Cree is a company that's obscure outside a fairly specific discipline, although anybody who is involved in LED lighting will be aware they're a huge outfit, manufacturing a large chunk of the world's high power LEDs. It's therefore an organisation that has a lot of relevance to Arri's presentation on LED lighting, although Cree apparently hasn't ever sent anyone to NAB before. It's therefore probably a bit cruel to name names, as this guy wasn't an official part of the presentation, but their man did make a particularly interesting statement, which was this:
"We've solved all the problems."
The technical low-down on what Cree seems to think has solved all the problems is actually quite interesting and relevant: briefly, the situation with LEDs is that they're subject to a lot of manufacturing variations. Usually they're selected into one of several bins according to brightness, colour and efficiency. This is usually done at the point where the tiny semiconductor chip that forms the actual LED device has been manufactured into a finished package. What Cree are now doing is testing these sub-millimeter-square chips independently, then assembling several of them into multi-emitter devices. It's an approach that apparently allows much finer grading of the finished devices, with better consistency from batch to batch.
This much makes sense, and multi-chip designs can achieve considerably less irregular spectral output; various white LEDs in combination is how the Arri L7 stuff works, although I'm not sure it's quite enough to substantiate the claim that they've solved all the problems. The Arri presentation itself was excellent. Not that it covered anything that most people on the forum won't already know, but it's nice to see these things practically demonstrated sometimes. Ryan Fletcher, Arri's product manager for lighting, actually had examples of tungsten, fluorescent, discharge and LED light sources and a spectrometer with which to show the actual spectral output in realtime. Of course, we know that tungsten has a smooth but blue-deficient output; white LEDs are a lump centred on the yellow and a spike in the blue; multi-chip LEDs look like a map of the Himalayas and fluorescent can be more or less anything depending what day of the week it was. Future presenters on that sort of subject take note: practical physics demonstrations are very, well, illuminating.
Now, trivia: did anyone else know that the Academy's first ever technical publication, in 1928, concerned the effect of then-new tungsten lighting on the rendering of various colours in black-and-white photography? It's not a new problem.
NAB: Arri and solid state lighting
No replies to this topic