What's new

DE razor geometry. A system for measuring aggressive razors.

Method of quantifying rigidity: This is an example of the Tatara razor. The blade cantilever (x=L=2 mm) is not measured, but you can clearly see how important the blade cantilever is as x is raised to the power of two. The karve razor needs to be slightly differently supported, because the top cap extends past the base plate support, creating a fulcrum support point.
1623678016157.png
 
Method of quantifying rigidity: This is an example of the Tatara razor. The blade cantilever (x=L=2 mm) is not measured, but you can clearly see how important the blade cantilever is as x is raised to the power of two. The karve razor needs to be slightly differently supported, because the top cap extends past the base plate support, creating a fulcrum support point.
View attachment 1281603
Good explanation. In this illustration, the blade cantilever is what the wiki has coined "free-end distance". If this measurement is used, then the linear regression classifier becomes

x*blade exposure + y*blade gap + z*blade stiffness where x+y+z=1

What do E, I stand for? Is P the pressure applied to the razor? Can the formula for blade stiffness be adapted for other razors by setting L different from x, let's say
  • x is either the length of the blade not sitting on a(n open) comb or the distance between base plate support end and blade cutting edge
  • L is as depicted in the figure
 
Good explanation. In this illustration, the blade cantilever is what the wiki has coined "free-end distance". If this measurement is used, then the linear regression classifier becomes

x*blade exposure + y*blade gap + z*blade stiffness where x+y+z=1

What do E, I stand for? Is P the pressure applied to the razor? Can the formula for blade stiffness be adapted for other razors by setting L different from x, let's say
  • x is either the length of the blade not sitting on a(n open) comb or the distance between base plate support end and blade cutting edge
  • L is as depicted in the figure
E is the module of elasticity, I is the moment of inertia and p is a unit force (not pressure)
One more model would be needed to capture the affect of the top cap protruding over/past the base plate. I can make an example.
 
Here is the first picture I took with the digital microscope at 1600 x 1200 of my Karve SBD
View attachment 1281531
Interestingly, I didn't use any additional lighting, these shots were taken using the LED light from the microscope at full power. The picture below was taken at slightly higher magnification, but the focus is very clear without any stacking.
View attachment 1281532

Holy pixels Batman. That's definitely way sharper than the prior 1mp images. What was really interesting about this was how much "non-resolution" parameters trump megapixels for measurement accuracy. I think 2MP is a massive improvement, and if the 100% of the photo is comprised of the target area, there's more contrast with the background, and there is a ruler for scale, we're basically at 0.1mm accuracy. Three 3 things impact measurement accuracy of the 1MP picture vs the 2MP picture:

1. Backlighting/Contrast: I was really surprised how much of a difference backlighting made. In the 2mp photo without backlighting/contrast the grey of the blade blends so closely with background under high zoom, it's hard to see where it begins and ends
2. Target Area %: In the 1 megapixel picture, the target area is 100% of the photo. The target area consists of the two points on the razor profile where shave plane is tangent to the profile, the top of the baseplate, and the razor apex. In the 2 megapixel photo, the target area is only 5%-10% of the photo, so although the resolution doubled, the number of pixels in the target area is much smaller. To describe the problem more precisely, the blade is 20 pixels tall in the 1mp photo, and only 10 pixels tall in the 2MP photo. If the target area was 100% of the 2mp photo, the blade would be 40 pixels tall.
3. Scale: In the 1mp picture, I could use the blade for scale with the assumption that it was 0.10mm (they actual range from 0.09-0.1mm). In the 2mp pixel photo however it's hard to see where the blade begins and the background ends. A steel ruler with 3mm visible would increase the accuracy in all cases. The more pixels in the object used for scale, the more accurate all measurements will be.

Taking a look at the pictures below show why the Plugable USB microscope will give even super high megapixel cameras a run for their money. These pictures explain it better than I do:

proxy.php


proxy.php
 
Last edited:
@Dovo1695
Rambling nonsense:
I may have failed completely at getting photos, but it appears some of my theories are proving to be correct.
Backlighting in itself may not be necessary. A semi enclosed makeshift box painted white might do. Or a translucent material so supplemental lighting con be moved around. Use your mobile's flashlight mode.
Early on, I made some calculations about how the width of the area being photographed compared to available pixels and determined it wasn't going to be great. As you said, if he can get it all the way to 1:1, blade width will be 45 thick. I calculated it will be closer to 0.75:1 max based on the width of the sensor alone. And blade exposure will be something less. The reason that matters more than calculations indicate is because of the optics and how jpeg affects those few pixels. I can see grey pixels 4 or 5 pixels above the blade, are those added or are the blue one below it the ones added? Is the real edge somewhere between? That could affect everything you measure 10%! If you are happy enough with the accuracy of the line with imageJ, then great.
A translucent ruler that some of these microscopes come with placed near the plane of focus would probably work better. It would be illuminated by the microscopes light source and not cause reflection like steel. Maybe painting your cut blades thickness a contrasting color if you have a dark background?
TLDR: In spite of my math focused brain, I am impressed with what this thing can do for the cost.
Did you order your microscope yet? :a30:
 
I may have failed completely at getting photos, but it appears some of my theories are proving to be correct.
Backlighting in itself may not be necessary. A semi enclosed makeshift box painted white might do. Or a translucent material so supplemental lighting con be moved around. Use your mobile's flashlight mode.

Actual backlighting definitely isn't necessary; there are a 100 ways to skin that cat. The highest contrast occurs when the background is white and is also light source. You can do this in one of 3 ways.

  • You can make the background a light source by making it translucent and putting a light behind it (pure backlighting)
  • You can shine a shine a light source directly on the background so it becomes a reflective light source.
  • Alternatively, you can do both at the same time.
I tried all 3, and I liked the results from doing both at the same time. Bear in mind I was using a cheap 2mp external webcam so the pics are lousy but it shows the contrast well enough. The first pic here is pure backlight, the second pick is backlighting with the light switch in the room turned on, the last pic is with no backlight and just the lights in the room turned on.

Backlight Only

proxy.php


Backlight + Room Lights
proxy.php


Room Lights Only
proxy.php


Here is my setup. The base is a 3 way mirror that I built for giving myself haircuts. The mirrors aren't really relevant to the backlighting, it just happened to be an object the right height that was in arms reach of the shop light I wanted to use as a light source. The light source is a shop light. The platform is a piece of 3/4" float glass I use for flattening woodworking planes. The filter is a couple pieces of legal paper. The camera is my 2mp external webcam. The wedges allow me to adjust the tilt of the razor. It took about 5 minutes to throw together.

If I was building something purpose built, I'd use a light source with a dimmer for the backlighting. I don't think backlighting is strictly necessary though.

proxy.php
 
Last edited:
I can see grey pixels 4 or 5 pixels above the blade, are those added or are the blue one below it the ones added? Is the real edge somewhere between? That could affect everything you measure 10%! If you are happy enough with the accuracy of the line with imageJ, then great.
A translucent ruler that some of these microscopes come with placed near the plane of focus would probably work better. It would be illuminated by the microscopes light source and not cause reflection like steel.

I agree 100%. I definitely think having a ruler that's on the same plane as the razor profile is critical to accurate measurements. If the thing you're using for scale is 10 pixels, and you're off by two pixels you're off by 20% on everything else you measure. If the thing you're using for scale is 100 pixels, and you're off by 2 pixels, then you're off by 2%.

Using backlighting and overhead lighting seems to improve the "reflectiveness" problem of metal quite a bit. Plastic rulers would work too though.


Did you order your microscope yet? :a30:

Yup. My kids are going to give it to me for Father's day. I can't wait to take it out into the backyard with them this summer and look at bugs. 😃
 
Last edited:
Yup. My kids are going to give it to me for Father's day. I can't wait to take it out into the backyard with them this summer and look at bugs. 😃
At least it gives you a chance to get your setup figured out and if the resolution proves too low you can use it for more important pursuits. :thumbup1:
 
I am thinking you could move the glass sheet to the bottom(raised slightly), put paper over the mirrors and aim the light in from the side at an angle. Remember light falls off at the square of the distance, you don't need a dimmer if you locate the light somewhere it can be move back. :a12:
 
Method of quantifying rigidity: This is an example of the Tatara razor. The blade cantilever (x=L=2 mm) is not measured, but you can clearly see how important the blade cantilever is as x is raised to the power of two. The karve razor needs to be slightly differently supported, because the top cap extends past the base plate support, creating a fulcrum support point.
View attachment 1281603
Calculation of stiffness and deflection for a razor head where the top cap extends past the base plate support (e.g. Karve, Rex etc) Values are not measured.

1623765451915.png
 
My USB microscope should come in the mail tonight. I haven't even taken a picture yet and I'm already looking to upgrade it. I might be pioneering a new acquisition disorder. 🤣

It seems much better than the one that comes with the Plugable microscope. I might have to buy one of these. Hard to beat for the price.

proxy.php

proxy.php
 
Last edited:
Oh man, now you are opening a box of worms! One of the reasons that I didn't like the microscopes on the cheaper end was the lack of a geared rail. Better throw a couple stick on bubble levels in the cart as well.:001_cool:
 
I got my microscope! I just took my first shots. First Impressions:
  • The stand is awful. Changing the focus requires physically turning a dial on the camera. Turning the dial moves the camera. When you let go of the camera, it moves again and changes the focus. That aftermarket stand is looking better and better.
  • The stand is awful part 2: Ideally you want a completely stable camera moving exclusively in the z axis (up/down). With the supplied stand, it moves on all 3 axis at once, making setting up the shot a nightmare. Touching it to refocus it causes you to have to start all over again.
  • I have no idea how @Rosseforp took the images that he did. There may or may have been magic involved.
  • Figuring out the lighting for this thing is going to be challenging. It's the light hungriest beast I've ever seen. There are lights on the microscope itself that can be adjusted, and with them 90% dimmed, but the room lights turned on, the camera interprets this as the "Utter Blackness: the complete absence of visible light".
  • Image stacking is going to be tough without a decent stand. The camera moves too much when focus is being adjusted.
  • Using Picolay to stack images is even easier than I'd hoped. It literally takes 20 seconds.
  • The microscope has a 1.5cm long transparent plastic cowl that prevents you from getting too close to the subject. The photos below are as close as it's possible to get without removing that cowl with a saw. It appears to be permanently affixed.
  • On balance, I've very pleased with it. I just have to figure out the lighting and figure out a more stable stand.
Here are my first pictures. The first two are the pictures I took, and the third is the focus stacked image. At any rate, the photos are trash, and the lighting is terrible, but it's a very close picture, and the focus stacking really worked! Focus stacking was a brilliant idea @Esox; I had never even heard of it until you suggested it. I think the clarity of the stacked picture will be dramatically improved when the camera is stable between shots. When the subject is moving during the stack, I think it makes it very hard for the software to do it's job. We shall see.

proxy.php


proxy.php


proxy.php
 
Last edited:

Esox

I didnt know
Staff member
I got my microscope! I just took my first shots. First Impressions:
  • The stand is awful. Changing the focus requires physically turning a dial on the camera. Turning the dial moves the camera. When you let go of the camera, it moves again and changes the focus. That aftermarket stand is looking better and better.
  • The stand is awful part 2: Ideally you want a completely stable camera moving exclusively in the z axis (up/down). With the supplied stand, it moves on all 3 axis at once, making setting up the shot a nightmare. Touching it to refocus it causes you to have to start all over again.
  • I have no idea how @Rosseforp took the images that he did. There may or may have been magic involved.
  • Figuring out the lighting for this thing is going to be challenging. It's the light hungriest beast I've ever seen. There are lights on the microscope itself that can be adjusted, and with them 90% dimmed, but the room lights turned on, the camera interprets this as the "Utter Blackness: the complete absence of visible light".
  • Image stacking is going to be tough without a decent stand. The camera moves too much when focus is being adjusted.
  • Using Picolay to stack images is even easier than I'd hoped. It literally takes 20 seconds.
  • The microscope has a 1.5cm long transparent plastic cowl that prevents you from getting too close to the subject. The photos below are as close as it's possible to get without removing that cowl with a saw. It appears to be permanently affixed.
  • On balance, I've very pleased with it. I just have to figure out the lighting and figure out a more stable stand.
Here are my first pictures. The first two are the pictures I took, and the third is the focus stacked image. At any rate, the photos are trash, and the lighting is terrible, but it's a very close picture, and the focus stacking really worked! Focus stacking was a brilliant idea @Esox; I had never even heard of it until you suggested it. I think the clarity of the stacked picture will be dramatically improved when the camera is stable between shots. When the subject is moving during the stack, I think it makes it very hard for the software to do it's job. We shall see.

proxy.php


proxy.php


proxy.php

Yep, that works. The background makes it difficult to see the resolution of the very edge of the blade. Was the razor sitting directly on the background or suspended above it? Just thinking loudly here but it might work better if the razor was placed on a sheet of glass or something clear, maybe 4-6" above a brightly lit white background. That might eliminate the texture from the background appearing and would help with the shadows. Given the extremely short focus plane it might not matter at all, but you get what I mean. Thinking about it more, the background texture I'm talking about only appears in the stacked final image so it might be artifacts caused by the software.

The camera seems to do a good job, the stacking works but the background makes it fuzzy and confused but you're definitely making progress.

Your R41 has room a shim! lol
 
Wow that is pretty cool. When I started this thread I did not predict the directions it would go. What is interesting in that picture is that there looks like there is space underneath the blade. Its tight up against the cap. Maybe further down the razor it touches the plate.
 
Your R41 has room a shim! lol

It really does. That's a 0.12mm gap. I finally have photographic proof that my R41 is a worthless piece of *$&%#!@ dog #*#*. It's only redeeming quality is that it's so poorly made, and so unpredictable, that it's the perfect shaving coach. It's easier to shave with piece of broken glass than my R41. :mad:

What is interesting in that picture is that there looks like there is space underneath the blade. Its tight up against the cap. Maybe further down the razor it touches the plate.

It does. This is the terrible side of the razor, the other side is just lousy. One side has play of a full 0.10mm the whole way across the blade, the other is more or less flush to the comb like a Fatip. You should see @Esox fatip. His was cast bent in a completely different way. The R41 is arguably the only razor on the planet that make the quality control at Fatip look positively cutting edge.

Just thinking loudly here but it might work better if the razor was placed on a sheet of glass or something clear, maybe 4-6" above a brightly lit white background. That might eliminate the texture from the background appearing and would help with the shadows.

I agree. These were just the test shots I took when I opened to make sure it was working, and to test the focus stacking. I set this up a few days ago to test backlighting, and I think it will help. Even with the backlighting, getting the lighting right is going to be a fair bit of trial and error.

proxy.php
 
Last edited:
I'm wondering if this is somewhat designed. The cap bends the blade and the blade is held tight to the cap through the blades spring tension. This is why they can get away using cast Zamak. The tolerances don't have to be that tight. It also might explain the blade chatter that many people complain about.
 

Esox

I didnt know
Staff member
You should see @Esox R41

FTFY :)

My Fatip is a work of art. My R41 on the other hand...

The comb is bowed outwards. Look at the bottoms of the center 4 teeth.

IMG_2792.jpg


Neither side of the base is square.

IMG_2796.jpg


The top of the comb is the same as the base of the comb. The razor will rock across the 4 center teeth.

IMG_2798.jpg


IMG_2786.JPG


Someone once blamed me for buying an R41 clone base and thats why its poorly cast. That however, wasnt the case. It came directly from Muhle.

IMG_2762.JPG


Sure it has great plating but even the plating thickness could be measured in thousands of an inch, not microns. Plating that thick can hide a lot.

Thats a problem with cast parts, not only razors but any cast parts. The pieces coming out of the molds are only as good as the molds themselves and, molds wear.
 

Esox

I didnt know
Staff member
It also might explain the blade chatter that many people complain about.

Some people might get a 'good' one. I think that if 10 random R41 base samples were measured, they'd all be different. Mine is virtually unusable. I can and have used it but it turns a 3-5 minute pleasurable shave into a 20 minute nerve wracking shave that I do not enjoy.

On the other hand, the 1930's Gillette New SC, which was machined, on measuring several different samples by a member were all exactly the same. Each one had the identical .023" blade gap. The New LC on the other hand, which was either cast or stamped, and the gaps across those samples a member measured varied by more than .017".
 
Top Bottom