Clip's Metallographic Microscope Photos

L

Lagrangian

Joined
Jun 25, 2011
Messages
400
Hi Everyone,

Clip, on the Spyderco.com forums, posted more photos from his metallography microscope. Normally I would not call attention to this, but the quality of images is exceptionally high (compared to most microscope photos in the internet knife community).

He just added some photos of his diamond Lansky diamond stones, and black Arkansas stone.
http://www.spyderco.com/forums/showthread.php?54864-Under-the-microscope/page5

If you missed the thread altogether, you can start at the beginning. It has lots of awesome photos of knife edges up to 1000x magnification on a professional microscope specifically designed for metallurgical work.
http://www.spyderco.com/forums/showthread.php?54864-Under-the-microscope

If you think this is cool, please encourage him by posting a "Thanks!" in his thread on the Spyderco forums.

Sincerely,
--Lagrangian

P.S. Some sample images from Clip:

------------------------- -------------------------
Knife Edge:
aW5RO.jpg


------------------------- -------------------------
Black Arkansas Stone:
mHHNr.jpg


------------------------- -------------------------
Lansky Diamond Stone:
EyAWt.jpg


--------------------------------------------------
P.P.S. It is interesting to combine these images with Komitadjie's Grand Unified Grit Chart and Mr. Wizard's plot of the GUGC data.
http://www.bladeforums.com/forums/showthread.php/856708-The-Grand-Unified-Grit-Chart
http://www.knifeforums.com/forums/showtopic.php?tid/904090/tp/7/
1316340420-Grit_Chart_1.png


P.P.P.S. Two additional length-scales for reference: Sharpness of a modern razor-blade, and the resolution limit of optical microscopes:

According to Prof. John Verhoeven, a modern razor is sharp to about 0.4 microns.
http://www-archive.mse.iastate.edu/...te.edu/static/files/verhoeven/KnifeShExps.pdf

Also, the resolution limit for most optical microscopes (ie: without doing anything super-high-tech) is around 0.2 microns.
http://microscopyu.com/tutorials/java/mtf/spatialvariation/index.html
 
Last edited:
Lagrangian,

This is crazy good stuff :D so many THANKS! :thumbup:

Can you explain what it means with 'multilayer' issue that you discussed with Clip? I'm a bit loss on what it was about and what was missing .. :confused:
 
Chris "Anagarika";10856682 said:
Can you explain what it means with 'multilayer' issue that you discussed with Clip? I'm a bit loss on what it was about and what was missing .. :confused:
Click to expand...

Hi Chris "Anagarika",

Sure, no problem. :)

---------------------------------------------------------------------------------------
TL;DR Version:

It is very hard for the microscope to focus on everything at the same time. Some things are too close to the lense, and other things are too far away. This means there are always parts of the image which are blurry. This looks bad and hides detail. One way to fix this, is to take many photos where each photo is focused at a slightly different height, and then to combine then using software. The result is a single image that is in-focus everywhere.

(Sometimes the collection of photos is called a "stack of images" or is described as being "multilayered".)

---------------------------------------------------------------------------------------
Long Version:

You know some of that "focus effect" you see in the movies and photography? It's the one where the camera focuses on one thing in the scene (maybe an actor), but then other things in the scene are out of focus. Depending on how the optics are set-up, it is often not possible to focus on everything at the same time: Some objects are too close, and other objects are too far away.

For cameras, the range of distances for which things are in-focus is often called the "depth of field". For example, for a given set of camera settings (focal length, aperture, arrangement of lenses in the camera tube, focus settings, etc.) stuff in focus might be, say, 1 meter to 3 meters from the camera. And for different settings, it might be 5 meters to infinity. For a camera, there are lots of things you can do to modify the depth of field, including using different types of lenses. In a microscope, you usually don't have so many options as a camera for changing depth of focus. I don't know exactly why this is; I think it has to do with technical details of microscope optics.
https://en.wikipedia.org/wiki/Depth_of_field

For a microscope, the higher the magnification, the shallower the depth of focus. In fact, at high magnification the depth of focus is just a fraction of a micron. This is so ridiculously shallow, that the top ridges of scratches in the knife can be in-focus, while the bottom trenches of scratches will be out-of-focus. The scratch looks blurry because you cannot put the entire depth of the scratch to be in-focus at the same time. This gets worse and worse as you increase magnification, so it becomes harder and harder for the microscope to have everything in focus.

Here is a table of depth-of-fields for microscope objectives from Nikon's website:
The "magnification" column is not total magnification, but just the magnification of the microscope's objective lense. (To get the total magnification of a microscope, you multiply the objective's magnification with the eyepiece's magnification. For example, if the objective is 50x and the eyepiece is 10x, then the microscope will have 500x total magnification. I don't understand all of the optics, but from what I've read, it seems the depth-of-field is mostly determined by the objective. Someone please correct me, if I am wrong.)
http://microscopyu.com/articles/optics/objectiveproperties.html
MXPWg.png


So what can be done? You could mess with the optics (changing the aperture is common). But a better solution is simply to take many photographs through the microscope, where each photo has a slightly different focus. In each photo, some parts of the image will be in-focus, while other parts will be out-of-focus. You then feed this "stack of images" to a computer algorithm that cleverly combines them into a single image where everything is in focus. There are lots of technical details here (really cool stuff if you like signal processing and/or computer-vision). You can imagine that the algorithm extracts the in-focus parts from each image, and then combines them into a single image that is in-focus everywhere. (In practice, the algorithm is lot more complicated and involves solving various "deconvolution" problems.) For technical reasons, the combined image is not perfect, so you see some slight artifacts from Clip's microscope, but overall the image quality is excellent.

What is super neat about Clip's microscope is that it has a feature that automatically generates this stack of images by robotically changing the focus and taking images repeatedly. It then feeds these into software and generates an image which is in-focus everywhere.

You can think of it as stitching together a panorama using your camera, where you take a gazillion photos that are then assembled by software. Except here, the panorama isn't horizontal or vertical, it's in depth (closer and farther from the camera). And instead of panning the camera to get multiple photos, you are focusing at different distances.

Hope that helps! btw, Nikon's MicroscopyU website is one of the best that I've seen in terms of explaining microscope optics. You can go there if you want more technical details, and/or are considering buying (or just using) a high-end microscope.
http://microscopyu.com/

Sincerely,
--Lagrangian

P.S. btw, you see Clip's 1000x images with the micron-markings? A standard optical microscope has a resolution limit of about 0.2 microns. So details which are smaller than 1/5th of the width of the two lines placed 1.0 microns apart, cannot be resolved. Smaller details end up looking blurry and lacking in contrast. From Clip's 1000x image, you can see that his microscope is getting close to the 0.2 micron resolution limit. More technical details are at Nikon's microscopy website:
http://microscopyu.com/articles/optics/mtfintro.html

P.P.S. Some of you may have heard about the Lytro camera, which basically can have a huge depth-of-focus. This camera basically takes many images at once, using a micro-array of lenses. This is very cool, but there are trade-offs, because currently Lytro has to sacrifice resolution because of the mircro-array of lenses. A single big lense would have better resolution, but less depth-of-focus. If you want to, you can think of Lytro as generating a stack-of-images with a single shot, because each micro-lense has a different focus. This stack of images can then be manipulated in various ways using the Lytro software. (The actual technical details are much more complicated, and I'm not familiar with them.) That being said, there is current research (I think at Stanford University) about applying micro-lense-arrays to microscopes.

P.P.P.S. If you want knife images that have a resolution better than 0.2 microns, then you need to go to non-optical microscopes, such as scanning electron microscopes. The best microscope images of knives I have ever seen, are electron micrographs that were taken by metallurgist Professor John D. Verhoeven in his technical report _Experiments in Knife Sharpening_ (2004). If you have never seen it, you can find it here:
http://www-archive.mse.iastate.edu/...te.edu/static/files/verhoeven/KnifeShExps.pdf
 
Last edited:
Hi Chris "Anagarika"

lol! :) I guess "over explaining" is better than "under explaining" ! :rolleyes:
In any case, glad you understand the answer.

Sincerely,
--Lagrangian
 
You must log in or register to reply here.
Back
Top