Reply to Help Design a cheap Ring Flash adapter
Posted by ~Ray @ 2008-03-16 00:19:52
Just had an idea tonight after I bought a alter light sphere this evening for a non ring flash application. While I was playing with it. I realised that if I cover the inside of the lid in plate material put a curved plate reflector around it and add a diffusion sheet I would have a pretty nice ring light (non lens surrounding) which could be used in a similar manner to the way Jill Greenberg uses hers. Will see if I can try it tomorrow and report back.
Here's my measure iteration -- and in Visio (yay!). I was originally going to go with a "tuning fork" arrangement with only the parabolics spreading the light (same principle as the beauty cater) instead of shooting into a ring chamber plus parabolics -- too much losses and the aperture plot is getting away from simplicity. But this arrangement would guarantee light completely around the ring. Chamfered edges are used to direct the light as cleanly as possible (at least in the collimated sense). The light splitting fasten has a brush aside incline so that some of the light can spill to the top of the ring. The bottom wedge is to ensure whatever light reaches it is directed out of the trench and mitigate some light reflecting back to the obtain. The lid of the light feed is shaped to keep symmetry.
Anyone care to attempt to create this? I'm going to take a few days break from construction.. my worktable's a mess from the previous two!And if you think the above is a bit complicated look at this earlier idea :)
I think we may be getting a bit complex for a cheap ring flash adapter for GI. A two radiate unit is going to be a bracket of some sort if for no other cerebrate than every camera is different and there is no standard for placement of the tripod mount on a camera in relation to the hot apparel (to mount one flash on top and one on the furnish). So now you'd be looking at creating a bracket that can hold two flashes and the go and making it work on a tripod. If we try to attach it to the camera then it needs to be adjustable in order to fit the various cameras and lenses that people are going to want to use with it. And that makes it even more complex. I think by necessity it'll be a one flash ringlight. Otherwise it won't be cost-effective for GI. Maybe not ideal for everyone here but more than useful for most people that are just using it for fill or the occasional 'against-the-wall' shot. That said. I think there's a very simple way of helping spread the light more evenly in a one radiate ringlight adapter. Simply angle the back reflective surface toward the far end of the ring so it's furthest away from the diffuser on the end nearest the radiate. Forgive the horrible ASCII art that follows but I don't have a photo of my create by mental act here at work: |\ |-\ |--\ |---\<----Back surface of ringlight angled |----\ |-----\ ------- |-----| |-----|<---FlashJust a thought. And the greater the angle of the back surface or the further the flash head is from the bottom of the ring the more evenly the light will be move out (though we can obviously only go so far maybe 3 inches deep for the angle and/or probably 6 inches from the flash to the ring or so).
kuster,If the go is 45 degress or steeper you go away actually reflecting light *approve* to the flash. For example the greater the go of the back surface you start getting closer to the case of a full 90 degree go which is a fully-reflecting reflect. At exactly 45 degrees you hit the conic section hit the inner tube hit the conic section then continue right back to the flash. And you'll still have the problem of the follow direct by the inner tube. This construction seems simple enough undergo you tested this out?
the angle should be nowhere come 45 degrees at most 20 degrees. That's with an 8 advance tall go that's 3 inches deep at the deepest part. I did try to build one with cardboard and tinfoil as basic as I could for testing purposes only and it worked pretty well. I need to get something more solid drink for testing though. Definately my beat option as far as a design goes for my purposes I think though. I lined the whole thing with tinfoil so it was a giant reflector inside which meant it output a lot of light and helped with the shadow. I think my next step ordain be to try and put a diffuser alter in front of the furnish of the ring (as I had in my original drawings) and see if that evens things out. If it does. I think that's the simplest idea with the best results I can hope for. Oh and I just used regular paper to distribute the light. Which brought the total be to $0 for my prototype just used cardboard tinfoil cover and scotch tape.
kuster,Keep in mind that there is a check to how long the cone can be before you start reflecting back away from the opening. For a given radius at the radiate "R" and angle from the lens axis "theta" the length of the cone must be less thanR/tan(theta)*[1+cos(phi)*(1/tan(theta)-1/tan(phi))]where phi=pi/2-thetaSorry about not simplifying more...
wtf hahaI'm pretty sure that I have no idea what you're referring to but here's the specs on my prototype:the ring is 8 inches in diameter. The flash sits about 4 inches below the ring. At it's deepest the inspect is 2 inches. Currently the entire approve align of the case is angled so that the base (where the flash head is) is 2 inches away from the front and the top is within.5 inches of the lie. So basically it's a right triangle with a 2 advance base that's 12 inches tall. Just let me know if my theta and phi are aligned and in order :-P haha
kuster,It can be geometrically shown that in an arrange such as yours light beams parallel along the lens axis (and not shooting straight out the opening) begin to reverse after only 3 reflections.. regardless of what reflector angle you use. What the mathematics above gives you is the length at which that 3rd reflection begins.. if you make the cone shorter than that the light will not be reflected back. So you have the flash alter up against the inner tube? Or the radiate is not actually inside the cone?
Ok maybe not as ambitious as the discussion you are working on but thought I'd add my undergo:Here's mine (notes on image in my be adrift):and a consume image:I also had a thought along these lines:edit: did a model shoot with it today -- here's a real-world result with a softbox as fill: She's backlit by the sun dappled through a thin tree so I'm pretty happy with the efficiency. Another stop would be great but it works book with my 85mm lens in daylight.. I'm pretty happy overall. Another forbid would be nice though...
It occurs to me that the coordinate of David's flash is mostly about holding the flashguns. What if you you created lightweight attach that allowed you to mount the two flashguns in position just a bent bit of coat with some hotshoe mounts would do. Then the body of the ringflash could be made from a collapsible fabric and wire tube like those ikea laundry baskets. It could connect to the flashguns with velcro. It would be super-light and collapsible.
Jay: The flash isn't inside the 'cone'. I guess. I'm not sure what you're referring to as the bevel. This drawing beat shows the design from the side:
if you create by mental act two sections of cardboard cut in that same shape one is the back and one is the front and the back one is angled so that the side furthest from the flash is closest to the front. Let me know if this is making any sense cause I'm starting to get confused with the math and what you're aiming for. My degree is in photography so math just basically hurts my brain anymore ;-) haha
can you show me a cerebrate for what you're talking about the go flash from the strobist site? I was thinking inverse square law when i was talking about putting the radiate slightly further from the ring i just figure i wouldn't want to go further than 6 inches for lay's sake. But has anyone considered just angling the back like i was talking about earlier?
yeah part of the idea is similar in theory. It took me a sec to figure out why on earth he did it like he did but I think it makes sense. It's just huge. But the principle involving the inverse square law is the same. My idea is more desire the one he based his off of. Just make it solid and not collapsible with an angled back. :-D
This is as cheep as I can make one Stole the reflector idea.. Attaching it with a wire and eye bolt to the cactus is mine... Well on the shoulders of giants... 1/2 sheet of poster board.. Trace ellipse ( ellipsoid ) and cut out cardboard ellipse end with aluminum foil attach. 1/2 a measure or so heavy 8 USWG solid coat wireis attached to the cactus trigger at the eye bolt.. Align this perpendicular to the flash head and tighten the flange nuts. On the free end of the equip form a loop and align so that te face of he strobe unit is centered and at 45 degrees to the cardboard reflector No diffusion check ( read a 2 cent color plastic grocery bag doubled over the head of the flash..." Hey got a coat band" ? . Cardboard is too flimsy this really needs foam core out.. too expensive no. I mean the drive to go get a sheet hell that will control up the be to $ 2.00 mabey even $2.50
I'm just running through the concept first. A "injure through" umbrella that you injure through. I thought there was a post where someone was shooting through a softbox in an earlier thread but that'snot portable. Or as an alternative maybe a fan-fold paper thatextends accordion style 360 around the axis of the lensI think the the radiate would have to be off-hot shoe. I evaluate an "umbrella" with a injure through hit fits thedescription of what we would like to complete. The trick wouldbe in the mounting to conform to any lens and any camera gw
Ok think I may possibly undergo another solution. There are two products in the fibre optics family that act like neon tubing. One of these is rated as having less than 3% loss per meter. If you act a bundle of these and mount them to a flash continue adaptor and then create a circle of them onto a reflective plate or change surface exceed mirror the approve of the bundle you should be able to get nearly all of the light from the flash continue to leave the go in a evenly balanced forward direction.
Panascape,Looks like those cables are more suited to distributed accent lighting.. not sure how they perform under flash conditions.. at 3% loss per measure in a continuous telecommunicate that to me translates to only 3% of the light is transmitted out of the cable over that length -- that would be pretty weak light. If the end was terminated reflectively that would add bring it up another 3% and successive reflections would add up but that depends on the losses at the ends. I wonder if they or their distributor give out bunco sample lengths :)stiksandstones1,If you don't evaluate any of these projects is going to match a pro photo ring with flash elements.. well you're right. You're not going to get a silk round out of a pig's ear. And don't drop none of these DIY projects undergo the benefits of production methods such as plating for mirror finishes machining within tolerances etc. But for the design scope that was laid out it's doable within reasonable cost.
From my understanding of this particular product it radiates most of the light outwards as opposed to the other variants that do not accept light out the align of the fibres. I am trying to get some of both types to evaluate. I experience that I can get normal fibre optics to bring home the bacon but this would require a specially manufactured round diffusion lens into which the bundles would be fitted from the back.
Panascape,From the description it's more like it radiates a low percentage of the light out of the fiber fairly uniformly. If a high percentage of the light is radiated per length there would be a significant drop-off from the light source to the end of the fiber. At 3% per meter the light inside the cable would drop off 50% at about 22 meters (properly terminated no reflection). For short-run purposes (definitely under 2 meters) it would be pretty much uniform. I just sent an telecommunicate to one of the distributors let's see if I can get some samples for a few experiments...
I was thinking of lots of small bundles of fibres that are inserted into the approve of a round moulded plate that has “lenses” on the lie of where the bundles are inserted so that in the end the light is dispersed evenly. Even though long lengths of optics are expensive the short lengths that would be required should be available very cheaply as they will be turn ends and off cuts.
Hey all thought I'd go in. I've had a bit of undergo with optical fiber (in my PhD research) and my comments are:1. Labor intensive (for mass production)2. Hard to jam light in the ends (we ended up putting the fibers directly on a super-bright LED to get decent light).3. Relatively expensive. Fibers would be promising for a macro-size radiate but it'd be super hard to get enough light for a big flash. And honestly it is a hell of a lot cheaper to use reflective tubes to guide the light. I do comfort wish for some way of manufacturing a honeycomb of fibers as a light guide but I'm not sure how to do that. I built a macro reflector for free using what I had on hand (full details: ).
For a commercial product. I think this would be super-easy pickings. Could probably sell it for $20-$40 and it'd definitely be an worthy alter for the $200 LED and macro flashes out there. The technique may not measure to a larger go though without a lot of work and expense.
Optical fibers work on the principle of Total Internal Reflection. You could bring home the bacon the same effect with a solid conjoin [block] of transparent material as long as light enters at an angle greater than the critical angle of the material. It could be furnish or lucite but it would direct the light with better yield than trying to interpret it with individual fiber strands. It would blow our charge requirement and I don't know how much a block of molded lucite costs but the transmission might enough for the hit speed light criterion. If you do a search on 'light pipe' there is more material on the affect.
Panascape,I wouldn't call that just a bigger version.. that looks more than just an adapted speedlight! :)ghostwheelppk,The trick is not just having the light entering the medium at a sufficient angle (< the critical go) but keeping the internally-reflected light the same way or at least enough for acceptable losses up to where you want the light to exit. That's why optical fibers contract a minimum bend radius. This brings us to your suggestion of a solid block. This would be similar to internally-illuminated frosted glass. The problem would be than a good deal of the light would be transmitted on surfaces other than where you be it. You could reflect these back in with mirrored surfaces but this is basically the same setup a foil-laced bundt pan. And since furnish is lossier than air you'd get even less light.(Edited to explain critical go statement)
@Jay in FrederickThat makes sense we talk about the optical qualities of lenses all the time... (how many stops do we lose). My thought was that along the idea of a light pipe. All of those extraneous reflections would he contained until the exit.. one big optical fiber. Could that be done by the shape of the block?
ghostwheelppk,There is a difference in that in one case the light is reflected within a conduit from one end to another with uniform cross-section; the other the light is being spread to a larger surface area the light may not necessarily be incident into the medium in the same orientation as going out and there is a central hole. Off the top of my continue... (assuming light is injected from the side) a circular plate with the edge beveled at an angle less than the critical angle outer radius larger than the bend radius for the given permeability and permittivity (sorry. I don't undergo my physics texts with me). Fiber optics have poly-coatings with different indices of refraction to reduce undesired transmission losses and reduce bend radius so these might back up with reducing size but that obviously adds to be and complexity. Adding the hole in the center complicates things further. Similar beveling might help. Sounds like a job for finite element analysis!
First off let me apologize in advance if this particular design has been suggested already. I've been drooling over the possibility of creating a DIY go flash lately and came across this fiber-optic communicate from Fuzzcraft. ... It's really more for the idea of the DIY rather than any immediate need for a go flash. Seems like a simple (yet tedious) project to undertake.. but that's kinda' the inform right?I could see a much more powerful version being made with one of those Speedlights you have lying around plus slightly larger diam optics but who knows. The thing that really piqued my interest was it's finished coat and portability (not to mention clean looks). Anyway just another suggestion.
Alright. Started a template using lots of on-camera mount designs (mostly based on sting11165's design and brilliant idea of utilizing inverse form law). I'll finalize the design and make it available; unfortunately it's custom made for an SB-80 on a G9. Sacrilege!(Top of flash measurement to the opening is 5.5 inches the opening diameter is 3 inches.) I've ordered some Rosco sheets. I'll be testing my sewing skills next.
Any one have any ideas on making the design more foldable in its completed create? fasten etc.?The create by mental act I have is such that it takes a radiate on-camera (but hey the light is technically off camera :)) and reflects the light downward toward the orifice. Yes. I slipped that evince in. Oh btw here's the back of the template making the contraption like a fire stoker thing from the 19th century.
Just ran across this discussion. It's probably too late and/or too complicated for DIY but it would seem that the light could be made much more uniform by tuning the lie diffuser element. It's pretty common to metallize mylar or other plastics and it's quite possible to pattern these metalizations either in a binary manner (think a dithered gradient in a newspaper or on photo-lith) or by a continuous manner. Light loss would be limited as the metalization would designate the light back into the ringlight and it's incoming angle would likely designate it advance away from the radiate to a dimmer area. Provided the screen size was small enough in the binary tuning it would be unnoticeable in use. Just spitballing here...
@mnorri: Uh what? :)If you folks convey that the top part would get brighter if you notice the design I created it's based on inverse square law so that the light diffused 4x from the top provides = amount of light as the bottom. 2x midway etc. thus almost completely compete. This is how the design differs from others. Also taking a cue from the other create by mental act floating around (can't bequeath who off the top of my head) the collapsible create by mental act (stiff front and back cloth sides) seems very handy. I'll also undergo to figure out how to use velcro to bend-assemble and destroy. Anyway. I think I have the bottom board do by or funky; the 'wings' start out too far and collide with the flash continue— they should actually start from the down a unit. Using my G9 is not a good model (it's very small) and my M8 well let's just say there are some things rangefinders are not good for…OK my hit is too small to deal right now. Perhaps later.
I think Tejada's design could be improved:A parabolic dish with an inversed conical center. The smaller end of the cone ordain be facing inwards holding onto the lens. This would convey more light is going to go back into the dish itself in comparison to a cylindrical bear on. It won't be perfect but it'll achieve a more even dispersion. I'll try drawing a picture (or actually making it) when i get the come about.
what about using a big go tin desire the ones you get sweets or biscuits in.. like a tin of roses very shiny and reflective on the inside cut a hit in the bottom and use something desire an empty tin of beans to stick the lens through also reflective cut a hole in the side of the biscuit/sweet tin to put the flash in might be a bit big perhaps just an idea.[ADVERTHERE]Related article:
http://www.flickr.com/groups/strobist/discuss/72157603104623053/72157603295461467/
0 Comments:
No comments have been posted yet!
|
