Day For Night - Additional Shooting Information

Day For Night - Additional Shooting Info :

30th August 2004

I am shooting a feature in little while and considering shooting day for night. I saw on web page a discussion of the Tiffen day for night filters. But I know there must have been discussions in the past on this subject. Does anyone know how to access them?

The location is a wooded area. I was thinking of shooting on an overcast day and using the light that filters in thru the trees as my "moonlight" and avoiding any sky or hot areas in the background.

Best,

Rick Lopez, DP NYC
web : www.lopezfilm.com

In his youth, Carter had believed everything was possible. Then in grief, he believed everything was impossible. And now ... he felt that when you had lived enough of your life, there was no difference between the two.

-- Glen David Gold, Carter Beats the Devil

Rick Lopez wrote :

>...considering shooting day for night...The location is a wooded area. I >was thinking of shooting on an overcast day and using the light that >filters in thru the trees as my "moonlight" and avoiding any sky or hot >areas in the background."

Most DFN effects involve about two to two-and-a-half stops underexposure. Some people use filters to provide a tint, that is often, but not always, a lavender-blue, as it mimics twilight and appears psychologically to emulate the mood of moonlight. Remember, though, that moonlight is a relatively hard light source, so that it creates sharp shadows, something you won't get on an overcast day. Bright, direct sunlight is usually better. You are right about wanting to limit visibility of the sky to a minimum.

Ira Tiffen
LA, CA

>Most DFN effects involve about two to two-and-a-half stops >underexposure.

Hi, Ira.

We will also be doing a DFN in a couple of weeks from now and will be doing a small test tomorrow. Since we don't have the budget to experiment all possible scenarios, can you advice on the following :

1/. Not underexposing the negative to the recommended two stops. I was thinking of obtaining a good dense negative and having the lab print down to take care of underexposure. I feel this can be more controllable (in our case of low budget) and less risky than doing the effect in-camera. Is there a reason why one should underexpose in-camera than in the lab? I may miss something and your advice could aid the production team.

2/. Is it better to remove the 85 filter to create the blue tint early on or have the lab perform this? The AC Manual warns against removing the 85 as "it is often impossible to restore normal color balance to the film." Again, we don't have the necessary budget to properly experiment on all possible scenarios but it would be great to know if the labs these days have the means to restore color balance with or
without filter.

Raymond Rodney Ocampo
Editor/Cinematographer
San Francisco, CA
www.oneframeoff.com

>obtaining a good dense negative and having the lab print down

Beware. By " a good dense negative" you probably mean something a little on the over-exposed side, which would probably print in the low to mid 30s. Printing - say - 2 stops darker to give you day for night would take the lights up to the range of high 40s. (It's 7 printer lights for each stop of camera exposure.)The top light is 50, so you are actually building in LESS control by straying towards that limit. Especially if you remove the 85, which will skew the balance of the negative and give the grader/timer no room to move.

Better to under-expose as recommended. (Why wouldn't you follow such a widespread practice?). The other thing this will do is give you reduced colour saturation, one of the most useful features of the "night" look.

If you remove the 85, you will have the blue look that you want. The lab would probably be able to restore neutral colours "more or less" if you lost faith and wanted day-for-day (after all it's common to leave the 85 off when there's not enough light to expose properly).

Dominic Case
Atlab Australia

Dominic, thanks for the info from a film lab's pov.

I guess it wouldn't make sense to create a fat neg and then have the lab print down by actually having the printer lights turn up (since they are working on a neg) to create the underexposure. This is good info to take with us for tomorrow's test shoot.

Raymond Rodney Ocampo
Editor/Cinematographer
San Francisco, CA
www.oneframeoff.com

One can also type Day for Night into camerafilters.co.uk and find pages on Harrison, Schneider & Tiffen Day for Nights with example images (except Mono Day For Night currently)

By the way, the Schneider DFN has no Magenta Basis, the theory behind the colour saturation is that non polluted night time vision of moon light thought cool, also has a tendency to accentuate part of the green frequency!

Regards

Carey Duffy

South London Filter Ltd
Tel.+44 (0)20 7735 1900
Fax: +44 (020) 7820 1718
Office Hours:
Monday to Friday 09.00am to 18.30 GMT
http://www.camerafilters.co.uk

Richard Lopez wrote :

>I saw on web page a discussion of the Tiffen day for night filters. But I >know there must have been discussions in the past on this subject. >Does anyone know how to access them?

Hi Rick,

I recommend that you take a trip to see Stan Wallace at The Filter Gallery, he has a nice assortment of DFN filters and can help you find one you like. He's in the building adjacent to Kodak across from the post office @ 350 West 31st in Manhattan. His number is 212.631.9177

Good Luck,

Anders Uhl
Cinematographer
ICG, New York

Raymond Rodney Ocampo wrote :

>1/. Not underexposing the negative to the recommended two stops. I >was thinking of obtaining a good dense negative and having the lab >print down to take care of underexposure...

Follow Dominic Case's recommendation. We have tried printing down a day for day shot that we later decided to use for night. It was evenly backlit, but it doesn't look convincing -- too much color saturation remained along with too much shadow detail.

Also, don't forget that the most convincing DFN is shot backlit or 3/4 backlit, so that the camera is looking at mostly shadows. High overhead sun or front light will tend to look like just an exposure error. Of course, this means that you are looking towards the hottest part of the sky, so you want to avoid sky. Polarisers won't work on this area of the sky, but grads can help if you just can't avoid some sky. You'll need something on the order of ND9 and bring the blend down across the horizon. This works only if your foreground action is below the horizon.

I have even cut a gel ND filter to conform to a mountain shape in the background and positioned it in the matte box to fit the jagged horizon line.

Wade K. Ramsey, DP
Dept. of Cinema & Video Production
Bob Jones University
Greenville, SC 29614

As for underexposing, I'd consider underexposing by one stop, not two. One stop is not risky for DFN and gives you some leeway in printing. Not underexposing at all might work OK -- but you could end up printing in the 40's or losing some bright detail in the frame if you have spotty sunlight in the woods.

As for not using the 85B filter, I've shot entire features on tungsten stock without the 85B filter and timed it back to normal. "Barry Lyndon" and "Heat" were shot that way too. It works fine; you may pick up a little color to the shadows and blacks, that's all. If you are worried, try using the Tiffen LLD filter -- it has no effective light loss, it cuts down on excess UV, and it partially removes some excess blue so your negative is a little less imbalanced than if no filter is used.

The thing to remember about DFN, especially in a shadowy location like the woods, is that once you darken the image enough to look like night, you have to consider how to direct the eye to what's important in the shadows. This may mean adding more fill light to certain areas to bring out detail. In a print, objects that are three to four stops under will tend to go black, so if your highlights are already one stop under, that doesn't leave many stops that your shadows can be before they go black (although of course, you probably want some black shadows). So you may want to use lights or reflectors to selectively punch up some dark areas so that there is still detail once you darken the image. Of course, if you shoot in overcast, this may be less of an issue since you already have a low-contrast image.

David Mullen, ASC
Los Angeles

>night time vision of moon light thought cool, also has a tendency to >accentuate part of the green frequency!

This goes hand in hand with the apparently desaturated look of night vision. At low light levels in the eye the non-colour rods take over from the r,g,b cones in the eye, leading to an increasing proportion of b/w vision as the light falls. Moreover, the peak sensitivity of the rods is in the green part of the spectrum, so greens appear as a brighter shade of grey than corresponding reds and blues.

More generally there is an overall shift in sensitivity towards the blue end of the (cones overall peak in yellow, rods in green) so all colours seem to shift slightly towards blue. (Look up the Purkinje effect.) This is a matter of vision rather than actual colour temperature.

In actual fact, moonlight tends to be a lower colour temp than daylight (it's usually quoted at 4000K) in other words it's redder, but because we tend to see night exteriors in comparison with even warmer artificial light (in house windows etc) it appears cooler. This is probably the reason we have adopted the conventional representation of moonlight as blue.

So I've just argued that moonlight is actually redder, visually greener, and psychologically bluer than daylight.

Next week, black is white...

Dominic Case
Atlab Australia

Dominic Case wrote:

>night time vision of moon light thought cool, also has a tendency to >accentuate part of the green frequency!

Just watched "Born Free" - sorry my 9 year old likes animal films in all shapes. Has some terrible day for night in it. I recommend seeing it for an example of what not to do.

Kind of looks like "pull the 85, and underexpose a stop or so" but its totally unconvincing.

If I remember correctly Lawrence of Arabia has some DFN which perhaps asks us to suspend a little too much disbelief as well.

In defence we have to remember the what they had to use for stock back in the mid 60's as well.

Mark Smith

Mark Smith wrote :

>Dominic Case wrote :
>night time vision of moon light thought cool

Not me, I was quoting Carey Duffy.

But following Mark's point about dodgy DFN, I recently worked on tinting and toning a restored Australian silent classic (The Sentimental Bloke, mentioned here not long ago).

Many of the evening or night sequences were clearly shot by day, judging by the strong sun shadows : but it's wonderful what a strong blue tint did to help the "suspension of disbelief". Colour is just another sign system or language that we use.

Dominic Case
Atlab Australia

Dominic Case wrote :

> Next week, black is white...

Well, it is if you're a negative sort...

Jeff "gray is always gray, though" Kreines

Mark Smith writes:

>Just watched "Born Free" - sorry my 9 year old likes animal films in all >shapes. Has some terrible day for night in it. I recommend seeing it for >an example of what not to do.

>If I remember correctly Lawrence of Arabia has some DFN which >perhaps asks us to suspend a little too much disbelief as well.

>In defence we have to remember the what they had to use for stock >back in the mid 60's as well.

In defence of the stock, it ain't the stock. Plenty of excellent DFN was shot long before faster films were produced.

Presumably you are watching a VHS or a TV broadcast therefore you are probably seeing what someone other than the DP -- and far removed from the picture making process -- has decided what the DFN should look like, not to mention the rest of the film. He or she may even feel they are doing the viewers a favor by making it easier to see in an over lit living room.

As Dominic and others have pointed out, DFN requires delicate handling at every step.

In any event, IMHO when it comes to depredations committed to films -- aside from colorizing -- pan and scan beats the hell out of correcting the N out of DFN.

Brian Heller
IA 600 DP

>If I remember correctly Lawrence of Arabia has some DFN which >perhaps asks us to suspend a little too much disbelief as well.

Having been out in the deep desert under a full moon, away from all artificial light sources, it DOES look a little like a bad day-for-night effect! It's surreal -- it sort of looks fake. Trouble with lighting the desert night-for-night is that there is no lighting unit powerful enough to light miles of desert with one source, so you end up with a spotty look where the "moonlight" just ends at some point in the background. So why is that more "realistic" than DFN? Probably because we know that the scene was shot at night, no matter how badly lit it is.

I think the DFN shots in "Lawrence" are pretty good, actually. They just get lightened too much for TV.

Modern effects allow you to take DFN work and make the sky realistically dark, even adding stars. This worked well for the early night scenes in "Cast Away" -- digital efx added a dark sky and a glow to his flashlight. But shooting during the day allowed the ocean to be "lit" for miles.

Recalling an old AC issue, the DP on "Jonathon Livingston Seagull" couldn't shoot birds and avoid the sky for its DFN scenes (and lighting the sky was not possible), so shot them on B&W stock with a red filter and then printed it on color stock with a blue tint. Never saw the results but it sounds interesting.

David Mullen, ASC
Los Angeles

>So I've just argued that moonlight is actually redder, visually greener, >and psychologically bluer than daylight.

>Next week, black is white...

Ha ha! Good point.

Rodney Ocampo
www.oneframeoff.com

> Next week, black is white...

Okay, you laugh, but I actually testified to this in court as an expert witness a while back.

Synthetic black fabric + overhead pot lights with lots of infrared + security camera with no infrared filter = black fabric appearing white on security camera video. Had to recreate this for the jury using both fabric samples and the actual clothing in evidence...

But the best visual aid was my TV remote, which looked like a flashlight on the security camera.

George Hupka
Director/DP
Downstream Pictures
Saskatoon, Canada

Dominic Case wrote :

>...In actual fact, moonlight tends to be a lower colour temp than >daylight (it's usually quoted at 4000K) in other words it's redder, but >because we tend to see night exteriors in comparison with even >warmer artificial light (in house windows etc) it appears cooler.

No, you already cited the reason it looks bluish--the Purkinji effect. As the illumination diminishes we lose our red sensitivity first, then blue, finally green. Under moonlight our eyes are "orthochromatic".

Mark Smith writes :

>Just watched "Born Free" - sorry my 9 year old likes animal films in all >shapes. Has some terrible day for night in it. I recommend seeing it for >an example of what not to do....

I don't recall what BORN FREE's DFN looked like, but if they were filming in the wide open spaces, those are among the worst conditions for success. Successful DFN depends upon having a preponderance of dark areas with minimal highlighting. So you shoot backlit as much as possible. In scenes lacking trees, buildings, etc., to form the shadowed areas you end up with wide open vistas of well lighted ground or pavement. The only thing worse than this is front lighting.

Wade K. Ramsey, DP
Dept. of Cinema & Video Production
Bob Jones University
Greenville, SC 29614

Wade Ramsey wrote:

>No, you already cited the reason it looks bluish -- the Purkinji effect. As >the illumination diminishes we lose our red sensitivity first, then blue, >finally green. Under moonlight our eyes are "orthochromatic".

Right, purely scotopic vision is colour-blind using only the rods (as opposed to photopic vision using the cones). In darkening situations the reds are the first to go - meaning that the "cheesy" blue night look is fully grounded in reality, if frequently overdone.

One of my favourite applications for DFN is creating an unlit look for night time or dusk interiors. Use a fast stock, silk everything so that there are absolutely no shadows, get as deep a stop as you can and go 1 under, throw a light fog filter on the lens. You get a soft, low contrast look that can be very sensual. Of course in this case you can let the scotopic vision of the audience take care of the color balance

Best Regards,

Anders Uhl
Cinematographer, NY
The DoP shop
http://www.thedopshop.com

Hi,

>I don't recall what BORN FREE's DFN looked like

Woodland exteriors. Ropey. Patches of bright blue sky visible.

Phil Rhodes
Video Camera/Edit
London

Under low illumination levels, the rods in our retinas are most sensitive to blue-green (cyan) part of the spectrum. I've always been kind of surprised that cyan (instead of blue) hasn't been widely used for DFN.

Jessica Gallant
Los Angeles based Director of Photography
West Coast Systems Administrator, Cinematography Mailing List
https://cinematography.net/

>Under low illumination levels the rods in our retinas are most sensitive >to blue-green (cyan)

But that doesn't make things appear cyan. The rods don't give any colour information. The only sense in which we actually "see" any colour is when the cones are operational, and as Wade points out, in slightly diminishing light, we lose the red cone vision first, making things appear cyan. Which may be a better reason to pose your question - (why don't we use a cyan filter for DFN, not a blue one?).

But Wade's argument :

>No, you already cited the reason it looks bluish -- the Purkinji effect.

Is a little brusque. I stand by the reason I proposed - a matter of comparison with lower colour temperature artificial sources in the scene, and a matter of "conventional" representation. Why can't we have more than one reason for an effect?

Dominic Case
Atlab Australia

Dominic Case wrote:

>...But Wade's argument :

>No, you already cited the reason it looks bluish -- the Purkinji effect.

>is a little brusque. I stand by the reason I proposed - a matter of >comparison with lower colour temperature artificial sources in the >scene, and a matter of "conventional" representation.

Sorry, Dominic, I didn't mean to be brusque. There isn't any reason why there can't be more than one reason. However, moonlight still looks cool even when there are no artificial sources. For example, I can recall many years ago driving through western Kansas at night under a full moon. Nothing but wheat fields, no other cars on the road. I turned off the headlights because the full moon was so bright I could see farther without them, my eyes accommodating to the dimmer illumination. The whole scene was cool cyan. You may have experienced the same thing in Oz out in areas that are wide open with little or no artificial sources. Tungsten sources in the scene will make the contrast more vivid, but the effect is there regardless.

Wade K. Ramsey, DP
Dept. of Cinema & Video Production
Bob Jones University
Greenville, SC 29614

Dominic Case wrote:

DC>
But that doesn't make things appear cyan. The rods don't give any colour information. The only sense in which we actually "see" any colour is when the cones are operational, and as Wade points out, in slightly diminishing light,

AU>
Right, and the best term to use for this is really not "scotopic
vision" but "mesopic vision" - where scotopic and photopic overlap. Since scotopic vision (rods) is colour-blind, it is the color *perception* of photopic vision (cones) that creates the blue effect.

DC>
I stand by the reason I proposed - a matter of comparison with lower colour temperature artificial sources in the scene

AU>
A very good point, and an interesting one. In artificially lit situations (lower color temperature) we can assume that photopic vision is carrying the weight and therefore the scotopic effect does not come into play, meaning a more neutral color sensitivity. However, natural night light (dark?) with a higher color temperature takes on a blue look due to color contrast. Or....

Best,

Anders Uhl
Cinematographer
ICG, New York

Given the recent discussion I thought this might be at least remotely topical :

http://www.riverrockstudios.com/riverrock/pages/dayfornight.html

In a fun to play with sort of way.

Ryan Liverman, editor

Neoscape, Inc.
700 Massachusetts Avenue
Cambridge, MA 02139