LED vs CFL for photographic lighting

[Mainecoonmaniac]

This is an interesting video. Very applicable for photographers.

https://youtu.be/EoAZ-u6hn6g

 

[Gerald C Koch]

An interesting idea to use a CD as an impromptu diffraction grating.

 

[Mainecoonmaniac]

I love it

An interesting idea to use a CD as an impromptu diffraction grating.

Call it McGiver or Ghetto. It's brilliant!

 

[wildbill]

Matthias is a genius.

 

[Old-N-Feeble]

Wow... that is interesting. I guess I need to replace all my CFLs with LEDs.

 

[DREW WILEY]

Hope you don't have color film in mind unless you think Halloween transpires all year long. The quicker LED technology gets fully up to speed and makes CFL's outright extinct, the better. Shove those ghoulish bulbs back in the coffin with Count Dracula.

 

[M Carter]

I think CFLs pretty much bit it for photography a while back (though you do see those eBay kits with piles of CFLs crammed into a soft box). Biax flos and long-tube stuff by Kino (and the knockoffs) are pretty mature tech now and have been go-tos for motion work for some time now. I've been pretty happy with my biax stuff (other than finding cases for 'em).

You can get a dual Biax for well under $200 nowadays - some of them may need some mild gelling to get right. But many current LEDs do as well.

LEDs are coming along nicely though, when you need a soft source, and we're seeing some interesting replacement "globes" for fresnels (but around $800 to replace one bulb… got a ways to go!) They only interest me for battery-only scenarios, but at this rate I'll likely get one or two for interview soft lights. Soft light is easy, and I find it pretty dull - I get more excited about HMI or HID fresnels and lensed pars.

For B&W film, I still prefer packs and heads. From mega power for large format product to shooting digital wide open, I can use all the same stuff for the most part. And there's a lot of good affordable monolights that are Bowens mount, so tons of modifiers for them, if someone can't afford a pile of Speedo gear. And I really love my tungsten stuff.

LEDs still lack the bang-for-the-buck for me I guess. All the young kids on the filmmaking forums are buying 'em like they're "the latest thing", but they're pretty limited and wouldn't be my first purchase.

 

[snikulin]

A good and educational video.
I especially liked this frame :

 

[Mr Bill]

Something I learned long ago, working for a large studio chain outfit, is that the energy-efficient fluorescent lamps, including CFL, are not anything you want to deal with in color photography. Some 20-some years ago, we found that we had problems with our color-correction booths (these are walled-in, neutral colored booths where test prints are evaluated for color and adjustments are called out).

We had a handful of lamps of lamps we tried out, including some nominal 5000 K units. Here was a surprising thing - after looking at an eco-friendly set of lamps, a "proper" set of lamps looked very dim, so much so that it seemed like something was wrong with them. After looking at spectral data, it was pretty clear that the eco-friendly stuff was so because it put most of its energy in the spectral ranges where human vision is most sensitive. But the tradeoff is that everything outside of those ranges doesn't look quite like it does under "natural" light, with a fuller spectral makeup. The problem is worse with the narrower-peak print dyes that give stronger colors, aka "wider gamut."

Anyway, my rule is that if it has the green eco-friendly band on it, don't even think about using it for either color photography or print viewing. I think there are problems with leds, also, with respect to good color, but for video or screen viewing, most people seem satisfied with them. But the makers really ought to find a phosphor to fill in the spectral range between blue and green; they'd be terrific then.

 

[M Carter]

Something I learned long ago, working for a large studio chain outfit, is that the energy-efficient fluorescent lamps, including CFL, are not anything you want to deal with in color photography. Some 20-some years ago, we found that we had problems with our color-correction booths (these are walled-in, neutral colored booths where test prints are evaluated for color and adjustments are called out).

We had a handful of lamps of lamps we tried out, including some nominal 5000 K units. Here was a surprising thing - after looking at an eco-friendly set of lamps, a "proper" set of lamps looked very dim, so much so that it seemed like something was wrong with them. After looking at spectral data, it was pretty clear that the eco-friendly stuff was so because it put most of its energy in the spectral ranges where human vision is most sensitive. But the tradeoff is that everything outside of those ranges doesn't look quite like it does under "natural" light, with a fuller spectral makeup. The problem is worse with the narrower-peak print dyes that give stronger colors, aka "wider gamut."

Anyway, my rule is that if it has the green eco-friendly band on it, don't even think about using it for either color photography or print viewing. I think there are problems with leds, also, with respect to good color, but for video or screen viewing, most people seem satisfied with them. But the makers really ought to find a phosphor to fill in the spectral range between blue and green; they'd be terrific then.

In the last 5 years or so, there has been a lot of progress on CRI (Color Rendering Index). The original Kinos, for instance, were regularly gelled with 1/8 minus green - now their "tru match" bulbs are considered better, but they lose some output since the phosphors and coatings have changed.

HMIs have been more reliably clean for decades, and good units give very high CRI light - but often at 6000 or 6500k. But those can be reliably corrected with 1/2 or 1/4 CTO (or just white balance to 6k and let the background go a little warm).

As far as whether something has a particular logo or band - all that really matters is CRI. And even in that case, there can be outliers. I have an 85 CRI HID setup that shoots just fine. I've seen LEDs listed as "95+" CRI that look all wrong.

I doubt you'll see much more work going into improving spiral bulb CRI for consumers (thought I think they should really warm those bulbs to match tungsten better - my wife hates how cold they look). And they just don't have the output for pro use - even the huge ones are pretty weak and fragile and hard to pack and hard to make a fixture for.

For pros, the development now seems to be focused on LED output and CRI (and the Chinese have been slowly improving their affordable HMI units). LED does have the potential to be the go-to daylight source, and there are new developments every year. They can get an amazing amount of lumens from a very few bulbs with very low heat and amp use, and we're really in the infancy of using LEDs for shooting. I imagine in the next 5 years they may replace the ubiquitous kinos and we'll likely see more hard-light and fresnel use as well (though fresnels and pars are nicely covered by HMI, the lower end of the pro market would likely prefer LED).

 

[Mainecoonmaniac]

When I shopped for lights for my professional kit, I bought strobes and a hot light kit. I used my strobes a lot for jobs, and hot lights for personal work. My strobes I haven't used in years, but I still use my Mole Richardson fresnels. Easy to use. What I see is what I get. When I shot chromes, I shot Fuji RTP. But LED lights for photographic applications is a real game changer. I saw HMIs at a photo convention one time. They're nice, but way too expensive. Also, a ballast is required. If the bulb needs replacing, it's also expensive. I don't know how close they are, but there are daylight LED bulbs for a fraction of the cost of HMIs. I would use CTB gels on hot lights if I needed continuous daylight. Now, I'd probably use LEDs.

 

[Mr Bill]

As far as whether something has a particular logo or band - all that really matters is CRI. And even in that case, there can be outliers. I have an 85 CRI HID setup that shoots just fine. I've seen LEDs listed as "95+" CRI that look all wrong.

I doubt you'll see much more work going into improving spiral bulb CRI for consumers (thought I think they should really warm those bulbs to match tungsten better - my wife hates how cold they look).

Hi, I used to think like you - that CRI was all that mattered. Long time ago, it probably was, but I think that manufacturers today know how to "game" the system. I came to this conclusion 10 or 20 years ago, after measuring spectral output of various fluorescent tubes that had fairly high CRI. After graphing the output, I thought, there's no way the CRI can be that high - but it was. I think they must be picking and balancing phosphors for the best score on the specific test objects used for CRI ratings.

Better rating systems should really be put to be use. But until then, color photographers should avoid eco-friendly fluorescents, regardless of how high the CRI is, because it's a rigged game. I don't think fluorescents are going to be improved for photographers, because having a fuller spectrum means that visually, the light output is lower, so what consumers are gonna be buying 'em?

"White" LEDs are much better in this respect, even if the CRI is not as good, because they have a more complete spectrum. But they still have problems - look at published graphs and notice the dip between blue and green. This is the main problem, and really good color probably won't happen until they find a phosphor (or some other method) to fill in that gap.

 

[DREW WILEY]

HMI lighting was horrendously expensive. CRI does not tell it all, becuase that is relative to how the eye allegedly sees something, not necessarily film. And any claim that some CFL ever gets over 75 CRI has a massive BS coefficient built into it for marketing purposes. You
get what you pay for. I did equip our pro color matcher here at work with special very expensive 98 CRI German fluorescent tubes made specifically for that kind of use; but even those won't prevent metamerism all the time. I've used the same bulbs inside a special lightbox
I designed for copying 8x10 chromes with a copystand. It did quite well, but certainly could never be perfect. The is nothing "e" about some
of this alleged e-technology. CFL's are marketing to last thousands of hours but often burn out in a week, contain toxic ingredients, and
give me migraines. I can't even be in the same room with them, and I'm certainly not unique in that respect. So I'm rooting for LED advances. Or we could just go back to tungsten bulbs for technical purposes.

 

[wildbill]

When I shopped for lights for my professional kit, I bought strobes and a hot light kit. I used my strobes a lot for jobs, and hot lights for personal work. My strobes I haven't used in years, but I still use my Mole Richardson fresnels. Easy to use. What I see is what I get. When I shot chromes, I shot Fuji RTP. But LED lights for photographic applications is a real game changer. I saw HMIs at a photo convention one time. They're nice, but way too expensive. Also, a ballast is required. If the bulb needs replacing, it's also expensive. I don't know how close they are, but there are daylight LED bulbs for a fraction of the cost of HMIs. I would use CTB gels on hot lights if I needed continuous daylight. Now, I'd probably use LEDs.

HMI's are $$$ but aside from carbon arcs, there's nothing that can do what they do. LED's have come a long way (including LED retrofits for your Mole fresnels) but they don't pack the punch of an HMI and certainly aren't a replacement for an 18K or 24K unit. It's just getting to the point where the color issues aren't there.
Putting CTB on a tungsten fixture results in way to much light loss for most applications.




There's a reason we have not/won't started using cfl's for motion picture work, they're garbage.

 

[Mainecoonmaniac]

Here's an interesting new product

HMI's are $$$ but aside from carbon arcs, there's nothing that can do what they do. LED's have come a long way (including LED retrofits for your Mole fresnels) but they don't pack the punch of an HMI and certainly aren't a replacement for an 18K or 24K unit. It's just getting to the point where the color issues aren't there.
Putting CTB on a tungsten fixture results in way to much light loss for most applications.




There's a reason we have not/won't started using cfl's for motion picture work, they're garbage.

This LED light consumes as much as my 200W Mini Mole but has an equivalent of a 2000W hot light. It also dims probably without a shift in color temperature. You can also change the color temp from tungsten to daylight also. Pretty amazing. Expensive now, but will be cheaper later I'm sure.

http://www.bhphotovideo.com/bnh/con...3a0ejAR-lZRoCjy_w_wcB&is=REG&m=Y&Q=&A=details

 

[M Carter]

Hey, I have low CRI stuff that renders beautifully, in testing and real-world. I've had high CRI stuff that looked like ass.

And yes, LEDs are a fraction of a cost of HMIs, if you compare a soft-light panel price with a hard-light HMI. But so far, to get an LED fixture that's equal output and look as a 575 HMI par or fresnel, you have to spend much more than the HMI. That will certainly change, and it will be awesome if LEDs can replace a 1.2k par or fresnel HMI for the same or lower cost. (1.2k being the end of the sweet spot for HMI as you have to get into dryer drops and generator trucks past that. But fixtures like Joker 400s and 800s are really great outfits, and stupid-expensive.)

CTB on a tungsten fresnel isn't much of a solution, either - you're cutting two thirds of the light's output. And still pulling the same amps and generating the same amount of heat. (1/2 CTB on a tungsten used for rim or hair looks very good in many color situations though).

For still shooting, this is usually a bit moot and academic - I don't need much kick for still shoots, and if I need to fight the sun I can mix strobes with fixtures. Doing things where you want motion blur hasn't required a ton of light for me either and I shot film stills for years with 650 fresnels and open faced units. Tungsten is usually 100 CRI after all.

Video and motion work is where I'm more hungry for photons, and where HMI and HID hard lights become tougher to come by.

I did post my DIY 150 fresnel conversion here though - equivalent to about a 500-650 fresnel output, but 5200k daylight for about $100 plus a donor fixture, parts readily available. And it looks very legit (no duct tape or PVC pipe). I'm going to make a couple more of those, really hits a sweet spot for color work and clients never notice they're DIY.

 

[wildbill]

This LED light consumes as much as my 200W Mini Mole but has an equivalent of a 2000W hot light. It also dims probably without a shift in color temperature. You can also change the color temp from tungsten to daylight also. Pretty amazing. Expensive now, but will be cheaper later I'm sure.

http://www.bhphotovideo.com/bnh/con...3a0ejAR-lZRoCjy_w_wcB&is=REG&m=Y&Q=&A=details

Never heard of that brand.
Mole, Arri, Dedolight, LitePanels, and others all have similar offerings. I checked out the new Mole 5K daylight led and Led 2k softlight a couple weeks ago. Don't look at the source! My eyes still hurt