Knowing how to determine how powerful of a light you need

D

DDirector

Veteran
Wondering how cinematographers/gaffers know how powerful a light they need to be to pull of a certain scene or shot.
Other than their experience, how would those less experienced know how powerful light would be needed?
How would we know which light to use? If a 1K be sufficient, 2K, 5K, 10K or 20K?

Typically I would do a location scout and see what the ambient light readings are with the aperture reading using a light meter.

But is there a way to help me know how powerful of a light I would need for whatever shot I'm trying to pull off?
Perhaps calculating the light to be used by its lux? and somehow working with that?
 
For hard, direct light you can use a photometrics table or chart. Arri has a great tool on its website. Lamp X at full flood covers Area Y at Distance Z to a brightness of # footcandles or LUX. The problem is when you want to use diffusion or bounce the light. At that point you need to use experience to determine needs. Remember that you can always get a bigger light and dim it down.
 
Mitch nailed it. Making light really soft takes a LOT of power for most situations. If I am going to punch a light through say a 6x6 full sail grid directly, versus making a 6x6 book light, I am going to need two or three times the output for the book light to achieve the same or similar light output.
Also, one thing rarely discussed but everyone instinctively knows is using egg crates to control spill is a good thing. The fact that a 40 degree egg crate considerably cuts your perceived output on the subject by a lot is rarely talked about but talk to any gaffer or DP and they know this from
experience. Also, placing big sources outside of windows is a popular thing for narrative shooting and unless your grips and gaffers are going to completely enclose your source in a Duvetin cocoon to block out ambient, you're going to bleed off a lot of that output with ambient sunlight.

In regards to your question, there are a lot of variables on set. You can always buy and learn the ASC cinematographer's manual and know the hard data but in the real world, especially for location shooting, that's what location scouts and light meters are for.
I also completely agree with Mitch about the fact that bigger lights and more out put is almost always better as long as you have the power and lighting control tools available on your trucks. There's nothing more frustrating than setting up a 4K HMI on a guess
and not having enough horsepower if it's a complex gripping job to locate a light where you need it, like outside a window on a second story building. Much better to have a 6K, 8K and dimmer, wire and diffusion to just get it to what you need than to have to re-rig.
Time is money on set and then you are just burning money with no tangible result to your project.
 
My lightmeter doesnt have lux or fc. Would I need it? I have the L758 but its not the cine version.
But still need to understand how I would use the lux to determine which light is right
I guess, how is lux useful
 
I think it is a hard question and the main answer is experience.

One of my weaknesses as a DP would be exactly that.. how powerful are big lights that I dont own and have not used?
What to spec for lighting a bigger space that might need huge or many fixtures from rental?

In my world (which has topped out at 2.5k hmi) the answer is lights are 'never bright enough'.

Personally I find a DSLR (in stills mode) to be a great tool for recces because Im too slack to write down a load of measurements and co-ordinates.

DSLR meta data sticks all that onto an actual photo.

On a recce Id set my DSLR at maybe 800ISO 1/50th and then stop down to a correct exposure.
I then have a picture of a scene knowing that the ambient is 800-T4 or 800T16 or suchlike.

I have hit the problem that 400ISO on a nikon stills camera = 800 ISO on an URSA.

I could then basically test my lights. - Maybe one of my panels through a 4.4 diff = T2.8 at 10ft or T11 at 1 foot

Ultimately you have to be able to 'smell' it!

Clearly you need to be across the inverse square law and other lighting basics.
 
morgan_moore said:
I think it is a hard question and the main answer is experience...Ultimately you have to be able to 'smell' it!
Click to expand...

I have to agree. There was a time many years ago when I was somewhat intimidated by "big guns" because I hadn't worked with them much, only that which could be plugged into house power. But once I got familiar with them, it became second nature to know how much power I needed. Daylight is a constant, and once you get used to how much you need to augment or overpower various levels of ambient light, it just becomes instinct. Same with bouncing or pushing through diffusion.

It's a little harder with LED units because they play by different rules than tungsten or HMI did--lenses and diffusers built into fixtures can alter the transmission level significantly, so going on pure wattage can be misleading. So, you have to get to know the individual units and learn their spread and output.

As Mitch and puredrifting noted, going "one bigger" gives you flexibility. I get antsy when we have "just enough" light because, sometimes you need to go to 11 unexpectedly. Likewise with the number of you units you have at your disposal. Units can break down (or fall down) and having exactly the required amount means you are vulnerable to getting caught out. I did just that recently on a low budget short film and sure enough, things changed and I had to scramble which resulted in wasted time and effort.

Everything about this craft follows a similar pattern: you experiment, you alternately fail and succeed, you learn, you build your knowledge base and eventually it becomes instinct.
 
Most people are answering that you learn what lights you need by experience, but you can also figure out what lights you need with an understanding of math and physics.

You need to know the wattage of the light, the type of light (HMI is, I think, 4x brighter per watt than tungsten, for example), and the distance away from what the light is hitting. With this info you can even figure out what the ratio of the light is going to be on the subject's face before you arrive on set.

Unfortunately I have forgotten exactly how to calculate all this info because I've only been shooting corporate work with dimmable lights for the past 10 years, but I used to know how to do this when I was a teenager. Looks like I need to revisit my old books.

To figure out how much your diffusion is going to modify the output of your light, you can run a test at home and measure the output of the light using a light meter with and without the diffusion.

---

But as Charles has said above me, I think most DPs pick lights intuitively based on a long history of trial and error.
 
FWIW there are a few different smartphone apps which have the photometrics of many of the more popular lighting instruments included.
For example: ARRI Photometrics App
 
All the popular incandescent fixtures have decent photometric data that indicates how many lux at X m, with Z beam angle - it's the only way we used to compare between different types - BUT - LED is far less well specified. The big (and expensive) makes have good reliable data, but anything imported from China has very questionable specifications - and they're frequently wrong, with their figures making no sense at all. Arri, for example tell you exactly how bright their fixtures are under many different operating modes. None of my Chinese kit makes any sense when you compare specs.
 
DDirector said:
My lightmeter doesnt have lux or fc. Would I need it? I have the L758 but its not the cine version.
But still need to understand how I would use the lux to determine which light is right
I guess, how is lux useful
Click to expand...

You need it because the lights have their output specs in lux or footcandles.

It's very simple math if you know the frame rate, shutter angle and ISO settings of your camera. The rule of thumb is shooting 24fps with a standard 180 degree shutter, 100 ISO at T2.8 on the lens requires 100 footcandles to get a "normal" key exposure. Shooting at 200 ISO? Doubling the sensitivity means the needed footcandles drop in half to 50fc. Each stop on the aerture ring is a doubling or halving as well. Knowing these basic relationships you can dial your way into exactly whatever lighting level requirements you may encounter.
 
Experience. Period.
You obviously can't play with an 18k on a condor the same way as you can with a 650w set up in your living room so you wait for some projects that have budget for the extras. That said, if you are like me, you always try to do more with less - on so many jobs i was maxing out avail power and the big lights. On one hand that extra 4k isn't breaking the budget but sometimes the required effort to set it up can. I always think of my main set-up and the accents lights (yeah, 4k or 6k as an accent...) and sometimes just settle without that accent...
 
I have nothing to add. I think everyone here has already shed a lot of light on the matter.
 
Nice pun, but there is still something to be said.

When photographing translucent objects - fruit, vegetable, milk, flesh, fat, skin, human etc. and want to have nice color bleeding inside or diffusion of light inside the shadow boundaries, forget about light meters, charts and tables.
 
At what ratio of direct/indirect light?

C'mon, it's easy to mechanically light the scene to given parameters.

To rephrase the question, which emotion would 10 sky panels evoke and which would you get with 30?
 
With 10 fixtures you might get the emotion of unhappiness at being fired for lighting the scene to T0.5/400ISO and the whole shoot being under exposed or needing to be shot at ISO 10000.

I think the main reason for doing the math is lighting big scenes or phantom high frame shoots or any scene that is outside the box of the work you usually do.

I know very well what my own panels do.. but would they light a stately home ext for a night shot? (actually yes!) but by taking that job without doing the math my ass was hanging out.

At least I tested.. https://twitter.com/sammorganmoore/status/509422075740561408

Here is another location where I felt underlit with my package.. but at least I had the nouse to request more lights.. not that I got them.
https://www.instagram.com/p/BuOgA0OA4Iw/?utm_source=ig_web_button_share_sheet

S
 
Last edited:
This Big Mick talk - Big Mick Hughes, Metallica's sound guy for years. I've been a couple of Tec events where he's been a guest speaker for the sound fraternity. Somebody ALWAYS starts to talk frequency response and dBs. He patiently talks it through.
What's your dB level at mix position? Loud
What if there isn't enough sub bass? Turn it up
What if at the venue you have used all your subs and it's still not loud enough? Get more
What if you can't get more? I've never been unable to get more?

He's just talking experience, not scientific data. In fact, he obviously knows exactly the answer the people want, but refuses to give it to them.

Experience has to be gained, by putting the time in. Mitch nailed it with the rules that work. You need experience to take the rules and run with them. I have a light meter, bought years ago. I have no idea where it is. Most times, all it did was say I was right, or explain why I was wrong. I never lit to it.
 
Great paul but..

How many skypanels for a 100X200ft warehouse?

If the shoot is too dark ill fire you. If there are unused but paid for lights on the truck - Ill fire you.
 
You must log in or register to reply here.
Back
Top