133 LPI Ronchi screen - math challenged

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Smokedaddy
Posts: 1953
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

133 LPI Ronchi screen - math challenged

Post by Smokedaddy »

I finally have my Universal BD Plan Epi-Illuminator up and running. Seem a few internal parts were screwed up. The aperture diaphragm section containing the aperture blades were misaligned, the field diaphragm sections ring did operate anything when I rotated it and was missing a part and the threads were stripped on one section preventing me from attaching the lamphouse. Thankfully I had a parts spare and had those components.

So I photographed a stack for a penny which came out fine but should of been a deeper stack. I was quite impressed actually at the quality. Then for kicks I tried a 133 LPI Ronchi screen I happened to have (shown below). I used a 5x BD objective, extension tubes and a 1.4x Canon extender on my 50D. The image below is a 100% crop, meaning I zoomed 100%, then cropped the image. Unfortunately it is something like 1276 px wide (or something) and I forgot the size limit here so I suppose that's a moot issue.

Question can a Ronchi screen be used for resolution testing, just curious? If so how.

Just to mess around I fired up imageJ and calibrated the images with my calibration slide. Since the colors obviously fade off so I wasn't sure how to actually measure the 'width' of one of the lines but came up with 9.185um and one pixel was 0.064um Does 'about' 143 pixels in width sound correct?

Image

hero
Posts: 72
Joined: Mon Jul 17, 2017 12:38 pm
Location: California

Post by hero »

If we assume:

1. The 100% crop image uploaded was 1276 pixels wide
2. The site automatically resizes images exceeding 1024 pixels in any dimension to fit within 1024 x 1024 px
3. The Ronchi screen is exactly 133 LPI
4. The definition of LPI is the number of black lines per linear inch
5. The EOS 50D has a pixel density of 4752 px / 22.3 mm = 213.094 px/mm.

Then I measured and estimated the average distance between one line to the next line as approximately 240 pixels, which gives us:

A. (1276/1024)*240 pixels = 299.063 px = 1 line pair at 100% crop
B. 133 line pairs = 1 inch real size = 2.54 x 10^4 micrometers
C. 1 pixel at 100% crop = 2.54 x 10^4 / (133 lp * 299.063 px) = 0.6386 micrometers real size = 0.0006386 mm real size.
D. The resulting system magnification is approximately 7.34868x

Smokedaddy
Posts: 1953
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

Post by Smokedaddy »

Thanks, I was trying to do it this way.

Image

-JW:

rjlittlefield
Site Admin
Posts: 23562
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: 133 LPI Ronchi screen - math challenged

Post by rjlittlefield »

Smokedaddy wrote:Question can a Ronchi screen be used for resolution testing, just curious? If so how.
As you've seen, the lines are far too wide to be used directly for anything other than checking magnification. After all, even a cheap 4X NA 0.1 objective can resolve over 300 line pairs per mm, versus your Ronchi at 133 line pairs per inch.

Of course you can use the fine texture of the Ronchi screen just like any other finely detailed subject, particularly for head-to-head comparisons between objectives.

But the most valuable use may be for evaluating chromatic aberration. It's a pretty good assumption that the Ronchi screen is B/W, no real colors in the target, so any colors that we see are due to CA. In the example that you're showing, there seems to be pretty strong green/magenta fringing, especially on the top bar. If I had that image, I'd be firing up Photoshop's lens correction filter to see if it could clean that up.

--Rik

Smokedaddy
Posts: 1953
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

Post by Smokedaddy »

Since I've never used this setup before I was looking for something to view and remembered I had this screen. It was mounted inside a 1.25" eyepiece tube supposably used to somewhat evaluate the quality of the optical train. I had a 1000 Lines/mm diffraction grating but couldn't find it. Not that it matters, I was able to remove the majority of the fringing.

Thanks ...

Image

I didn't stack enough images for this puppy.

Image

Image

-JW:

rjlittlefield
Site Admin
Posts: 23562
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

The CA correction looks very good.

--Rik

Post Reply Previous topicNext topic