"Open source project seeks to create next-gen photographic lighting"
- Thread starter Bern Caughey
- Start date
Chris Adler
Veteran
Great idea.
While the idea is good, LEDs are just not a great source of light without a LOT of work.
LED light falls off pretty quickly with distance unless collimated and focused. The T-type LEDs that they plan to use, and that most LED panel people use, are very poor at directing their own light. They use internal "lenses" and a poor reflector behind the die. This is one reason that LEDs look so blindingly bright(lux) but don't actually illuminate very well(lumens).
This means that while numerically electrically efficient, their power to usable light ratio is not good.
Using the higher power LEDs is slightly better as their design is more conducive to this application and you can get collimators that fit them nicely as well as achieving up to 95% power to usable light ratio(5% light lost in refraction).
Also, their light peaks are poor, with most being in the yellow and blue ranges with almost no red.
i found while designing robotic vision systems that I could mimic more natual light by adding a couple red and IR LEDs with an array of white LEDs.
If I were them, I would look at using collimated power LEDs instead of the cheap LEDs. This is not about AMOUNT of light, it's about the QUALITY of the light.
LED light falls off pretty quickly with distance unless collimated and focused. The T-type LEDs that they plan to use, and that most LED panel people use, are very poor at directing their own light. They use internal "lenses" and a poor reflector behind the die. This is one reason that LEDs look so blindingly bright(lux) but don't actually illuminate very well(lumens).
This means that while numerically electrically efficient, their power to usable light ratio is not good.
Using the higher power LEDs is slightly better as their design is more conducive to this application and you can get collimators that fit them nicely as well as achieving up to 95% power to usable light ratio(5% light lost in refraction).
Also, their light peaks are poor, with most being in the yellow and blue ranges with almost no red.
i found while designing robotic vision systems that I could mimic more natual light by adding a couple red and IR LEDs with an array of white LEDs.
If I were them, I would look at using collimated power LEDs instead of the cheap LEDs. This is not about AMOUNT of light, it's about the QUALITY of the light.
Norman99
Well-known member
Interesting project. I just went through helping design and prototype a very simplified version of a led on-camera 3 color video light after over 2 years of doing my own research and opinions of what an ideal led light should look like. In theory everything seems possible unless one goes through the process itself, from inception to a finished product and every hurdle along the way. From what I understood is there will be a mind sharing of vast information with the project and there will be many ideas, revisions, concepts etc. One must always remember what the ultimate goal should be, and it is very easy to get carried away when many different opinions are readily available. Nothing wrong with that, but I think it's like going forward by designing a flying car that does everything including hover, rides on water and has a freshly brewed cup of coffee ready for you in time for the morning commute. Too many things at once to consider. A concept idea is most always vastly simplified by the time it is ready for mass production. I stuck to basics and minimal components necessary to get the unit done, but only after finding out the needs of the end user.
Absolutely.
Innovation is always key in a product like this. there are tons of LED panels that are designed in the same exact "me too" kind of way and they all suffer from the same problems. Mostly that problem is getting light from point A to point B without excessive design. I found that a pair of 5W LEDs could do a much better job illuminating an object 20ft away than 200+ T-type general purpose LEDs could. Brute force was not the answer in my case and I don't think it's the answer in the design of LED panels for video cameras either.
Innovation is always key in a product like this. there are tons of LED panels that are designed in the same exact "me too" kind of way and they all suffer from the same problems. Mostly that problem is getting light from point A to point B without excessive design. I found that a pair of 5W LEDs could do a much better job illuminating an object 20ft away than 200+ T-type general purpose LEDs could. Brute force was not the answer in my case and I don't think it's the answer in the design of LED panels for video cameras either.
Norman99
Well-known member
Innovation and cutting edge technology is great. Hope they also consider the final selling price of their product, as too often a fine product will fail because of it's unaffordable price. No use drooling over a product you can't afford. Make it affordable ( not cheap ) and they will come.
Hi, Morten from Rift Labs here. Sorry for being late to the party!
The objective is actually not to design the simplest possible light. That has already been designed. Neither is it to get maximum lumens for the lowest possible price. That is not our game. Sometimes you need the lightest, sometimes the most powerful, or the smallest, or the cheapest. Those are already out there on ebay, on B&H or whoever your favourite dealer is.
We are out to explore how to make video lights better by putting some smarts into the light source.
@Norman99 I agree on the risk of feature creep. The cunning plan is to think "platform device". A relatively basic light that can be expanded by firmware updates.
Pricing is a real headache. Should we rip out stuff, use lower quality materials and inexpensive manufacturing in order not to "fail"? Or is that the exact thing that makes such a product"fail"? I guess that is impossible to know in advance.
I'm curious about what you consider an affordable price.
The objective is actually not to design the simplest possible light. That has already been designed. Neither is it to get maximum lumens for the lowest possible price. That is not our game. Sometimes you need the lightest, sometimes the most powerful, or the smallest, or the cheapest. Those are already out there on ebay, on B&H or whoever your favourite dealer is.
We are out to explore how to make video lights better by putting some smarts into the light source.
@Norman99 I agree on the risk of feature creep. The cunning plan is to think "platform device". A relatively basic light that can be expanded by firmware updates.
Pricing is a real headache. Should we rip out stuff, use lower quality materials and inexpensive manufacturing in order not to "fail"? Or is that the exact thing that makes such a product"fail"? I guess that is impossible to know in advance.
I'm curious about what you consider an affordable price.