My view on LED lights at DVinfo.net

Go Back   DV Info Net > The Tools of DV and HD Production > Photon Management

Photon Management
Shine an ever-loving light on you.


Reply
 
Thread Tools Search this Thread
Old June 1st, 2009, 10:20 AM   #1
Regular Crew
 
Join Date: Dec 2006
Location: Tampa-Orlando, FL
Posts: 124
My view on LED lights

I haven't had the time to spend much time on this board lately, we are having an unusual busy year but I've seen many posts on LED lights and also a lot of frustration. After 40 years in this business I've seen products come and go, some of the good ones stay and the not so good ones dissolve into darkness. What I don't like is to have to readjust my techniques to accommodate a new product, manufacturers have to listen to us, give us gears that we can use an integrate with everything else we have, LED has not done that.

This is one of my posts on another boards on the LED subject. This board is driven by suppliers so I hope Chris keep it, this is the only way we have to communicate with manufacturers, the other is not buying their products.

Quote:
I donít usually like to work with an on camera light but many times I have no choice if thatís what the client needs. I have at least six different on camera lights but the one I use the most is the Anton Bauer, and the only reason I like the AB is because in one second I can change from the tungsten to the HMI heads.

For the last few days I worked (or at least I tried) with an on camera LED light; the one made by Litepanel, the large rectangular not the smaller one, I donít know the model numbers. The client sent it to me; apparently someone sold him as this being the most versatile light. I donít mean to be critical of new products but I need stuff to make my job better an easier, the financial welfare of the manufacturers is not my concern, but I can assure them that if they make a product that I can use Iíll test it, buy it and endorse it.

The LED light is daylight balanced so I tried outside under different conditions and it was useless. I use my Anton Bauer UltraDayLight (on camera HMI) on similar situations with good results especially by backlighting the subject; in direct sunlight no on-camera-light will do the job.

I then tried inside under daylight conditions. Beautiful north-light (no direct sunlight), all I needed was to fill a little on the shadow side. Iíve done this shot many times in the same place and used the Anton Bauer as fill, I even had to diffuse it a little because it was giving me too much light and reducing the modeling ratio. At 8 feet from the subject the LED light wasnít even visible, on or off it didnít make any difference, a little catchlight in the eyes maybe, but that was it.

Now I moved indoor on the convention floor where the lighting there was close to $3200K balance. I put the conversion filter supplied with the light and now I have the correct colors but Iím working with half the output. The problem now is that the existing ambiance light is more powerful that my LED light, what I usually look for is for my subject to be slightly brighter that everything else, not much just a stop. Couldnít get that out of the LED.

The last of the problem is that the light with the batteries is very heavy, considerably heavier that everything else I ever used and throws the camera way off balance.

The problem with all these new LED lights is that we live in a 3200K world when we are inside and 5600K when outside. All LED lights are useless outside so why make them in 5600K color balance in the first place when apparently there is a 3200K LED technology out there; placing CTO conversion filter on a LED will make the light even more useless.

The other problem, especially for those of us that work for networks is that often we have to work side by side with other TV photographers (gangbang pressers) we all have our little lights on but they are all the same color temperature, 3200K indoor. Put just one light in the mix with a different color temperature and everybody videos will be off color, thatís a good way to make enemies among your colleagues. Inconsistency across the products from the different manufacturers is the biggest problem that I see with LED lights.


LED lights that can help our jobs in certain situation and the power saving is important but not important enough to alter the versatility and the results. Although Iím not sure how power efficient they are because I didn't get the chance to try them out.

As far as color goes Zylight has perhaps the best solution with a single light that changes color temperatures. I'm not sure if it's available but Lowel has also been showing for the last 2 NABs a less sophisticated than the Zylight but a similar LED light that can gradually change from 5600 to 3200K by employing two banks of LED for each color and with a dissolver that can change the intensity of each bank. These are what I consider the best solutions yet. Using color corrective gels is a fix for using the wrong equipment and not a working solution.

Manufacturers listen to us please, we need something functional.
__________________
NinoMedia Productions - EFPLighting.com Advanced Lighting and Gripping Workshops
Nino Giannotti is offline   Reply With Quote
Old June 1st, 2009, 01:14 PM   #2
Major Player
 
Join Date: May 2007
Location: KLD, South Africa
Posts: 983
You do get the option to buy 3200K LED's most people opt for the 5600K. I've been told, you don't get one light that does it all. On camera lights are mainly used in emergency lighting situations, one needs better lights to make a difference. I wish for a 500W pocket light that runs on fusion, anyone? Then again I can't afford any half decent lights and use hardware work lights :)

I would love to hear what you think of the Coollights LED600?
Nicholas de Kock is offline   Reply With Quote
Old June 1st, 2009, 04:57 PM   #3
Major Player
 
Join Date: Jun 2006
Location: Shenzhen, China
Posts: 781
He talking about on camera lights. The Cool Lights LED 600 is in another class, size and weight and not at all considered to be an on camera light--its something equivalent to a 650w fresnel. So not intended as the small lightweight source used as on camera lighting and shouldn't be compared and contrasted or reviewed as such. Lots of people use the 5600K LED 600 outside in some actual daylight conditions and can get a decent fill out of it so there is no comparison at all to any small on camera light.

As a manufacturer, here are my thoughts on the subject, many of which I've said before. I don't know any other manufacturers that actually chime in on these boards so I may be alone here covering that side of things.

LEDs themselves are evolving and as they do evolve the instruments we make out of them will too. We know of the shortcomings and actually write about them telling people of them before they buy. For this reason we have customers that are entirely satisfied because they know whatever the limitations of LED arrays ahead of time. We also don't think its possible to make on camera LED arrays that are all that great so we mostly work on producing the larger capability instruments.

I can't speak for or influence what other manufacturers do other than by creating very competitive products that are popular and doing things the right way myself. You have to know the limitations of the different technologies and work around them if you want to get the advantages that LEDs give you.

1). Given that LEDs are like "miniature spotlights", which are not useful by themselves, but in combination with a lot of other LEDs they become useful (just as one pixel on a screen would never be useful), you're making a new kind of light out of a bunch of small spotlights so you do have a new kind of source that has to be used in other ways if you agree with the advantage this gives you.

This doesn't lessen the utility of the light it just means its a different kind, to be used in different ways and knowing the limitations. No one light in the toolkit can do it all. You wouldn't want to carry a 650w tungsten fresnel on a large car battery array down the street while filming a walking interview. You can do this with one panel like an LED600 on a lithium v or ab type camcorder battery. Its entirely usable in a situation like this where you've got another person to act as the "human light stand" so to speak. You could use a small DC HMI par for this kind of thing, but is it as affordable? If you rent, I guess so.

I don't view an LED on camera light as being all that useful outside at this point so that's why we concentrate on these larger type of units which are very usable when used as intended.

Not until you have one point source large wattage LED will you be able to construct LED sources that mimic the behavior of fresnels with one single shadow and all that comes with that. We've experimented with some large LEDs (30w and higher) but they aren't ready for this kind of use for a number of reasons.

2). Daylight LEDs still have color rendering limitations and aren't full spectrum. There are workarounds (like using filters) for this until we get a better LED. When you're in a "power challenged" area, you may find that something like an LED panel is really the only viable choice given its size, weight, price and time it can run on batteries.

3). RGB or dual color panels are fine but in effect you may only have half or so of the capacity of the panel producing light depending upon the color temperature or single wavelength color you dial in. You can also take care of many of the CRI issues with a dual color or RGB type panel but at what cost? It's expensive if you need microprocessors to help you dial in a color temperature.

I've polled people about this capability of a "dual color" panel (which would be less expensive than the RGB type) and they like it until they realize that to produce 3200K you may only have half the LEDs firing and only half firing to give 5600K as well perhaps depending upon what the upper range of the panel is.

Most customers that would want a large LED array want all the power going out all the time. Thus, that's why daylight only large panels are all the rage now that they're becoming semi-affordable. People like the LED 600 and they also want a 1200 and even a 2500 which we have in the works. We also have an LED 256 coming not intended for on-camera use but rather as a compliment to an LED kit as a small spot to be used as a back / rim / hair light.

This output capacity issue is also true for the more expensive RGB type panels. Each LED has a blue, green and red chip. Say you wanted a red single wavelength output to mimic a light with a red gel. Well you only get 1/3 the output of the panel because only the red chips on each LED fire up. Thus, its not entirely as you might think or as advertised in some cases. Since most will want to use different color temperatures of white light for the work they do (from 3200K to 5600K), you would need a huge and expensive array of RGB LEDs to make anything large enough to be usable once you believe that on camera lighting won't be enough for your uses.


All that being said, we're experimenting with a bi-color 1200 and 2500 panel now and will have a limited offering of that to see how people like it with the full knowledge that all 3200K only uses about half the output capability of the panel.
__________________
Richard Andrewski - Cool Lights USA - RED #114
http://www.coollights.biz

Last edited by Richard Andrewski; June 1st, 2009 at 05:44 PM.
Richard Andrewski is offline   Reply With Quote
Old June 1st, 2009, 05:19 PM   #4
Regular Crew
 
Join Date: Mar 2004
Location: NY
Posts: 186
Quote:
Originally Posted by Nino Giannotti View Post
Inconsistency across the products from the different manufacturers is the biggest problem that I see with LED lights.
I agree this is a serious issue because of some LED light's particular color cast. Two 32k LED lights from different manufacturers can read +/-300k. This could be workable but even if both read 32k you may get a different color cast on each one that can't be read by current meters. What instruments manufacturers are using to calibrate their LEDs to 32k and 56k is of concern. What are they using? Which industry-wide standard set of metrics is being followed across all manufacturers? Because this is not the case with color meters and cameras. With Tungsten and HMI everyone knew what the temperature was going to be. With LED calibration we have a different spectrum with variables to control and this is dependent on who is supplying their LED diodes. Some of these diodes are coming from suppliers more experienced with the architectural lighting market, rather than film and television.

I'm looking forward to an opportunity to test the Arri LEDs to see if some of these concerns are addressed. They're only available to rental houses right now. I tested the Zylights for 32k color cast. You can read the results in another thread.

As far as lumens you may want to try the Zylight Z90. I found it has more power than the panels. The Litepanels on-camera lights were good for creating the market but not my preferred light due to a few issues.
Max Allen is offline   Reply With Quote
Old June 1st, 2009, 10:10 PM   #5
Regular Crew
 
Join Date: Dec 2006
Location: Tampa-Orlando, FL
Posts: 124
Seeing that the conversation went from on camera LED light to the larger units I'm re-posting this that I originally posted on another board. Again my view is based only on established products, something that has been around for a few years.

I have hundreds of professional around the world that count on me to give them a honest opinion, this is the way I see it.

Quote:
This NAB report is a little late coming but paying customers have priority. My time at NAB this year was limited so I didnít even bother to look at prototypes of lights as probably most will never make it to the market. I opted to concentrate my time to those manufacturers that have been around for few years and most likely will be there next year too.

Until recently the last bastion of simplicity in this business, untouched by computerized technology were lights; simple and economical lights. For those who were under the impression that lights were expensive, just wait.

For the last few years at NAB weíve seen a dozen of LED light manufacturers, most are oriental companies hoping to find US distributors. 3 US companies have consistently shown LED lights, those are Zylight, Litepanel and The Prompter People. Iím talking about larger lights, not the ones on-camera. Please note that this is my opinion, itís the way I see it. I would also like to say that there will be LED lights in my van in the very near future because they can be very useful for certain situations, also because Iím a gadget freak and I love toys.

So what are LED lights good for? The most beneficial use that I see of this new technology is the ability to set up lights quickly and considering that most LED lights can also operate on standard video batteries we can place them anywhere without the worry of finding an outlet and carrying long extension cords. LED lights also consume less energy and today this could be a significant feature.

Although I have yet to use one, the problem that Iíve seen at NAB where I had the opportunity to see them all is that they have a very limited reach. It wasnít a fair test because of the powerful overhead lights but at over 10 feet the lights were barely noticeable and overpowered by existing lights. These are also the types of conditions that LED lights would be most useful such as using them on a trade show floor or in an arena for half time reporting.

From someone who is making a living with good lighting the first problem is the high cost of technology. A single 12x12Ē LED fixture can cost as much as a complete light kit consisting of several lights. I have over 50 lights in my van, if I had to replace them all with LEDs I would also have to invest into an armored truck and probably hire a security guard.

Zylight at NAB had perhaps the most sophisticated LED light on the market, a jewel of a light. Just like their 50 and 90 models, the new fixture is packed with features. You can continuously adjust the color temperature output as well as a separate compensators for working with off color light sources like fluorescent lights, the unit also has a full range dimmer. The new Zylight also can match color output via wireless with other similar Zylights fixtures or the smaller 50 and 90 light. What a dream light, but at a MSRP of around $3400, and if you need a battery the recommended one is the AB cine starting at about $1400 Iím afraid that for me it will have to remain a dream.

Litepanel also has two new lights, one that can be adjusted from 5600K to 3200K and another that can be adjusted from flood to spot. In the past these features meant having to buy separate lights. These are basically two lights in one, they have two sets of LED lights, thereís a dissolver on the back that gradually switches between the two sets of LEDs. The Litepanels are a bit more reasonably (sarcasm) priced at around $3000 for a single 12Ēx12Ē light.

There are also others manufacturers that make relatively inexpensive LED lights, the Prompter People has two LED lights in the $600 to $800 range. These lights have a color temperatures around 5600K +/- 5%. any color changes will have to be done the old fashioned ways, by using corrective gels.

Next is the very important quality of light. By comparison to existing lights all currently produced LED lights are very small the largest being approximately 12Ēx12Ē. A light this small will not give a satisfactory ďwrap-aroundĒ, something so important in order to get a soft transition between highlights and shadows particularly when photographing people; it would be like using a XXsmall chimera soft box as a key light. If you wish to create a light the size of a small Chimera you have to cluster together about $12,000 worth of lights in order to obtain a 2íx2í light source, and this is for the key alone. Add a fill and a hair or back light and now for a simple 3 points light you are looking at about $18,000; 3 Lowel Rifas will do the same job for about $2500. Of course we are not done with lighting yet, we still have to light the background and we have no clue if LED lights will play nicely with other lights, like fresnels, as color temperature will be inconsistent. If anyone is considering mixing light sources with LED I would strongly recommend a good color temperature meter (about $1200) and a good supply of corrective gels.

The first rule for a good photographer is the quality of the lighting, convenience is an added bonus. If manufacturers want LED lights to make a significant dent into the ENG/EFP arena theyíll have to start pricing it on line with existing lighting fixtures. Personally I really donít mind using inexpensive gels attached with old fashioned clothes pins.

The most significant use of LED fixtures has been outside the ENG/EFP arena. Manufacturers have realized that the cost is too high for our market and they are going after clients with deep pockets. Displays, studios and the rental market appears to be the areas of success for LED lights. At the NAB show Canon had a lens demonstration set illuminated with at least 20 Litepanels, that would be about 40K worth of lights.

At NAB I also had a conversation with the marketing manager of Kino Flo, the venerable manufacturer of fluorescent fixtures. Last year they had a very sophisticated LED light on display but this year the light wasnít there. He explained to me that the LED project was put on hold for many reasons, the most important being the ďlumens vs dollars ratioĒ, in few words the cost of a good LED fixture is grossly disproportionate when compared to other existing compatible fixtures and wasnít addressing the needs of the core of their business, the ENG/EFP clients.

The reason that Iím posting all this is because too many people, mostly those who have no clue about lighting are talking about LED being the state of the art replacement for all existing lights, thatís salesman talk of course. The point is that price, quality and versatility are still a long way off for ENG and EFP needs.

As a veteran photographer I would like to see the manufacturers create products around my needs with quality as the first priority, rather than me having to adjust my skills downward to accommodate their products.
__________________
NinoMedia Productions - EFPLighting.com Advanced Lighting and Gripping Workshops
Nino Giannotti is offline   Reply With Quote
Old June 1st, 2009, 11:58 PM   #6
Trustee
 
Join Date: Jan 2004
Location: Scottsdale, AZ 85260
Posts: 1,538
Quote:
Originally Posted by Nino Giannotti View Post
Seeing that the conversation went from on camera LED light to the larger units I'm re-posting this that I originally posted on another board. Again my view is based only on established products, something that has been around for a few years.

I have hundreds of professional around the world that count on me to give them a honest opinion, this is the way I see it.

Nino,

I understand your concerns and have shared many of them. However, it's also true that the energy use profile alone makes color-balanced LEDs an excellent new option once you get to know them.

Yes, as small LED Obies and similar on-camera fill lights, they're are nearly useless.

Only when you get up to something close to a foot square does the output make them practically useful.

And yes, most of these lights are insanely overpriced. Paying $1500 - $2000 for a single panel is nuts.

My advice is to try what I did. I ordered two of Richard's CoolLight panels in 3200k. Fitted with Sony V-lock mounts (Anton Bauer is available as well) these have become two of my go-to locations lights for the very reasons you indicate. The ability to move a fixture anywhere it's needed and achieve fast, usable light nearly instantly. The cost was a MUCH more reasonable $400-500 per unit (accessories depending) so one could assemble a 3 light kit for the cost of ONE of the competing panels.

Yes, they work a bit differently from what you're used to. For instance their barn doors work as expected in the first 50% of their travel distance - cutting spill outside the beam pattern - but beyond that, instead of cutting light like a similar barn door on a tungsten fresnel, they mask rows of lights such that you get a venetian blind effect. It's just how they work.

But those small inconsistencies are no match for the fact that a 650 watt tungsten fresnel, pulls 10 TIMES the power of a 65w LED array like the Cool Lights unit. And yes, in this size, the light is extremely usable.

In fact, in my corporate practice, I'm noticing more and more national retailers chasing energy savings by installing high efficiency fluorescent overheads that output color-friendly warm light at about 3300 degrees kelvin. In these settings, the 3200 degree LEDs provide an EXCELLENT face fill and/or rim light. Again with all the advantages of weight and portability.

So yes, LEDs aren't for everyone. But the technology is coming along very fast and knowledgeable manufacturers like Richard are making it much more affordable than before.

So I'd recommend trying not to let your previous experience with what this technology is NOT suitable for - color your opinon about the whole class.

For what it's worth.

Last edited by Bill Davis; June 2nd, 2009 at 05:30 PM.
Bill Davis is offline   Reply With Quote
Old June 2nd, 2009, 03:00 AM   #7
Regular Crew
 
Join Date: Mar 2004
Location: NY
Posts: 186
How's it going Nino?

Quote:
Originally Posted by Nino Giannotti View Post
If anyone is considering mixing light sources with LED I would strongly recommend a good color temperature meter (about $1200) and a good supply of corrective gels
From having tried to use the Z90 as a variable temperature light to match spotlights and daylight corrected Source 4 Tungstens in major venues for off-light calibration of multiple cameras in a switched environment, I can tell you that using a color meter will not work because color meters are not designed to read the LED spectrum. So matching temperature based on color meter readings is a hit or miss, in my experience. Charting a variable offset scale between meter readings, a particular camera, the LED light and the type of continuous filament source you want to match is a starting point. The Z90's temperature readout scale will also not necessarily match what a meter and your camera reads. Another reason here is that meters are designed to read for film, not digital sensors. I have not tested the Sekonic in this regard.

Like you, I also like the Zylight innovations but there are hurdles to overcome with these new tools. At this point I'm not confident that LED's will behave like other lights in terms of throw anytime soon. I think they are another tool in the kit at this point. If only the temperature calibrations get nailed down 100%, having them play nice with other lights will be a big step.
Max Allen is offline   Reply With Quote
Old June 2nd, 2009, 05:13 AM   #8
Regular Crew
 
Join Date: Mar 2006
Location: Minneapolis, MA
Posts: 84
I have been interested in LED lighting for years. I do think things have progressed significantly over the last few years. It wasn't that long ago that the idea of lighting things with a panel of low power consuming LED's was science fiction. I do agree though, that LED technology is not yet ready to become a main workhorse of this trade.

I like to follow the technology of LED's themselves, and there has been no slowdown of ideas on making more powerful, more consistent LED's. Richard has made some very good points & informed us quite a bit on the different challenges of making LED's into a light package that can be used in production. It seems to me that the real challenge is in making LED's that can put out enough light so that a useful panel does not need the hundreds or thousands of LED's to create enough light.

I won't bore you with the details, but there are tecniques being developed that will eventually result in diodes that emit orders of magnatude more light than current versions. As these diodes become available, manufacturers will be able to create more powerful panels. This would make ideas like panels with multiple sets of led's for adjusting the color temp more viable, because even at half output it would create a useful ammount of light. Also the idea of a single LED that could compete with a small incandescent bulb would make the on camera light with a single source finally possible.

One issue I am not sure will ever be resolved is the CRI problem. My understanding of CRI is that basicly in addition to the "primary" color temp, a high CRI light puts out lots of light above & below that color. LED's have a major disadvantage, because they emit light at a specific frequency, and almost nothing above or below that wavelength is created.

It would take someone who knows lots more about filter technology than myself to figure this out, but would it be possible to pass light from an LED source through some sort of filter that would "convert" some of the light to varying wavelenghts? If this could be done I am sure it would be at the cost of some light at the primary color, but just maybe it would not cost so much that the level is too low to be useful.

Another method would be banks of LED's that each put out slightly different casts, but banalced in such a way that the overall cast is of the correct color. I suspect this mathod may already be in use to some extent.

All in all, the current products are definately way too expensive. As LED's develop, prices should drop, but it may be a very long time be before anyone chooses between a Lowell & an LED panel based on which is less expensive.

I guess my point is that LED's are coming, but they still need some time to get ready for their big broadway debut. I do think there is light at the end of the tunnel.
Justin Deming is offline   Reply With Quote
Old June 2nd, 2009, 05:23 AM   #9
Wrangler
 
Join Date: Jan 2002
Location: Los Angeles, CA
Posts: 6,781
Quote:
Originally Posted by Nino Giannotti View Post
The problem with all these new LED lights is that we live in a 3200K world when we are inside and 5600K when outside. All LED lights are useless outside so why make them in 5600K color balance in the first place when apparently there is a 3200K LED technology out there.
What about an interior with uncorrected windows? Here HMI's and daylight flo's usually rule the roost, and ambient light levels are generally higher, so having a daylight balanced unit like the 5600K Litepanel Mini makes reasonable sense, depending on how you are using it of course.

I've had one for a couple of years that I occasionally use onboard my Steadicam as an eyelight--I've replaced the knob with a little 32 pitch gear and mount a Preston motor to it, then hand the remote iris to the DP so he can feather the output as needed. In this application, it's never the primary exposure so it's OK that the level is a little down especially when corrected down to 3200. I've been thinking about trying the Zylight 90 as a more flexible option, especially since the Zylight guys have offered to make me a custom one that dims all the way to 0 (the standard model does not) which is critical when passing by reflective surfaces. However I'm a little underwhelmed by the current softbox attachment (Chimera and speedring, a bit clunky) and would like to see something a bit more custom.

FYI I used an interesting product called the Kisslite from Gecko; it's a small LED ring light with integrated filter trays and eyebrow. You can switch the LED clusters on or off in quadrants, plus dim the whole works and all of this locally or remotely. It comes with both 3200 and 5600 modules and can be field swapped quickly. It's not cheap but it fulfilled a need. We used it full-time on the recent ABC series "In the Motherhood".
Attached Thumbnails
My view on LED lights-kisslite.jpg  
__________________
Charles Papert
www.charlespapert.com
Charles Papert is offline   Reply With Quote
Old June 2nd, 2009, 06:23 AM   #10
Major Player
 
Join Date: Jun 2006
Location: Shenzhen, China
Posts: 781
Quote:
Originally Posted by Justin Deming View Post
One issue I am not sure will ever be resolved is the CRI problem. My understanding of CRI is that basicly in addition to the "primary" color temp, a high CRI light puts out lots of light above & below that color. LED's have a major disadvantage, because they emit light at a specific frequency, and almost nothing above or below that wavelength is created.

It would take someone who knows lots more about filter technology than myself to figure this out, but would it be possible to pass light from an LED source through some sort of filter that would "convert" some of the light to varying wavelenghts? If this could be done I am sure it would be at the cost of some light at the primary color, but just maybe it would not cost so much that the level is too low to be useful.
I'm optimistic that problem will be cracked eventually. Fluorescent tubes were primitive at the start too but they discovered rare earth phosphors and other tricks to balance them out to better color rendering. Its a good enough simulation of full spectrum that it works.

Color temperature and CRI are measured on two axes that are mostly unrelated to each other. Color temperature is a scale of shades of white light with a bias toward red on one end and blue on the other. Some people mistakenly describe the red end as yellow but its "red" because of the ideas behind color temperature being all about a theoretical metal object being heated from red hot to blue hot, etc.

The other axis which is a balance between green and magenta bias is not classically thought of as CRI but that's what it ends up being in practical use if you know something about lighting and think about it.

People confound color temperature and CRI but as I said, they really have no relationship for all intents and purposes as long as the light is relatively full spectrum-like in nature. You can have a bad CRI with any reasonable color temperature in middle range and a good CRI as well.

A light with a CRI of 99 would be like daylight, tungsten or carbon arc. They're full spectrum and what all other electronic sources (fluorescent, HMI and LEDs) are measured and benchmarked against in both color temperature and CRI. They are thus balanced between magenta and green. The CRI scale then becomes a measure of how far off a particular source is from one of these benchmarks. 99 being damn close and 1 being about the worst you can imagine.

A light with a CRI of 90 has only a very slight bias toward green or magenta. A light with CRI 60 has quite a bias toward green or magenta. Its mostly toward green though. All these sources will tend toward green unless they are well engineered for CRI through whatever means necessary. Pushing it toward magenta through some means is the common way to fix the issue. Why magenta? Because magenta is across from green on the color wheel. You use the color across from another color to counteract that color and null it out. Push it too far and you end up with bad CRI thanks to too much magenta.

You can have a 5600K light with a good CRI, well balanced for green/magenta and a 5600K with a bias towards green and low CRI. They're still both color temperature 5600K though. Some color meters may be fooled though by the green bias if they aren't well setup to read that axis.

You can fix CRI with several means. 1). Simply balancing the phosphors in tubes with more magenta. 2). With correction filters placed on the lights themselves. 3). With white balancing (if its available to you) which not only does a great job of sorting out what color temperature you're working at and making white appear white but also balancing the green/magenta axis as well. We're lucky with digital mediums because for the most part we don't have to worry about these things and there are easy corrections in most cases. Things can get a bit confused for the white balancing when you have different types of light mixed together, but that's another story and problem.

Single wavelength colors like pure red, blue, green, etc. are not measured in color temperature but are just measured by their wavelength (nm). Again color temperature only has to do with white light and how its color bias tends towards.

As for color meters and their accuracy, I have a Kenko KCM 3100 and I think its pretty darn true for most of the sources I've tried it on. It agrees closely with what the sprectrometer / integrating sphere readings are, which is what LEDs, HMI and Fluorescent are measured with at factories producing them. It not only does a good job of telling you where you are on the blue to red bias of white (color temperature) but it also does a good job of indirectly telling you about CRI when it mentions where you are on the green/magenta axis and what kind of correction filter you may need to get you to balance. If you derive CRI from this it would be a guess though since it doesn't directly tell you the CRI value.

An integrating sphere and spectrophotometer are the only instrumentation I know of that can truly tell CRI.
__________________
Richard Andrewski - Cool Lights USA - RED #114
http://www.coollights.biz
Richard Andrewski is offline   Reply With Quote
Old June 2nd, 2009, 10:22 AM   #11
Trustee
 
Join Date: Jun 2006
Location: Burbank
Posts: 1,811
I will suggest that for run and go location production with a tiny crew and one or two actors, the LEDs are perfect.

Available light is largely used. The LEDs can be main light, fill, background, shadow casters, etc. They work outside but also down in parking garages, inside cars, in alcoves, and so forth.

I have a couple of Richard's lights. They are fully dimmable, and there are 4 bank switches. They operate on AB batteries. I supplement these two with a prompter people cambat light. The the lights can be setup in seconds. And the can be carried around as continuous fill, etc.

We are doing a location web series, and no other lights really can work the way and as fast as we are using these.
Jack Walker is offline   Reply With Quote
Old June 2nd, 2009, 02:38 PM   #12
Regular Crew
 
Join Date: Jul 2005
Posts: 110
In addition to all of the points covered, light output stats of leds are confusing and frustrating. There seems to be a desperate need to compare light ouput to a known entitiy i.e. tungston bulbs. The problem is that Lumens, Watts and even candlepower are not an effective way to measure output. Each LED is an individual unit that has a degree of spread. A 1 watt LED with a 20 degree spread will output stronger than a 3 watt at a 60 degree spread. A longer throw LED (20 degrees) is great at distance but will blowout everything up close.
When a commecial lighting company creates a light, it seems as thought they are trying to be all things to all demands and as a result not many have "hit" it yet. I made my own with long throw Crees and have been using them for well over a year. Yes they are powerful but they are easily controlled with diffusion material. Oh and I rarely used them on-camera.
Gary Moses is offline   Reply With Quote
Old June 2nd, 2009, 04:20 PM   #13
Trustee
 
Join Date: Jan 2004
Location: Scottsdale, AZ 85260
Posts: 1,538
Quote:
Originally Posted by Gary Moses View Post
In addition to all of the points covered, light output stats of leds are confusing and frustrating. There seems to be a desperate need to compare light ouput to a known entitiy i.e. tungston bulbs. The problem is that Lumens, Watts and even candlepower are not an effective way to measure output. Each LED is an individual unit that has a degree of spread. A 1 watt LED with a 20 degree spread will output stronger than a 3 watt at a 60 degree spread. A longer throw LED (20 degrees) is great at distance but will blowout everything up close.
When a commecial lighting company creates a light, it seems as thought they are trying to be all things to all demands and as a result not many have "hit" it yet. I made my own with long throw Crees and have been using them for well over a year. Yes they are powerful but they are easily controlled with diffusion material. Oh and I rarely used them on-camera.

Gary,

I have a suspicion that this LED verses older technologies is going to turn out precisely like the debate I went through when all the "Professionals" argued that early DV was inadequate for pro work since it didn't have everything that the industry was used to in its current workflow. They were technically correct - early DV didn't have professional timecode, it's color capabilities weren't robust enough for keying, and even the picture didn't have the technical resolution of high level analog video of the time.

In the end, NONE of that mattered. The benefits of Digital eventually pushed the old analog world aside.

And the fact that LED has the proven ability to generate useable photons at 1/10th the cost of tungsten technology (and for that matter that color balanced fluors can do so at a 70% energy savings) will - in the long run - be the final arbiter.

LED is very young technology. But not for long.

What they don't do well at this stage of development will fall away. And sooner rather than later tungsten will go the way of the time base corrector.

I don't like it anymore than anyone else. I have a significant investment in tungsten gear. But that doesn't mean I can't see the writing on the wall.

At some point everyone WILL move to incorporating LED lighting technology in video workflows. You get to decide whether that will be earlier or later in the transition. That's the only choice I really see. It's the math. A 10 to 1 advantage in anything is just too big to ignore.
Bill Davis is offline   Reply With Quote
Old June 2nd, 2009, 05:01 PM   #14
Major Player
 
Join Date: Jun 2006
Location: Shenzhen, China
Posts: 781
Interesting thread.

Indeed Bill, that's a great analogy. Beating on a technology won't make it move ahead any faster than actual circumstances will allow it to move. It will advance at its own pace and as knowledge about it can. And there's the cost curve and acceptance curve that all things go through as well before they reach a stage of plentiful commodity and the pricing falls. We are in the early days of this kind of lighting, but there are companies that are providing the pricing pressure already right now (hehe). You will have people that don't like change and will rail against it in any case.

I remember when I started down the LED product development path in 2007 I wasn't terribly enthusiastic about it myself. At the end though, I came away shocked and amazed what those little solid state spotlights are capable of. The more I tried to apply old ways of doing things to the development though, the more frustrated I was.

So, its like anything else. Your ability to use a new technology is only commensurate with your ability to exercise limiting beliefs. Said another way, you have to open your mind and also be open to change the way you do things. Not expecting to apply old ideas to new technologies. That just brings more frustration as you try to do things in old ways using new tools.

Do the techniques of lighting change? Not really, we'll still have 3 point lighting, playing with light and shadow, hard and soft light (for drama or no drama), backlighting, key and fill; but the rules about how you do it may change as the fixtures change. So expecting an LED fixture to behave exactly like a single point source fresnel will bring plenty of frustration. Fluorescent, HMI/metal halide, and of course tungsten are far from dead. There will be plenty of people that prefer those for a while. We still sell quite a bit of that, but the LED products fly off the shelves. I got the message. They want LEDs and the benefits they bring.

In one long sentence I can sum up why I personally like LEDs. They are far more powerful than lumen (or lumen per watt) specs can ever tell because they are individual solid state spotlights themselves (complete with source and lens) and should not be compared to light bulbs at all but to other types of fixtures. Bring your light meter and you would be amazed what you can get out of LEDs once you open your mind to it.

I wrote a blog article on some of my discoveries about LEDs during the development process for our first LED product. It shows why the writing is on the wall.

http://www.coollights.biz/wordpress/archives/72
__________________
Richard Andrewski - Cool Lights USA - RED #114
http://www.coollights.biz
Richard Andrewski is offline   Reply With Quote
Old June 2nd, 2009, 05:14 PM   #15
Regular Crew
 
Join Date: Jul 2005
Posts: 110
I totally agree. It's the future and it's good. Just a little immature. I converted my tungsten setup to daylight cfl's plus I use the leds. It appears that most of my shooting these days has more ambient daylight then 3200k. If I need more light than that, I converted all of my tungsten with dichroic filters just in case. I love the low electricity draw and lack of heat in this new technology. In a pinch I can comfortably run my setup off my 1200 watt inverter in my car.
Gary
Gary Moses is offline   Reply
Reply

DV Info Net refers all where-to-buy and where-to-rent questions exclusively to these trusted full line dealers and rental houses...

Professional Video
(800) 833-4801
Portland, OR

B&H Photo Video
(866) 521-7381
New York, NY

Z.G.C.
(973) 335-4460
Mountain Lakes, NJ

Abel Cine Tech
(888) 700-4416
N.Y. NY & L.A. CA

Precision Camera
(800) 677-1023
Austin, TX

DV Info Net also encourages you to support local businesses and buy from an authorized dealer in your neighborhood.
  You are here: DV Info Net > The Tools of DV and HD Production > Photon Management

Thread Tools Search this Thread
Search this Thread:

Advanced Search

 



Google
 

All times are GMT -6. The time now is 06:41 PM.


DV Info Net -- Real Names, Real People, Real Info!
1998-2017 The Digital Video Information Network