Author: Pieter de Vries

The green screen should be set at around 3 meters from the subject and the larger the screen the better. Try to not compromise on this distance from the subject, even if it means choosing another location. 

This gives you separation from the screen and reduces the amount of green reflecting off the screen and onto the back of your subjects. This green contamination is a real no-no.

Light your screen separately. Lowel Tota-Lites are good for this however KinoFlos or Lowel Caselights are also very good. I position one to each side and run them through a dimmer to control the output. The plan is to get an even spread on the screen with each of the lights washing out any creases in the fabric if that's what you are using. Lighting the screen so that it's one stop darker than the subject is a good rule of thumb.

Once you are happy, light your subject separately using all the modelling with the light and shade that you would normally use. Keep in mind that strong back-light on the hair and shoulders can exaggerate the fact that’s it is a green screen shot when the final composite is done. When you have finish lighting the screen, turn off the lamps lighting your subject so you can check to see if any green is reflecting back onto the backs of the subject. A small amount of back lighting will generally take care of this problem. It's important to make sure that there is no reflected green from the screen falling on the back of your subject – if there is, it will come back to haunt you as the production manager coughs up the extra dollars for the editor to clean up the problems.

Even lighting

This was a talk on lighting basics I gave to filmmakers about to shoot their first film for the annual Tropfest Film Festival held in Sydney in February each year. There are helpful suggestions on which lights are best for particular shooting situations (duration 05:22)

Sony RoughCut lighting talk from Pieter de Vries ACS on Vimeo.

There are a number of different styles of tripods available. Regardless of whether it’s the telescopic type or the dual tube type, always try to use a tripod/fluid head combination that has a "levelling head". A tripod that has a adjustable rising centre post is designed for stills photography – it has no place under a video camera.

Video and stills tripods consist of two parts – a tripod and a fluid head, however it is common to refer to the whole unit as a tripod or a tripod system.

Tripods with ultra-lite telescoping tubes are popular for use with video cameras – the height range is impressive and you can operate from almost ground level to over the top of crowds.  Setting up a tripod can be a little like setting up a deck chair, so it’s best to get in the habit of doing things in the right order. For this tutorial, we're using the Miller Solo DV 2-stage telescopic tripod which is the one of the best tripod systems available.

Stand the tripod on the floor, un-tighten the pan/clamp and raise the pan/tilt handle parallel to the ground. This gets it out of the way so you can proceed to the the following steps.

1. Lift the tripod and cradle it on one arm. Loosen the (larger) upper concentric lock on one leg and extend the tube to the approximate height that you’d like to work then re-tighten it. If you want to set the tripod to a higher position, then extend and lock the lower tube first then the upper tube. You can stand the tripod on the one extend leg to check if the height is right for you.

2. Once you’re happy with the height, hold the tripod as before and extend the other legs to match the length of the extended tube. Make sure that you tighten all three concentric locks. 

3. Grip two of the legs with both hands near the top of the tripod and place the foot of the third leg to the ground in front of you. Keeping that tripod leg in place and take a step backwards, drawing the other two legs towards you and placing them on the ground. Be sure that you pull the legs out to the limit of the leg angle limiters or built-in spreader.

4. Using a Miller DS10 system, you should firstly mount the rectangular camera plate onto the base of the camera with the supplied attaching screw. The screw should be tightened and there should be no slipping movement – the locating pin on the plate should prevent this.

5. Slide the attached camera plate (now securely attached to the base of the camera) into the recessed channel on the fluid head and make sure that the yellow securing lever has fully engaged to lock the camera onto the head. Tighten the smaller yellow slide channel lock on the side of the head to ensure the camera doesn't slide forwards or backwards in the channel.

6. Set the panning arm angle. Find a position where the pan/tilt handle feels comfortable. I find it handy to have an index mark on the head to find this position in the future. Try a few positions to see how it feels, but once you find the sweet spot, you should place an index mark on the head so you can angle the arm just where you like it every time. You’ll find that it feels comfortable and familiar when it’s set at the same position each time. 

7. You are now ready to level the tripod. You should do this by gripping the top camera handle in one hand, and loosening the main locking adjustment cup underneath the tripod head. You should now be able to roll the fluid head with the camera and adjust all angles so that the small spirit level indicator on the head is in the centre of it's circular index marks. At this point re-tighten the adjustment cup.

8. You now need to balance the camera on the head. This is done by loosening the black tilt lever on the side of the fluid head. This will let you tilt the camera on the tripod. You will probably notice that the camera will tend to naturally tilt backward or forward depending on the where the camera is seated on the fluid head. You need the camera to sit level to the horizen, so undo the yellow slide lock (near the camera mounting plate) and slide the camera forward or backward so that it sits in that horizontal position. Tighten when balanced.

You are now ready to set your tilt and pan drag settings with the small rotatable knobs. This is a personal preference, but when set correctly you should be able to make smooth slow or fast speed camera moves without jerking, and be able to start the move from a stationary position without having to apply excessive force.

Sequences recorded inside vehicles at night will only look convincing if they are under-lit. 

Setups that look fake, are usually the result of using too many lights, or by placing them in the wrong position –  less is more. Lighting vehicle interiors should be approached with a low key look in mind and these has to be a balance of artificial and ambient light to get realistic results.

What will you learn? 

This tutorial will show you the type of light fixtures to use, where to put them, and the mechanics of light placement in an on vehicles. It will additionally give you the tips and the techniques to get that lighting balance right so you can shoot a convincing night driving sequence or in-car interview.

A quick way to shoot the sequence but with no over-the-top expenses. In particular, let us take a scenario, where you have to record an interview with the driver of the vehicle at night. The interior of a vehicle is a dark place at night, and without some lighting you will not get an image. So, how do you get the illumination that you need?

Firstly, finding a shooting position. The position from where you put your camera is going to help you find the position of your lights. Handholding the camera while sitting beside the driver is not going to give you the best result, and here is why.

1. the camera will shake and sway as you roll with the movement of the vehicle
2. it will be an unflattering side view
3. the lens will be too close to the driver
4. you will be seeing straight out of the driver’s side window into the blackness of the night

you position yourself in the front passengers seat next to the driver, you should be able to sit the camera on a CineSaddle (or something similar) and hold it onto the dash. This will give you a far nicer angle.

On many video cameras, it is possible to twist the LCD panel so that you can see the image, and keep check on framing and exposure as you go along. This dashboard camera position is at approximate thirty degrees (the four o’clock position) to the driver. This is a good position because it is not a full profile view, and the area immediately behind the driver is not the blackness of the night. Helpful, especially when there is very little street light.

A slightly front-on angle works nicely for a number of other reasons.

1. the camera is not too close to the driver
2. the camera is locked to the movement of the vehicle
3. the driver is not in full profile
4. you can see some of the vehicle interior

What natural light sources are there? 

There are a few light sources already at play both inside and outside, but they are subtle and not usually outputting enough level to get a usable image. 

What are these sources?

1. the glow from the instrument cluster directly in front of the driver
2. street lighting
3. the lights of other vehicles
4. general night exterior ambience

These are the light sources that we will use, however they need to be re-created with your own lights. Essentially, you have to boost these levels.

A quick rewind

• there has to be a balance of artificial and ambient light to get realistic results 
• the camera positioned on the dashboard is best 
• turn the LCD panel so that you can see the image
• Lighting source #1 – the instrument lights

Instrument lights at night

There is always some illumination from the vehicle’s speedometer area illuminating the driver’s face, however, it is not usually bright enough to give you the ”base“ light level needed. You will have to recreate that, and here is how to do it.

Simulate the light from the instrument cluster to ensure that the light level is the minimum required to get a decent exposure. It should have enough output to sustain the scene, regardless of the exterior brightness; brightness levels that are sure to vary as you travel.

Here are a few ways to simulate the effect of that subtle instrument glow. 

Bounce the light from a battery light into a bounce board. Cut to size, then fit a small rectangle of white card into the instrument area. Place a battery light low and out of shot, but in a position where it is shining directly into the white bounce card. 

A low-wattage 12 volt battery light will be too bright for this purpose, so to reduce the output: 

1. take a sheet of Blackwrap Photofoil and roll it into a cone shape 
2. tape it to the front of the lamp head
3. squeeze the Blackwrap to allow just a small amount of light to shine through

Position the battery light. The light bounced from the white bounce card should suggest the subtle soft glow onto the face of the driver from the instrument area – similar to what might normally be happening. Providing that it is not too bright, it will be convincing.

Warm up or cool down the output. A sheet of CTO gel (Colour Temperature Orange) will add a warm quality to your light. You could even consider going for a blue CTB gel (Colour Temperature Blue) for a modern high-tech look. 

If lack of space around the instruments is a consideration, then wrap the white bounce card with the gel and leave the battery light gel-free. The result will be the same. You may need to devise a way of attaching the battery light in a position where it is not in your shot. One suggestion, is to use gaffer tape to secure the light to the top of the steering column.

Rosco’s LitePads are ideal for these set-ups. They use safe LED technology to output bright ultra-soft daylight. They are lightweight with an unusually slim form factor. Installing them is straight forward, not only because they are a compact all-in-one fixture, but because they occupy very little space.

Place the LitePad in the instrument area facing the driver

Rosco’s LitePads are safe and easy to manage in the cramped interior spaces of vehicles. They use narrow gauge 12 volt cable, useful in this situation, as it can be easily gaffer taped to surfaces in the vehicle and connected to a 12 volt accessory outlet. 

It is easy to over-light inside a vehicle and remember, you are trying to simulate the muted soft light of the instrument panel. Two sheets of Blackwrap Photofoil wrapped around the LitePad will cover the surface of the light and provide a way to decrease the light level.

Fit a sleeve of Blackwrap at both ends of the pad, leaving just a narrow slit of light in the centre. Slide both sleeves away from the centre to allow the right amount of light to sneak through. Sliding the Blackwrap in or out becomes your light dimmer.

In some ways this is a better solution to using a dimmer or ND gel. LitePads throw a wide, hard-to-contain spread of illumination, and it is hard to limit the light from spilling and bouncing all around the interior of the vehicle. This narrow slit of light is more centred on the subject, and it enhances what we are trying to achieve. 

Rewind

• fix a small rectangular white bounce card into the instrument area
• a sheet of CTO gel will add a warm feel to the light
• sheets of Blackwrap provide a way of reducing the brightness 

Lighting source #2 The door/rim light. Another fixture can be used to lift the overall interior level, without the prospect of over-lighting. This second light can be mounted to the exterior of the vehicle, in this case on the roof. 

The purpose of this second light, is to simulate the characteristics of illumination that comes from street lamps or just general suburban street ambience. But it does more than that, and has the potential to be a constant and motivating source. 

Filling it out. The soft light coming from the instrument panel, creates a slightly spooky feel, because the angle of the light is low. This second light is used to provide contrast and a little fill-light at the same time. Importantly though, it injects the spikes of high contrast associated with lighting at night. 

It is best located at a high angle, where it can throw light onto the inside of the door, spilling onto the drivers lap. Some light will also fall onto part of the face and shoulders. Falling as back-light in the scene, it will have minimal effect on the exposure level on the face of the driver. 

Catching the light. One situation where it does have an effect, is when the driver turns and looks out of the window. The exterior mounted light will catch and rim-light the side of the face. This is just the way it happens in reality. It is a good reason to ask your subject to turn occasionally and play to that light. 

Turn it off. The door/rim light will not be so convincing when the vehicle is passing through locations where there is little if any outside street light. You could switch it off in this situation, and let the instrument lighting carry the scene. This rim and door light, can be left switched on, or it can be triggered remotely, turning it on and off, or dimming it up and down to give the effect of passing by the street lamps as you move through the street light.

Can I simulate the odd colours of the street lights? 

You can use a small battery light or another LitePad. It could be tungsten or daylight balanced – it depends on the colour that looks best in the particular location.

Rosco Cinegel 3152 Urban Vapour can be used to simulate Sodium Vapour street lamps. To mimic the blue appearance of street lighting, typically Metal halide or Fluorescent, you can add Rosco Cinegel 3204 1/2 CTB or 3202 Full CTB to the light source being used. You will hear the gel rattle in the wind as you drive, so make sure the gel is taped tightly.

Attaching this light is simpler than it seems. 

1. You will need two Manfrotto Magic Arms, two Super Clamps and two Manfrotto Pump Cups.
2. Attach one of the Pump Cups to a flat area on the roof of the vehicle preferably just above the driver’s side window
3. Connect a Magic Arm and secure the light to the opposite end
4. Lock the arm in a position where the light is not seen the shot, but skirts the door interior and the shoulders of the driver
5. Locate an attachment point for the second Pump Cup. It is often most secure when it is sucked onto a glass surface – car windows are perfect
6. Attach the second Pump Cup – in this case, to the rear driver’s side window
7. Attach another Magic Arm and fit a Super Clamp to the opposite end
8. Brace the first Magic Arm close to the light attachment point with the Super Clamp and lock the Magic Arm
9. Make sure that the Magic Arms and light are not extending too far from the side of the vehicle (150mm at most)

Use gaffer tape to run the cable from the light to a point where you can switch it off or on. It will be useful to have this control.

Finally, driving sequences at night look better if there is a reasonable amount of suburban or city light in the streets. It helps to justify this simple lighting setup.

Call it integration converging or merging, High Definition Digital Video continues to power on a DSLR camera.

There are many good reason to like the Canon 5D Mk3. One is that it answers the question as to why digital video cameras have not shared sensors of similar specs to those in digital SLR's. There is no reason now, and with the Canon 5D and 7D along with Sony's A7s, they do. Specialised training in the use of Digital Single-Lens Reflex cameras is now available through PDV Digital Cinema Workshops.

The Canon 5D Mk3

I wonder if Canon’s engineers knew what they were starting with the 5D M11 back in 2008? The video feature has all the hallmarks of being a something that was added on simply because it could be easlity done.

DSLR's are not designed to shoot a lot of handheld video, however it is an amazing engineering achievement and people have gone nuts over it – with good reason. You have the ability to record 1920 X 1080 full HD resolution onto a full frame sensor in a compact stills camera.

I’d like to take a look at some of the more practical aspects of this integration that may get lost in the excitement. If you plan to shoot HD video with this camera, there are a number of things that have to be considered.

Stay focussed

Video displays are larger than ever with resolution that has never been higher. Digital cameras continue to capture even crisper and cleaner images and essentially you have in your hands a lightweight and compact device with the potential to capture near 35mm cinema quality movie footage.

Even with today’s pace of development that is quite an achievement however there are a number of important technical considerations. The difference between shooting stills and recording video images is considerable. Some differences are technical, others creative, but each has some influence on the other. What does this mean if you plan to use a camera similar to the 5D MkII for serious video work'

Full frame sensors

The primary difference and the one that will impact the most on you, is the resulting shallow Depth of Field (DoF) characteristics in the recorded images – characteristics similar to those shot on 35mm film by motion picture cameras such as those designed and manufactured by Sony, Arriflex, RED and Panavision.

Razor thin Depth-Of-Field is one of the reasons why large format images look cinematic and movie-like and having that capability in a small still camera marks a turning point for cameras like the Canon 5D Mk3. Stills cameras have included some basic option for video capture for a few years so why this full frame stuff is different.

Different because the Canon 5D Mk3 uses a CMOS sensor approximately 36mm x 24mm in size, larger in size than the image area exposed on double perorated 35mm motion picture film.

If you happen to set focus just a tiny bit off the mark, the image will look out of focus on one of these big screens – these soft shots will not make the cut .

It means that precise focussing has inched it’s way up and become one of the more essential technical skills you will need to master if you are to get the most from a DSLR camera. Focus, and the selective use or it, can now be used as an effective mood-generator in your video work.

Photographers have always had a fondness for the artistic benefits and the pitfalls of shallow DoF as defined by the aperture setting (as determined by shutter speed) in concert with lens focal length and image format size. This awareness, carried over to shooting video on the 5D MkII, will most likely come as second nature.

Videographers on the other hand who have come through the “Mini-DV Revolution” will be right at home shooting scenes where everything, from the those irritating specs of dust on the front element to the mountains on the horizon is in sharp focus.

Keeping an eye on things

Remember that you will be recording action that may be constantly changing its distance from the camera. With the Canon, the Live-View screen at the back of the camera is the only way to monitor the video recording without going to an external display (which I recommend).

It’s fine for framing and to some extent setting exposure, but it is not a that accurate or convenient way to deal with critical things like focus. Firstly, it’s hard to pick correct focus on the wide angle shots and I suggest that you use the feet/meter distance markings on the lens to set focus to the right distance.

Nothing is worse than covering action a few meters from your shooting position only to find the lens focus distance was set to infinity. If it’s fast and furious action, you may not pick it. Stay on a telephoto shot and let soft focussed objects to pass through the foreground. They can reveal a perfect moment and you’ll be amazed how good this can look in HD.

When you use the telephoto end of the zoom, focussing becomes more challenging. Telephoto lenses compress the frame – drawing the foreground and the background in together resulting in a composition that has multiple focus planes.

Use shallow depth of field to:

If your video camera has a Zebra function then it’s handy to understand a little about it. It’s a way of setting the correct exposure. 

The slanted striped pattern (hence, Zebra) that you see is set to overlay on areas of the viewfinder/LCD image that are of particular brightness – commonly somewhere between a value of 70 and 80 IRE* of the whitest white that’s possible to record. There are two basic types of zebra display, 100 IRE and another that’s often adjustable, 70-80 IRE.

Notice that this electronically generated striping plays on different parts of the image as you close down or open up the iris. This is because as you lighten or darken the image, elements in the scene fall into this 70-80 IRE value. Caucasian skin highlights occur at about 80 IRE, so you should adjust the iris to display the zebra pattern on the highlighted areas of the subject’s face. Just for practice, try this:

Frame the camera on someone as if you were shooting a portrait. Making sure that the Zebra is set to 70 IRE (or “percent” as it’s sometimes written), turn the iris adjustment so that the zebra striping is active all over the face. Now close down the iris so that the striping starts to disappear. 

Leaving a stripe or two on a highlight area – a cheekbone or top of the forehead is usually the best setting. The skin tone is now correctly exposed and this ensures that everything else in the scene is also correctly exposed.  

Another useful way to use Zebra function is to use it as an exposure meter. With your camera set to Auto mode, frame the camera at the scene to be recorded and note on which parts of the scene the Zebra stripes are active. Switch the camera to Manual operation and adjust the iris control so that the Zebra stripes overlay in the same place that they were in Auto. 

This way, your manual exposure setting is going to be at least in the right zone. I usually then close down another stop to give the picture a little more saturation. Remember, the slanted striped pattern is not recorded on the tape.

If your camera doesn’t have a Zebra exposure control but you’d like to have manual control of the iris, then this is what I suggest: with the camera set to Auto mode, frame it at the scene to be recorded, make a note of what the overall brightness level looks like in the viewfinder. 

Switch to Manual operation and adjust the iris so that the image brightness matches what you saw in Auto mode. In it’s basic form you’ve used the Auto mode as an exposure meter.

© 2013 Pieter de Vries ACS

I always try to have the sections of the document required to be filled out by me, completed before arriving in the country. Usually on the plane as I'm about to land. I find that it's good to have the document already out and in full view of customs officers giving directions to the area for processing.

Have the rest of the crew follow you and they will usually slide through the procedure in your 'wake'. You will probably know more about the procedure that the customs officer in some countries. 

Have the number of items listed (eg 1 thru 86) written clearly in that part of the document and insist on stamps where they are needed. Have the total value of the carnet written in the local currency of the country you are entering. This tiny bit of thoughtfulness impressed the hell out of tired and weary Customs officials. Typo's can be corrected on the way out with a pen and stamped by the customs people there. 

Don't accidentally fill out and tear off an exportation voucher instead of an importation voucher when entering the country. Done that! We lost a carnet in Italy (house keeping threw it out!) and had to wait a week to get the equipment from customs in Milan while we waited for a hard copy of the carnet to be DHL'd from Australia. Could have been worse – a week of sitting about in Moderna, Italy drinking coffee etc! 

Some production companies ask you to reduce the value of your items listed on the carnet document in order to reduce the bond they have to pay to the authority. I suggest that you don't agree to do this as in the event of an insurance claim, these marked down values can be used as a guide to the replacement costs. 

Finally, allow 3 hours to get yourself and the gear out of Tel Aviv. You will be subjected to an extensive interview with your equipment laid out in front of you. Your interviewer may be no older than 24 and be very pleasant and apologetic.

© 2013 Pieter de Vries ACS

You should always be aware of the background when you set up an interview. If the area behind the subject is relevant and appealing, you might wish to take in as much of it as you can. On the other hand, if it’s unsightly, distracting, or bears no connection to the topic of the discussion, you may wish to exclude much of background. In this case, a narrow field-of-view will hide a world of nasties. How can you do that?

The zoom lens is a very nifty inclusion on a digital video camera and there is a direct connection between the focal length used (the degree of magnification) and the resulting field-of-view. Framing your subject where you use the telephoto end of the zoom lens is the best way to get a narrow field-of-view. 

Place the tripod further away from the subject so that you make use of the telephoto end of the zoom. In the viewfinder, the size of the subject would be the same as would be the case on the wide angle end of the zoom, however, what is included in the background will be greatly reduced. 

A once distracting background will also now be beautifully diffused mainly due to the fall off in focus that accompanies a greater magnification by the lens.

© 2013 Pieter de Vries ACS

For the smallest file size, the ideal subject would probably a talking head against a plain and boring background in a noise free studio. This is not very practical so please consider the following guidelines.

• Try to shoot on a tripod when possible. If you shoot video with the camera in your hands, every slight move you make is translated into change between frames and generates more data to compress. The more similar each frame is to the next, and the one before it, the better the final product.  It’s all about keeping the data rate lean.

• A word about wardrobe. Stripes, dots and checks are not so good. When you can plan ahead, ask your talent to wear low contrast solid colours. You might be tempted to set your subject up in front of a tree swaying gently in the breeze, but you’ll need a lot of file size to capture the movement of the leaves online. Try finding a stationary background that can be easily compressed and still look good.

• High contrast busy backgrounds are also unhelpful. Plain walls will compress better than busy wallpaper

• If you don't have a choice of location then shallow depth of field will come in handy. Open the camera's aperture (f2.4 – f5.6) and place the camera further away and increase the focal length to reduce depth of field. A neutral density filter (ND filter) will help to prevent over exposure. Your subject can be kept in focus while the background is blurred, reducing the amount of detail that has to be compressed.

• High contrast lighting is also not useful. Use a soft-light indoors or put a diffuser in front of a hard light. Shooting at midday in bright sun will cast dark shadows under chins and into deep set eyes

• In a web browser’s small viewing window, close-ups give the viewer a better chance of working out what is going on than wide shots.

• In addition, make sure you use large enough titles, as the screen will most likely be reduced to a 360×240 (or smaller) size. You can size down your preview window in Premiere or Final Cut Pro to that size and that will help you preview your logos and titles

• To obtain the cleanest and clearest sound the microphone needs to be very close to the subject. An on-board microphone does not achieve this so an external microphone and a skilled sound recordist will produce the best results. Minimize background noises. Good clean audio produces better results in compression

Source: www.urbanfox.com