Yeah, that's a great idea for use with a fixed light source.
Yeah, that's a great idea for use with a fixed light source.
“You often feel tired, not because you've done too much, but because you've done too little of what sparks a light in you.”
― Alexander Den Heijer, Nothing You Don't Already Know
Here's the results of the experimentation I went through to obtain an evenly illuminated light source... I'll break it into separate posts as there's a limit of 4 pics per post...
Attempt #1 - styrofoam box with a 45 degree reflecting bend
I took the time to adjust the angle of the reflecting bend so that the intensity of the reflected light beam aimed for the center of the final opaque light source. Having said that, the results were the worst for edge darkness. I tried lining the outside of the styrofoam box with aluminum foil, which didn't seem to help. I also tried using diffuse light before it entered the box (which made no difference).
Please excuse the drill holes in the opaque acrylic. It was the only piece i had on hand at the time (overlapping the internal dimensions of the box). Here's an unedited capture of the light source:
Dimensions of the internal styrofoam tube are 18cm x 12cm
And Here's the same light source with the levels crushed to within 10 RGB values:
Attempt #2 - mirror box with prismatic acrylic sheets
I put together a box using 98% reflectivity aluminum. My thought was that the internal reflections would create a perceptually wider light source. Here's the box:
Here's the view inside the box:
This attempt was probably about equally unsuccessful... Try as I did, I could not eliminate a distinctive hot spot in the middle of the acrylic. I tried an additional layer of prismatic acrylic in the middle of the tube, but it had little discernible benefit. For what it's worth also, the 98% reflectivity aluminum also seems to reflect colors better from the red end of the spectrum, so there's a slight tinge to the color spectrum.
Here's an unedited capture of the resultant light source.
Dimensions are 18cm x 12cm
And here's the same capture with the levels crushed to within 10 RGB values
Attempt #3 - Acrylic sheet set over a masked beauty dish
This seems to hit a sweet spot.
Here's a pic of the masked beauty dish. The mask has a diameter of 21.5cm & the beauty dish is 70cm dia.
Here's a side view of the setup:
By varying the distance between the masked beauty dish and the acrylic, the shadow it casts can be increased (by reducing the distance), or decreased (by increasing the distance).
Here's a pic of an unedited capture of the light source
Dimensions are 30cm x 30cm
And here's the same capture with the levels brought to within 10 RGB values:
This was definitely the most satisfactory setup. If you blur the unedited capture (there should be some bokeh involved in DSL capture anyway), the RGB values are within 1 of 255 values -- very acceptable, and it would not be hard to tweak the setup (I did not bother as I figured I would do that with the final design).
The edge darkness in the acrylic is due entirely to the sides receiving no light cast on it. This could be fixed with bevelling the acrylic or maybe introducing a white reflector to surround the setup. I feel confident that the unevenness in the light source could virtually be eliminated, possible with a larger mask on the beauty dish, possibly with an opaque mask or by playing with the distance--there's many options for tweaking
Last of all, mainly to satisfy my curiosity, I tried narrowing the diameter of the beauty dish by cutting a mask for the outside (32cm diameter) like this:
Then I re-ran the tests I had done with the larger diameter beauty dish... The result was that overall evenness of the illumination suffered. I tried several times, but the overall setup proved much more finicky, and definitely not worth the time...
Great info. I wonder if there is way to do this with PS. I also wonder if a 'dust off' Nikon shot would work. I am not that handy with these procedures. Nor do I want to buy Capture 1. I spent enough with Adobe CC and my D750.
And no, I don't think all of us know much about this.
Thanks for sharing!
Tin Can
Good stuff, Bert. Thanks for sharing your results.
My suggestion, and it's admittedly a shot in the dark, would be to get the best uniformity you can and lock it down. Then make an image of the light source at the intended brightness and invert it, then print that on transparency meant for overhead projectors, and secure it to the diffuser. With some trial and error that could effectively negate any residual unevenness.
Are you designing this to be a non stitching platform? Single capture of larger formats? Please let us know what your approach is.
Hi There!
I always had the best result from building a integrating sphere with leds as lightsource. What you need is a sphere with a diffuse interior. The hole should only be 5-10% of the surface area so you need quite a big one for 1:1 . The smaller tiles obviously needs a smaller backlight if the light is static with the camera and the neg moving and so the sphere can be smaller. I have used several 1-3 watt leds and it works very nice. Its best to flash them only during exposure I find since that means I can get away with smaller heatsinks. Here is a picture of a commercial one http://goo.gl/lS2wdI . Its easy to build, you need to find a sphere at the right size and right material, then split it in two and paint it with the mattest whitest paint you can find. Make a hole for the camera and a hole or several for the light(s). Then put it back together. If you find one with a nice white matte interior that works well you can of course skip the splitting with is nice.
http://s12.photobucket.com/user/barr...here6.jpg.html
http://s12.photobucket.com/user/barr...here1.jpg.html
Bookmarks