LED illumination

[Graham Stabler]

After making some first successful macro shots for my object-movie the bulb blew in my twin fibre light source. More are on order but I took the opportunity to explore LEDs. Philips do the Lumiled range including some complete with collimating optics. I ordered some as they had next day delivery and they provide enough light and look surprisingly white compared to a lot of white leds. The next issue is holding them, I'm hoping to create a neat way of illuminating in a flexible way without spending a fortune. If they prove to provide good enough lighting in terms of colour it might something others are interested in. I thought we might compare notes.



This is used to direct coolant (also seen on gorilla pod) but could be used as an adjustable mount.



What about a fully adjustable led ring-light, or line-light or L-light.

Graham

[Graham Stabler]

An expensive option but very neat, single knob locks the central hinge and both ball joints. This is a modular version.

www.noga.com

[gpmatthews]

I was wondering about adapting one of the flexible USB reading lights that you can find on the web - I have one that I may cannibalise to repace the USB connector with something more convenient and then replace the light with a high intensity LED

[acerola]

I also think about LED lighting. Now I use a Fenix LED Torch on a Flash bracket. It is only one LED and I need at least one more. This Flexible hose is promising.
I just googled a similar solution with it.

http://www.instructables.com/id/Desk-Sq ... s-improved./

[lauriek]

If it's the only lighting you're using then the colour shouldn't matter, just shoot a known white/grey card and set your WB according to that... It's worth doing that anyway as these LEDs never seem to be quite white, even if they look it!!

[Graham Stabler]

I don't know why I didn't think of this before, I just bought a Gorillapod and it is quite easy to separate the sections. They are very flexible and have tighter bend radii than the hoses. I should end up with many adjustable mounts. They don't have holes down the middle so the power will have to be separate but that's OK.

[Graham Stabler]

Graham, the pound shops might be worth a look for those.

Laurie, I take it these cards are available form typical photography suppliers?

Graham

[lauriek]

Yes you should be able to get a standard grey card from just about any decent photographic supplier. Kodak do them, and Lastolite do some fold up cloth ones like a small version of their reflectors. I imagine there are plenty of options these days.

I don't personally use one as I currently shoot flash most of the time so I know the temperature, I just set 5600k in raw processing parameters. If I have a shot with dodgy lighting I just try to correct it by eye as best I can, sometimes this can be hard work!

[microcollector]

I use LED lights for my mineral photography, a couple of multi bulb floods from Leica and a ring light from Orled. I use the gray card to set a custom white balance and it works very well.

Doug

[Graham Stabler]

OK so here is what I have come up with. To mount the led to the flexible leg required a little piece of aluminium with three tapped holes. Two allow the led to be bolted to the aluminium and the third allows a countersunk screw to hold the final segment of leg. This segment was drilled out to 3mm and M3 screws were used. The aluminium plate also provides some extra heat sinking.

In the other end of the lone LED worm is a neodymium magnet, glued in with metal filled epoxy.

The other option happened by chance, the metal base was in a cupboard here and has a 1/4" Whitworth hole tapped in it, the tripod is essentially mounted inverted (complete with quick release). I may make a better base with the hole at a 45 degree angle to allow for a greater range in heights for the arms.

Wiring will follow, I have drilled out all segments but may keep it on the outside for flexibility of configuration.

These LEDs are VERY bright!

[lauriek]

These LEDs are VERY bright!

What's the wattage on these? One of my torches has a 3W luxeon LED and that one's pretty bright, around 100 lumens iirc, I've seen 5W LEDs more recently, is that what these are? Do you know how many lumens they put out?

[Joseph S. Wisniewski]

If it's the only lighting you're using then the colour shouldn't matter, just shoot a known white/grey card and set your WB according to that... It's worth doing that anyway as these LEDs never seem to be quite white, even if they look it!!

I normally go a step beyond that.

No matter what I did with gray cards (or a pricey ceramic white reference standard) I could never get the reds and oranges where I wanted them with white LEDs.

White balance only sets two parameters that determine the location of the "center of the color universe" (typically in terms of two numbers: color temperature and a red/green "tint" component). That's fine, for light that "works like" sunlight or incandescent light, light that's produced by a glowing hot "black body".

A light source like a white LED or a fluorescent light (even the "high CRI" ones) is more complicated. You can get the white "center" of color right, yet the oranges are too red, the greens are too blue.

I set my lighting up in a slightly more complex fashion: shoot a Gretag Macbeth "ColorChecker" card, a card with a good gray reference and 18 good color references scattered all through the "color universe". I then use the new Adobe "DNG Profile Editor" to compute a raw profile that contains "regional corrections" for all the different neighborhoods in the color universe. This can be applied easily through Lightroom 2 or Adobe Camera Raw in PS CS3.

The end result is as close to "perfect" color as one can get, which is a wonderful thing for botanists, mineralogists, etc...

[rjlittlefield]

Regarding color, remember that ultimately you're stuck with the fundamental problem of metamerism. Different materials can look exactly the same under one illumination, yet different from each other under another illumination whose spectrum has bumps in different places. No amount of calibration can remove this effect.

The calibration procedure that Joseph describes is very powerful. It'll do an essentially perfect job of putting the ColorChecker's 18 sample materials exactly where they ought to be in color space. Whether it will also put other materials in the right place depends to some extent on luck of the draw regarding bumps in the spectra.

To get the colors as close as possible to what a human would see under, say, midday sunlight, there's really no substitute for starting with light that has the same spectrum as midday sunlight. Barring that, a spectrum that is as smooth as possible is to be preferred. The last time I looked, white LEDs still had a tall narrow spike in the blue. That opens the door to residual color errors even with the most careful calibration, whenever the color standards are made of different materials from your actual subjects.

--Rik

[g4lab]

Well I was going to keep my mouth shut about LEDs because I don't think Graham is that worried about color perfection in his bee movies (oops no pun intended but that is a little funny ) I think he is more interested in the fact that LEDs can be strobed on and off in microseconds. (Although I wonder if the phosphor square that makes the blue or UVA light , white has a longer relaxation time compared with a plain LED)

But since both Joe and Rik started it I am going to jump in even if it changes the direction of the thread a little bit. On the gemology online forum there has been alot of talk about LEDs and members there have needed to be reminded that they really are not suitable for most gemological purposes even though some firms making microscopes that are for the purpose have started using them. Including some that REALLY should know better. The HRD, Hooge Raad Diamenten or high diamond council of Antwerp sells a nice scope that has an LED ring illuminator. A ring of high wattage LEDs probably starts to exceed the brightness of the usual 150 watt quartz halogen fiber optic ringlight.

But that does not cure the pitfalls outlined by Joe and Rik. Even a high CRI number if you can get one (not on LEDs but on the similarly popular fluorescent lights) and IF its an honest number may not tell the tale.

In the case of diamond and colored gemstone grading depending on where the spectral bumps are both in the lighting and the object being looked at can cause all sorts of difficulties.

Most educated people have heard that diamonds are color graded under north daylight. But that is a very imprecise and ambiguous term.

http://www.palagems.com/gem_lighting1.htm
http://www.palagems.com/gem_lighting2.htm
The above links are a thorough exposition of the subject which I greatly like because they happen to agree with my own personal prejudices.
I like tungsten filament bubbs with or without quartz halogen cycles. I also like carbon and xenon arcs. And for photography only but not spectroscopy there are a few metal halide bulbs that can deliver acceptable CRI if you did a procedure like Joe describes. They are more intense and efficient than the aforementioned sources. But they have spikes. But as the tech lamp companies continue to tune them up for projection TV I expect they will continue to improve.

http://www.ganoksin.com/borisat/nenam/s ... source.htm
Here is another post on the subject but it mentions Ott Lites which I think are just a big sales gimmick as are many other "natural" lights. JMO

http://www.adamasgem.org/giafluor.html
The gentleman that has this page makes a spectrometer for gemology and is very aware of lighting sources.

There is a brand of QH MR-16 lamps available with a blue daylight correction filter on the lamp capsule. They are called SOLUX lamps and might be worth a look.

I think white LEDs are great for flashlights and bike lights and any battery powered application where their great efficiency is an advantage. I even
have an LED ringlight made by the Swiss firm of Volpi who are very big in the microscope fiber optics business. But after buying it on fleabay I don't actually use it. It would be great for monochrome machine vision or use by some one in the electronics surface mount rework business. But for gemmo specimens I probably won't ever use it except if I need to go portable for some reason.

[rjlittlefield]

I wonder if the phosphor square that makes the blue or UVA light, white has a longer relaxation time compared with a plain LED.

That's a really good question!

As I understand it, in most white LEDs the blue spike is direct LED emission, while the rest of the spectrum is filled in by secondary emission from a phosphor [ref]. Presumably the blue spike will die out quite quickly when the current cuts off, but the rest of the spectrum may last quite a while longer. The phosphors used in CRT displays typically have persistence of several milliseconds, but there's no good reason to think that the "phosphors" used in white LEDs will be the same. In fact, http://en.wikipedia.org/wiki/Phosphor notes that

White light-emitting diodes are usually blue InGaN LEDs with a coating of a suitable material. Cerium(III)-doped YAG ... is often used; it absorbs the light from the blue LED and emits in a broad range from greenish to reddish, with most of output in yellow. The pale yellow emission of the Ce3+:YAG can be tuned by substituting the cerium with other rare earth elements such as terbium and gadolinium and can even be further adjusted by substituting some or all of the aluminium in the YAG with gallium. However, this process is not one of phosphorescence. The yellow light is produced by a process known as scintillation, the complete absence of an afterglow being one of the characteristics of the process.

White LEDs can also be made by coating near ultraviolet (NUV) emitting LEDs with a mixture of high efficiency europium based red and blue emitting phosphors plus green emitting copper and aluminium doped zinc sulfide (ZnS:Cu,Al). This is a method analogous to the way fluorescent lamps work.

Putting all this together, it sounds like the answer is "It depends".

Perhaps the bottom line is to be aware that there's a potential problem here, ask about it first, and look for it as soon as the LEDs in question actually get delivered.

--Rik