How the world has changed and how LED is the way forward.

[Adam Frisch FSF]

In today's world of commercial cinematography - whether that's features, TV, commercials, documentaries or music videos, one thing is certain: There is rarely the time and manpower available to use the conventional lighting equipment I cut me teeth on. This becomes clearer and clearer for every project I do. I can't recall a single job in the last 5 years where there has been enough time to set, build and control light they way it's supposed to be done. Or the way we used to. This even on big budget commercials. There's the director who's come up through the new school and used to shooting, now, now, now with no lights on a 5D with available light. They look at you with that face that goes "what's taking so long?". Producer is circling you wondering the same thing. You shoot so quick that you can never get ahead and start lighting the next set. You're step in step with the director and producer. The minute you step onto the next setup, they're right behind you asking "how long before you're ready to shoot?".

 

For me, it's time to say goodbye to lights and procedures that can not help me be ahead of the curve or speed things up.

 

Let me give some very recent and random examples. I was at a NASA type facility where we could load in 8.30 in the morning and had to have a hard out by 5pm. It was very complex setups where I basically had to attach 8x8ft single Kino tubes to a wall behind a centrifuge that spins at 10G with lots of air moving. We also wanted a red warning light spilling down over a control panel in the background. I had suggested clamping and securing a LED panel above it to some pipes, but that proved to be too weak after we gelled it. So the gaffer suggested hiding a 575W HMI above the pipes.

 

Took forever to get the bigger light secured up there compared to a panel. By clipping gel to the doors, we now had a ton of spill everywhere, ruining the illusion this was a real red warning light. Cue gaffers trying to black wrap the whole lamp, but still small sliver of spillage coming out everywhere. Wrap more. Still spill. Now lamp overheats and shuts down. Remove blackwrap. We get it going again. Still too bright, stick scrims in there. They smoke. Now the gel has been upset and more white leaks, gel burned in the middle and colour is now pink in middle and uneven, back up again, lamp can't be pointed straight down or it overheats… Probably lost a good 20 minutes on that lamp alone.

 

Next a shot in the control room. Manage to hang some Kinos from ceiling without them being visible in shot by the skin of my teeth. Too bright compared to the window in background. Turn them to low setting. Still too bright. Switch off tubes. 4, 3, 2,… OK, 1 tube is enough. But now the quality of light is not as good as the source has become smaller, they have sharper and ugly shadows on their face. OK, bring out the ND9 gel, oh, it's on the truck 10 minutes away, run, run. Get it on, turn 4 tubes on and shoot after having lost 15 minutes.

 

And it's not just on smaller lights, it's the same on huge rigs, too. Big soft box/moon box for night exteriors or shooting cars? Cue 4hrs of rigging by a team of 4 guys to get that soft box up in the air….

 

You get the idea. This is just one of a thousand examples like this that we have to deal with everyday. There just isn't time to halt the shoot, break out a scissor lift, drive onto set, have a gaffer go up to scrim the big HMI, drive out and start over again. That should be on a dimmer or remote. Today I need self contained units that can dim quickly, are not bulky and ideally can also change colour temp. If it's a soft light, then it's built like a soft light - not a hard light that's been adapted by means of bulky chimeras etc. And I simply don't have time to break out gels to counteract a tungsten dim, or breakout ND's, nets or scrims to hang on lights.

 

For me, it's time to say goodbye to:

 

Chimeras (always spilling, bulky and sagging)

Kino tubes (can't be dimmed, unwieldy, heavy, hard to hang)

Small HMI's (can't be dimmed, spill problems, can't easily be gelled)

Joker Bugs (always spilling everywhere, nasty, time consuming to change lenses)

Vista Beams (too bright for their size, heavy, not easy to use)

etc

 

The future is LED. Either as panels, phosphor exciting or fresnels/PAR's. I love using the Arri L7C light - easy, fast, dimmable, colour changeable and I can't wait for bigger units in this series. Same goes for those 1x1ft LED panels that are battery powered. Use them all the time - just chuck them in there as a little edge light or clamp them onto some plumbing as a toplight etc. I use the Celeb light from Kino all the time, and the bigger one is a perfect replacement for the Kino 4x4 tubes. I love the Rifa tungsten light. Although it colour changes when you dim it, this is less of a problem when you're in a tungsten world, I find. I have my gaffers build covered wagons I chuck on the floor. And with the new LED household bulbs in them, they dim without colour change as well. I will still continue to use old school self contained favourites like the tungsten zip lights and big fresnels.

 

Time has moved on. I hope within the next decade we'll see 18K HMI lights replaced by similar LED units that draw a quarter of the power. Can dim all the way down, can change colour temp.

[Phil Rhodes]

I thought it was just me (until you got down to the L7C part, which definitely isn't me).

 

LEDs are unlikely to become significantly more efficient than HMI in the short term, unfortunately. An 18K HMI equivalent would, currently, have to be an 18K LED, which is some way from practicability.

 

One of my most commonly-used pieces of equipment are actually stuff I made in the shed years ago - smaller, more portable fluorescents than a Kino could ever manage. I'm always astonished at how heavy they manage to make those, all packed up for transport.

 

P

[Jon Kline]

Personally, I'd usually keep my bulky fluorescent lights or HMIs. We're a long way off from LEDs with a reasonable CRI. They're incredibly useful in specific applications, just like any light on the truck. Right now, they're just not right as a key or fill for anything living or organic.

[Stuart Brereton]

Every job I do is on a tight schedule. 15 day features or 1 day commercials, there's never enough time. That said, I work the way I always have, with the lamps I've always used, and never had a producer or director complain.

 

Every time I've used LED lighting, I've been unimpressed. The only exception was the L7, which I thought was a useful lamp, although I wouldn't necessarily carry one on the truck.

[Mark Kenfield]

You really need to look into the very, very latest of LED fixtures in order to get the good stuff - the newest Fresnels and Remote Phosphor panels remove the multi-source nature from LED fixtures (a BIG deal IMO) and the very best units now have colour quality to comfortably match (and in some instances exceed) that of fluorescents. Older LED standards like the 1x1s aren't even in the same ballpark now.

 

I've moved from Kino Divas to Area48 remote phosphor panels for my softlights, and the LEDs win in virtually every way. Output is higher, spill is much more controllable, colour is superior, the fixtures are more robust, you can change colour temperature much quicker, and you can run them remotely via batteries.

[Phil Rhodes]

 

 

Every time I've used LED lighting, I've been unimpressed

 

Why so?

 

You can lose the multi-source problem with LED by using diffusion - which is essentially all a remote phosphor device is doing (albeit diffusing the short-wavelength drive light before it hits the phosphor). Yes, there are inefficiencies in doing this, but it's the same situation all round.

 

Battery power is a really nice trick with LED. What sort of situations have you done that in?

 

P

[Keith Walters]

I've just taken delivery of some advanced production samples of "Quantum Dot" AC LED lights.

 

Just about all current white LED designs consist of a blue LED crystal with a Yttrium Aluminum Garnet (YAG) crystal glued on top. The YAG fluoresces yellow and when mixed with blue LED light produces the illusion of white light, but with some pretty savage spectrum holes.

 

These look just an old-fashioned incandescent lamp filament, but what they actually are long strips of blue LED chips soldered together, and covered with a "quantum dot" phosphor powder that's applied like paint.

 

I haven't been able to establish exactly what Quantum dots are, but they are apparently made by some sort of crystallization process. They can be made to produce just about any wavelength, and Sony currently use them in their latest LCD TVs to produce ultra-pure red, green and blue sources.

 

If they're made smaller than a certain size, they produce more or less true white light, and can be applied as a powder. The "filaments" are made of about 20 blue LEDs in series, which gives an operating voltage of about 70V. Two strips in series gives an operating voltage of 140V, which I presume allows them to operate directly off 110V AC mains. (Because they have to work from 240V mains in Australia and Europe, the samples we have use a miniature Switchmode Power supply to generate 140V).

 

A few experimental photos I've shot show that they do seem to have a considerably better CRI than conventional LED lamps. I'm going to try to do some comparison images and post them here.

[Adrian Sierkowski]

I'd be much appreciative to see those

[Adam Frisch FSF]

The new type of Phospor exciting LED's like the Cineo TruColor series is a huge step up. But even with the multi LED units we use most of the time, with a little diffusion nobody can tell it's multi point or not full spectrum. Hey, less spectrum might even be an advantage sometimes. It's a bit like the talk about megapixels vs film resolution when digital started. People will choose ease of use and low price over quality, as was evidenced in that battle. Ultimately LED is easier, cooler, draws less and it's good enough.

 

I'm a convert.

[David Mullen ASC]

What's missing are just very powerful and punchy LED's for the equivalent of a fire starter parcan or a 6K HMI with a spot lens.

 

Also some of the tricolor LED's can produce odd reflections in surfaces where you still see three different colors being reflected.

[David Mullen ASC]

I've also found that creating strips of LED's using things like lite ribbon is not quite the same as using a bare KIno tube because the little LED bulbs are quite directional. If you stick them behind a door edge to light someone, they don't spill both out and to the side as well as a fluorescent tube does.

[Keith Walters]

Also some of the tricolor LED's can produce odd reflections in surfaces where you still see three different colors being reflected.

That's exactly right. Your eye is quite tolerant of brightness variations across a TV screen, but it is exquisitely sensitive to colour shading errors (referred to in TV parlance as "White uniformity errors").

 

The spectral requirements for LCD backlight illumination are actually the exact opposite of those desired for scene illumination. You get the most vivid colour rendition on an LCD panel when the "white" light source actually consists of extremely narrow Red, Green and Blue "spikes".

 

To do this, higher end TVs sometimes do use a mixture of red, green and blue LEDs, but getting them to produce a uniform white is horrendously difficult. They need to be closely matched and perfectly aligned, otherwise you get annoying tinted patches.

 

Most LCD TVs just use conventional White LEDs.

 

However the "white" LEDs in the latest Sony panels use blue LEDs coated with quantum dot "paint" containing a translucent mixture of red and green quantum dot powder. This produces the desired "Spikey" white spectrum without any problems of directional shading.

 

It's interesting that with more or less the same construction but using quantum dots below 35nm, you get almost pure wide-spectrum white light. So even though both types of White LEDs will look the same, their modes of operation are fundamentally different.

 

Quantum Dot nanoparticles are set to be "The Next Big Thing". Not only do they improve Light Emitters, they actually improve the absorption capability of solar cells, and they can potentially be used to produce radically new dyes for paint or clothes. Bizarrely they're even showing potential as a cancer treatment. They can be made to be preferentially absorbed by tumours, which will then make the tumour susceptible to blasting with laser light of the exact quantum dot wavelength, while the surrounding tissue is not affected.

 

Here are some photos of the new type LED lights. These have the potential to produce the sort of "Punchy" illumination David is talking about, since the light emitting surfaces are constructed more like conventional filaments.

 

Here you can make out the individual LED elements

[Mark Kenfield]

Fascinating! I don't suppose the development of lighting technology has ever been as rapid and as exciting as it is at the moment. It's a great time to be making movies.

[Keith Walters]

Fascinating! I don't suppose the development of lighting technology has ever been as rapid and as exciting as it is at the moment. It's a great time to be making movies.

The Holy Grail for lighting is making Quantum Dots that are directly excited by an electric current, rather than indirectly by blue LEDs.

Potentially they could be made with even higher efficiency than LED lights, and have a spectrum that just covers visible wavelengths and nothing else.

It might also be possible to manufacture them in large evenly-illuminated sheets, so you could have interior lighting that closely resembles sky-scattered daylight.

[Guy Holt]

Let me give some very recent and random examples. I was at a NASA type facility where we … wanted a red warning light spilling down over a control panel in the background (by) securing a LED panel above it to some pipes, but that proved to be too weak after we gelled it. So the gaffer suggested hiding a 575W HMI above the pipes. Took forever to get the bigger light secured up there compared to a panel. By clipping gel to the doors, we now had a ton of spill everywhere, ruining the illusion this was a real red warning light. Cue gaffers trying to black wrap the whole lamp, but still small sliver of spillage coming out everywhere. Wrap more. Still spill. Now lamp overheats and shuts down. Remove blackwrap. We get it going again. Still too bright, stick scrims in there. They smoke. Now the gel has been upset and more white leaks, gel burned in the middle and colour is now pink in middle and uneven, back up again, lamp can't be pointed straight down or it overheats… Probably lost a good 20 minutes on that lamp alone.

 

The way forward is not to be found in new technology like the latest LEDs, but in valuing experience and respecting the craft of lighting. As a gaffer with 30 years under my belt, I recognize the delays Adam experienced as nothing to do with the limitations of older technology but the inexperience of the gaffer. Every one of the issues recounted by Adam that caused a delay were rookie mistakes, starting with the choice of the wrong light at the outset (an LED by the way), to hanging the right light in the wrong position, to not knowing how to control a light without sealing it in black wrap. The problem is not with older technology, the problem is that producers do not value the role of gaffers & grips in production enough to pay them a living wage so that they can mature in that capacity and support a family. The result is that you have inexperienced guys and gals filling these positions and making rookie mistakes like those described here.

 

The reason Adam can’t get ahead of his schedule is that he doesn’t have the support under him that he needs. For example, I did an HBO doc project with DP Dyanna Taylor where I practically never was on the same set with her. That’s because she insisted that the producer pay a decent rate so that she could have a gaffer with sufficient experience to light ahead of her so that when she arrived on the set with the Director it was 90% lit. In this fashion she was able to cover her day in a reasonable time and produce good production values (she is a 2x Emmy Winner). There is no replacement for experience and if there were a silver bullet, it is certainly not the latest generation of LEDs. IMO the output of LEDs is still too anemic, the color rendering still too poor, and the form factor still too limited to be of much use outside a few specific applications in which they excel (battery operated car rigs for instance.)

 

In the hype over the “promise” of LEDs, let’s not lose site of technological advances in lighting technology that are available today and contributing more to production today than “remote phosphor” or “Quantum Dot”, or “TIR” LEDs (Litepanel’s latest” ever will. I refer to the introduction of power factor correction in more and more lights, the introduction of the MAX reflector and new HMI power classes by ARRI, the introduction of Inverter generators by Honda, the introduction of other alternative lights sources like Light Emitting Plasma (LEPs) that in my opinion hold greater promise than LEDs.

 

The reason why, IMO, LEPs hold greater promise than LEDs is because they are a power factor corrected full spectrum daylight point source that is a natural for Fresnel lenses and will in the near future scale up to outputs comparable to 4-6 kw HMI Fresnels. Kino Flo is also incorporating power factor correction into more and more of their product line so that they draw half the power and generate much less harmonic noise than their previous models. What is also making a real difference today is the fact that because of this greater efficiency you can run all the lights needed for reasonably good production values on a night exterior on the 100A output of paralleled Honda EU6500s. Or, for day Interior/Exteriors you can run a light (the ARRIMAX M90) with the equivalent output and light quality of a 12kw HMI Fresnel also on paralleled Honda EU6500is Inverter generators (use this link for details). What you can say good bye to is diesel tow plants since everything else you need can be plugged into house power these days.

 

As great the promise of these other technologies, it will always come down to the experience and creativity of the individual handling the light. As Adam’s experience clearly illustrates, in the hands of a rookie, the latest and greatest lighting technology is not a guarantee of success.

 

Guy Holt, An “Ornery Old” Gaffer, ScreenLight & Grip, Lighting Rental & Sales in Boston.

[Michael LaVoie]

I have to second Guy's opinion on this. I am currently trying to negotiate the G&E requirements on my new feature and I'm running into resistance on the concept of a pre-light team. Prelighting means having enough lights and stands to send out to the next location while you're shooting the current scene. Admittedly not in everyone's budget but it is the way it's traditionally done and on a feature it's very important when there's company moves.

 

With the flexibility of Raw and the higher ISO's people seem fine to go out with no large lights and very little crew. It's only when the available daylight runs out and they have to wrap cause there are no HMI's that people realize their mistake when they need to suddenly add scenes to other days. LED's aren't the answer. It's production taking the whole idea of lighting seriously and budgeting for the time, crew and equipment that is required.

[Phil Rhodes]

The way forward is not to be found in new technology like the latest LEDs, but in valuing experience and respecting the craft of lighting.

 

Well, yes. But on the other hand, the practical reality is that nobody has the slightest respect for what I do on the odd occasion I do actually get to shoot things these days, and there's not a lot of point in trying to stand on ceremony in that circumstance. So, while I'm not particularly proposing LED lighting as a solution to the problem, the choice I often have is to work quickly or not work at all.

 

We are not all 30-year experienced gaffers on big US productions. In fact, only a very small minority of us are, and it isn't reasonable to take the position that everyone should work that way. It isn't possible.

 

P

[Adam Frisch FSF]

Guy, none of these were inexperienced. I was however in Philly so the crew was new to me and perhaps not as on it as they are in LA or NY. Whenever I work here in LA or NY I work with the most experienced crews there are. Top of their game. But. Sometimes (and more often than not) there isn't that time, money, locations and patience (new generation of directors) anymore. We can fight it, but ultimately we'll all be unemployed if we don't go with the times.

 

Another pet peeve of mine is that lighting hasn't caught up with the sensitivity of the new digital cameras yet. And what I mean by that is that lighting manufacturers are still stuck in "brighter is better" mode. I don't need brightness anymore, I need adaptability and customisability - and size. I'd much rather have a twice as big turnkey soft source that's ten times less bright, then the opposite. Hard light is fast, soft takes time.

 

Just setting a big bounced indirect light in a location, flagging all the spill, crating it with Lighttools etc etc is a 30min deal involving gaffers and grips. When all that was really needed was a bigger custom made light. Add 30 min here, 10 there, 25 there and all of a sudden you won't make your day and you don't get hired again. Don't even get me started on doing big soft boxes at night for car commercials etc - they literally take hours and hours to build. And then they're always too bright, too heavy, can barely be moved etc. When again, all that was needed was something simpler and less powerful in the same size.

 

It's not only a huge problem from a time perspective, but what ends up happening is that because these things take so much manpower, time and resources, the producer leans on you and ultimately veto's you and now both your creative work and the film suffers. In today's naturalistic cinematography, mainly shot on locations, we can't still treat it like a build on a set where there's tons of room, tons of juice and we have a full swing gang. I think it's been 2 years ago since I last shot on a built set on a stage.

 

Everyone benefits from simplicity.

[Guy Holt]

I'd much rather have a twice as big turnkey soft source that's ten times less bright, then the opposite. Hard light is fast, soft takes time. Just setting a big bounced indirect light in a location, flagging all the spill, crating it with Lighttools etc etc is a 30min deal involving gaffers and grips. When all that was really needed was a bigger custom made light... Don't even get me started on doing big soft boxes at night for car commercials etc - they literally take hours and hours to build. And then they're always too bright, too heavy, can barely be moved etc. When again, all that was needed was something simpler and less powerful in the same size.

 

Then it sounds like you would like the Mactech 960 Sled. It is basically a 4x4 box with 24 – 4’ 40W LED “Tubes.” The tubes are like the ones now being manufactured to replace T-12 Fluorescent tubes in overhead fluorescent fixtures, but at specific color temperatures (3200 & 5600K.)

 

As you can see from these stills from the show “Chicago Fire”, you can set one up on a Roadrunner stand and wheel it around – no more pushing lights through a 4x frame of diffusion. The best thing about the Mactech 96 is that it draws only 9 Amps, so you can run them off putt-putt generators.

 

 

In fact this entire set of “Chicago Fire” could probably be powered by the combined 100A output of a couple of Honda EU6500s operating in parallel. But, like any soft source they still a Lighttools grid and still need to be boxed in with flags to keep them from spilling all over the place. And, as you can also see from these production stills from “The Secret Life of Walter Mitty” below, it still takes the same number of grips and electrics to rig them on a condor and they are just as cumbersome to move around in this application.

 

 

My point being that it is not necessarily the fact that a light is LED vs. Tungsten that makes it faster to use. Rather the manner in which LEDs are typically used (run and gun) is a faster style of production.

 

 

As these pictures clearly illustrate, when they are used in the same manner as other light sources to obtain similar production values (say a soft top light over a large area) they take just as long to rig.

 

Guy Holt, Gaffer, ScreenLight & Grip, Lighting Rental & Sales in Boston

[Adam Frisch FSF]

I saw the Mactech LedSled at Cinegear and was very impressed. There was also some guy that had built a Dino-type light with LED MR16 bulbs. It was huge and it had a tremendous punch. I could light up a wall 40ft away in bright daylight. Not that I normally need that power, but it's certainly an alternative for a night light far away since it's a bigger source than a big fresnel.

 

But Guy, the Mactechs you pictured are kind of not as cumbersome to rig. What would the alternative have been to getting those three wide sources on a crane traditionally on Walter Mitty? 3x6K HMI's with a Chimera in front? Probably not, but with a 4x8 soft frame rigged in front of them or each with individual 4x4" frames of diffusion perhaps? Both those options take longer to rig, not to mention running the amperage to run such big units.

[Freya Black]

Hey, less spectrum might even be an advantage sometimes. It's a bit like the talk about megapixels vs film resolution when digital started. People will choose ease of use and low price over quality, as was evidenced in that battle. Ultimately LED is easier, cooler, draws less and it's good enough.

 

I'm a convert.

 

I think there will be a lot of changes like this over the coming years.

I think that things will evolve down the new paths and over time will get somewhat better as the technology matures, however essentially "cheaper" and to a lesser amount "more practical" will drive everything. The rules that held for camera technology will also apply to lighting. Everyone is very aware that most people won't be able to notice the difference and if they are able to notice the difference, won't care. Unless it is a field where they are especially trying to put out a feeling of quality then the compromises will be made and continue to be made and given that even advertisting is now shot on rather nasty video cameras I'm not hopeful.

 

There will be a race to the bottom with a few things at the top and no middle ground.

 

Freya

Edited September 5, 2014 by Freya Black

[Chris Albert]

I work in news magazine and I'd love to move away from the Joker 200/400/800s I use, the ballasts go down, bulbs break and crap out, don't strike etc etc. However there's nothing (I think) at the moment that can work as hard sources, use in S/M Chimeras, pancake/chinese lanterns etc.

 

Is there anything out there that can do the job of a Joker Bug style HMI? I wish someone would make a bug LED/Plasma.

[Mark Kenfield]

Plasma certainly seems like the next step for solidstate point-source lighting, with LEDs (preferably remote phosphor ones) covering everything else.

 

Problem is there's only the one 276w plasma bulb currently available (or at least available for film/tv fixtures. once they can get 500w, 1000w (and hopefully at some point, 2000w) bulbs going, then we're in for a real treat... imagine a 4k HMI equivalent fixture that you could plug into a wall or run off a tiny suitcase generator? That would really revolutionise things.

[Phil Rhodes]

Is there anything out there that can do the job of a Joker Bug style HMI?

 

Sort of, but not really. The most powerful single LED junctions commercially available right now are a few watts each, and anything larger is made of clusters, so it's hard to make a small enough source that it's really a hard light.

 

There are theatrical profiles (er, ellipsoidals, projector lights) which do vaguely similar things at similar power levels, but the cooling requirements mean that they're physically quite bulky and often use forced-air cooling.

 

We're not quite there yet. The industry is keenly aware that it's wanted and it will happen at some point.

 

There actually are kilowatt-level LED clusters but they're intended for warehouse lighting and don't have the colour precision or packaging style for film and TV work. Soon, soon...

 

P

[John Miguel King]

The Rotolight Anova is absolute fantastic.

 

Very high CRI, runs off an Anton Bauer Dionic for hours, it can be controlled from a phone or ipad, and it can be the first link on a wireless DMX universe. The very app can even take a colour temp reading, and it's also a DMX sequencer!

 

I used two of them yesterday on a 100% exterior shoot running around London under pouring rain. I rigged a 12x12 tarpaulin to keep the talent dry, backlit it with two of these at 4000 K. It looked like summer and I didn't need a genny. Just two batteries!

I call this awesome.

What is more, if you call them with any questions, the guy guiding you is the very own inventor. Word on the street here is that all the rental companies that bought into Litepanels are deeply regretful for doing so.

Edited October 14, 2014 by John Miguel King