Lens that accentuate chrimatic aberration for colorful effect.

I was wondering if there are any lens that distortion colors on purpose to add color.  I've seen the prism lens and filters, but I am thinking of something more like a fisheye lens than a lens ball.  Specifically, not making multiple images or facets, but excentuating a normal image with CA.

I imagine soething could be achieved digitally by tweaking a CA correction the wrong way.

SmoothOperator wrote:

I was wondering if there are any lens that distortion colors on purpose to add color. I've seen the prism lens and filters, but I am thinking of something more like a fisheye lens than a lens ball. Specifically, not making multiple images or facets, but excentuating a normal image with CA.

I imagine soething could be achieved digitally by tweaking a CA correction the wrong way.

I was thinking about this more.  Most CA looks especially bad because it's just a blue and red edge.  Maybe to accentuate the CA a lens could make a better spread separating more colors more uniformly, sort of rainbow edges.

gordonpritchard wrote:

Here are two examples:

You attach a bellows to your camera and then attach a standard magnifying glass to the bellows.

Adjusting the bellows allows you to focus. You can also use two magnifying glasses for a stronger effect. You can also make an aperture out of black paper to increase depth of field. If you cut the aperture as a shape - e.g. a star - the the bokeh will be in the shape of the star.

That is a surprisingly nice result. Kind of like fisheye, your eye can tell something isn't right, but it doesn't matter, because colors. I might have to try the diopter/magnifying glass idea with my bellows.

One potential strength of such a lens might be low light and high dynamic range.  If you aren't constrained by color reproduction, you could have a really fast lens.

Back lit tree branches are a tough shot with a conventional corrected lens, I've come to like the blue and red tinges in my woodland sunsets.

I wonder what something like this 100mm diameter 100mm focal length lens would do.  I calculate it would be an f1, if you could figure out a way to mount it without stopping it down.

http://www.novatech-usa.com/Products/Optics-Educational/LCC104.html

I bet it would be soft and colorful.

gordonpritchard wrote:

SmoothOperator wrote:

I wonder what something like this 100mm diameter 100mm focal length lens would do. I calculate it would be an f1, if you could figure out a way to mount it without stopping it down.

http://www.novatech-usa.com/Products/Optics-Educational/LCC104.html

I bet it would be soft and colorful.

That linked lens is actually a reducing lens and not a magnifying lens so it wouldn’t work.

I'm not sure this one would work with a bellows or helicoid 50m focal length 100mm diameter double convex.

https://www.amazon.com/PSC-2426929-Double-Convex-Diameter/dp/B01N5Q7WZQ/ref=sr_1_6?s=industrial&ie=UTF8&qid=1542328418&sr=1-6&keywords=100mm+convex+lens

gordonpritchard wrote:

SmoothOperator wrote:

gordonpritchard wrote:

SmoothOperator wrote:

I wonder what something like this 100mm diameter 100mm focal length lens would do. I calculate it would be an f1, if you could figure out a way to mount it without stopping it down.

http://www.novatech-usa.com/Products/Optics-Educational/LCC104.html

I bet it would be soft and colorful.

That linked lens is actually a reducing lens and not a magnifying lens so it wouldn’t work.

I'm not sure this one would work with a bellows or helicoid 50m focal length 100mm diameter double convex.

https://www.amazon.com/PSC-2426929-Double-Convex-Diameter/dp/B01N5Q7WZQ/ref=sr_1_6?s=industrial&ie=UTF8&qid=1542328418&sr=1-6&keywords=100mm+convex+lens

In my experience cheaper is better. As long as it's glass.

I mis read that last one it was 50cm focal length.

This one is an F1 100mm lens double convex.

https://www.ebay.com/itm/1pcs-Optical-Glass-100mm-Diameter-Focal-Length-100mm-double-convex-Lens/152355211014

Now how to attach to a bellows.  It might be too big for a bellows.  I use a 105 focal length enlarger lens on my bellows, and it will focus to infinity with room to spare.  I think even a 75 or 80 mm lens would, but its only an f5.6.

Maybe a short helicoid and some sort of adapter.

SmoothOperator wrote:

I was wondering if there are any lens that distortion colors on purpose to add color. I've seen the prism lens and filters, but I am thinking of something more like a fisheye lens than a lens ball. Specifically, not making multiple images or facets, but excentuating a normal image with CA.

I remember once returning a screw-on front-end 1.7x TC because it did exactly what you are asking for, because it wasn't right for the lens I was attaching it to. It was a Canon 1.7x 58mm-thread TC, I believe, on my Sony F707. I bought the Sony TC recommended for the camera, and it didn't do this.

I imagine something could be achieved digitally by tweaking a CA correction the wrong way.

Most software CA correction only corrects mean red band, and mean blue band, so it never really full corrects by original wavelength, which can only be guessed by a more clever correction, anyway, and wrongly assumed to be the same for every photon that charges a photosite, or equally distributed in a white pixel.

I would keep searching for an optical solution, with its analog treatment by wavelength.

I guess to really appreciate Chromatic Aberation, you would need to limit other types of distortion and aberration like spherical aberration.

https://www.diyphotography.net/homemade-lenses-whats-all-the-fuzz-about/

One of these aspheric lenses might give colorful images,

https://www.edmundoptics.com/f/Plastic-Aspheric-Lenses/13968/

petrochemist wrote:

SmoothOperator wrote:

SmoothOperator wrote:

I was wondering if there are any lens that distortion colors on purpose to add color. I've seen the prism lens and filters, but I am thinking of something more like a fisheye lens than a lens ball. Specifically, not making multiple images or facets, but excentuating a normal image with CA.

I imagine soething could be achieved digitally by tweaking a CA correction the wrong way.

I was thinking about this more. Most CA looks especially bad because it's just a blue and red edge. Maybe to accentuate the CA a lens could make a better spread separating more colors more uniformly, sort of rainbow edges.

CA can't give you full rainbow edges. the most you can get is a portion of the spectrum the reds on one side & blues on the other.

The more you spread the colours the less you'll see of the central normal image.

You could try a plain ruled diffraction grating. That will give you a series of images in different colours with each colour displaced according to it's position in the spectrum. More finely ruled gratings will spread the colours more. Note diffraction filters as sold for photography are slightly different, typically having several gratings, and probaly with bigger spaces between the lines.

A diffraction grating is neat.

I've been thinking about the a reason for a lens which doesn't correct for CA(or even accentuates it to exist).

One scenario is colorful scenes in low light.

The theory would be if you are shooting things that have mixed pan colors( whites, tans, browns, some purples), you'll end up with muddy looking scenes.  If you are shooting things where the colors are relatively pure, say tones of red or green, then CA won't even show up.

John Sheehy wrote:

SmoothOperator wrote:

I was wondering if there are any lens that distortion colors on purpose to add color. I've seen the prism lens and filters, but I am thinking of something more like a fisheye lens than a lens ball. Specifically, not making multiple images or facets, but excentuating a normal image with CA.

I remember once returning a screw-on front-end 1.7x TC because it did exactly what you are asking for, because it wasn't right for the lens I was attaching it to. It was a Canon 1.7x 58mm-thread TC, I believe, on my Sony F707. I bought the Sony TC recommended for the camera, and it didn't do this.

I imagine something could be achieved digitally by tweaking a CA correction the wrong way.

Most software CA correction only corrects mean red band, and mean blue band, so it never really full corrects by original wavelength, which can only be guessed by a more clever correction, anyway, and wrongly assumed to be the same for every photon that charges a photosite, or equally distributed in a white pixel.

I would keep searching for an optical solution, with its analog treatment by wavelength.

So the corrected term is aplanatic an example of which is the Petzval lens, that corrects SA and coma but not CA.

Ah.