Optimal aperture position?

Started 2 months ago | Discussions thread
OP E Dinkla Senior Member • Posts: 1,840
Re: Side-by-side images with laser-printed apodizer

ProfHankD wrote:

fferreres wrote:

ProfHankD wrote:

There are really three issues for printing:

  1. Optical transparency of unprinted areas. This is where inkjet doesn't do so well, because the material needs a coating to hold the ink.
  2. Optical density and continuity of the printed areas. Inkjet makes relatively smooth shades, while laser output tends to be clumpy and less continuous (discrete particles of toner). Laser toners vary wildly, but some can produce higher transmission density than any inkjet will give; some (mostly older) lasers also produce solid-shaded regions that have markedly higher density at the edges than in the middle of a region, which makes them unusable for making apodization masks.
  3. Spectral blocking. Organic dye is generally transparent in NIR, so you need pigmented ink for an inkjet. Most black laser toners are pretty good, but it does vary. Ideally, you want carbon black in the ink/toner.

Thanks! I feel little dumb as it had not occurred to me I could just make different sizes and print all of them on one sheet. It's obvious, unless after it's mentioned. Are there any other fun shapes to play with that make sense /only work near the aperture (I liked the black spot in middle to leave a lens with the most aberrated part only too)?

Do you know if there's any name for a tool to cut perfect circles? I guess I will start apreciating lens centering after this experiments.

None of this is particularly difficult, but the printed masks tend to be pretty miserable quality. Here's an example shot of a Christmas tree (out of focus) using a laser-printed 300DPI apodizer vs directly shooting at f/1.7 using an old Minolta Rokkor 50mm f/1.7:

Notice that the apodized OOF PSFs look rather grainy, contrast is dramatically lower, and there is still deformation of the OOF PSF by vignetting (and the above is a roughly APS_C crop). In other words, you see problems #1 and #2 described above plus the vignetting I also warned about.

Now I have to make it clear that LED Christmas tree lights are NOT good point sources, so some of the flaws above are due to that. For comparison, here's the same scene shot with the perfectly apodized Sony STF 100mm (and uncropped FF):

From the above you should note that the circles that have a dark ring around a bright core are the result of lenses on the LEDs interacting with the apodization... the very smooth ones are what you get from better approximations to point sources.

Nice examples!

I guess there are trade offs with any choice of printing technology for this purpose.  For example my HP Z3200 wide format 12 ink inkjet printer in B&W gloss mode will use 3 monochrome pigment inks that are perfectly neutral. No color ink is added to make the B&W print neutral. That is quite unusual for inkjet printers, the carbon pigments used are normally warm, dye grey/black inks are not neutral either and not stable in time. I can use it on transparent PET media with an inkjet emulsion that is not too thick but still not perfectly transparant.

For alternative photo printing processes the photo inkjet printers with dye inks usually make better gradients on the films to use in contact printing than pigment printers do. Neutrality there is less important as their task is to block actinic light for the process and that is often better controlled with the color inks than with the monochrome inks in the ink set. Yet for the camera filter purpose even a small percentage of colored dots might influence the image.

Using analogue thin emulsion B&W lith film with a developer that delivers a softer image may be the best choice, the PET film there has more optical transparency and the silver halide image will be quite neutral. It may still be possible to print a negative radial gradient with an inkjet printer and use it in contact to expose the B&W lith film. Some trial and error steps to get the gradient right in the lith film.

Stochastic halftoning gradients made on an image setter could be an answer too, film like the lith film but the gradient still made with high density black spots. At least the stochastic distribution will reduce moire compared to normal halftone. I do not know whether the raster point size will not introduce other optical phenomena.  However there are few companies with image setters for film these days.

Met vriendelijke groet, Ernst
No photographer's gear list is complete without the printer mentioned !

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow