When do we need focus stacking?

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

wpl
Posts: 22
Joined: Thu Jun 28, 2012 9:43 am
Location: New Mexico, USA

When do we need focus stacking?

Post by wpl »

Why do we need to use focus stacking for photographing a bug but not a tiger? Some will say that it's the large magnification required for the bug photo. But suppose the camera is smaller than the bug? Then we are not in the macro regime any longer and all should be good. This is an extreme example and probably will not really work but it suggests that a small sensor is better than a large one. Actually, this is not true but what is true?

Image

For an object with a depth comparable to its transverse size, we want RDOF to be no less than 1.0 in order to make everything appear sharp. For a low-relief object like a nontilted coin, we want RDOF to be no less than about 0.02.

As an example, consider RDOF=1.0, n=3000 pixels across the field, and a wavelength of 550 nm. Using these values in Eq. 4 and solving for h_o gives us 1.8 m as the smallest object for which we can produce a sharp photo. That's pretty large, definitely more like a tiger than an bug.

Note that RDOF does not depend on sensor size nor on any other length associated with our equipment. The wavelength is the only length scale. The reason we can't get a sharp photo of a bug is that its size, measured in units of the wavelength, is much smaller than the number of pixels. That is a strange result, not what I expected. Is there something wrong with this analysis?

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

This seems OK, at a quick glance. Agree an interesting result.

As you've noted everything depends on lambda, the wavelength of light, which is constant regardless of the subject, lens or sensor/pixel size. Lambda does vary with color and material medium tho.

This is why electron microscopes have such good resolution, the wavelength of an electron is 1.23nm, so over 400 times shorter than green light.

Will be interesting to see what the optical gurus have to say!!

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

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

Post by rjlittlefield »

I have not stepped through the math, but your summary of the result certainly sounds right.

DOF and resolution on subject are both determined by lambda and the subject-side NA.

Sensor size does not matter, at least until the sensor gets so small that resolving pixels on the sensor adds to the challenge.

In electron microscopes, another effect of the short wavelength is the scope can run with a narrow aperture (large f-number) while still retaining good resolution on sensor. This effect is what's responsible for the large DOF of SEM images, compared to optical.

--Rik

wpl
Posts: 22
Joined: Thu Jun 28, 2012 9:43 am
Location: New Mexico, USA

Post by wpl »

Here is how I think about this now. There are only two lengths involved, the object size and the wavelength lambda. To get good resolution, the ratio size/lambda must exceed n, the number of pixels across the field (I am ignoring small numerical factors here). If, in addition, we want good depth of field, we need another factor of n. Therefore, the ratio must exceed n^2, which is the number of pixels.
According to this simple theory, an object the size of a bug meets the first criterion but not the second. A sharp photo can be obtained with focus stacking. A tiger meets both criteria and doesn't need stacking. Just snap the photo. An object much smaller than a bug will meet neither criterion and we are out of luck. Our only hope is to use a shorter wavelength.

Post Reply Previous topicNext topic