Cheap diffuse LED light source?

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: Pau, rjlittlefield, ChrisR, Chris S.

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Cheap diffuse LED light source?

Post by bugbear »

I would like to build a co-axial light rig, similar to this one.

http://www.crime-scene.com/store/A-6061.shtml

My initial design will use normal glass (not a 50% mirror).

In truth, it's not much more than a box section, light at one end, black "light dump" at the other end, 45 degree mirror in the middle, and holes above and below the mirror!

My present difficulty is the light source. Commercial COAX rigs, sold for machine vision use LED "grids".

link

I've looked at camping lights:

72 LED light

and considered arranging a fairy light strip into a grid format;

fairy light string

Does anybody have any other suggestions?

BugBear (cheapskate)

rjlittlefield
Site Admin
Posts: 21207
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

It looks to me like one of the camping lights should work. You might prefer one of the other form factors that Amazon offers, rather than the rectangular block. Or if you want something bigger, maybe gang a couple of the blocks. For this use, I suspect the fairy light strip would be less satisfactory because of low power and lots of wires.

If your subject is very shiny, then having a grid of undiffused LEDs may give prominent grid-textured reflections. That could be muted with a simple paper diffuser placed an inch or two away from the LEDs.

What sort of subjects do you have in mind shooting with this?

--Rik

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Post by bugbear »

rjlittlefield wrote:It looks to me like one of the camping lights should work. You might prefer one of the other form factors that Amazon offers, rather than the rectangular block. Or if you want something bigger, maybe gang a couple of the blocks. For this use, I suspect the fairy light strip would be less satisfactory because of low power and lots of wires.

If your subject is very shiny, then having a grid of undiffused LEDs may give prominent grid-textured reflections. That could be muted with a simple paper diffuser placed an inch or two away from the LEDs.

What sort of subjects do you have in mind shooting with this?

--Rik
Yes, I had intended to us a diffuser of some kind; my prime targets will be medals (I'm doing genealogy) and tool marks (one of my hobbies is old woodworking tools).

I'm currently using a couple of old coins (coin dealers junk bucket) as handy targets for testing.

BugBear

Peter De Smidt
Posts: 233
Joined: Thu Jan 05, 2012 9:10 am
Contact:

Post by Peter De Smidt »

I used RGB LED strips from Lee Valley http://www.leevalley.com/US/hardware/Pa ... 3629,70322 to make a light source for a scanner. Image

It is very bright, and you can control the red, blue and green channels separately. If you don't need that control, white strips are available.

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Post by bugbear »

I've found (widely available) 48 LED "dome light" bulbs designed for the car accessory market.

http://www.dhgate.com/product/48-smd-50 ... 16601.html

They seem to be around 6x5 cm, and run off 12V.

And they're cheap. Around 3-6 GPB. 4 of these would provide a large light source.

On the diffusion front, would people recommend:

* paper/cloth diffuser ?
* a sheet (or two...) of textured/privacy glass ?
* a textured reflector ?

Since I'm not using a proper beam splitter, I do need quite a lot of light, so I think the latter two sound brighter (less lossy) than the first one.

BugBear

Chris S.
Site Admin
Posts: 3580
Joined: Sun Apr 05, 2009 9:55 pm
Location: Ohio, USA

Post by Chris S. »

Bugbear,

I’ve been intending to build a coaxial lighting unit for my rig, too. When I get around to doing this, I don’t think I’ll bother using a diffuse emitter of light. Rather, I’ll look for a very bright single source, shining through an IR-blocking filter (if the source is hot), and then diffuse it, using a lens to place the light controllably on the diffusor and facilitate the use of an iris to control light intensity. I will probably experiment with two diffusion layers—well separated spatially—as I’ve often found that double diffusion produces a far better diffusion effect than single-layer diffusion. Of course, double diffusion does greatly attenuate the light.

As to what material to use for diffusion, I’ll bet I’m not the only photographer who constantly lifts found objects up to a light to see what the effect is. When I walk into a store, security probably gets a call: “Attention in isle 29—a strange man is staring at the ceiling through Tupperware.” For your project, how about a couple of pieces opal glass at $5 each?. Or perhaps just one piece would suffice?
my prime targets will be medals (I'm doing genealogy) and tool marks (one of my hobbies is old woodworking tools)
I can't help adding a truly crazy idea that I have on my own project list--mentioned at the risk of inciting massive scope creep in your project: For your intended use, you might find great value in building your axial light to perform reflected DIC. (Admittedly, this is a bit like adding, “Why not just use cold fusion?”) :twisted:

Also, while your plan to use regular glass will probably work well, if you look around, 50/50 mirrors are not very expensive.

Cheers,

--Chris

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Post by bugbear »

I making myself thoroughly confused.

I am trying to draw the design of my rig, aiming to get everything quite
close together, with an eye to minimising the size of the rig (for convenience) but also to minimise the size of light source needed, and size of mirror. Both of the latter will serve to reduce cost (*).

Most of the diagrams in net land (1) draw rays of light emerging in a cone from the light source, and spreading out as they progress via the mirror and subject to the lens. This is lovely.

But if you draw the opposite diagram, tracing the cameras "field of view", it appears that the light source you need is much larger than the one normally draw in (1).

Messing around with a round mirror (the ultimate bad subject) as the subject subject confirmed this.

I am struggling to work out how to draw a "correct" diagram and thus
work out how to make the right as small, neat and cheap as possible.

Most of the coin guys just use a free standing mirror and a free standing light; the only "rigs" I've found are:

http://www.school-of-digital-photograph ... chive.html
http://www.forensicphoto.ca/equipment.htm

I also want my rig to shield the subject from ambient light; since the coaxial light isn't bright, this is important (I've done all my work on coaxial light so far in a room at night with the room light OFF).

If anyone can help, even with principles or sketches, I'd be grateful.

BugBear

(*) half silvered mirrors appear to come in very many sizes and qualities; I'm struggling to find "the right one" for my purpose; the lab grade laser ones are small and expensive (due to their very high quality), the big ones used for teleprompters are too big and expensive (due to their large size); I have yet to find an "ok" quality mirror in the 3-6 inch size area; for the moment I'll continue with plain glass and long exposures; my subjects don't dodge).

rjlittlefield
Site Admin
Posts: 21207
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

bugbear wrote:Messing around with a round mirror (the ultimate bad subject) as the subject subject confirmed this.
...
If anyone can help, even with principles or sketches, I'd be grateful.
Yes, illuminating a mirror is quite challenging!

The following diagram comes from Theodore M. Clarke's "Brightfield Illumination of Large Field Sizes", HERE:

Image

The key element is at the bottom of the diagram,
Field lens achromat one focal length from the illumination aperture diaphragm and the aperture of the macro lens

Light cones normal to field plane
The field lens redirects light so that central rays coming from the illuminator strike the mirror perpendicular to its surface, then get reflected back along exactly the same path, being split at the mirror so that part returns to the illuminator and part goes to the entrance pupil of the imaging lens.

This arrangement requires a field lens that is somewhat larger in diameter than the largest subject. A high quality achromatic closeup adapter should work well, but unfortunately it won't be cheap. Surplus Shed has a variety of large achromats that should work OK too.

--Rik

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Post by bugbear »

Thanks for the reply - I'm using a mirror as a "worst case" proof grade coin;

Since Axial coin photography can be considered as photographing the light source in a "poor mirror", I thought I'd use a real mirror - this gives a VERY good view of the light source (if your camera lens will focus out that far; I removed the close up filter so that it would; I could then evaluate light source size and position).

I am flattered by your diagram, but the scenario I'm trying to model is the usual coin 1, as diagrammed here:

Image

Albeit with a diffuse constant source, not a flash.

Setting this up in a dark room, using "large" mirrors and "full size" soft boxes is simple, but fiddly.

My reason for wanting to draw out the various rays is to make a "rig" (like the one shown for portable forensic use) that is a neat and small as possible.

My initial sketches show that one ideal wants a high magnification lens with a long focal length; this makes the working distance high, which in turn reduces the angle of view, which decreases the mirror and light source size (which essentially have to "site" in the angle of view, which increases in size with distance.

One also needs to withdraw the light source beyond a baffle, so that the subject does not pick up any direct (non axial) light from the source.

Image

Attempting to optimise all this is as yet beyond me.

BugBear

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Post by bugbear »

I have sketched my arrangement, using a long working distance to narrow the "view cone" of the camera.

Image

The bright green rectangle is the (diffuse) light source, the tourquoise triangle arrising from it is the ambient light from the source, and the sandy triangles represent the view of the camera.

This is very depressing; the subject is only 55mm (a 2" medal) and the working distance (camera to subject) is 284 mm (which would not even require the use of a close up adapter).

The required diffuse source size is still 125mm.

I will attempt to "jiggle things" to get the size down, but it doesn't look promising; as the working distance comes down, the view cone gets fatter.

My goal is to get "everything" small, notably working distance, mirror size and light source size (although I would like subject size to as large as possible).

The dream would be to have a design parameterised in term of working distance and subject size.

BugBear

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Post by bugbear »

I am making progress (I think);

herewith a diagram of an "easy" case, where the working distance is around 150mm, and the subject 26mm.


Image

The "view cone" bounces twice, once at the reflective subject, once at the mirror.

There's lots of "room" for the mirror and light source; the sloping green line runs from the far side of the subject via the "view cone" and shows where the light source must be baffled (the black vertical line) to avoid directly (non axially) lighting the subject.

BugBear

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Post by bugbear »

Now, I think I've managed to draw the "limiting case" for a given working distance (150mm in this example). I've also come up with the idea of reflecting most of the diagram via the axis of the reflective subject; this slightly simplifies the drawing.

(I arrived at the drawing incrementally, by "binary chopping" on the size of the subject. The constraints are that the light source can be no bigger than the working distance, and the subject must be as large as possible).

Image

If that's right, I should be able to come up with a formula for it all, based on working distance.

BugBear

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Post by bugbear »

Wow! If my geometry/maths (which is rusty) is right, it all boils
down to:

The maximum subject size which can be axially lit using a light source which fits within the working distance is 1/3 the working distance.

BugBear

ChrisR
Site Admin
Posts: 8575
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I may be missing something obvious here, but if you could remove the problem of the turquoise(ambient/spilled) light then you could move the gizmo nearer the subject and get your 3 down somewhat. (Depends on meaning of "fits within the working distance" :? )
You might be able to do that by using a studio light type "honeycomb" baffle on the light source. I think the image of it, reflected in the coin would be sufficiently out of focus that you wouldn't see it.
Either solid (metal) baffles, or transparent light guides, could do the trick.

An array of tightly packed narrow angle leds may work similarly, if the off-axis transmitted light were sufficiently low that the ambient light level would become low enough to not be a problem. A standard led "video" light looks right but no doubt they're wide angle leds.

Image


The other thought is that I'm not sure why you need coaxial light. As long as you fill the "dark hole" with light, then the rest can come from a diffuse souce in much the same way as a light tent around the lens. Or in your case, illuminate the bottom of the box. The join would be out of focus?

bugbear
Posts: 67
Joined: Thu Mar 18, 2010 7:39 am
Location: UK

Post by bugbear »

ChrisR wrote: (Depends on meaning of "fits within the working distance" :? )
You post demands careful thought, but I'll just answer your question; by "working distance" I just mean the distance from the lens to the subject.

I would like my rig to be a box that will (obviously?) fit in this distance.

It is easy to set thing up when you use unlimited space (although you can use a LOT of space)...

I was hoping to be neater and more compact that this...

https://www.flickr.com/photos/dmswart/295945998/

BugBear

Post Reply Previous topicNext topic