Light House Effect
Published : 30th October 2008
We are going to build a exterior set in a studio where we want the light from a light house sweeping over. The set will be about 160 by 180 feet so I think I will need quite a large lamp (about 12k sun). The beam should take about 3 seconds to travel over the set and it should come every tenth second.
I have done a rig before with two 6k in a rotating 4-sided mirror but it didn’t work as well as I hoped. The strength varies depending on angle and the rig got quite big.
Have anyone done anything similar? Advice would be appreciated.
As boring as it sounds, a guy sitting on a ladder panning the light one way, and then tilting up and panning back and then panning again, while boring and stupid, will work.
I am unsure though... does the lighthouse light illuminate the set or does the beam of the lighthouse just scan over the top of the set in the air?
If it is the beam that you want passing over, you might want to consider a large Xenon light or one of the large Beam Spot projectors (like the big mole ones) or even a cluster of Very narrow Par bulbs all rigged parallel to each other.
In general, the bigger and the farther away the light, the less difference between the beam strength at one end of the set and at the other end of the set
If you are rigging a tungsten source, it is relatively easy to set up a slip-ring arrangement to have the light rotate continuously in one direction. You can also rig the cable for the light up to the grid above you ... the further up you tie it the better, and you can have it twist one revolution for every revolution of the lamp... it will take a long time to twist up too much to be able to continue...but then you have to untwist it
and twist it back the other way before you start shooting again.
Hope this helps
High End Systems just released the Showgun this past summer. It is arguably the brightest automated luminare. I've included a link to the web page. I know a few rental houses in LA have them, I'm not sure about overseas yet. I think you will be pleased at the specs this light has to offer.
The only expense you incur labour-wise is the initial programming and I would imagine that most of the popular automated lighting desks have already built the library for the Showgun (ie: Hog III, Grand Ma, Avo Pearl etc)
Rolf Carlbom wrote:
class="style21">>>A Xenon will probably be the way to go. The problem with having a guy on pan duty is that the rig >>will be used every day for six weeks.
Yikes! Then you really do need a rig. I'd try making a box of 3 or 4 mirrors (four-foot size) and hang it from a turntable. Point the light of your choice away from the set, at the mirrors, tilted up very slightly, so that the reflected beam goes just over the top of the light.
The turntable should be variable speed, so you have options for wider and tighter shots.
Reliable tungsten Molebeams (20K) may eliminate the hassle of sometimes troublesome Xenons. The HMI Molebeams are great as well.
Does the beam touch the set or touch anything visible, or does it just light the air (presumably with smoke pumped into it) above the set? the reason I ask is that there are some ways of doing this with existing technology... but it depends on how coherent the beam needs to be etc. Actual lighthouse beams are created (or were in the olden days) with a light source at the focus of a very big segmented Fresnel lens. and a large reflector ... reflector and lens revolving around the light source... a big beam spot projector would give you a better beam with more parallel edges for the same amount of watts than a big Fresnel... but par cans working together might work just fine... it has in the past for me. It depends on whether you need to see a splash on something or whether you are just lighting atmosphere.
If you want to automate it, and if the beam does not have to be really tightly defined, (or even if it does) there are pan and scan mirrors that have been around in the theatrical world for at least fifteen years... I used them in the early nineties.
They are designed to work with PAR cans. You can take any DMX board and control mirror pan with one channel and mirror tilt with the another channel. With three channels per fixture and VNSP or NSP globes in theatrical par cans pointing straight up into these colour-frame-mounted pan/scan mirrors, you could build an effect cue on any modern memory dimmer board that would pan three or four of these 1k or 1.2k tungsten PAR beams across the set in unison (so they look like one beam)... dim them out or not... pan them back to the starting point with the beam pointed either down or up to avoid the set, and then pan across the "sky" again. I actually did this sort of thing back in 91 or 92... technology must still be lying around somewhere. I used one on a baby 10k with a snoot, but the heat got to be a problem... but if you have not used 1.2k VNSP tungsten par bulbs, you are in for a treat.
There are a lot of ways to skin this cat, different ways will spend money in different places... you can build an electromechanical rig that pans the lamp or rotates it around with a slip-ring to power it... you can use this mirror system pan and scan... I think the thing to do is do a quick test of different types of light sources and atmosphere to see what is going to create the look you want... and then move on to the automation phase...
LA based VFX DP/supervisor
A Xenon will probably be the way to go. The problem with having a guy on pan duty is that the rig will be used every day for six weeks. I thought it would be nicer and probably more cost effective to build something automated.
Xenotech/Strong make a 7k xenon with moving mirror head I think they were built for a U2 tour.
Its called the razorhead Also you can get sky tracker type units that will rotate endlessly
The light will hit the set. I’ve done a fast 3d render with three rotating lights which seems right but maybe four lights per turn would be better. http://www.questmedia.se/fyr2.avi
The rig has to be at least 100 feet from the set and should give a 5.6 on 200ASA and be in 5600K.
Are there pan/scan mirrors that are big enough?
Shouldn’t it be at least 20 inches to reflect the entire beam from a molebeam or xenon? If this would work I think it would be perfect since you can control angle and speed.