Description image

Why Hollywood Will Never Look the Same Again on Film: LEDs Hit the Streets of LA & NY

cleantechnica led street lighting lights lamps sodium vapor mercury clean green la los angelesAfter Michael Mann set out to direct Collateral, the story’s setting moved from New York to Los Angeles. This decision was in part motivated by the unique visual presence of the city — especially the way it looked at night. Mann shot a majority of the film in HD (this was 2004), feeling the format better captured the city’s night lighting. Even the film’s protagonist taxi needed a custom coat to pick up different sheens depending on the type of artificial lighting the cab passed beneath. That city, at least as it appears in Collateral and countless other films, will never be the same again. L.A. has made a vast change-over to LED street lights, with New York City not far behind. Read on for why Hollywood will never look the same again — on film or otherwise.

Here’s the trailer for Collateral if you need a visual refresher:

Mann chose to shoot HD because of how the format rendered the story’s setting. Considering that Collateral takes place over the course of a single night, its portrayal of LA’s nocturnal landscape is integral to the film. Due to the city’s recent retrofit of over 140,000 street lights, that nocturnal landscape has changed forever.


Environmentally speaking, this is a good thing — though it’s easy to get a little nostalgic to put the implications of this retrofit into perspective. In a sense, every night exterior LA-shot film previous to this change is rendered a sort of anthropological artifact, an historical document of obsolete urban infrastructure.

Justin Gerdes has written a ton for Forbes about LED lighting and its benefits to cities big and little, including the following in September: “On June 18, Mayor Antonio Villaraigosa announced [PDF] the completion of the first phase of the project, with LED fixtures installed on 141,089 street lights.” He goes on to say (my own emphasis):

The City of Los Angeles estimates it will see at least $7 million in electricity savings and $2.5 million in avoided maintenance costs annually with the switch to LED street lights. Street lighting can account for up to 40% of a city’s electricity bill, according to Eric Woods, writing at the Navigant Research blog. The LED fixtures used in Los Angeles, which include Cree’s XSP series and LEDway series, Hadco’s RX series, and Leotek’s GC series, consume about 63% less electricity, and last much longer, than the high-pressure sodium (HPS) fixtures they replaced.

The ecological and economic benefits are rather astounding — but the transition to LEDs has another unique effect, and one that’s specific to filmmakers. Let’s first take a look at what LA has ‘lost’ before we check out what it’s gained, and why it could make a difference on camera.

What Makes Artificial Lighting ‘Realistic’?

Simple_spectroscope prism cri color rendering index full spectrum white light artificial lighting source tungsten incandescent versus fluorescent

The interesting thing about non-tungsten artificial light sources is that they often produce a non-continuous or incomplete spectral output. This can affect the appearance of certain colors under that output. More simply, you can’t really put colors back in that weren’t there to begin with, even by gelling such a light source or color correcting in post. At left: emitted spectrum of incandescent versus compact fluorescent lamp, courtesy of Wikipedia user Timwether. Notice the non-continual spectral output emitted by the latter (bottom).

Color Rendering Index or CRI attempts to rate a lamp’s ability to approximate an ideal, continuous spectrum source. 100 is the highest possible CRI score, though some types of artificial lighting — ahem, many traditional street lights — can actually score in the negative numbers. For filmmakers, CRI is a yardstick of limited usefulness because a simple number such as “85″ can’t explicitly tell you which points along the spectral output may suffer, or even how many such points that output may have. On the other hand, a spectral power distribution graph of sufficiently high-resolution will tell you pretty much everything you need to know about a lamp’s spectrum, but information may not be readily available.

It isn’t necessarily simple, easy, or inexpensive to get non-tungsten lighting technologies to perform at the level filmmakers want and need. This is why HMI and Kino Flo fixtures do a very good job and don’t come at a low price. That being the case, it’s not hard to imagine that old-school street lights may not hold up so well by comparison — for better or worse.

The Light That Was

There are two very common types of artificial lighting sources used for street lighting: sodium-vapor and mercury-vapor lamps. If you’ve ever lived anywhere that gets lit up at night, you know sodium-vapor lamps. Aside from being the key to an absolutely fascinating old school film compositing process, sodium-vapor  is one of the most common types of street lighting. The distinctive orange glow cast by this type of light is efficient, functional, and cost-effective  – or at least was in 1933. Interestingly, low pressure sodium-vapor lamps cast a virtually monochromatic spectrum of light. Apparently this can have several beneficial effects on night driving, but is downright hilarious to see plotted for spectral power distribution:

Light Wavelength Chart

High pressure sodium lamps look less silly when compared by chart, but still don’t hold a candle (I went there) to the spectral power distribution of “full body” light sources such as tungsten lamps or sunlight. They’re also common enough for filmmakers to need lighting gels which help mimic their distinctive output (see also the Collateral writeup in American Cinematographer). Mercury-vapor lamps, on the other hand, produce better color rendition than either of their more efficient sodium brethren, though their CRI is still poor. Being related to our beloved fluorescents, their cast is generally a ‘cool white’ with a blue-green dominance.

Understandably, none of these guys are ideal for the reproduction of natural (never mind pleasant) skin tones. As such, sodium and mercury-vapor lamps are mostly relegated to streets, parking lots, industrial locations, and the like, but in turn give such locations their starkly unique, artificially lit recognizability. At least, until now.

The Light That Will Be

To see how it all affects ‘the big picture’ both past and present, check out the incredibly striking before and after below (photo courtesy Los Angeles Bureau of Street Lighting):

cleantechnica led street lighting lights lamps sodium vapor mercury clean green la los angeles

According to the Forbes article, the same residential LED fixture going for $432 a piece in 2009 went for $245 by the end of 2012, with its output boosted from 42 to 81 lumens per watt and a life expectancy improvement from 80,000 to 150,000 hours. If I’m doing my math right, that’s about 17 years of life if the lamps were running constantly.

los angeles la led street light conversion complete

How much of an upgrade these LEDs will be over their sodium or mercury-vapor ancestors in terms of performance will likely vary depending on the fixture. Whether it’s an upgrade at all, I think, is a more clear-cut matter: in terms of color rendition, the above comparison speaks for itself. The answer is: yes. The LEDs should very well prove a benefit to existing-light photography — better for the environment, and in nearly every case, better for cinematography. Whether the new look is visually preferable is a subjective matter, but you probably won’t find as many proponents for the old one.

At left is a geographical representation of the program’s progress, courtesy the LA Bureau of Street Lighting (click for full PDF). It’s visually apparent that filmmakers exposing by the existing street lighting of Los Angeles will be picking up very different imagery, even in many of the same locations, as they might have before. It’s also interesting to consider how Collateral might look if it were shot today (or tonight, I should say), advancements in digital acquisition aside.



We’re all here for the same reason: to better ourselves as writers, directors, cinematographers, producers, photographers... whatever our creative pursuit. Criticism is valuable as long as it is constructive, but personal attacks are grounds for deletion; you don't have to agree with us to learn something. We’re all here to help each other, so thank you for adding to the conversation!

Description image 128 COMMENTS

  • I think it’s pretty exciting! Can always make it yellow in post, but like they said, colors missing from the spectrum can’t be added back in.

  • Don’t forget about the dreaded flicker effect from LED lighting. If ever shooting with these LEDs I would go from 48 shutter speed to 30 because the flicker is noticeable at 48 and even more so the higher the shutter speed is.

    • Clark Nikolai on 02.2.14 @ 5:03PM

      The flicker is only on the cheap LED lighting. Hopefully they’ll use better quality ones that don’t flicker.

    • I tested the LED streetlights in West Seattle by scanning the lamp at a low shutter speed. There was no flicker at all. The previous sodium lamps flashed at half frequency as fluorescent, probably 60 Hz. I could see it in a 1/6 second exposure of moving cars front lit by a service station with a bit of flare from a street lamp.

      It depends on the power fed to the diodes. Cheap 120V Xmas lights that cut the voltage by placing several diodes in series look like they flash at 60 Hz. (I tried to feed them DC and at 60V they were unexcited. It may be that peak to peak 170V might be what matters instead of 120V RMS.)

      Lifetime on LEDs depend as much on the other components. If the LED lasts years but the power supply lasts months maintenance will be high. The household LEDs have cheap noisy switched power supplies, test with an AM radio tuned off station. Your mics or cables might pick it up.

      Also the light from fluorescent varies depending on quality. Hallways and stairwells are likely to get the economy bulbs. Office space might have better and art space better yet. A cheap spectroscope will show this. It sure was disappointing to look at my fluorescent back lit workstation monitor with it. Oh well, it works as good as the others.

  • Larry Vaughn on 02.1.14 @ 11:50PM

    What does shooting in HD rather than SD have to do with color rendition?

  • Larry Vaughn on 02.1.14 @ 11:52PM

    Residential LED fixture going for $432 a piece in 2009 went for $245 by the end of 2012. Who pays $245 for a residential LED anything?

    • I think residential is meant for the (public) area of the light to be used, not in actual homes.

  • Larry Vaughn on 02.1.14 @ 11:58PM

    What does shooting in HD rather than SD have to do with color rendition?

    Or was the author talking about HD vs film?

  • Paul Stephen Edwards on 02.2.14 @ 12:09AM

    I’ve shot under LA’s LED lighting numerous times. No flicker.

    • but if you d shoot slowmotion you would get a flicker, no? because you also have to adjust the shutterspeed and angle accordingly.

      • Paul Stephen Edwards on 02.2.14 @ 1:35PM

        I’m not a slo-mo guy. :)

      • Emiliano Ranzani on 02.3.14 @ 10:02AM

        Shooting slow-motion with available light at night?

      • In the slow motion cinematography world LED’s have actually becoming a saving grace. At full output, something like a 1×1 LitePanel does no show any flicker at variable high frame rates. Sodium vapor lights do flicker something terrible, so from the slow motion stance, this might actually be a good thing.

        • Isn’t that because LED panels are designed to work with batteries, so they have to convert AC to DC? Street lights would be connected directly to AC, so it may be different.

    • LEDs themselves, are inherently flicker-free at any frame rate (speed). HOWEVER, it is the inexpensive power supplies that make them ‘flicker’ at high frame rates. I am certain our city used the cheapest foreign-produced power supplies they could find… ?

  • I know some people will feel nostalgic about the old light. But the new will have it’s own special feel. Colors are more distinct, everything isn’t just different shades of yellow-brown, and details are more visible. With higher K’s in cameras also bringing out more detail in low light this opens up new possibilities. The glass is half full, and with campaign. Smile! :-)

    I’m not really much of a believer that governments are changing over to these lights because of the environment. It’s about the huge savings. But it is good that this particular money saving is getting mercury light bulbs permanently out of use. I always thought it was telling that the twisty straw florescent bulbs that were supposed to be saving the environment, the bulbs that came to be known as Al Gore light bulbs, had to be handled in special ways when disposing of them because they had very hazardous levels of mercury—they were saving the environment with those bulbs….? Got it.

  • I lived in one of the neighborhoods that switched over. It took some getting used to. The LED’s are around 5000k I think, very odd seeing cool bluish light flooded on streets. Camera wise it looked pretty great, but because the LEDs are set up in multi source arrays the shadows that tree branches and other things are pretty crazy!

  • The article is in error. You can always color correct and add the so-called sodium vapor look back in. Clean white lighting is the easiest thing to degrade in post. It’s much harder to take OUT the sodium vapor look if you wanted the night scene to look clean.

    I think that whole look was ugly and overrated anyway. I think there’s far more interesting things cinematographers can do with night looks, and the color temperature of the city lights is the least of their worries. This is a meaningless article written by a non-technical person who didn’t bother to get a pro DP to comment on the color temperature of sodium vapor lights (2700 degrees K) vs. LED lights (roughly 5500 degrees K). It ain’t a big deal.

    • You’re right. It’s aesthetics and I think the new look that we will be seeing is just as good. Nice, clean light. It’s usually a treat to watch a night street video lit with white light in the first place.

    • This new look to city streets will make it easy for them to teal and orangeify everything, the background is already cool; just light the actors with tungsten, set the white balance to about 4K and you’re set – orange skin and blue/teal lights.

    • Where did he claim you couldn’t put the sodium vapour look in through colour correction ? I can only find the part where he mentions you can’t put colours back in which weren’t captured in the first place, which isn’t the same thing. I think you’re confusing his assertion that LA’s streets will no longer have the sodium vapour look right out of the box (which he is saying) with the assertion that you’ll no longer be able to get the sodium vapour look at all (which he doesn’t actually say).

      • Actually, you CAN put different colors back in during post. It all depends on the time available and the skill of the colorist and VFX people involved. We radically change stuff all the time during color correction, particularly when retakes and pickups are done weeks or months after principal photography has wrapped, even at different locations. It’s done far more often than you think.

        The original article claims that LA will lose the look because the lighting has changed. It hasn’t, in that you can get almost any kind of look that the DP wants. If the original author had bothered to contact a DP and have them comment on the piece, it would’ve added much more authority to the content. This is what happens when non-technical people try to tackle technical areas that involve artistic judgement and taste.

        • I won’t weigh in on putting colours back in post because I’m not qualified but as a reader, again, I’m not sure the article’s saying what you think it is, or at least seem to think it is. What I took from it, and what I meant by “out of the box”, was that LA’s nighttime streets are now going to look different from the previous sodium-vapour look – to the eye. So if you’re trying to reproduce the “look” of LA’s streets (as Mann was in Collateral), then you’re not going to try for the sodium look anymore, with any technique – gels, grading or otherwise – since it no longer reflects the visual reality of the city. And that, at least, isn’t something you’d need to consult a DP about, obviously. I can’t speak for the author, of course, so I could be wrong – Dave, if you’re reading this, feel free to jump in and clarify.

        • Dave Kendricken on 02.3.14 @ 9:24PM

          Hey Vidiot, here are words from David Mullen of the ASC discussing the part of my post with which you seem to be taking issue:

          “Sodium Vapors have missing wavelengths so they cannot be fully corrected — all you end up with is a muddy, desaturated image if you try. Plus if you try and white balance to a color temp much lower than 3200K, you are already pushing the blue channel quite heavily. At the most, you could try just adding a little coldness to take out some orange. But white balancing under the sodium lights isn’t really going to work well, I’d try putting some 1/4 CTO on a tungsten light pointed at a white card and white balancing to that — that would add a little blue into the sodium color but not cancel it completely.

          As for gels, there are lots of combinations people try. Rosco just created two new gels for this purpose: Industrial Vapor, which is an ugly brown-green gel that supposedly matches real sodium, and Urban Vapor, which is an orangey-yellow gel that looks like most people’s impression of sodium vapor without as much green. But you could also try various gel combinations, like Full CTS + 1/4 Plus Green, Apricot, etc.”

          Here’s the source of this quote:

          My point was not that LA could not be color corrected or lit to match a certain look, but that its natural ungraded look has changed, which it has. The aesthetic implications of this are subjective. That and incomplete-spectrum lamps behave differently than natural or full-spectrum sources.

          • You misunderstand R. David Mullen ASC’s message. He’s talking about the difficulty of color balancing under Sodium Vapor lights. He’s 100% correct. I’m saying, if you had to, you could *easily* take today’s white LED lights and make them yellow in color correction. Just pull a highlight key and add yellow. Easy and trivial to do in color correction. We could recreate the Sodium Vapor look in seconds.

            The article implies that we’ve *lost* something by the changeover to white LEDs. My point is that the new lights are actually better and more natural. We can always degrade streetlights and make them uglier in post; what we can’t do is take ugly lights and make them look *good* in post. The article implies that the change is bad; I argue that the change is good.

            Mullen is a very, very good DP and was robbed when he didn’t get nominated for the much-slammed SMASH last year. Extremely well-lit show. Content… not so good. Lighting was great.

  • Ah, but how will these new lights effect the hookers on Hollywood Boulevard?

    • More natural flesh tones should ensure higher revenue streams and save on make up to ‘white balance’…

      • You guys have won the internets today.

      • Actually, warmer tones are better for hookers, since warmer colours helps to increase blood pressure, increase the pulse rate and increase the physical sense acuteness.

        This is why the sex industry use a lot of red lights, as this helps them to excite their clientèle and thus get more business.

        Put this to a test yourself: have your partner and you have an intimate encounter in a room with blue or cold light, see how off-putting that would be.

  • SydneyBlue120d on 02.2.14 @ 3:41AM

    If You want to deeg deeper into LED color lighting fidelity, You can check this paper from OSRAM:
    Light and Color Methods of Achieving High CRI with LEDs

    • Thanks for the link! Do you have more beginner explanations which don’t go too deep into scientific terms (as far as not explained)?

      • SydneyBlue120d on 02.3.14 @ 12:23PM

        I do not have any other documents beside that, sorry.

      • ok, so light is measured in 2 main areas: color temperature and Color Rendering Index (CRI). The color temperature is what the color of the light is from reddish to yellow, white, blue. The CRI is how well you can differenciate colors. HPS has low color temperature (yellow) and low CRI (can’t differentiate colors well in the light). LEDs have low to high color temp (yellow to white to blue) and usually high CRI (even the yellow colors).

    • This is an interesting and informative read but I think it primarily deals with how human eyes and brains perceive color, not how digital sensors perceive color. As the paper points out, CRI is not a very good judge alone of the color fidelity of a light source for human vision. As it relates to sensors, it’s an even poorer indicator, from what I understand of the AMPAS solid state lighting project:

      More meaningful and relevant measurement tests are being devised for solid state light sources.

  • Not only do I find that the new street lighting is not as pictural as sodium, but also, filmmaking considerations put aside, 5600K daylight lighting at night is bad for our biological clocks, and it seems like nobody ever gave a thought about this issue at the Bureau of street lighting.

    • nothing beats falling asleep while drunk driving under sodium vapour lights.

    • You’re right. Blue light stimulates our brain and keeps us awake for longer. Maybe it’s linked to the fact that ambient daylight light is blue or something.

  • I agree with most people on this thread. Sodium is a pain in the neck and Post Productions let us choose weather we want it yellowish or not. For the gentleman talking about flicker. He may istake LED with CFL light bulbs who are actually flicking. And even some of them are flicker free now :

    • Daniel Mimura on 02.11.14 @ 7:22PM

      LED flickers and so does CFL. Even the kinoflo CFL’s are for regular frame rate only.

  • What they need is some of those bluetooth led lights, then they can change the colour to whatever they like ;-)

  • I’ve been trying to find before and after videos of the lighting but it’s been impossible. I can’t even find videos of the current LED lighting. Can anyone point me in the right direction?

  • I miss the gas lamps. Damn that Edison!

  • marklondon on 02.2.14 @ 1:58PM

    This and the green bike lanes have vexed production in LA the last year or so. The bike lanes have been repainted slightly less flouro green but we still spend a lot of time painting them out in post. Dear LA: if you want to know why we have to shoot 90% of car ads in Vancouver, thats a BIG reason.
    Have been on a couple sets where the street lighting was turned off and ‘replaced’ with tungsten or lower ‘practicals’.
    On my low budget stuff I’ve had no problems at 24fps with the new LEDs. Haven’t noticed anything at 60 either but that may be because I will have probably lit that with a fixture.

    • Yeah, it’s funny that bike riders that pay zero in registration fees take up so much money and space on the streets.

      • Note enabler on 02.2.14 @ 10:34PM

        Los Angeles, home of the worst traffic in the country, doesn’t benefit from bicyclists?

      • Yes, it’s those bike riders’ fault we have a lot of traffic and no new freeways. It isn’t as though the taxes those bikers pay could be enough for new freeways especially since billions are constantly thrown away on feel-good policies and programs to appease hysterical masses. Every tax dollar should primarily go towards new streets/freeways, and if anything/whatever is left should go to other things.

  • Connerkward on 02.2.14 @ 2:52PM

    In my city, the streets are beginning to change from a very unique looking orange to a blue LED look. I can’t decide which is better… Orange seems like horror lighting, the white light of the LEDS seems to remind me of the “blue night” films of the 80′s.

  • they shot collateral on a Viper. part of that was because the camera was one of the first HD cams with really good low light performance that let the camera see the ambient light far better than conventional film at ISO500. it also had 4:4: output as another factor.

    Not sure if 800 was around when they where shooting, or that you could push process a 1/2-1 stop at the increase of contrast.

    as for LEDs @5000K. if you want the warm, though maybe not pure yellow look of So Vapor lights, you can do the reverse cheat. gel your FG lights at 5600K with CTB to get 8500K or so. set your WB to that and the LED’s will go warm. I’ve done the reverse as well. I did a nite ext. shoot w/o any HMI’s but wanted the cool blue look. I also didn’t have a ton of lights or power to be putting up 5K’s with CTO. what I did instead was go straight 3200K tungsten, but then cheat the white balance down to 2200K, as low as the camera would go. This turned the 3200K pleasing cool. Flashlights in the scene came in as neutral so it worked perfect and I got max light from the 1K’s I did have up. so you need to understand WB – its relative and you can push things around once you learn to separate FG from BG light.

  • Nice article. Thanks.

  • I used to think that the streets there were paved with gold … now I realize it was all lighting.

  • Thank god. I’ve always thought sodium vapor light looks disgusting

  • I’m myself am looking forward to the switchover to LEDs for night shooting. I realize there is an entire, younger generation that will wax nostalgic about the yellow sodium vapor look, but growing up in both the neighborhoods and films of the 1970′s thru mid-80′s, for me this is a wonderful return to “the good ‘ole days”. The mercury vapor streetlights that were everywhere when I was a kid was reflected in the coolish blue nighttimes of HALLOWEEN, ESCAPE FROM NY, WARRIORS, THE TERMINATOR, etc. The similarity between LEDs and mercury vapor lights is a welcome change. I’m looking forward to taking advantage of nighttime available light for that “film noir in color” that a cool night lighting scheme creates…

  • Oh gee, what is creative Color Grading?

  • What a bizarre article? Cities look the way they do at night because of a compromise. Not because anyone CHOSE to use light sources that were not full spectrum. It’s what we’ve grown accustomed to, but clinging to it for subjective reasons is pure nostalgia.
    Yes, in time it will make film, video and still images immediately dated. Just like film grain and scratches, the shredding of video without a global shutter, and the effect of 3:2 pulldown on broadcast video. Asking whether we should keep the old compromises for the sake of continuity is simply sentimental foot-dragging.
    If you want to make retro movies, make sure in the future you carry some sodium vapor tinted gel to wrap the LED street lighting…

  • As a video camera rental house for the pass 5 years, LED is here to stay. Indeed looking at the power savings and duration, we will see more places using LEDs. Years later LED will look realistic and tungsten unrealistic because we will live in a world lit by LED. Good that you highlighted this early.

  • Hank Roberts on 02.3.14 @ 2:11PM

    Nobody cares that the American Medical Association warned several years ago about adding this amount of blue light to nighttime lighting. Just another Cassandra, those people.

    American Medical Association:
    ?Adverse Health Effects of Nighttime Lighting” — “adopted policy recognizing that exposure to excessive light at night can disrupt sleep, exacerbate sleep disorders and cause unsafe driving conditions. The policy also supports the need for developing lighting technologies that minimize circadian disruption and encourages further research on the risks and benefits of occupational and environmental exposure to light at night.”

    Research sources:
    Am J Prev Med. 2013 Sep;45(3):343-6. doi: 10.1016/j.amepre.2013.04.011.
    Adverse health effects of nighttime lighting: comments on american medical association policy statement.
    Stevens RG, Brainard GC, Blask DE, Lockley SW, Motta ME.

    The American Medical Association House of Delegates in June of 2012 adopted a policy statement on nighttime lighting and human health. This major policy statement summarizes the scientific evidence that nighttime electric light can disrupt circadian rhythms in humans and documents the rapidly advancing understanding from basic science of how disruption of circadian rhythmicity affects aspects of physiology with direct links to human health, such as cell cycle regulation, DNA damage response, and metabolism. The human evidence is also accumulating, with the strongest epidemiologic support for a link of circadian disruption from light at night to breast cancer. There are practical implications of the basic and epidemiologic science in the form of advancing lighting technologies that better accommodate human circadian rhythmicity.

    More at:
    Prof. Abraham Haim</a

    Dangers of exposure to “white” light
    Exposure to light at night: link with obesity?
    Light at night and cancer
    Seasonality and seasons out of time
    Exposure to artificial light at night and prostate cancer

    Hey, who cares?

    • Use a tungsten gel in your bedroom … otherwise, I’d prefer not to have the motorists falling asleep while navigating the city streets.

  • Spend $141M to save $7M. Smart.

  • Does anyone know how the incandenscent ban will affect lighting production in general? Are specialty lights an exception or should I run out and buy all the 500w floods I can find??????

  • Dave – The emphasis that you’ve chosen in this article is rather poor. It causes the casual reader to believe that LED lighting will have a poorer color quality than the outgoing HPS lighting. As I’m sure you know, that couldn’t be further from the truth.

    In every respect the LED lighting is a closer representation of natural daylight. Unlike how this article reads, LED lighting provides a fully continuous spectrum (it is those nasty mercury-vapor based, and sodium based lamps that are discontinuous). Additionally, the higher Correlated Color Temperature of the LEDs more closely mimics that of natural daylight.

    So, yes, while the “nostalgic appearance” of HPS lighting will be lost, it will be going the same way as the oil and fuel burning streetlights of the 19th century. Except, in this case, replaced by something that not only is more efficient and more reliable, but something that actually has better and more natural color rendition!

    Take a look at this comparison between the incandescent-tungsten based lamp, natural daylight and an LED. The LED is actually a much closer simulation of the daylight color content, is it not?

    I hope I’m not coming off to harshly, as I do appreciate the “look” that you are discussing in this article. I just wish that the presentation of this vastly improved technology (LED streetlights) was done in a more careful and holistic manner. Its not just the environmental and economical reasons that set LEDs apart, it is their color characteristics as well!

    Disclaimer: I previously worked for one of the manufacturers involved in the upfit of the City of LA, and I am currently in the business of helping others “see the light” (yes, I went there too).

    • Dave Kendricken on 02.3.14 @ 9:33PM

      Hey Kyle, it was not my intention to to imply that LEDs have a poorer color quality than sodium-vapor lighting. I felt that my inclusion of low pressure sodium’s very limited SPD — as well as a link to a high pressure sodium SPD chart and other comments in the conclusion — would illustrate this sufficiently. I have made some corrections to help emphasize the point a bit better.

      • Dave, your article absolutely did not imply that LED would have poorer colour quality than the HPS lighting. From your explanations of the CRIs to the included before and after shot and your remarks upon it, I cannot for the life of me understand how anyone could have come away with that impression (by the way, I sighed heavily when I saw you felt compelled to add, “The answer is: yes” to the statement “in terms of color rendition, the above comparison speaks for itself.”). It seems to me that the article has come under attack by commenters who didn’t read it very carefully or simply applied their own meaning to it wholesale without much regard for what was actually written down. So, to balance the scales: it was a good article, deftly-written, and thank you for it.

  • What’s the view on the future of plasma lights for the high end/pro level production? (I know, it’s not what this post is about entirely but I am being curious)

    • long way out. LEDs are what’s up right now. Costs are constantly going down, efficiency is going up. plasma is seriously out of control expensive… as in unaffordable for anyone.

  • Maximus Moretta on 02.3.14 @ 11:25PM

    Just put a sodium vapor filter on it. ;-) people are so afraid of change.

  • Thanks for the article. It looks according to the PDF map that the downtown area has not been converted yet. Do you think that is intentional and at least partially due to the fact that LA Studios and others want to shoot there and maintain the Sodium-vapor look?

  • At least where I live the LED retrofits aren’t living up to their advertised rated life. In just a few short years many streets here are being converted back to high Pressure Sodium and Metal Halide as the LEDs fail. Typically in mid summer and mid winter is when the most LED fixtures fail. Temperature has no bearing on the performance of HID lighting such as HPS

  • This is great for the city and of course it will prompt new studies in lighting techniques and theory. Of course I think movie makers should be more concerned with making good movies first.

  • I warned people about this years ago. You can’t make an efficient led that has broad spectrum, because even though you can, it gets its efficiency by matching peak efficiencies in the human eye and leaving the other frequencies alone. Though you might be able to get the city to swap out the light modules for such broad light, or strap on such lights temporarily). There will be shifts in colors, but what is there is authentic anyway, as seen to film authentically.

    Collateral was not authentic anyway, they shot much with the electro luminance panels of some sort that I also was looking at using for devices too at the time, in the car, that is the strange glow. They were used as backlights on some sorts of LCDs and watches. Still nice look, but not authentic.

    • “Authenticity” has nothing to do with it. As an inherently manipulative medium, film — and digital — are nothing more than another method of telling stories. The authenticity comes from the truth of the story being told, not the lighting or any other production values, which are there to serve the narrative. The lighting in “Collateral” worked very well for that particular story, but future stories will work just as well under the glow of LED lighting.

      The story itself is the cake — the lighting (and I say this as someone who has spent 35 years in set-lighting) — is just the frosting.

  • I agree with some comments that this article is misleading in a way that will make the reader believe LED is a poor alternative.

    LED has dramatically improved recently, and you can find many different types.Some of them at 5000k with a CRI <95 on all spectrums. This improves quite extensively the outcomes of any shooting.

    Should the wanted end result by washed out images, this is easily achieved in post rather than the contrary.

    Finally, any big production set will easily lit up by their own means or temporarily change individual lights of an entire street.

  • Daniel Bryant on 02.10.14 @ 7:03AM

    Sodium lights are old school? Yeah right. Take a look at “Terminator”. Notice any orange lights? How about “Blade Runner”. Any there? That’s no studio lighting— the orange lights aren’t there because they weren’t there at all. High pressure sodium lights weren’t even installed until the late 80s, and weren’t commonplace until the 90s.

    The *real* old school street lights ran on mercury vapor, and had a green-blue hue (and the color corrected bulbs were cool white). *This* is the type of lighting you see in classic movies set in LA, as they were used from the 1940s to the late 1980s (and some are still in use today); the large majority of the colour film era.

    Night having an amber hue is a recent phenomenon that’s only been around for about 15 years or so. Either the author is a 15 year-old, or he doesn’t pay attention to this sort of thing.

    Bemoan it all you want, but the modern LED streetlights are closer in colour to the original mercury-vapor lights that lit up the sky at night, and we can finally close the book on the orange sky that’s plagued us for the last 15 years.

    Induction lights are even closer to the classic mercury vapor look (and cheaper, and just as energy efficient). I wish more cities were adopting induction lamps, but LEDs have that “sexiness” that induction doesn’t seem to have for the “green scene”. A real disappointment for those of us who enjoy the night sky, because the induction lamps have lower light pollution than LEDs. Pomona, CA has installed the induction fixtures, but I don’t think much of anyone else in the Southern California region has.

  • Daniel Bryant on 02.10.14 @ 7:16AM

    Bad research defines a lot of this article.

    For a continuation of my comments above, the lights that were invented in the ’30s were low-pressure sodium lights, and are in fact, still more efficient than any other light source out there. The problem is, the hue is a monochromatic yellow that makes it impossible to differentiate colours, and much of the light produced is “scotopic”, meaning it only activates the rods in your eye, and not the cones, which are responsible for most of our vision.

    The second error in this article mentions a negative CRI. It’s impossible to have a negative CRI. CRI is on a scale of 0 to 100. Low pressure sodium lights have a CRI of 0, meaning it is not possible to differentiate colours at all.

    Despite their invention, low-pressure sodium lights remained a niche product until the 1970s when the energy crisis forced many cities to install the low-pressure sodium lights to save on energy costs. However, since the light was not very effective, most cities switched back to the mercury vapour lighting they had been using before.

    In the mid to late 80s, high pressure sodium, which was developed in the 70s, became affordable for cities to install, and by the late 90s, most fixtures has been converted to high pressure sodium. These are the amber lights that this article refers to, and they haven’t been around for anywhere near as long as this article infers. High pressure sodium is different from low pressure sodium in that it has a higher CRI, meaning that it is possible to differentiate certain colours under it.

    Really dude, if you’re going to write an article, get your stuff straight!

      • Daniel Bryant on 02.19.14 @ 8:29AM

        “Although the CRI definition allows negative numbers they are often rounded up in literature to zero.”

        So, yes, you are correct. I apologize. The equation for calculating CRI can produce a negative number, but generally the scale is considered to be from 0 to 100, but since the formula used to derive CRI does not technically end at zero, it’s possible to get a negative number. I’m sorry about that; I’ve never actually read a negative CRI before since most round up to zero.

        However, I still suggest you take a look into mercury-vapour lighting and its effects on film. MV was around until the early 90s in most areas, and is responsible for some really wonderful night scenes in movies. Check out films like Taxi Driver, the Terminator, Blade Runner, etc. Those use the ambient mercury vapour lighting from streetlights.

        Mainly my dismay was that HPS is getting so much love here when it’s only been around for a very short time, while mercury vapour lights have lit some of the greatest scenes in film history. The light it provides is both sublime and a bit eerie at the same time. I think the loss of MV on our streets was really the greatest loss to cinematography, and it doesn’t get anywhere near the recognition it deserves, despite many cinematographers attempting to emulate the MV look with tungsten filters.

        I feel like those attached to tungsten filters are really missing out on what made the classic movie scenes look the way they did simply because they assume the visual effect was achieved using filters, when in fact it was all real.

        I’d love to see an article on here about mercury vapour. Maybe it would inspire more cinematographers to get that classic look in real life rather than using a color-correcting filter.

        • Daniel Bryant on 02.26.14 @ 9:21PM

          the lack of reply illustrates why i chose to go on the offensive. nobody mourns the loss of mercury vapour street lighting. what a shame, and a loss to the film industry.

          • I miss the MV night scenes like we really saw in the 80′s. The “Terminator” is indeed a good example. I also remember lots of T.J. Hooker night scenes shot in L.A. and Hollywood taking advantage of the mercury vapor lighting. It is unfair the bad rap MV lighting got because of lack of maintenance. These lamps did not “burn out” like the incandescents, or “cycle” like a HPS lamp at the end of lifespan so many of them were producing low light levels while consuming the same amount of energy. I witnessed changed out MV luminaries still containing 20+ year old original lamps still working, but dimmed out due to overuse and lack of re-lamping on time (8 years). LED luminaries do however produce white light, but also produce a lot of glare and still have high failure rate requiring complete $400 fixture replacement vs. $40 lamp.

  • On Hawaii Island, in Hawaii, we will continue to use the narrow spectrum yellowish lights, not sure if they have LEDs in that spectrum. The reason is to prevent light pollution. Our island is the premiere site for astronomical viewing with many of the world’s most power telescopes here. By using narrow spectrum light the astronomers can just put a cut filter on that specific wavelength and get a clear view (would be clearer without the filter but hey). So for us it’s not about saving energy with sodium vapor, it’s always been about light pollution. We’ve been dealing with this nasty available light for awhile now and will continue.

Description image44 pingbacks