Mixing Chamber Design

[distributed]

Recently I have become interested in the design of mixing chambers for providing diffuse illumination in an enlarger. I am a total newcomer to the area, so I thought I'd start with a simple experiment.

I have built a simple paper version of the LPL mixing chamber shown at the bottom of the page here: https://theonlinephotographer.typepad.com/the_online_photographer/2010/09/lpl-co.html

Clearly, using thin paper is not an optimal choice for actually working with the thing, but it was what I had at hand and so I went for it. You can see the result in the pictures attached to this message. The base area is 6x6 cm and covered with tracing paper, also not an optimal choice, but what I had at hand.

I put the paper mixing chamber on top of my negative carrier, shone an LED into it and measured brightness on the baseboard. I measured more than one stop of falloff between the center and the corners, very clearly too much.

I can't claim that I expected my cheap construction to work well. What surprises me, though, is that it uses what I think is the same geometry as the linked LPL design which certainly has to be much better in order to be actually used. I cannot work out what makes the big difference. Maybe someone here has an idea?

Performing some more research, I stumbled across this page: http://www.glennview.com/durst.htm
The author builds all kinds of contraptions for darkroom use, pretty amazing stuff. On the linked page, there are multiple mentions of "variable density diffusion sheets", the author even makes these sheets himself. This got me thinking: Is it possible that mixing chambers like e.g. the LPL one only achieve even output at the last moment, through the mentioned variable density diffusion sheets? I understand them to be roughly equivalent to a diffusion sheet with an inbuilt center filter. Can someone familiar with the design of mixing chambers, ahum, enlighten me?

 

[DREW WILEY]

Making an efficient mixing box takes some experimentation. You can temporarily use something like white styrofoam sheet as a temporary mock-up before going to more serious material. Putting paper or cardboard in there sound like a good way to burn down the darkroom ! If you want to pay someone else to make it, yes, Glenn Evans would be a good choice.

The diffuser below the box is a second question. You need to tailor it to match the illumination of your preferred enlarging lens at its typical working aperture. Acrylic sheeting can be ground convex to do this - a tricky task if you've never done it before, so you might want to have Glenn do that kind of thing to, provided you first measure the amount of falloff, which itself takes some description how to properly do.

 

[distributed]

Thanks, @DREW WILEY, for your insightful answer.

What kind of material would be ideal for the interior in your experience? I will pay attention to heat effects. Fortunately thermal power levels are rather low with LED lighting.

I imagine that working an acrylic sheet into the desired shape is art form. Do you think that it's common design practice for mixing chambers to include a variable density element at the output?

 

[Nicholas Lindan]

What you are trying to make is an Ulbricht Sphere - https://en.wikipedia.org/wiki/Integrating_sphere

The geometry doesn't have to be a sphere, any closed surface will work, especially as photography requirements are quite loose.

The key to the whole thing is that the inside of the box/sphere has to be a highly diffuse and reemissive (reflective) surface. If you would take your paper mock-up and paint the interior with a bright-white matte spray paint you would find it works well. You can buy special paints for optical work, but pure white latex house paint - the sort that is then tinted at the paint shop - works just as well at visible wavelengths.

If the integrating box is well designed there is no need to add a diffuser. This does, however, require the box is quite a bit larger than the negative size.

For a diffuser, if needed, try theatre 'gels' https://www.bhphotovideo.com/c/product/43770-REG/Rosco_RS11611_116_Filter_Tough.html

 

[distributed]

I haven't thought about this way, as an approximation of Ulbricht sphere, but it is a fitting idea. I appreciate the conjured image

Thanks for the tip about the latex paint, this sounds like something very manageable. I'll have to look around a bit for the diffuser material, unfortunately it's hard to come by as cheap options as B&H in Switzerland. It's always useful to know a direction to search in, though.

Interesting you should mention that you can also get by with an well designed and oversized box instead of with some variable density element. Judging from my measurements I would say my cheap paper mockup could be even enough for 24x36 negatives, even though it is way to dim.

 

[btaylor]

A couple of thoughts on the diffusion material- it is a standard material on film shoots, so you should be able to find it at film or theatrical supply houses in any country.
Also, when I put together an LED light source for my 8x10 enlarger I found that an additional layer of diffusion material solved my unevenness problem with LED hot spots. One layer was spaced off the diffusion about an inch, then another an inch further out from the first layer. It solved the hot spot problem. I don’t remember the strengths of the diffusion- they usually come in whole, half and quarter strength. I got all three and experimented- ended up with even illumination and not too terrible light loss.

 

[Kilgallb]

I used a cylinder with a LED pot light at the top and a diffuser at the bottom. The pot light had a built in diffuser. The side walls are painted white. I call it my coffee can enlarger light. It fits over top of my old beseler 45mx.

 

[distributed]

@btaylor Good point about which stores to find suitable materials in. I also suspect that a second, offset layer of diffusion will take care of a lot of unevenness. You'd however want a rather transmissive material, otherwise you're losing a lot of light. I will have to invest some research into this. For the conversion of an enlarger to LED lighting I have used light shaping diffusers. They look like they would have quite interesting characteristics, especially high transmissivity. However I only have some samples left and as a private person I cannot order them.

@Kilgallb I like the name! What's the ratio of height to diameter in your design? Assuming good reflectivity of the walls I suppose that a longish tube, be it of round or recangular cross section, is a good design.

 

[Nicholas Lindan]

@btaylorYou'd however want a rather transmissive material, otherwise you're losing a lot of light.

In theory, a diffuser won't cause loss of light. Light hitting a spot on the diffuser has an equal chance of going off in any direction. So half of the light hitting an ideal diffuser will go straight back to the mixing chamber proper and half will go on to the lens-side of the chamber - ideal opal glass. No light is lost (theoretically) as light bouncing back into the chamber illuminates the chamber and bounces around until it has a chance to hit the diffuser and get another chance at exiting.

I started out with an old Federal enlarger. The lamp housing, er, chamber, was a steel cylinder about 4" across and 12" long. The inside of the cylinder was painted with matte silver paint. The diffusion system was made up of an opal enlarging bulb at one end of the cylinder, a sheet of frosted glass about 4" above the negative and a sheet of opal glass just above the negative carrier. This enlarger was a series of 3 mixing chambers: 1) the PH211 opal enlarging bulb itself; 2) the chamber between consisting of the bulb, 8" of cylinder and the frosted glass; 3: the bottom chamber consisting of the frosted glass, 4" of cylinder and the opal glass above the negative. Each chamber 'leaked' its light into the next one.

 

[RalphLambrecht]

Recently I have become interested in the design of mixing chambers for providing diffuse illumination in an enlarger. I am a total newcomer to the area, so I thought I'd start with a simple experiment.

I have built a simple paper version of the LPL mixing chamber shown at the bottom of the page here: https://theonlinephotographer.typepad.com/the_online_photographer/2010/09/lpl-co.html

Clearly, using thin paper is not an optimal choice for actually working with the thing, but it was what I had at hand and so I went for it. You can see the result in the pictures attached to this message. The base area is 6x6 cm and covered with tracing paper, also not an optimal choice, but what I had at hand.

I put the paper mixing chamber on top of my negative carrier, shone an LED into it and measured brightness on the baseboard. I measured more than one stop of falloff between the center and the corners, very clearly too much.

I can't claim that I expected my cheap construction to work well. What surprises me, though, is that it uses what I think is the same geometry as the linked LPL design which certainly has to be much better in order to be actually used. I cannot work out what makes the big difference. Maybe someone here has an idea?

Performing some more research, I stumbled across this page: http://www.glennview.com/durst.htm
The author builds all kinds of contraptions for darkroom use, pretty amazing stuff. On the linked page, there are multiple mentions of "variable density diffusion sheets", the author even makes these sheets himself. This got me thinking: Is it possible that mixing chambers like e.g. the LPL one only achieve even output at the last moment, through the mentioned variable density diffusion sheets? I understand them to be roughly equivalent to a diffusion sheet with an inbuilt center filter. Can someone familiar with the design of mixing chambers, ahum, enlighten me?



View attachment 301704
View attachment 301705

all I can suggest isto get ahold of a Meopta mixing chamber. They are made of Styrofoam and have amazingly equal illumination. Their design seems to be a winner!

 

[Kilgallb]

@btaylor Good point about which stores to find suitable materials in. I also suspect that a second, offset layer of diffusion will take care of a lot of unevenness. You'd however want a rather transmissive material, otherwise you're losing a lot of light. I will have to invest some research into this. For the conversion of an enlarger to LED lighting I have used light shaping diffusers. They look like they would have quite interesting characteristics, especially high transmissivity. However I only have some samples left and as a private person I cannot order them.

@Kilgallb I like the name! What's the ratio of height to diameter in your design? Assuming good reflectivity of the walls I suppose that a longish tube, be it of round or recangular cross section, is a good design.

Mine is 6.5” diameter and 7” tall.

 

[Javier González]

I tried once to build a 6x6 mixing chamber for a Durst 305, just upscaling the design of the 35 mm one. But I failed to obtain a uniform illumination on the negative plane, the corners were darker than the center. The culpruit is the diffusive material I used; the original one is thicker in the center an produce a very even illumination. Recently I have read this can be sort out by using an acetate printed with a radial density gradient pattern above the diffusive plate; some try and error would be neccesary I guest.

 

[Nodda Duma]

I second Nicholas’ approach of viewing as building an approximation of an integrating sphere. A couple of additional points:

1) ensure no direct path between light source and output port. Typically means a 90 degree angle or greater between the two, with a blocking baffle or feature between the two.

2) use white spray paint to flock the interior surfaces. The pigment used for your paint should be titanium oxide for uniform reflectance across the color spectrum (check the paint’s MSDS).

 

[distributed]

Thank you all for the helpful information. It will helpful in my future experiments

@Javier González Do you have a link for that? Are you maybe referring to my printed center filter?

 

[DREW WILEY]

True light integrating spheres seem to reduce the overall luminance quite a bit. And a real mirror box is a little bit trickier to design than a white one. If the heat is not unreasonable, white styrofoam sheet like that food containers are moulded from seems to work well. I once used Gatorboard plus white paint. Official (if old-school) spectrophotometer paint comprised of strictly barium sulfate pigment is very expensive - around $200 per half pint last time I looked. But I stumbled onto something at the art store which was around 95% barium sulfate, and only 5% titanium dioxide, for around $10 for the same amount. Haven't seen any color shift yet. No, that kind of product does not exist in spray cans.

(Jason - MSDS sheets won't give either exact proportions or reflectance ratings; but the highest reflectance I've ever seen for titanium dioxide was around 92%; and most paints are now blends of white pigments anyway. CaCO3 "adulterant" pigment discolors over time).

As far as diffusers between the mixing chamber and carrier - try to get the light as diffuse as possible via the chamber alone. Then what I like to use as the diffuser is thin (1/16th inch) "Sign White" Plexiglas or comparable other brand acrylic. It transmits more light than ordinary white acrylic, and dramatically more than opal glass.