounced off of an 6'x6' UltraBounce, background lit with the practicals mostly, maybe another bounce as well:

 

 

In the gym during the homecoming dance, with a soft frontal key from a Chinese Lantern (with a daylight photoflood in it) mounted to the dolly for a 360 track, with the actors (stand-ins here) on an opposite platform mounted to the dolly so we all go around together:

 

 

In the principal's office set, my Key Grip Anthony Gamiello (on the left) hanging a 4'x8' beadboard for an HMI Source-4 ("joe-leko") bounce (the light not yet centered and doored off to match the card):

 

[Justin Hayward]

Looks good.

 

On the last one, was that because you couldn't get lights out the window and you had to keep the source high enough that it was out of frame or because you wanted a softer wrap on the top of everyone?

 

Thanks

[David Mullen ASC]

Being a second floor room, it would have required a pretty big condor to get the lights higher than the windows, so bouncing from inside was cheaper and faster, plus it would normally blend with the real ambience coming through the windows.

 

But actually in this case, I lost the daylight... and had to paper the windows and white them out with an HMI on a condor at window level, so the high bounce acted as a soft backlight while the windows were just whited-out (you can see in the photos that the last two windows on the right side already have the paper on them).

[Satsuki Murashige]

Great stills. Love the camera crew pic -- cargo shorts and New Balance sneakers are the standard camera crew uniform, I guess. ;)

 

So you're using a Spectra meter now, David? How do you like it compared to the Minolta? And why such a big light (5K bounce) for the medium shot in the frat house still?

 

Congratulations on wrapping the feature, BTW.

[John Holland]

David would like to know how you got on with the Arricam ? your first time i think ? John.

[David Mullen ASC]

I used a 5K bounced because that is the reverse angle -- the first set-up looked the other way into a big empty living room, so I lit the big space with a 5K bounce. So when I turned around to shoot the reverse close-up at the archway, the bounce still worked OK on the face so I just went with it. I could have used something smaller and closer of course.

 

My Spectra works fine, my Minolta worked fine, I don't see much of a difference except that the Spectra allows me to input frame rate.

 

The Arricams are great cameras. I did find that for some reason, when it was hot & humid, the Fuji stock tended to rub against the Arricam mags; we spent the first week slapping a lot of mags or swapping them out. Don't know why mags can't be built another millimeter wider internally, why it has to be so snug against the roll...

 

Also, I did find the color tap of the LT to be dimmer than the one on the ST. And changing the LT from studio to Steadicam mode was longer than it takes on a Millenium XL, plus it caused the video tap's focus to go out everytime we switched between the two modes, because of how you have to take apart the viewing system. But otherwise, the LT is a great handheld camera.

 

Loved the Zeiss Ultra and Master Primes. The ability to use Zeiss T/1.3 Master Primes may be the ultimate deciding reason between Panavlsion and Arricam, unless Panavision adapts them for Panaflex use.

[John Holland]

I have only had experience with the Arricam ST twice on shoots loved it. Interesting about LT .

[Max Jacoby]

David

 

Did you notice much difference between the Master Primes and Ultra Primes?

[David Mullen ASC]

David

 

Did you notice much difference between the Master Primes and Ultra Primes?

 

Mechanically, they seem and feel different in terms of the markings and gearing (more marks on the Master Primes).

 

Optically, they were similar though I did feel that the Master Primes were a bit "clearer", almost like looking through one sheet of glass, or not looking through anything at all. It's deceptive because when you look at them opened-up to T/1.3, it just feels like you're looking with your own eyes, not through a lens, yet of course, the depth of field is ridiculously shallow. Yet the image doesn't feel "swimmy" the way it does on an old Super-Speed wide-open, just "clear". I suspect that the 100mm and 150mm Master Primes shot at T/1.3 at night will become known as "AC killers"...

[Max Jacoby]

Optically, they were similar though I did feel that the Master Primes were a bit "clearer", almost like looking through one sheet of glass, or not looking through anything at all.

That was my impression as well, they are incredibly neutral, it's like you are looking at the scenes with your own eyes.