Multi-source and multi-shadow issues

[Darell]

New thread so I can stop hijacking the Creebar one!

I've built a 6-emitter desk lamp with five white and one red emitter. I will soon take that red emitter out - the idea was that it would warm up the resulting light, but the red "halo" that my whole office now has is a bit odd... and as you'll soon see, it makes the multi-shadow situation even more entertaining.

This is the modified lamp in question... shows the emitter configuration.

More on that mod here:
http://darelldd.com/light/desk_lamp.htm

Here are the shadows on the desk. Notice the red and the blue ones...

And a video of the shadow situation:
http://darelldd.com/light/video/shadows.wmv

OK... so the discussion about why those shadows appear red and blue and grey... all at the same time, is fascinating to me, so I wanted to continue it here. I also hoped to discuss the means of combining all these beams together to make a unified, single-shadow situation. Either shining light through a diffuser, or bouncing the light to mix it better.

 

[Darell]

Some great info from Don:

Multiple light sources will give you those shadow repeats if the source is bright and concentrated like those LED's. If you move the sources closer together the lines willget closer together. Obviously the closer the item casting the shadow is to the sources the more distinct the lines.

I think something like the MCPET coupled with diffusing film where you both reflect and bounce the light as well as diffuse it will help considerably.
---------------------------------------

The blue region is where the red LED has no coverage; its shadow. You will see this with your naked eye no problem. The eye and camera see the total light as "white" so in a region where the red has been removed you see a blue color relative to the white. Our eyes as well as the camera integrate all light received into a composite. When you remove a spectral element of that composite you will have a different integration and this will result in a different perception. Since a color has been removed the difference perceived is in terms of color. It is easier to understand the various regions of different grays because obviously these regions define areas of partial shadow from some of the "white" sources. Since the shadows represent areas of less intensity but of the same spectral distribution, we see areas of grey (is grey a color? Show me grey on a spectral chart or graph). We see gray relative to white. There is no gray light source here. We see blue relative to white and there is no blue light source here. We perceive the red, blue and grey regions in the image due to the absence and nature of the absent light in these regions, relative to the area where all of the light is present. This image is a result of perceiving that which is not there, as much from what is there.

I don't pretend to really understand this and my explaination is an attempt but not to be taken as gospel.

Consider a spectral graph. Light is defined and understood as energy that travels in waves or as particles (photons). We can measure this energy and the wave lengths. We know the wave length of red, green, blue, IR and UV and all the other colors of the spectrum. There is no wave length or region in the spectrum for white light. We perceive white when the integration of the available spectrum meets certain conditions. In this sense, there is no such thing as "white" light and I think this is one of the reasons there is so much to do about various light sources and the claims of "white".

It's all relative and there are no absolutes. I think this image here is a great image to consider and allow us to consider just what white light is!!

EDIT:
I missed the comment about covering up the red and seeing the blue disappear or in my perception become a gray. I think approaching this transition is a better way to consider what is happening. In the region with no red, there is no red;duh! The red band just beyond that shows a relative difference being that of red now being present. We see this red relative the the adjacent are void of the red and in the absence of red, this shadow seems blue. If you block out either the red or the blue portion the remaining portion almost becomes gray with just a hint of tint. This obvious difference in color that we perceive is due to relative comparisons and not so obvious in other relative comparisons. When we shine two flashlight beams on a wall, we see obvious differences in tint and yet these differences are not so obvious when one of the lights is used as a sole source of illumination.

 

[McGizmo]

I had another thought or example of blue light as it is something I see daily and the light source in question is likely one that we would accept as the standard for "white light". I am referring to the sun. In the clear water over here, the red and longer wave lengths of light are effectively filtered out of the water rather quickly as the light passes through. In less than 10' of "clear" water, the red has been essentially removed and a camera can't pick up a red object as red; it's just dark. Most UW photography uses artifical light sources (flashes often) to provide full spectrum illumination as well as color returned from the subject. These sources need be close by to keep from loosing the reds.

A white sandy bottom 6' down looks blue or cyan in color. Is this because the light source is blue or because the sand is blue? Nope. It appears blue because the red has been removed or filtered out.

Swimming pools are often painted white yet they appear blue for the same reason. If you spend any time in a pool, you adapt to the light and perceive the walls as white. Our eyes adapt to not only levels of light but spectral conditions as well and balance or normalize the view for ideal descrimination and intrepretation or so it seems to me. I am rambling here.......

On topic of aleviating the multi shadows, bounce and blend is the key. Darell's sample there looks to have an effective disbursing reflective material but it is behind the sources of light if I am viewing it properly?!?! I would guess little light actually encounters that reflective surface.

 

[Darell]

Good stuff, Don. But really... there's no water on my desk. Did you watch the video? Fun to play with the shadows of your hand.

On topic of aleviating the multi shadows, bounce and blend is the key. Darell's sample there looks to have an effective disbursing reflective material but it is behind the sources of light if I am viewing it properly?!?! I would guess little light actually encounters that reflective surface.

Yup, you are correct. The backs of the emitters are glued to the reflector - not used as a reflector but as a heat sink now. That reflector did a FANTASTIC job of scattering the original FL light source. Now I'm starting to wonder if I shouldn't try attaching some cross-pieces of metal to that center hump that runs the length of the reflector - and firing the emitters back up into the reflector. Might be just the ticket, though I'm not thrilled about removing all these things AGAIN! I'll bet it would smooth everything out. Heat sinking wouldn't be as good, of course - as I'd want to use some narrow metal cross pieces to avoid shadows from those.

 

[McGizmo]

Darell,
I didn't watch the video but I have played with a couple RGB sources and had one set up under a fountain where the water stream itself played with the light and caused never ending shadow dances in a sense.

 

[Darell]

OK... at this point I'm taking votes... Would I do best to remove all emitters, turn them around and fire into the reflector (cheap and hard solution) or leave everything as-is and buy the $200 diffusing material and snap it up in there behind the grille (easy and expensive solution)?

Man, this thread is yet another great example - once you break the flow of a good off-topic project, it just dies on the vine.

 

[Kiessling]

Why not design somethnig that is made for LEDs?
All the solutions you propose have the problem that your fixture isn't ideal, and for a project this complex and work-intesive I'd only go with the best.

Not that I had a solution, but I think you're trying to cure the problems of your basic assambly with lots of work and money while better starting from scratch


Another idea to avoid those shadows: ceiling bounce. A room light can work this way, and I don't theink there's a better diffusuor.

bk

 

[Darell]

Hey Bernie -

Thanks for stopping by.

Why not design somethnig that is made for LEDs?

Mostly because I lack both the skills and the resources to build a desk lamp *fixture* from scratch.

All the solutions you propose have the problem that your fixture isn't ideal

Quite true. Retrofitting with LEDs basically sucks. Sucks for car indicator lights (a never-ending project with me) and sucks for fixed lighting for many reason (inherent bulb insulation, directional lighting, etc). When fixtures are designed for LEDs they are ALWAYS better. But as it stands now, I don't have much choice but to use what is already available. It isn't like I'm Don or something!

I think you're trying to cure the problems of your basic assambly with lots of work and money while better starting from scratch

Yup. That's pretty much sums up my life in a nutshell. The good news is that this is my hobby... and even if nothing works and I spend lots of money, I normally at least learn something valuable (even if it is simply, "don't ever do that again.")

Another idea to avoid those shadows: ceiling bounce. A room light can work this way, and I don't think there's a better diffusuor.

Yeah... if I wanted to light the whole room... and if the ceiling weren't 8' away (16' round-trip!), I'd be all over that. In fact that's exactly what I've done for general lighting in the master bedroom. I have a shelf up near the ceiling. I put two Crees and one amber firing into the ceiling. Works REALLY well, and the color of the paint softens the light even more. But for my desk, I only want task lighting, and it needs to be aimable and bright. Can't do that when I bounce off the ceiling.

The one thing that works GREAT for my desk lamp? The UI! I haven't ONCE needed it in a tactical situation.

 

[Kiessling]

The UI! I haven't ONCE needed it in a tactical situation.

If it's not tactical, it ain't worth it
Is it black at least?


As I am not Don either I don't really have good suggestions for you and can only offer unqualified side remarks. :shrug:

bk

 

[kenster]

OK... at this point I'm taking votes... Would I do best to remove all emitters, turn them around and fire into the reflector (cheap and hard solution) or leave everything as-is and buy the $200 diffusing material and snap it up in there behind the grille (easy and expensive solution)?

Man, this thread is yet another great example - once you break the flow of a good off-topic project, it just dies on the vine.

Darell, this a great topic for a thread! :thumbsup: I am ordering some of the diffusion material and you are more than welcome to some samples of the various materials I recieve to try both options as long as you post pics with your opinions for folks to enjoy and learn from your experience.

Ken

 

[Darell]

Hey Ken -

Thanks for the generous offer. Which angles are you choosing? I'd considered "going in" with somebody else on a set of the diffuser sheets, and would be happy to do that with you! I'm thinking that something around 30 degrees would be a good starting point. Probably the wider the better for my project. So far I've done NOTHING... except build another light from scratch to make Bernie happy. Pictures to follow (this is NOT a replacement for the desk lamp, but is for our electronic keyboard. Still, I may try something similar for the desk.

 

[Darell]

Here's my next multi-emitter project completed... Very simple, two emitters driven off the keyboard's power supply through an nFlex. And on this one, I don't care about shadows. At least not until my daughter complains anyway!

I think if somebody were to make me a well-cooled head that could take a cluster of four-six emitters so I could bunch them up, and put ONE small piece of diffuser over it, I'd be in business. I don't need that huge head from my desk lamp. the more I think about it, the more I realize that big head is in my way more often than not. I love the minimalist way this turned out. TOTALLY out of the way.

(And thanks again to Don for helping me find all these great LL parts over the years!)

 

[McGizmo]

The diffusing film is expensive but depending on where it's placed, you may not need that much of it. You might find that 1" disks of it mounted slightly above the LED's converts the tiny hot sources to larger not so hot sources. Such an approach would likely need an additional window or barrier above as well. I made a proto flashlight some time ago where I had a slice in the reflector perpindicular to the Z axis. The idea here was you had a colimated light until a LSD disk was pivoted in place about 1/4" above the LED. This disk difussed all light direct from the LED as well as that which had been colimated by the reflector below the disk. The result was a good flood.

In your sample light here, you might consider clustering the whites about the red much closer together and then placing a disk of diffusing film above the cluster sort of unifying and blending the individual sources into a more homogenious (sp) source and larger and more singular in nature. :shrug:

I need to get a set of samples of the film myself......

 

[Illum]

Darell, your desklamp reminds me of [in my opinion] Brlux's overkill work light
quored from http://www.candlepowerforums.com/vb/showthread.php?t=153728

your 5 emitter light versus....

 

[kenster]

Ahhhh.... Now we are getting somewhere with some great ideas & thoughts on dealing with this shadow issue! A red used colser with more LED`s I have planned along with diffussion was one of my hopeful solution ideas??? Using smaller pieces of this expensive material I hadn`t thought of but I do like that idea very much, of course. :thumbsup: Also interested in the keyboard music lighting for my little Yamaha and my Kurzweil 250 honey of a Keyboard as well.

Darell, I will order some diffusion kits in various degrees and send you a PM when they arrive.

Ken

 

[souptree]

Slightly (well ok, very) off topic, but I do sometimes enjoy night hiking with a red SF L1 and a blue L1 on either side of my head. Especially on trails with lots of shrubs and low trees, it makes you feel like you're in a 3D movie! Very strange and interesting effect, and well worth checking out, IMO. I know you guys are trying to get rid of the weird artifacts, but I enjoy them! Admittedly, not on my desk when I'm trying to work....

 

[McGizmo]

Darell,

I still don't have a good rule of thumb provided by the experts but I am trying to adhere to 7sq inches of surface area per watt of LED for thermal relief. I realize that this depends on a lot more than just surface area. Convection, conduction and radiation plays a role as does the added thermal load of a driver perhaps. For long lived lumen maintenance, I don't want to experience fixed lighting LED systems getting hot!!! I have some of those cool (in design and looking) screw in LED replacements in a couple can fixtures as well as in my referigerator but damn!!! They get too hot to touch! I don't buy 50k hours or whatever on those puppies!!

 

[Darell]

Darell,

I still don't have a good rule of thumb provided by the experts but I am trying to adhere to 7sq inches of surface area per watt of LED for thermal relief. I realize that this depends on a lot more than just surface area. Convection, conduction and radiation plays a role as does the added thermal load of a driver perhaps. For long lived lumen maintenance, I don't want to experience fixed lighting LED systems getting hot!!!

Indeed. I'd so love to have some soup-can shaped lamp head that had all kinds of fins hanging out the back. I guess it won't be long before we have several commercial choices. But I want to build them now!

I have some of those cool (in design and looking) screw in LED replacements in a couple can fixtures as well as in my referigerator but damn!!! They get too hot to touch! I don't buy 50k hours or whatever on those puppies!!

I bought a couple of the "warm" 3x cree screw-in units. Probably the same as you got. I tried one in an enclosed down-light. Five minutes and I worried about it catching on fire. Damn! No way am I leaving it in there. I'm using them for oustide security lighting now. Not on for very long. Again, we have the retrofit problem where the sockets are insulated from the fixture by design!

 

[mahoney]

Possible solutions to the desklamp problem:

To avoid the multiple shadow problem you'll probably have to diffuse all the sources.

Inexpensive diffusion material can be had at your local theatrical lighting or video lighting supplier. Don't have one? Go be friendly to the technical manager of the local theatre. Gel sheets are never the same size as the holders on the lighting instruments, there's usually scrap or old cuts in a pile somewhere that could be had for some form of inexpensive barter.

Alternately, pry all the sources off and mount them much closer together. This would be easier if you'd mocked up the lamp first using heatsink adhesive tape. I tried Bond-Ply 100 in .005" thickness as heatsink/adhesive/slug isolation on my LED home lighting project. So far so good, I'm still monitoring for signs of bond failure, but I figured if my idea did not work out in general, I could pry the emitters off and save them for something else

 

[Darell]

Gel sheets... are never the same size as the holders on the lighting instruments, there's usually scrap or old cuts in a pile somewhere that could be had for some form of inexpensive barter.

Well, I'll have to check into that. Maybe even talk to Brock!

Alternately, pry all the sources off and mount them much closer together.

Yeah, my original idea was to have this big, fat spread of light by spacing them out. Not really sure what I was thinking as that is certainly not needed. But it felt silly having that huge head, and bunching the emitters all in one spot. I have plenty of surface are on my "heat sink" but I still worry about getting them too close together for heat reasons.

This would be easier if you'd mocked up the lamp first using heatsink adhesive tape. I tried Bond-Ply 100 in .005" thickness as heatsink/adhesive/slug isolation

Well, how about that. I've never heard of this. Where would a civilian like myself get my hands on something like this? I'll get on Google right after posting here...There are SO many times that I wish to mock something up, but end up just making a mess of things, and finally gluing them all down. I've tried using thermal grease, and something will invariably pop off and cook itself to death.