View Full Version : Make a center filter using photoshop

22-Feb-2001, 04:09
Another fancy idea. I wonder if there is a tool in photoshop allow you to draw a gray circle then fade into a white backgroud, and then print it on transparency to achieve same CF effect.

Thilo Schmid
22-Feb-2001, 06:35

you can find a brief description of one possibility under http://www.luminous-landscape.com/panorami.htm However, Radial Gradients in Photoshop will not allow an exact simulation of the cos4-falloff

William Leviit
22-Feb-2001, 06:52
Create a gray circle and then fade the edges using the gaussian blur filter. By changing the pixel blur you can adjust the amount of "fade" you achieve.

Dave Mueller
22-Feb-2001, 10:18
If the purpose is to overcome light falloff, what about shooting a grey card, then scanning and then changing the brightness of the image in Photoshop? If you convert the image to a negative (from whatever the input was), wouldn't you would end up with a perfect correction pattern for that lens?

James E Galvin
22-Feb-2001, 11:42
I thought of this too. Looking at a center filter, there is no visible image str ucture: no grain at all, even under a microscope. Almost as if Schneider used black glass and melted it in. I have no idea how they did it. Mak ing one will have the usual grain of the printer. This will absorb just fine, but will also scatter some light. There is also the problem of any thickness variations of the film, it is not an optically ground plane parallel. I suspect you will lose contrast from the scattering, and sharpness from the poor optical surfaces. An easy first experiment: try a shot through a grey transparency, and see what it does to the image. I'll be interested in the results. If it works, then work on the radial gradient tool.

David A. Goldfarb
22-Feb-2001, 12:25
I've tried this: made a target of several radial gradients of different degrees of gradation, copied onto 8x10" film to produce a 3x3" image that I could use in a filter holder, and couldn't get one that was satisfactory. The contrast was usually too great or the gradation on film uneven. Maybe I just need a better printer or need to experiment more, but in any case, it will take a number of attempts. I also tried defocusing slightly to smooth out the printer matrix and any banding, but that seemed to make things more uneven. If anyone does it successfully, I'd be very interested to know.

I might try just taking an image of an evenly illuminated surface with the lens I want to correct, producing a negative with the right amount of falloff, then copying the negative in the right scale to produce a 3x3" filter on film.

I think the way glass center filters are made is by grinding a neutral density filter to produce an even amount of falloff correction then bonding it to a piece of glass ground in the opposite way to produce an optically neutral filter.

22-Feb-2001, 14:27
David's idea is even better. The light transparency adapter for flatbed scanner is quite even. Hold it against your lens front so you can custom make CF for different lens. Exciting!

James E Galvin
23-Feb-2001, 11:55
Some experiments last night: I have negatives from zone calibration with large a reas of grey card. With a single lamp across the room, I put the negatives in front of 6x30 binoculars, to get an aperature and magnifi cation in the range of a photo. The effects I saw were also visible with the naked eye. With a tri-X neg, the lamp is sharp, but a diff use glow obscures shadows. The glow is much less with a Portra neg, but you are stuck with an orange filter. This is scattering or diffr action. With film from a Linotronic printer, I get multiple images, diffraction from the halftone pattern. I tried printing grey from an Eps on 1270 on Tektronix Phaser transparency film, a very clear film that doesn't hurt the image that I could see. The ink is still wet an d running this morning. I examined my Schneider IIIa CF, took it out of its cell. I can see no seam at t he edge, if this is two pieces of glass cemented together I can't see it. I see the reflection from the back is dim in the middle, bright at the edge. I think the filter is a coating. I don't know why it is 1/4 inch thick. The only hope I can see is to put the homemade filter in the film holder in cont act with the film. But if you use front motions or shifts, you have to put it in off center. I use the Schneider CF on a 90/6.8 Grandagon and also with a stepup ring on a 65 /8 SA, it works fine, as opposed to older threads saying never use a stepup ring, and use only on the lens it was made for.

William Nettles
25-Feb-2001, 15:43
re: Centering filters- There is a standard lens formula (based if I remember correctly on the Square Root of 5-I could look it up), that predicts light fall off based on the angle off center-this is why if affects wide angle lenses more than normal focal length.

To make a centering filter on a computer/inkJet printter for use in the camera I don't think will work. The transparency film that is available has a surface for the ink to stick to that is about as blurry as the frosted back of a tansparency sleeve. Also there is the dot pattern--which even on a clear substrate would act as a diffuser. (Cinematographers have available a very nice diffusser that is made from black particles but is very expensive--softens without the glow)

What you can do on a computer is to make a mask for use in an enlarger. View Camera Mag describes methods to make masks for B&W printing. I've done this and they work-though they are time consuming. You have to saandwich the elements from the Light source to the lens as follows. On top the mask(transluent inkjet film), than 1/16th inch translucent plex then the neg, Lining up even the tightest masks is not at all difficult. A Centering filter made from a scanned camera neg (focused at infinity) of a blank surface should be fairly simple-though I'm betting you could just make a 1" circle in the middle of a 4x5 and then guassian blur the whole image. (Should probably bleed out to white at 3" diameter)

I own two centering filters. One designed for my Rodenstock 65mm and another Schneider IV? that I got with my used equipment. The 65mm centering filter works fine on the 90mm/f8 and the larger Schneider works on a 120mm lens, I tape them to the front of the lens.

I've learned that anything close will work just fine even for critical work and that Centering filters do make a significant improvement (architectural stuff/white walls).

Clive Kenyon
30-Jun-2001, 15:09
Hey, I'm new here so don't laugh!

Would it be possible to make a center spot filter on a glass negative in the old fashioned way by photographing a grad' grey circle on a black background?


Paul Schilliger
30-Jun-2001, 15:58
Clive, your idea is not bad, but I think the layer of gelatin will act as a soft ening and diffusing filter and produce some interesting effects. Maybe it would be worth trying the other way around (c lear center), this would perhaps produce an interesting portrait and still life filter! Just a thought... never t ried. The best way of going around a center filter is by using a Photoshop technique after the scan, but a center fil ter, when needed, is a worthy investment. They not only produce an even illumination of the film, but they als o protect the lens, from the stray light and parasite reflexions that occur inside the lens with very wide angles i n bright light.

Clive Kenyon
1-Jul-2001, 05:13
Paul, I take your point - but if you can project light through a negative to give sharp results on an enlarger, then why will it not work on a camera?

Answers on a postcard please.


Paul Schilliger
2-Jul-2001, 04:28
Clive, industrial films have extremely even and accurate layers and even so, the y would produce a diffusion effect if put in front of a taking lens. Older enlargers were equipped with sets of optica l condensers to direct the light and minimize the diffusion effect and produce sharp images. The paper's contrast wil l also reduce the diffusion effect. I am not aware of any modern standard glass plate films available on the market, s o making them yourself by painting a sheet of glass with photosensible emulsion would be far from the standards of a good optical filter, without underminding your skills! But as I said earlier, it could be worth a trial. Why don't you shoot a grey card on a B&W negative, use the neg as a neutral density filter and report here of your findin gs? (Was that too long?)

Sal Santamaura
2-Jul-2001, 10:47
T-max 100 4x5 glass plates are available at around $12.00 each.

Clive Kenyon
2-Jul-2001, 17:50
Well, that's me done folks. I supplied the idea - it's up to you now!


Mike Parrish
29-Aug-2004, 20:23
Has anyone tried the following:

1. Expose a piece of film against a blank wall with the lens at infinity and f/22 to get the proper pattern. *The formula used to calculate light fall-off may not be very accurate with modern lens formulas.

2. Scan the processed film.

3. Blur the image just enough to kill any grain or noise.

4. Adjust contrast and density as needed.

5. Use an inkjet printer on a blank or UV Wratten gel filter.

6. After testing for accuracy, sandwich the gel filter between optical glass with the proper cement. *I believe many high quality filters were made this way in the "early days", just not center filters... until now :-)

Paul Butzi
30-Aug-2004, 09:08
"Paul, I take your point - but if you can project light through a negative to give sharp results on an enlarger, then why will it not work on a camera? Answers on a postcard please. "

It won't work the way you expect because the sharp results you get when enlarging are the image of the film focused on the paper, and not a point source simply shining through the negative. That's why you need an enlarging lens.

Take a duffer piece of film, fix out all the silver, and dry it. Or, just look through the thin part of one of your negatives. Does it seem like you're looking through clear glass, or some frosted material?

Mike Parrish
30-Aug-2004, 21:06

Actually, gelatin filters carry the very best optical characteristics. Heliopan/B&W sandwiches/glues gel filters between optical glass for many of their very expensive filters even today. So quality isn't the issue unless optical anomolies are introduced by the ink from the printer or if the alignment/gluing are poorly done.

If there is a printer with black ink of optical quality, this should work. That's all I really meant to say. If anyone out there is aware of such a beast please let me know.

I've seen the highly prized Heliopan center filters made in this manor, though I don't know how they make the graduated center. It makes you wonder, doesn't it?