LED Light Options

[nyvz]

I'm constantly on the lookout for new lighting technology that will change the way I work, and for the most part LED seems like it has a lot of potential, but almost all LED lights designed for film production seem to be pretty seriously underpowered or overpriced or both. For example there are the many <50W panels that are marketed saying they have 500W-1000W of incandescent output but really that is comparing to a light heavily diffused and wasted in a softbox, and these <50W LEDs at best have only the output of a 150W-300W fresnel at the same beam angle. They are useful when you want to run lights on battery or need daylight balance, but they usually lack in color rendering and some are still >$1k despite barely outshining a 300W fresnel while rendering much worse shadows...

Then there are a few bigger LEDs like the new ARRI's and Creamsource's which seem to put out about as much light as a 1K, but they are in the $3k-5k range...

One thing I've been looking at lately are RGB LED products such as the inexpensive ones used in LED ribbons which are pretty interesting because you can get 72W of SMD RGB LEDs that are small enough to mount anywhere on a ribbon for only about $20 for a 72w 5 meter kit with remote controller. These are cool in that I mounted them on a coroplast panel that weights nothing tape them anywhere and run on a cheap 12v li-ion battery for hours and can put out any color you want and controlled remotely. On the other hand their light output is quite underwhelming, partly because the "72W" strip only runs around 40W at the rated 12V, because it uses resistors to adjust voltage which just waste a fair percentage of the power, and have no lenses so they have none of the punch LEDs are known for with their commonly quite narrow lenses. These are pretty cool but are probably better suited for installations in sets, though I am tempted to try to build a big flexible blanket light from the 4 I bought

Those ribbons got me interested in RGB LEDs since I was surprised how good they could look on camera despite being very cheap LEDs. I expected narrow spectrum R G B leds being combined would render odd colors since clearly parts of the spectrum must be missing, but then they seemed quite good perhaps because cameras can only see R G and B through similar filters anyway. This made me wonder if there are cheap, bright, small LED fixtures out there that could compete with 400w-1.2k HMIs or more.

That leads me to what I really wanted this thread to be about, though, is that there seem to be lights marketed as DJ lights that are designed to replace PARs, that pack many bright RGB and RGBW (+white/daylight) or RGBA (+amber/tungsten) LEDs in one, inexpensive, DMX unit, and I've seen ones online that are anywhere from 3W to 720W. The largest ones I have seen that are easily available from american sellers at reasonable prices that are probably no bigger than a 300W fresnel are $250 18x10W units that pull about 180W and with photometrics that seem to show them competing tungsten fresnels of about 650w and should compete with ARRI's LED L7-C Fresnel lights at 1/10th the price since it pulls a similar amount of power for its LED emitter. Looks like there are some LED PARs that also have adjustable zoom (aka beam angle) though those might cost more. There are also units that use 24 10W LEDs and then panels that use as many as 72 10W RGBW LEDS that might really start to compete with HMIs.

Personally I wouldn't mind a non-focusable LED as long as it came in a very narrow beam angle, since it's much easier to diffuse a narrow spot light to spread the beam than it is to intensify and narrow the beam of a flood light.

Anyone have any experience with these or other relatively inexpensive LED lights that are worth looking into?

 

[Moseph]

I wouldn't really trust a mixture of red, green, and blue LED's to create real white. It's possible, but to me it sounds like you got lucky with the ribbon that you got. Were you testing it out on a person? They, as well as the LED PAR cans, seem like they would be great for lighting sets, but I would be concerned about skin tones. Also, I've seen the PAR cans flicker pretty severely when dimmed (through their own DMX control, not caused by a TRIAC).

MacTech has a great product, I've used their 960 extensively. It's 24 4' LED T12 retrofit tubes, so it's a little bigger and heavier than a maxi brute. Small or inexpensive? Definitely NOT. But they do make smaller versions and they're surprisingly color accurate for LEDs. No ballast or even any power supply required, you can literally make a harness for the tubes out of zip cord and add-a-taps that plugs right into the wall.

 

[Norman99]

"This made me wonder if there are cheap, bright, small LED fixtures out there that could compete with 400w-1.2k HMIs or more."
There are but they are not cheap, as they are in the custom made to order prototypes. Also consider that the beam angle is that all important factor that many manufacturers seem to shy away from when it comes to specs. Sure, a narrow , almost useless beam angle makes for a pretty bright lux figure approaching and even exceeding a 1.2K HMI if the led optic calls for that. I have a flashlight with one led that throws a 4 degree beam 500 yards, but if that light is diffused to say, a 30 degree beam, it would not be capable to properly light up a subject 10 feet away. The best I have meddled with in my designs is a 12 degree spot using 12 Cree XM-L2's on the Ledil regina optics. That gives out an incredible lux rating of 22K at a distance of 24 feet. I can always diffuse it wit a Macrolon Lumen XL panel. Best part is battery operated from a liion 14.4V V mount. It surely does surpass a1.2K hmi, but nobody sells it. Just to show what leds are capable of, oh, by the way those 12 leds are in 90 CRI 3000 Kelvin.

 

[clint lealos]

I know I sound like a broken record on here, but those remote phosphor LED lights really impressed me... Super bright, single source, high CRi rating (90+), in a small battery powered package. They cost around $2K though, so I can't say they are cheap. I hopeful that will change soon enough. Here are a couple links:

http://www.cineolighting.com/node/37
http://brothers-sons.dk/area-48-soft/

Both these lights were so strong you couldn't look directly art them even when diffused, which isn't really the right word. I saw them both in person at NAB this year and was quite impressed. Good luck with your search!

 

[nyvz]

When I compare LEDs to tungsten or any other light, I always compare at the same beam angle, though even that is difficult given different beam shapes and different ways of measuring or locating the beam edge, but certainly better than manufacturers who often compare narrow LEDs to broad tungsten lights...

Even ARRI's fancy new 160W LED fresne is huge and expensive and barely replaces a 300-650W fresnel...

I just read an article about Zero Dark Thirty that mentioned they extensively used Creamsource's LED panels which the DP equated to nearly a 1.2k HMI and the manufacturer compares to an 800W HMI. So someone is making big LEDs, too bad they are $4k+

I do want to test RGB LEDs more in person, they sure seem like they offer a lot of flexibility since you can put out any color at the press of a button. I haven't dont really extensive tests but plan to do some more soon, though I was also impressed with the color rendering out of a newer zylight unit I used briefly that seemed to use RGB LEDs.

That Mactech LED looks pretty awesome though I would wonder if it really offers anything two Image 85s stacked wouldnt? I can imagine its probably also $4k+ and probably far too large and heavy and inflexible for me to ever justify owning... Also seems kind of odd to build a unit around LED strips built into T12 retrofit bulbs, that must add a lot of unnecessary bulk, no?

I'd be more interested in a small light with a lot of punch since I'd rather a small light that is easy to carry or store that could be bounced or diffused to turn it into a large softlight. I'd prefer a light that could be used as a hard narrow sun source for interiors and could be diffused into a big softlight when needed. I get obsessed with having versatile lights since I prefer to own as little as possible.

I guess it could be a while before any LEDs that compete with 1.2k HMIs end up in the <$500 range, though I imagine at least a light that might compete with a 1K openface and is naturally daylight balanced could go a long way.

It's true that dimming quality could be a big issue with cheap RGB LEDs. The strips I have have no flicker even at shutters beyond 1/1000th with all colors at full, but as soon as you adjust (dim) any color they start to flicker at most shutter speeds other than 1/100th and 1/50th sec. This is common in many cheap LEDs as I understand it and is caused by the cheap LED controller that came with the LED strips probably dim using 1/100th sec PWM... I imagine I could avoid this by using a higher quality controller with >1k/sec PWM. I noticed some of the RGB LED PARs I've seen have specs of >7.8khz PWM, flicker free so that looks promising and I should probably only look at lights that are marketed as high PWM or flicker free.

I looked at those remote phosphor LEDs and they do look impressive in their color rendering and quality of soft light. On the other hand, besides being very even at the source, they seem rather high priced and may only provide marginal benefit compared to a fluorescent lighting unit of a similar size... You are stuck with the color temp of the phosphor panel you put in, and they are quite expensive and about the same size as ziplight or biax fluorescent fixture. Aren't there already enough 2'x1' soft lights. It probably has a higher output than most similar sized lights, but the overall difference doesn't seem worth the price to me. I don't doubt it is to someone though. I tend to be most interested in small bright units with controllable beams and large soft units that make it easy to get big soft lights. I don't do a lot of studio lighting, and I need to stay very portable so I don't care so much about medium sized broad soft lights that are too big to put in a softbox, too small to make a large soft key out of, and too broad to get much punch out of.

 

[TheDingo]

The Zacuto Plazma-light looks pretty amazing and will sell around $1,300 when it's out.

 

[nyvz]

The Zacuto Plazma-light looks pretty amazing and will sell around $1,300 when it's out.

Weren't they supposed to ship a year ago? They showed them off at NAB 2012 and they never came out... also microplasma panels aren't nearly as power efficient as LED (microplasma is ~30-40lumens/watt vs 60-100lumens/watt for LED). Seems pretty expensive considering the company they get their plasma panels from (Eden Park, I think) lists 1'x1' plasma panel lights as about $100-200 (presumably for finished light fixture and not in bulk from the way it's presented on the table on their website).

 

[TheDingo]

Weren't they supposed to ship a year ago?

Zacuto showed prototypes at NAB 2012, which was thicker and clunkier looking than what they had at NAB this year.

also microplasma panels aren't nearly as power efficient as LED (microplasma is ~30-40lumens/watt vs 60-100lumens/watt for LED)

Zacuto claims their light is significantly brighter than an LitePanel 1x1, especially once you factor in the diffusion which the LitePanel needs to be as soft as the Zacuto light.

Seems pretty expensive considering the company they get their plasma panels from (Eden Park, I think) lists 1'x1' plasma panel lights as about $100-200 (presumably for finished light fixture and not in bulk from the way it's presented on the table on their website).

If they are indeed the same light ( I have big doubts about this ), then what's stopping anyone from coming out with their own light at half the price of the Zacuto?

I think Zacuto is taking their time to get things right, and so far it looks like they are on the right track. I can't wait to read a proper technical review of an actual shipping product, especially in terms of brightness and CRI performance.

 

[nyvz]

Zacuto claims their light is significantly brighter than an LitePanel 1x1, especially once you factor in the diffusion which the LitePanel needs to be as soft as the Zacuto light.

I'm sure it does put out more light than a diffused 1x1, but with a 1x1 you can take off the diffusion and get more punch than you would have from a plasma light if you need it. Plus there is a huge market of LED panels that outperform litepanels 1x1s in various ways (especially price). I was looking at buying some CN600's which are basically 1x1's from china for <$250. Why would I get the $1300 plazma when I can get similar LED unit that can put out more hard light or be diffused and put out about the same amount of light while drawing 1/3 as much power for 1/5 the price?

I don't really get Zacuto's table comparing their Plazma with 1x1's. Why does it say 1x1 lumen ouput *as measured* but not for their light? I would expect a 40W LED like a 1x1 to put out much closer to 4000lumens. And Zacuto claims 2200lumens? Seems awfully low, since I would expect that from about a 120W tungsten light and the specs say the Plazma draws about 140W which works out to about 14 lumens/watt for the daylight one. So it is no more efficient than halogen? Doesn't look right. It is normal for actual lumen readings out of a fixture to fall well below a light emmiter's rating, but you can't really mix measured and specified numbers in a comparison.

 

[TheDingo]

I'm sure it does put out more light than a diffused 1x1, but with a 1x1 you can take off the diffusion and get more punch than you would have from a plasma light if you need it.

Zacuto is claiming much higher light output compared to a LitePanel 1x1 with NO diffusion, so you may be several stops dimmer once you add diffusion.

Plus there is a huge market of LED panels that outperform litepanels 1x1s in various ways (especially price).

I just bought a few F&V 5600K LED lights, which are a great deal. They are bright, can be battery powered, and the CRI seems to be OK so far.

Why would I get the $1300 plazma when I can get similar LED unit that can put out more hard light or be diffused and put out about the same amount of light while drawing 1/3 as much power for 1/5 the price?

For me it will be mainly...

1- High CRI color ( better than 90 CRI, maybe as high as 95 CRI )

2- The big difference in diffused brightness. I have literally bolted the diffuser to the face of my cheapo F&V 1x1 lights, because I know I will never shoot without the diffuser on. ( LED light is extremely nasty without diffusion )

3- Proper CTO and CTB filter behavior. LED lights have such a mis-shaped spectral response, you can't just slap on a filter and expect it to behave the way it would with natural daylight or tungsten light. Try and filter most 5600K LED lights down to 3200K and you suddenly have to deal with a huge amount of green contamination, and your brightness has dropped more than 2 stops. Most bi-color LED lights are a mess because the performance of the 5600K and 3200K LED emitters is very different, so brightness fluxuates and color shifts start to show up.

4- I expect to be able to rent the Zacuto lights when I need more lights for larger jobs. ( I will own a small kit for basic corporate set-ups, and then rent for bigger things )

I don't really get Zacuto's table comparing their Plazma with 1x1's. Why does it say 1x1 lumen ouput *as measured* but not for their light?

I don't think the manufacturing spec on the Zacuto light is finished yet. Likely they are still finalizing the exact composition of their panel, so there's no point in publishing the exact spec until it's ready to ship, which sounds like 3-4 months from now.

I would expect a 40W LED like a 1x1 to put out much closer to 4000lumens. And Zacuto claims 2200lumens?

I think you will get close to 4,000 lumens with the SPOT version of a 1x1 LED light, which is not a nice even light compared to a plasma light. I expect to be able to shoot with the Zacuto light 2 feet from my subject and get a beautiful soft "window" like light spread, where I would have to diffuse the crap out of an LED light to use it the same way.

I expect the Zacuto lights to prove themselves when they start shipping, so there is not much point in comparing them to LED lights until real independent testing happens.

 

[Moseph]

That Mactech LED looks pretty awesome though I would wonder if it really offers anything two Image 85s stacked wouldnt? I can imagine its probably also $4k+ and probably far too large and heavy and inflexible for me to ever justify owning... Also seems kind of odd to build a unit around LED strips built into T12 retrofit bulbs, that must add a lot of unnecessary bulk, no?

The output from these LED tubes is significantly higher than Kino Flo T12 tubes while drawing less amps (40w per tube, like a regular output T12) as opposed to 75w. And all the bulk really is necessary in order to dissipate the heat. They don't radiate heat the same way tungsten lamps do but the back of the tubes do get extremely hot, the LED's would fail if there wasn't enough surface area for a heat sink.

Also, being in the same shape of a T12, you can buy the tubes alone and put them in a 4 bank Kino. You can literally buy the fixture shell alone (the reflector won't do you any good with these) and make a harness that plugs straight into the wall. At full power it will draw 160w as opposed to 600w like a traditional 4x4 Kino would.

 

[nyvz]

Zacuto is claiming much higher light output compared to a LitePanel 1x1 with NO diffusion, so you may be several stops dimmer once you add diffusion.

...

I think you will get close to 4,000 lumens with the SPOT version of a 1x1 LED light, which is not a nice even light compared to a plasma light. I expect to be able to shoot with the Zacuto light 2 feet from my subject and get a beautiful soft "window" like light spread, where I would have to diffuse the crap out of an LED light to use it the same way.

Are they? sounds unlikely given the specs they list. Even if 1x1's somehow only put out 1200 lumens, 1200 lumens emitted at a 50degree flood beam will still illuminate a subject at the center of the beam brighter than a 2200 lumen microplasma light at a 120+ degree beam...

why would a spot give out so many more lumens? Isn't it just the same emitters with narrower lenses? The lumen output should be nearly the same unless there is much change in the efficiency of the lenses. Are you thinking of lux?

For me it will be mainly...

1- High CRI color ( better than 90 CRI, maybe as high as 95 CRI )

2- The big difference in diffused brightness. I have literally bolted the diffuser to the face of my cheapo F&V 1x1 lights, because I know I will never shoot without the diffuser on. ( LED light is extremely nasty without diffusion )

3- Proper CTO and CTB filter behavior. LED lights have such a mis-shaped spectral response, you can't just slap on a filter and expect it to behave the way it would with natural daylight or tungsten light. Try and filter most 5600K LED lights down to 3200K and you suddenly have to deal with a huge amount of green contamination, and your brightness has dropped more than 2 stops. Most bi-color LED lights are a mess because the performance of the 5600K and 3200K LED emitters is very different, so brightness fluxuates and color shifts start to show up.

4- I expect to be able to rent the Zacuto lights when I need more lights for larger jobs. ( I will own a small kit for basic corporate set-ups, and then rent for bigger things )


I don't think the manufacturing spec on the Zacuto light is finished yet. Likely they are still finalizing the exact composition of their panel, so there's no point in publishing the exact spec until it's ready to ship, which sounds like 3-4 months from now.

...

I expect the Zacuto lights to prove themselves when they start shipping, so there is not much point in comparing them to LED lights until real independent testing happens.

True, I'm sure its not worth judging these products before there is more information on them. Has Zacuto released the spectral output of their lights? Hard to say how they will perform with gels, hopefully it will be better. The Eden Park microplasma whitepaper lists their CRI around 80-85 compared to LED at 80-87 (who knows where they get that since I've seen LEDs claiming from 60-95 CRI), but it doesn't sound all that promising.

 

[clint lealos]

I looked at those remote phosphor LEDs and they do look impressive in their color rendering and quality of soft light. On the other hand, besides being very even at the source, they seem rather high priced and may only provide marginal benefit compared to a fluorescent lighting unit of a similar size... You are stuck with the color temp of the phosphor panel you put in, and they are quite expensive and about the same size as ziplight or biax fluorescent fixture. Aren't there already enough 2'x1' soft lights. It probably has a higher output than most similar sized lights, but the overall difference doesn't seem worth the price to me. I don't doubt it is to someone though. I tend to be most interested in small bright units with controllable beams and large soft units that make it easy to get big soft lights. I don't do a lot of studio lighting, and I need to stay very portable so I don't care so much about medium sized broad soft lights that are too big to put in a softbox, too small to make a large soft key out of, and too broad to get much punch out of.

You should take a closer look. The Trucolor LS is 9.5" x 11", weighs less than 5lbs., with 248 lux @ 3 meters, and can be powered by a camera battery (same with Area 48). Pretty sure you can't do that with a ziplight and no fluorescent can compare with that kind of output in such a small package. It sounds like the perfect light for what you do. It is also dimmable with zero color shift. You can put a soft box too and easily replace a 1K tungsten lamp. I hear you on the price though. There is a group buy on Reduser right now that can you save you 25% on the Area 48 light. It ends up around $1600. Also, if you like those Mac LED lights you should check out FilmGear lights. Their LED tubes look exactly like Kino tubes, so no multiple source shadows, but still very bright.

 

[nyvz]

Yeah, I've seen the group buy and the Trucolor, but I'm not really in the market for a $1500 light especially when it may be worth $200 in a couple years when the now rare remote phosphor technology is more common. Also, doesn't the Trucolor end up with pretty similar specs to something right in between a Kino Barfly 200 and 400 light? The barfly looks to be about half the price, right? Obviously they are different in many ways, but they seem fairly comparable.

 

[TheDingo]

why would a spot give out so many more lumens? Isn't it just the same emitters with narrower lenses?

It's proportional to the area you are lighting, just like it would be with a tungsten fresnel light, except that the beam angle is fixed.

For example, an ARRI T1 ( 1000 watt fresnel ) light at 10 feet and a beam angle of 8 degrees has an illumination value of 1634 foot-candles.

Spread the beam out to 54 degrees and the illumination value drops to 194 foot-candles, but it now covers a much larger area.

Has Zacuto released the spectral output of their lights? Hard to say how they will perform with gels, hopefully it will be better.

In one of the NAB articles I devoured over the past week, one Zacuto rep stated that their new light will have a "normal" CTO/CTB filtration response, unlike the current LED lights.

I don't expect the Zacuto lights will be perfect, but it sounds like they are moving in the right direction, and at a price that is more affordable than the alternatives.

...All I am looking for is a portable daylight balanced soft light with good color that I can use when shooting outside or with window light.

 

[nyvz]

It's proportional to the area you are lighting, just like it would be with a tungsten fresnel light, except that the beam angle is fixed.

For example, an ARRI T1 ( 1000 watt fresnel ) light at 10 feet and a beam angle of 8 degrees has an illumination value of 1634 foot-candles.

Spread the beam out to 54 degrees and the illumination value drops to 194 foot-candles, but it now covers a much larger area.

Yes sounds like you are mixing up luminous flux and illuminance. Or you seem to be talking about illumination, but your previous statement I commented about was talking about luminous flux (You said a 1x1 spot would have 4000lumens compared to 1x1 flood's 1200lumens or so, which I took issue with). Yes illumination (lux or lumens/m^2) does change depending on whether a focusable lighting fixture is spotted or flooded, but its luminous flux (lumens) does not change much because luminous flux of a light system refers to the total amount of light it puts out. As you focus a light into spot, it theoretically pushes the same amount of light into a smaller space so it has a narrower, brighter beam, but the total amount of light is approximately the same, minus changes in efficiency of the reflector/lens when operated at different positions. As you focus a light, the luminous flux of the bulb or emmitter certainly doesnt change, since it does not know what the lens/reflector is doing.

In one of the NAB articles I devoured over the past week, one Zacuto rep stated that their new light will have a "normal" CTO/CTB filtration response, unlike the current LED lights.

I don't expect the Zacuto lights will be perfect, but it sounds like they are moving in the right direction, and at a price that is more affordable than the alternatives.

...All I am looking for is a portable daylight balanced soft light with good color that I can use when shooting outside or with window light.

Sounds interesting, though I'm sure you can tell I'm skeptical. It is possible Zacuto has developed their own microplasma tech, though its much more likely they buy it from someone else who actually already makes such things. By the nature of the fact that plasma is not 100 CRI and probably does not exactly replicate spectral output of tungsten or daylight, I would expect that plasma would be subject to some similar issues with gels as LEDs. Plasma may have fewer color spikes which will likely help, but then I've never heard any LED manufacturer admit their lights don't work properly with gel filters either

Also I hardly imagine a 2200 lumen lighting fixture with a 120degree beam angle will be of any use outdoors. If these numbers are accurate, they can be used to approximate illumination, and at 1m that light would have an illumination of 700lux, and even closer at 0.5m it would have an illumination around 2800lux. By comparison, outdoor daylight is about 10,000-25,000 in the shade and 32,000-130,000lux in direct sun (from wikipedia), so the Plazma would barely read at all even in the shade a foot or two from a subjects face. And that is generous since the 120degree beam angle is based on 50% illumination and in reality, a super soft light like that gives off its light at nearly a full 180degrees.

In my mind, we don't need softer LED/Plasma sources, we need punchier ones. They are so underpowered compared to tungsten and HMI options, that their real benefit is in their reflector/lens efficiency. A big part of LED's benefit is in their very efficient fixed lenses which really help push all the emmitter's light into a tight beam to compensate for the overall low luminous flux of the emitters. That's part of the reason manufacturers can even consider claiming their units put out more light than a tungsten unit of 10-20x the wattage when the LED emmitters have at best 4-5x the luminous efficiency of tungsten.

 

[Norman99]

OP
"This made me wonder if there are cheap, bright, small LED fixtures out there that could compete with 400w-1.2k HMIs or more."

The 1X1 panels certainly are not cheap, small and definitely don't approach 400-1.2K HMI's.
furthermore the Zacuto plasma needs a power supply ballast of some sort to operate it, I believe. Making it a larger unit.
How many lux @ 3ft ( diffused) are you looking for and how small is small for you ?

 

[TheDingo]

In my mind, we don't need softer LED/Plasma sources, we need punchier ones

The Zylight F8 LED Fresnel looks interesting, but I'm not keen on paying $2,500 per light. ( same goes for the ARRI L7 LED Fresnel )

Perhaps if LED technology improves this price will eventually be cut in half and then it's affordable to a much wider audience.

 

[Mitch Gross]

If you want focusing, punchy LED-based lighting fixtures, then I suggest you look at the new LED fresnels from Mole. They just debuted at NAB. These lights use Quantum Dot technology for their LED emitters, yielding a very high CRI. That plus the control of a classic Mole fresnel makes them very attractive. Available in tungsten or daylight, in Tweenie, Baby and Junior sizes. They even sell retrofit kits to convert old tungsten fixtures over to LED. The current shipping units are DMX but later they will have non-DMX versions for less money.

http://blog.abelcine.com/2013/04/11/nab-2013-mole-quantum-dot-led-fresnels/

 

[TheDingo]

Mole QD LED Fresnel Info