nico wrote: >> Do I understand correctly that your goal is 16 x ~30 lumen avgerage? > > nope, my goal is to have enough lumen per led.
You didn't explain what you consider enough, or what you want to light, so it's difficult for anyone to provide a really good answer. This can be very frustrating for people wanting to help.
> in the link above, they say 108lm...
Yes, that fits. It's three LEDs per package, R, G, and B, and they're about 30 lumen each on average. Considering each individual colour's lumen may actually be much more relevant than the total or the average, all depending on what you want to light and how you want to light it.
>>> http://wn.com/VU_LM3915 >>> do you think this kind of material is controllable via DMX ? >> The question is unclear: The video demonstrates an audio based LED >> controller. You ask if the "material" can be controllable by DMX. >> I would answer no, the LM3915 is controlled by audio. > ok, > rephrasing gives : with what kind of DMX controler would you > achieve that.
To achieve *what* exactly? Generate DMX signals from an audio signal, or controlling some small LEDs from a DMX signal, or both, or neither? You write "DMX controler" which would be the unit generating the DMX communication, ie. the lighting desk - I'm not sure that's what you mean? Please try to be much more specific about your idea? I'm thoroughly confused.
Are you using the video as an example of some concrete technical parameters (LM3915, audio input, small LEDs, etc..) or just as a conceptual reference, instead of saying "I want to blink LEDs" ?
Specifically, the video shows nothing related to DMX, and nothing related to driving high-power LEDs, which is what I think your question was really about.
If I was lighting an "installation" and had some budget I would design my own LED driver with a small microcontroller to control it. The microcontroller would receive DMX from the DMX controller, or use some other protocol from whatever controller there was.
I am absolutely not suggesting that this is the way everyone else should do it, quite the contrary, it only makes sense for someone who likes to develop things and also has the right experience and toolbox - otherwise the project will quite easily become much too expensive, or take much too long, or both.
> i think i will go with links previously posted on this thread
All right! I'm glad that you've gotten some helpful suggestions! :)