It's easy to intuit that imperceptible variations in the surface would cause a pattern in the reflection, but the convex shape making it distance-independent, that's the real genius. Amazing what they could invent thousands of years ago.
Perhaps they were polishing convex bronze mirrors and noticed that imperfections produced this projection effect. Then generations of trial and error to make recognisable images...
@@poppyrider5541 not that surprising to me. There's a lot of information we have about our eyes and light which draw the connection. Eyeglasses pretty naturally come from telescopes because those with poor eyesight realize their eyesight is better when viewing through an adjustable telescope. This doesn't affect your eyesight.
In your first example of an indent in the mirror (2:30) you actually created a retroreflector and the outgoing light would stay parallel to all the other rays.
I noticed his error at 2:30 as well... really makes me cringe. Considering all the complex things Steve has educated us on, it is a little painful to see such a mistake being made. Mistakes happen though, so I just hope that Steve will correct the video rather than knowingly continue to give young impressionable minds the wrong idea on how light reflections work.
@@FloodAnxiety you guys are such pedants. He's clearly using some arbitrary geometry to explain a larger topic and all you can think about is the fact that it happens to be 90 degrees. It's like your brain can't move on until everything is as YOU deem correct. This is not about "young impressionable minds" to you, you're just finding an excuse to be outraged and "cringed".
As this is now the top comment about it, I'll reply here as well! Yeah, I'm kicking myself! The faces of the divot should be at 22.5 degrees not 45. I've used the new corrections feature on RUclips to put a note on the screen at that time stamp. Though I'm not sure it's showing up for everyone. It's in the pinned comment too. Thanks for the heads up.
I imagine in manufacturing, the reflected image is simply masked on the mirror surface then etched with a mild acid which would slightly change the relief on the mirrored side. Once etching is completed it's simply cleaned and polished. This would account for the misalignment and non-matching images. This is also similar to the process which constructs the silicon wafers you mentioned for transistors.
Yep i agree acid etching is an excellent way of putting this detail in. The same process that is used for etching DIY circuit boards could be used here
I was thinking the same thing. As a DIY 3D printer maker,I imagine that I could modify a 3D printer to use an inkjet cartridge to spray a solution that could etch bronze or brass. This may also be within the capabilities of a laser engraver.
Laser engraving,acid etching,3D printers. They’ve been made for hundreds maybe even thousands of years 🤣 pretty sure they didn’t have anything like that back in the day how are the real ancient ones made 🤷🏻♂️
I could easily imagine that these originated with the embossing/casting affecting the surface hardness and polishing like Steve describes. Then some enterprising person figured out how to recreate and improve upon the effect with etching. Mirrors where the image didn't match the embossing might have been a surprising/special variant of the "trick". The etching method is likely more reliable and easier, though perhaps a bit technically more complicated to do. I suspect most magic mirrors makers switched to using it once that secret was leaked or they figured it out. A speculative "just so" story to be sure, but seems rather plausible to me. Or maybe someone spilled lemon juice on a mirror...
Grand Illusions! That's a shout out on this channel I never expected to see. Those toys and trinkets that Tim shares from his collection are often a wonder to see.
It's not a hologram though. I really wanna know where I can get more real transparent holograms. Most of the keywords are contaminated by those cheap glittering film covered stuff.
I sent the company who made that one a message asking if they are avalible for purchase. The company is Rayform and they have a jewelry line called The Rayy. Very expensive stuff, so I’m not sure if they normally sell something like a refractive image cube or something, but once they get back to me I’ll post another reply here. I’m curious how Steve got his though.
The ones you can by now are not made in the same way as they were traditionally. The "original" way was decently simple as bronze, while being a metal, is soft enough that surface-differences on one side can easily pass through to the other side when stress is applied - aka when polishing. That can intentionally be enhanced further but just casting it with a relief and then polishing will already give you this result. But to produce them now there are easier methods than painstakingly and slowly polishing them. As many have said - etching is a very simple and fast method. You could also imprint them with a stencil.
Actually there is one more thing missing in this explanation: The part about the optical illusion. How come, that there are large dark areas with light edges of the same brightness as the thinner dark lines? (or: how are there really large dark areas at all) All the reflected light from this large dark area would have to be focussed in the light edge around it, making it much brighter than light edges of thin dark areas. Answer: There are no "large" dark areas. Our brain interprets them as such. The inner "dark" part of the "sign of the zodiac" is actually as bright as the "light" part of the inner circle. Simply pause the video (in the right frame), make a screen shot, paste it in "Paint", cut out a part of the dark and light area each and paste them on a white surface. You will find that they have almost equal brightness. The "edge effect" really creates just pairs of bright and dark lines. Our brain additionally adjusts the brightness of the areas in between.
Like when a total solar eclipse is shown in a film using crude special effects, and there is a solid black disk where the sun was. In a real eclipse I think it would not be darker than the sky, just look darker due to the contrast with the surrounding bright corona. The film effects overdo this.
Just tried this-- there's a clear difference in the color values in as many screenshots as I could pull. What you're describing is still real, but... the video doesn't really allow it to be perceived.
One method I've seen at least, involves engraving the image in reverse on the back of the mirror, placing the back of the mirror on a hard, flat surface, and then polishing the front of the mirror to its final mirror finish. The pressure of polishing the front causes the metal above the relieved areas to distort slightly. Upon then bringing the back to mirror polish as well, the image becomes imperceptible to the human eye by direct observation from either side, the image only revealed in light reflected onto a surface.
@@JB-fh1bb Seems like that would work. The context accompanying the above-mentioned demonstration was, and please take this with a grain of salt as I am only vaguely recollecting hearsay, that this technique was used by early adopters of Christianity in Asia. That around the 5th or 6th century as Christianity was spreading to Eastern Asia, followers could carry a mirror that would appear perfectly ordinary front and back, but would cast a secret image in the light it reflected. I'm not saying this is true and I can't cite any related sources, but hopefully someone out there does know and can *shine some light* on this.
@@leecarrell6952 I remember one was shown in one of the two Hidden Christians series of Japanology from NHK. Christians were persecuted in Japan, so they had to keep their practices secret. Though now I'm hearing Japanese magic mirrors are made differently from the Chinese, with the last traditional maker being Yamamoto Akihisa.
That is how they were originally made. There are several methods though, obviously, as ones where the patterns don't match obviously can't be made that way. They could be etched, as well. Leaving an acid (lemon juice, for example) over a mask on the mirror side can also get similar results, and would allow differing patterns.
"So if you think your life is complete confusion Because you never win the game Just remember that it's a Grand illusion 'Cause deep inside we're all the same We're all the same..."
During WWII, my dad spent some time in India, and these mirrors were on sale in the markets. He ordered one that reflected "I love you Edna" for my mum, and he picked iot up two or three hours later. They were made of brass, and brass is a work-hardening metal. To produce the image, they simply tapped out the shape with a hammer and a sheep's foot chisel (a small one, of course), which hardened the surface on the image, then they polished it with metal polish like Brasso. That wore down the softer parts, which had not been beaten, more than the work-hardened parts. I imagine that getting the image right with a hammer and chisel was a real skill - but in such places as India and China, there are many people with finely honed skills.
The optical physics of this is fascinating, but I'm more excited by the realization that this is what the Mirror of Twilight was supposed to be in TLoZ: Twilight Princess.
Actually I'm thinking about the mirror shield in OoT and MM. Yes, there's a decal on the surface of the shield, but it actually doesn't match the shape that is reflected.
Regarding semiconductor manufacturing, there is even more crucial role for this technique in some processes - mask alignment. For each manufacturing step a different pattern needs to be transfered onto the wafer surface so the first mask usually contains alignment targets for all the subsequent masks.
4:43 it might be more intuitive to portray the light rays as bright lines on a dark background :) So bunched up lines would actually be lighter. Regardless, wonderful video as always, thank you!
@@SteveMould I just tried watching that section of your video again with a color inversion filter enabled on my phone, and I saw the aliasing you were talking about. The slightly angled lines looked all blobby. No idea why that would happen with white on black but not the other way around. I could sure use a Steve Mould video explaining it. 😉
Aside from acid etching, you can also just stamp the surface then polish it back; if you light it while polishing in stages you can do it by hand without killing the image (I've seen this done with backed foil). Etching would probably be more consistent for mass production though.
Maybe they force a bunch of these mirrors together to impress the shape into the next, and then possibly only use softer polishing pads, that don't flatten like hard pads.
At first, I thought they probably transferred the image on the back by making the mirror, and then placing it on some stiff rubber and giving it a really hard whack, slightly bulging out the areas that were raised on the back. After you pointed out that the images' don't always line up, though, I bet the etching trick is probably correct.
I remember long ago watching a documentary that included how traditional magic mirrors were made in Japan after the technology was imported from China. The workshop would cast the mirror, with the image on the back. To bring out the image on the mirror side, months would be spent scraping the mirror surface. The maker would feel the surface, scraping off finer amounts, eventually polishing and apply a nickel finish. The presenter did mention that false backs could be added to conceal the original image, they didn't say how. My source for this is the 2005 series "What the Ancients Knew" s01e05 "The Japanese".
I read once of medieval Japanese using the hidden image to hide a crucifix on their person, during the period where christianity was outlawed. The explanation given was that the thinner sections were carved away by a draw knife stile of polishing, with an iron knife on a bronze mirror, to produce the fringing pattern.
As a machinist, I have seen many examples of how cold-formed metals have a lot of stress residing in them, especially if they were never annealed or stress-relieved after forming. For example, a smallish rectangular bar of common "cold-rolled" steel such as type 1018 will have stress that causes it to visibly warp outward like a slight banana peel effect if you split it down its centerline for 3/4ths of its length on a bandsaw. It has always been my observation that this stress is higher near the surface of such cold-formed bars. So, say you had a bar that measured 1/4" thick and 3/4" wide, you could greatly reduce the amount of spreading you got at the bandsaw cut, if you removed just 1/16" from each edge, making it into a 1/4" by 5/8" bar before splitting it. I suppose that if the mirror blank was "coined" by stamping under great pressure, with a considerably raised pattern on what will become the mirror surface, then there could be a planned pattern of distortion when that raised pattern was machined off later. Depending on how you machined it off. Of course, if this type of mirror dates back thousands of years, then cold forming seems unlikely on something so large. But similar stresses could be set up in the metal by having considerable variations in thickness created by traditional sand-casting techniques, and then at some point quenching the metal from a high temperature. So that is my guess: They did not rely on the raised pattern on the back of the mirror, but rather a raised pattern on the FRONT, which was then machined off.
If you want an explanation of those from someone who makes them, look for Matt Ferraro "The Physics of Magic Windows". It includes going into the math needed in order to create the surface that has the 2nd derivative that matches an arbitrary image.
I was just thinking "this looks like Tim's collection" and then you mention it literally is his😂 I love his channel so much, so it’s a joy hearing his name!
Took me a while til I realized the light wasn’t passing through the disc. You’re pointing the light from the wall to the shiny side. Then the light bounces back towards the wall
We infer that the image on the wall comes from the embossing on the back. It's only arbitrary that they match. Just done so you know what image you're about to display.
I was yelling "etching!" at the screen right from the beginning. It's THE go-to technology even today for creating shallow features after all, easy to control by time of exposure, and there's nothing high tech about it that couldn't have been done just as well in more primitive times.
The obvious solution to me as well as something I could produce easily is to simply stamp it then buff it off until it's almost gone. Imagine taking a coin and polishing it for fun, which I'm sure was done even way back when. You think you're done, but see the sun's effect on it still shows the image of the coin. Magic! you show your friends and the process turns into a product. This would be an extremely likely way this would pop up. Nobody would intentionally etch something for another purpose with an image they couldn't even see
I think it would be cool to have a magic mirror that reflects the patterns from an actual silicon circuit etching. The contrast between an ancient parlor trick and contemporary technology could make a nice accent light.
@@nefariousyawn The issue with chips is that the designs are too small, to the point where they diffract light rather than just reflect it, This is why you see colorful rainbow patterns reflecting in them, the patterns are splitting white light into its component wavelengths.
@@kunjupulla like the comments explain, all they needed to be able to do is stamp or etch (whether physically or chemically) the image and then polish it to the point where the details were no longer visible to the eye but a reflection would spot it.
When you said it’s “maybe etched then polished” I immediately felt confident that this is the way they do it. I bet they have a mask and they use cotton buds or similar to apply a slight acid of some description and immediately neutralise that acid
Did you know in the early days, instead of toilet paper, people used corn cobs as the primary butt whiping utility...then sears would sell corn cobs in bulk
my theory is that it's just embossed onto the mirrored side and they just polish it into a convex shape but the polishing medium is quite soft & flexible allowing it to retain the embossing in the mirror surface
Maybe this is the same principle to hide a person behind a shield. All is about the magic of the light. Incredible artifact, thanks to show us this stuff.
Once upon a time, I made my own wave chamber. They used to be popular in the 80s and 90s. You pour alcohol and food coloring into a glass container. In my case, it was a mayonnaise jar. Fill the remainder with kerosene, and you have a chamber that will slowly form waves as you tilt it from side to side. Commercial ones had a long, thin chamber mounted on a motor that would tilt it left and right, automatically forming the waves. The peculiar nature of my wave chamber I only discovered by accident, when placing it on my desk, near a small desk lamp with a tiny light bulb (think auto turn signal type.) As the light shone down upon the boundary surface (which had a much larger surface area than the commercial product; 5-inch diameter mayo jar) between the alcohol and the kerosene, it refracted in a similar way upon the wall behind it. You could see caustics in a similar way that sunlight shining through a pool causes caustics upon the bottom and sides of the pool. It was truly mesmerizing, as it appeared holographic in nature, as the surface deformed with wave formation while the strong, single-point light source shone through it. I could rotate the jar and see the shapes changing as if a hologram was being rotated. It fascinated me. I can't imagine anyone else has had the same experience, because I have never seen any reference to this type of phenomenon since then. Maybe I'll try searching on RUclips. I love your curious videos, Steve. Thank you for dislodging this 41-year-old memory. Cheers.
I seem to recall there's a description of how to make these in "Dunninger's Complete Encyclopedia of Magic", Steve should definitely get a copy, it would keep him busy for weeks.
The traditional method of making these was polishing a piece of metal by hand on top of a stone with the image you wanted to project. Each one would take days or weeks to make. But I doubt that’s how they make these modern versions you used in the video.
Heavy embossing creates pretty much the same situation. Of course, it seems like someone figured out how to recreate/improve the effect with etching, and it probably didn't take long for that to become the preferred method.
I just love these discussions. Even knowing how something works, it can still SEEM magical, knowing that people have been making these things for centuries before they knew exactly how they worked.
I believe it is - or was originally - done by having some design on both sides, with one side then being manually sanded and polished, stopping just short of erasing the image completely. The progress could always be checked by holding a light to it, not only just when it's complete.
you are probably on the right track with the differential cooling. If I had to take a guess the thinner parts that cool slower have a harder micro structure due to larger crystals. When polished, this would result in the minuscule raise to surface compared to the material that cooled faster.
There is an article and a video in RUclips where a Japanese artisan explains how at least the traditional process works. The image is indeed produced from the casting on the back side. Then the mirror surface is just ground to exactly right thickness so that the small changes in the metals inner structure become visible. Also Wikipedia article "Chinese magic mirror" has a reference to a book that gives a similar description of the process.
I think that the way that they make it is with masked polishing. you polish the whole mirror surface with the base curve preset, then you mask the image and polish just a litle bit more, the reflected image is then the controlled defects in the polished surface
My guess is that after it is cast an image is stamped onto the back of the mirror, creating a pattern of differing density in the metal. Next, the back would be sanded with a flat, rigged abrasive to create a perfectly flat surface. If it is then polished with something with a bit of flex to it (not unlike a foam sanding block), the less dense areas will abrade slightly faster then the more dense areas. Also, the flex in the polisher would create a slightly concave surface, giving the mirror a slight convex surface.
Agreed. This seems the most logical to me. I belive an enchant would be less reliable, due to the very minor pollutants in the casting with the miniscule variation in the solution.
9:00 I love it when new insights and technologies allow for such beautifully intricate and advanced designs to be created onto otherwise rather mundane and simple materials.
Some places have souvenir "penny presses" that take a penny and use mechanical pressure to flatten and re-emboss it with a new image. Some of these are two-sided, but older ones left the back of the penny flat. However, even flattened, it's possible to see what that side of the penny originally looked like, and still read the minting date or text that was on that side of the coin. Perhaps there's something similar happening here - An image is placed on the disc with a normal casting/minting process, then pressed into the elliptical-dish press to flatten out the image. If the disk is only pressed part of the way, the image from the original casting would still be present, just much flatter
I don’t think so. I’m pretty sure the penny effect is from the plating and/or oxidization. This means that parts of the metal have different colors. When you “smash” it you get patterns. This is somewhat similar to “Damascus” steel.
you cast it, you pregrind it, you press it tightly onto a stone with a leather lining, you polish it, done. magic. relying only on tension while casting wouldn't work because all metal acts inhomogenously in its crystalline structure referred to temperature induced tension, so your result would get wobbly
with a mismatched image it is pretty much the same precedure, but you reshape it afterwards for example. there may be a connection between mismatched mirrors and less "resolution" you are getting because of thicker metal and things. best thing about doing it like this is you can use your grinding stone to retain the convex shape.
the grinding stone could even be the casting shell. you make some kind of wooden stamp you hold into the molten metal to get the upper surface, let it cool off, then you literally lap your casting surface and your workpeace out on each other to equalize their surfaces, then you take that leather piece, hydraulic press on dead center, polishing, done. :) they could have used massive stones back then for example to get that pressure
Earned a sub, very nice video with outstanding explanation of the thematic, pls don't stop doing stuff like this, we need more people like you in this world! :)
Absent powerful calculations and clear understand of photon behavior, it is mind-bending to consider the effort, trial, error, and precision required to create the techniques required for these magic mirrors.
I wonder if there is any similarity to how Schmidt Cassegrain corrector plates are made? Apply vacuum to the embossed side, polish the mirror side flat, then release the vacuum. You can imagine that a suitable technique might result in some parts of the mirror being ground down more or less, resulting in a faceted mirror surface.
I agree. Considering how hard it's to compute caustics even in 3D rendered images, it'll be interesting to learn how those glass pieces are actually created computationally.
I'm very satisfied that the real object deals with the interaction of a sheet of light hitting a plane, and the animations deal with a line of light hitting an edge, continuing your trend of making a 2D version of a 3D thing to more easily explain it
The instructions I've seen involved work hardening the bronze by use of a punch, then polishing flat. There would be tiny raised areas where the hardening has occurred.
10:20 Not only is there no smallest uninteresting integer, but that absence proves by induction that there are no uninteresting integers at all. If there were, there would be a smallest member of that set. The same proof you mention also works for "most uninteresting integer", as that property too would be very interesting.
What if you had a concave metal "stamp" of the desired image that you pressed the soft brass into and then polish it so it's almost invisible. Or maybe an image made of a clear lacquer that distorts the light is covered with one that doesn't so you can't see the difference? Try a solvent on it!
I'm astonished how you pull up random things from all parts of the world and manage to baffle me more each video. This channel is just pure (technical) magic to me. I absolutely lost it in the end with that clear card. Seriously, I was screaming "WHAT?!?!?!?!" at my screen - a kind of physical reaction that I rarely have when watching any video! Thanks for bringing that kind of magic into my life, Steve! I'm sitting here, being just baffled as to how somebody can scratch all the itches in my brain so perfectly....
Once again I find your video topic overlapping with the world of ultrahigh power laser driven inertial confinement fusion reactors. That last transmissive version you showed is exactly what we do to the laser beams right before they fall onto the frozen hydrogen target. We don't use a lion image... as you may have guessed, but rather a random pattern of bright and dark areas across the whole area of the beam. This essentially destroys the phase coherence of the beam and DRAMATICALLY increases beam intensity uniformity or "smoothness" by suppressing laser speckle induced by sub-beam or intra-beam beamlet interference as it is focused from a third of a meter in diameter down to a tenth of a millimeter at the target capsule surface. This increase in beam intensity uniformity is incredibly important in suppressing the in-flight formation of Rayleigh-Taylor hydrodynamic instabilities of the 100 million Kelvin plasma as it collapses to a few microns in diameter. We call these optics "distributed phase plates" or DPPs, and without this technique the maximum fusion neutron yield on the machine would probably be something like an order of magnitude lower. Interested to see the next video.
This reminds me of the Cross Atlantic Communications Cable problem The length of the cross Atlantic cable meant that, for traditional telegraphy, you needed very high voltage for your comms link, which burnt through the cable's insulation. This was solved by using a Mirror Galvanometer. The very small changes in the rotating mirror meant that much lower voltages could be used.
I imagine the way these things are made are as follows: 1) The design is with nothing on the the mirror side. 2) The 'projection' is painted on the back half using clay (in much the same way that sword smiths in Japan have been 'staining' metal for centuries). 3) The device is cooked in the forge so that the clay stains and thus weakens a few microns of metal underneath. 4) The clay is then removed during the polishing process, taking stained/weakened metal with it. Not only would this likely achieve the desired affect, it would also serve as the perfect indicator for when to stop abrasive polishing ("as soon as the design disappears"), and it is perfectly doable with tools and techniques we've had for thousands of years. Acid etching is also a good theory, but I worry about how far back we're able to date the use/concentrations of acids.
i am a geologist specialising in trace fossils but also have a solid backgrownd in minerals and how they form the theory is on point but it will take alot of calibration to work out the right timing and thickness to produce variable crystal sizes that result in the correct level of deflection. Very impressive this is the first thing that really impressesme from a te4chnological point of veiw this is the same as the mechanism from the mediterian. it suggests that there is a lost knowledge or something like if you used this to make call signs, and you have a amplifier then you could send this 160 km on a clear day.
I remember hearing stories about forensics recovering VIN numbers from cars that had them sanded off in an attempt to anonymise them. The stamping of the plate changed the properties in the metal making it much harder wearing.
The basic mirror shape, with the design on the back, was cast flat, and the convexity of the surface produced afterwards by elaborate scraping and scratching. The surface was then polished to become shiny. The stresses set up by these processes caused the thinner parts of the surface to bulge outwards and become more convex than the thicker portions. Finally, a mercury amalgam was laid over the surface; this created further stresses and preferential buckling. The result was that imperfections of the mirror surface matched the patterns on the back, although they were too minute to be seen by the eye. But when the mirror reflected bright sunlight against a wall, with the resultant magnification of the whole image, the effect was to reproduce the patterns as if they were passing through the solid bronze by way of light beams. quoted from wikipedia
Enjoyed the video, and the exploration of what is behind a phenomenon I'd never seen before. But, to be honest, the teaser for the future video about transparencies looks amazing! Definitely returning for that one.
The trick is this: Cast the mirror in a mould which has the rough 3D image on its convex side. Using a press with a corresponding concave button and a flat punch on the top; press the casting's initially convex side until it is flat against the punch. This will in turn push the casting's initially flat side down into button. Now, the image side will be flat, and the blank side will be convex. Polish the convex side to a mirror finish, and voila! The mould used for the casting was made the same way. A plug is crafted with the 3D image engraved into it in its expected flat appearance. Using that plug, a mold is formed. Then the mold is pressed into a convex shape, thus distorting the image such that; when a casting is made from it and later flattened, the 3D imprint is guaranteed to flatten back into the original artistic form.
The most interesting effect of these mirrors is how they lead the one to think that there is a causal relationship between the reflected image and the pattern on the reverse of the mirror. No matter how many times one is told or even says, "looks can be deceiving," one often finds one's self fooled.
Its videos like this one that reminds me how much I love your content as well as your good friend Matt Parker. Found both your channels at the same time, thru number/computerphile many years ago.
I assume it was cast, stress relieved, and then cooled (or heated) for polishing. I would assume the polishing is done chilled. When it returns to room temperature, the thickness differences cause very minor distortions. Perhaps it needs to be an original image on the polished surface side vs the distortion due to the thickness on the back side. The thermal change before polishing should still translate after the thermal delta to room temp is "released" But I think management of casting stresses would be quite important
This phenomenon is basically the reflective equivalent of water caustics, the light patterns projected on the bottom of the swimming pool by the water whose surface is definitely not flat.
Speaking as someone familiar with metallurgy and historical metalworking techniques- I feel like this was probably done right in the casting process. The bottom of the mold would likely be stamped by a 'master' mirror, and then once the metal was poured a weighted cap with the design on the bottom surface would be placed on top of the still-molten metal. I feel like the molds and caps were likely made of a fine clay or pottery to ensure minimal surface defects. Think of how records are made- there's a master, and then copies are created based on that This would explain the inconsistencies in the surface image vs. the image on the back, and would be a much less complex task than trying to etch it as others have suggested. Speaking from personal experience, it's hard to achieve the kind of smooth, sine-like curves required for this mirror using a chemical etchant. It would cause the etched surface to become a matte finish and require a lot of polishing to achieve the smooth transition between peaks and valleys, even with the miniscule variation in depth. Perhaps that was how the master was created, but I feel like the majority of these had the images cast right in to them
when polishing and lapping diamonds for record cutting styli we found that the temperature at which we grind would affect the surface flatness ( fractions of a micron ) when its service temperature differed from its cutting temperature. In this way we could grind concave or convex surfaces on an essentially flat lapping skaife. As the surfaces we are grinding are less than 0.1 mm^2 laser reflection would reflect individual crystal lattice dislocations
If you support the structures on the back, seal around the perimeter, and pull a vacuum, the glass will deform where it is not supported while you polish the front. Once the vacuum is released, that beautifully polished front surface will deform as the glass springs back into shape.
Hmm. I thought it might have something to do with those one-way dark mirrors that are hard to see through from one side, and then reflecting off the engravings on the back. Guess this makes more sense though
The easiest way is to make a cast, and then polish mirror side in high temperature of the piece. Cooling down will shrink material proportionally to the thickness (back shape) making small mirror surfaces deviations.
Regarding the magic mirrors with mismatched images, it's possible that the piece was cast in a mould with an obverse (the mirror side) that has differences in thickness matching the buddha image. It's also possible that the mould has some sort of inlay of a material with higher/lower thermal conductivity and/or specific heat capacity (e.g. copper inlay over exterior portions of a steel mould). In this way, the mould itself could strongly influence the cooling time of the obverse, mirrored face of the piece, and thus the crystallographic properties.
A hypothesis on how they're manufactured: During the polishing process, the plates are heated/cooled such that thicker parts of the plate have a different temperature from thinner parts of the plate. This creates thermal stress, so that some parts expanded more than others. Once the thermal stress is relieved, the polished surface is deformed, revealing the thickness gradient. Edit: this is very close to what's being described is the video, so :shrug:
The first thing I thought of was uneven tension in the metal, I know when people talk about casting (and glass making) that sharp edges and uneven cooling create high and low tension areas in the material. I suspect thats why the castings are so deep, you could pour water or oil into the shallow parts to increase the temperature difference.
According to Dunnigers Complete Encyclopedia of Magic most of these trick mirrors are made by etching the image desired, then polishing the surface. To get the effect shown here, you could etch the entire design as opposed to just the outline.
It's easy to intuit that imperceptible variations in the surface would cause a pattern in the reflection, but the convex shape making it distance-independent, that's the real genius. Amazing what they could invent thousands of years ago.
I couldn't have said it better.
or at all.
Perhaps they were polishing convex bronze mirrors and noticed that imperfections produced this projection effect. Then generations of trial and error to make recognisable images...
It's amazing they managed this but didn't work out how to make glasses.
@@poppyrider5541 not that surprising to me. There's a lot of information we have about our eyes and light which draw the connection. Eyeglasses pretty naturally come from telescopes because those with poor eyesight realize their eyesight is better when viewing through an adjustable telescope. This doesn't affect your eyesight.
@@thugpug4392 I;m pretty sure the idea of lens' to correct eyesight came from ppl looking through a glass of water. I might be wrong.
In your first example of an indent in the mirror (2:30) you actually created a retroreflector and the outgoing light would stay parallel to all the other rays.
Right, it should be a shallower angle instead of a right-angle at the back.
I noticed his error at 2:30 as well... really makes me cringe. Considering all the complex things Steve has educated us on, it is a little painful to see such a mistake being made. Mistakes happen though, so I just hope that Steve will correct the video rather than knowingly continue to give young impressionable minds the wrong idea on how light reflections work.
@@FloodAnxiety you guys are such pedants. He's clearly using some arbitrary geometry to explain a larger topic and all you can think about is the fact that it happens to be 90 degrees. It's like your brain can't move on until everything is as YOU deem correct.
This is not about "young impressionable minds" to you, you're just finding an excuse to be outraged and "cringed".
please educate me. I don't understand retoreflector and all the parallel light stuff. Also "shallower angle" by Barney Laurance. Thanks!
As this is now the top comment about it, I'll reply here as well! Yeah, I'm kicking myself! The faces of the divot should be at 22.5 degrees not 45. I've used the new corrections feature on RUclips to put a note on the screen at that time stamp. Though I'm not sure it's showing up for everyone. It's in the pinned comment too. Thanks for the heads up.
I imagine in manufacturing, the reflected image is simply masked on the mirror surface then etched with a mild acid which would slightly change the relief on the mirrored side. Once etching is completed it's simply cleaned and polished. This would account for the misalignment and non-matching images. This is also similar to the process which constructs the silicon wafers you mentioned for transistors.
Yep i agree acid etching is an excellent way of putting this detail in.
The same process that is used for etching DIY circuit boards could be used here
I was thinking the same thing. As a DIY 3D printer maker,I imagine that I could modify a 3D printer to use an inkjet cartridge to spray a solution that could etch bronze or brass. This may also be within the capabilities of a laser engraver.
That's what I thought too
Laser engraving,acid etching,3D printers. They’ve been made for hundreds maybe even thousands of years 🤣 pretty sure they didn’t have anything like that back in the day how are the real ancient ones made 🤷🏻♂️
I could easily imagine that these originated with the embossing/casting affecting the surface hardness and polishing like Steve describes. Then some enterprising person figured out how to recreate and improve upon the effect with etching. Mirrors where the image didn't match the embossing might have been a surprising/special variant of the "trick". The etching method is likely more reliable and easier, though perhaps a bit technically more complicated to do. I suspect most magic mirrors makers switched to using it once that secret was leaked or they figured it out.
A speculative "just so" story to be sure, but seems rather plausible to me. Or maybe someone spilled lemon juice on a mirror...
Grand Illusions! That's a shout out on this channel I never expected to see. Those toys and trinkets that Tim shares from his collection are often a wonder to see.
Okay that ending hologram looking thing is so freaking cool. I'd love to see you go into a deep dive into hologram type things.
I want one. That was gorgeous
I know! I was like 🤩 Can someone please tell me what even to Google!?
It's not a hologram though. I really wanna know where I can get more real transparent holograms. Most of the keywords are contaminated by those cheap glittering film covered stuff.
I sent the company who made that one a message asking if they are avalible for purchase. The company is Rayform and they have a jewelry line called The Rayy. Very expensive stuff, so I’m not sure if they normally sell something like a refractive image cube or something, but once they get back to me I’ll post another reply here. I’m curious how Steve got his though.
@@sphygo Intriguing! Looking forward to hearing their response.
The ones you can by now are not made in the same way as they were traditionally. The "original" way was decently simple as bronze, while being a metal, is soft enough that surface-differences on one side can easily pass through to the other side when stress is applied - aka when polishing. That can intentionally be enhanced further but just casting it with a relief and then polishing will already give you this result.
But to produce them now there are easier methods than painstakingly and slowly polishing them. As many have said - etching is a very simple and fast method. You could also imprint them with a stencil.
Maybe that's how they were discovered
Actually there is one more thing missing in this explanation: The part about the optical illusion.
How come, that there are large dark areas with light edges of the same brightness as the thinner dark lines? (or: how are there really large dark areas at all)
All the reflected light from this large dark area would have to be focussed in the light edge around it, making it much brighter than light edges of thin dark areas.
Answer: There are no "large" dark areas. Our brain interprets them as such. The inner "dark" part of the "sign of the zodiac" is actually as bright as the "light" part of the inner circle.
Simply pause the video (in the right frame), make a screen shot, paste it in "Paint", cut out a part of the dark and light area each and paste them on a white surface.
You will find that they have almost equal brightness. The "edge effect" really creates just pairs of bright and dark lines. Our brain additionally adjusts the brightness of the areas in between.
On a related note - the fact that the moon looks so bright white to us may be an additional clue as well.
Well I'll be damned...
Like when a total solar eclipse is shown in a film using crude special effects, and there is a solid black disk where the sun was. In a real eclipse I think it would not be darker than the sky, just look darker due to the contrast with the surrounding bright corona. The film effects overdo this.
Wow! I was wondering about that, thank you!
Just tried this-- there's a clear difference in the color values in as many screenshots as I could pull. What you're describing is still real, but... the video doesn't really allow it to be perceived.
One method I've seen at least, involves engraving the image in reverse on the back of the mirror, placing the back of the mirror on a hard, flat surface, and then polishing the front of the mirror to its final mirror finish. The pressure of polishing the front causes the metal above the relieved areas to distort slightly. Upon then bringing the back to mirror polish as well, the image becomes imperceptible to the human eye by direct observation from either side, the image only revealed in light reflected onto a surface.
What it they polished the mirror first, then turned it over and indented the back enough that it affected the polished surface on the other side?
@@JB-fh1bb Seems like that would work. The context accompanying the above-mentioned demonstration was, and please take this with a grain of salt as I am only vaguely recollecting hearsay, that this technique was used by early adopters of
Christianity in Asia. That around the 5th or 6th century as Christianity was spreading to Eastern Asia, followers could carry a mirror that would appear perfectly ordinary front and back, but would cast a secret image in the light it reflected. I'm not saying this is true and I can't cite any related sources, but hopefully someone out there does know and can *shine some light* on this.
@@leecarrell6952 I remember one was shown in one of the two Hidden Christians series of Japanology from NHK. Christians were persecuted in Japan, so they had to keep their practices secret. Though now I'm hearing Japanese magic mirrors are made differently from the Chinese, with the last traditional maker being Yamamoto Akihisa.
That is how they were originally made. There are several methods though, obviously, as ones where the patterns don't match obviously can't be made that way. They could be etched, as well. Leaving an acid (lemon juice, for example) over a mask on the mirror side can also get similar results, and would allow differing patterns.
Great to see Grand Illusions here! I'm only just now realizing there's probably a few videos you could make using Tim's magic toys. :D
"So if you think your life is complete confusion
Because you never win the game
Just remember that it's a Grand illusion
'Cause deep inside we're all the same
We're all the same..."
I was watching the preview and saw Grand Illusion mentioned in the subtitles, I love Tim.
@@TechGorilla1987 Right, back into the lamp, genie.
@@bertjesklotepino I'll simply follow the river Styx.
Heh
During WWII, my dad spent some time in India, and these mirrors were on sale in the markets. He ordered one that reflected "I love you Edna" for my mum, and he picked iot up two or three hours later.
They were made of brass, and brass is a work-hardening metal. To produce the image, they simply tapped out the shape with a hammer and a sheep's foot chisel (a small one, of course), which hardened the surface on the image, then they polished it with metal polish like Brasso. That wore down the softer parts, which had not been beaten, more than the work-hardened parts.
I imagine that getting the image right with a hammer and chisel was a real skill - but in such places as India and China, there are many people with finely honed skills.
The optical physics of this is fascinating, but I'm more excited by the realization that this is what the Mirror of Twilight was supposed to be in TLoZ: Twilight Princess.
And the pair of Chinese mirrors in Shenmue.
@@SianaGearz The phoenix mirror
@@bazahaza And the Dragon Mirror.
"I'm looking for some sailors..."
Actually I'm thinking about the mirror shield in OoT and MM. Yes, there's a decal on the surface of the shield, but it actually doesn't match the shape that is reflected.
Regarding semiconductor manufacturing, there is even more crucial role for this technique in some processes - mask alignment. For each manufacturing step a different pattern needs to be transfered onto the wafer surface so the first mask usually contains alignment targets for all the subsequent masks.
4:43 it might be more intuitive to portray the light rays as bright lines on a dark background :)
So bunched up lines would actually be lighter.
Regardless, wonderful video as always, thank you!
Yes! I tried that but the aliasing made a total mess of it. Really odd and annoying.
@@SteveMould could just make your lines thicker?
@@SteveMould
I see, that makes a lot of sense. Well, was a minor nitpick anyway. It's great as is!
@@SteveMould I just tried watching that section of your video again with a color inversion filter enabled on my phone, and I saw the aliasing you were talking about. The slightly angled lines looked all blobby. No idea why that would happen with white on black but not the other way around. I could sure use a Steve Mould video explaining it. 😉
@@hollt693 this would be great!
Aside from acid etching, you can also just stamp the surface then polish it back; if you light it while polishing in stages you can do it by hand without killing the image (I've seen this done with backed foil). Etching would probably be more consistent for mass production though.
Maybe they force a bunch of these mirrors together to impress the shape into the next, and then possibly only use softer polishing pads, that don't flatten like hard pads.
These mirrors are rare they are crafted one at a time and the polishing process is complex.
At first, I thought they probably transferred the image on the back by making the mirror, and then placing it on some stiff rubber and giving it a really hard whack, slightly bulging out the areas that were raised on the back. After you pointed out that the images' don't always line up, though, I bet the etching trick is probably correct.
Exactly the same thought 👍
It's actually hand scraped in and polished. There was a special about the mirrors about 20 years ago showing how they are made
I remember long ago watching a documentary that included how traditional magic mirrors were made in Japan after the technology was imported from China. The workshop would cast the mirror, with the image on the back. To bring out the image on the mirror side, months would be spent scraping the mirror surface. The maker would feel the surface, scraping off finer amounts, eventually polishing and apply a nickel finish. The presenter did mention that false backs could be added to conceal the original image, they didn't say how.
My source for this is the 2005 series "What the Ancients Knew" s01e05 "The Japanese".
I read once of medieval Japanese using the hidden image to hide a crucifix on their person, during the period where christianity was outlawed. The explanation given was that the thinner sections were carved away by a draw knife stile of polishing, with an iron knife on a bronze mirror, to produce the fringing pattern.
As a machinist, I have seen many examples of how cold-formed metals have a lot of stress residing in them, especially if they were never annealed or stress-relieved after forming. For example, a smallish rectangular bar of common "cold-rolled" steel such as type 1018 will have stress that causes it to visibly warp outward like a slight banana peel effect if you split it down its centerline for 3/4ths of its length on a bandsaw.
It has always been my observation that this stress is higher near the surface of such cold-formed bars. So, say you had a bar that measured 1/4" thick and 3/4" wide, you could greatly reduce the amount of spreading you got at the bandsaw cut, if you removed just 1/16" from each edge, making it into a 1/4" by 5/8" bar before splitting it.
I suppose that if the mirror blank was "coined" by stamping under great pressure, with a considerably raised pattern on what will become the mirror surface, then there could be a planned pattern of distortion when that raised pattern was machined off later. Depending on how you machined it off.
Of course, if this type of mirror dates back thousands of years, then cold forming seems unlikely on something so large. But similar stresses could be set up in the metal by having considerable variations in thickness created by traditional sand-casting techniques, and then at some point quenching the metal from a high temperature.
So that is my guess: They did not rely on the raised pattern on the back of the mirror, but rather a raised pattern on the FRONT, which was then machined off.
the lion at the end is the coolest thing shown in the video lol, i hope he makes a sequel for it :)
If you want an explanation of those from someone who makes them, look for Matt Ferraro "The Physics of Magic Windows". It includes going into the math needed in order to create the surface that has the 2nd derivative that matches an arbitrary image.
I was just thinking "this looks like Tim's collection" and then you mention it literally is his😂 I love his channel so much, so it’s a joy hearing his name!
Took me a while til I realized the light wasn’t passing through the disc. You’re pointing the light from the wall to the shiny side. Then the light bounces back towards the wall
Wow, yeah I was confused the entire video.
same same
@Levi Tate
he jumped the gun, but I wouldn't say it was terrible. It eventually made sense 😅
We infer that the image on the wall comes from the embossing on the back. It's only arbitrary that they match. Just done so you know what image you're about to display.
@Levi Tate I agree, this video was extremely frustrating
I am truly astonished. How in the world did they make these 1000 years ago !?
This is amazing today for goodness sake. Thanks for a cool video!
I was yelling "etching!" at the screen right from the beginning. It's THE go-to technology even today for creating shallow features after all, easy to control by time of exposure, and there's nothing high tech about it that couldn't have been done just as well in more primitive times.
The obvious solution to me as well as something I could produce easily is to simply stamp it then buff it off until it's almost gone. Imagine taking a coin and polishing it for fun, which I'm sure was done even way back when. You think you're done, but see the sun's effect on it still shows the image of the coin. Magic! you show your friends and the process turns into a product. This would be an extremely likely way this would pop up. Nobody would intentionally etch something for another purpose with an image they couldn't even see
I think it would be cool to have a magic mirror that reflects the patterns from an actual silicon circuit etching. The contrast between an ancient parlor trick and contemporary technology could make a nice accent light.
@@nefariousyawn The issue with chips is that the designs are too small, to the point where they diffract light rather than just reflect it, This is why you see colorful rainbow patterns reflecting in them, the patterns are splitting white light into its component wavelengths.
Interesting... But how did they do it 1000s of years ago???
@@kunjupulla like the comments explain, all they needed to be able to do is stamp or etch (whether physically or chemically) the image and then polish it to the point where the details were no longer visible to the eye but a reflection would spot it.
When you said it’s “maybe etched then polished” I immediately felt confident that this is the way they do it. I bet they have a mask and they use cotton buds or similar to apply a slight acid of some description and immediately neutralise that acid
I’ve never heard of anything like this but it’s so cool! I love that I never know what I’m going to learn about when you post 😅
Did you know in the early days, instead of toilet paper, people used corn cobs as the primary butt whiping utility...then sears would sell corn cobs in bulk
Engineering degrees usually take an extra year to complete than science degrees
Melanin is found in your brain
The transparency video you teased is gonna be awesome, I’ve never seen anything like that
my theory is that it's just embossed onto the mirrored side and they just polish it into a convex shape but the polishing medium is quite soft & flexible allowing it to retain the embossing in the mirror surface
how would the embossing remain if they ground a flat surface into a convex one?
@@mfultimate Just emboss it after adding the convex shape?
1. Convex Shape
2. Emboss
3. Polish
No wonder, glad you and Tim know each other, it's like finding out two of your distant friends are actually related.
You don't spoil the Magic by explaining how it's done. The explanation on how it's done is part of the magic. Science is a type of wonder
Maybe this is the same principle to hide a person behind a shield. All is about the magic of the light. Incredible artifact, thanks to show us this stuff.
Neat that Grand Illusions supplied you with some mirrors.
That aside, it's crazy to think how this was found out thousands of years ago.
The effect is from polishing on an textured surface.
Once upon a time, I made my own wave chamber. They used to be popular in the 80s and 90s. You pour alcohol and food coloring into a glass container. In my case, it was a mayonnaise jar. Fill the remainder with kerosene, and you have a chamber that will slowly form waves as you tilt it from side to side. Commercial ones had a long, thin chamber mounted on a motor that would tilt it left and right, automatically forming the waves. The peculiar nature of my wave chamber I only discovered by accident, when placing it on my desk, near a small desk lamp with a tiny light bulb (think auto turn signal type.) As the light shone down upon the boundary surface (which had a much larger surface area than the commercial product; 5-inch diameter mayo jar) between the alcohol and the kerosene, it refracted in a similar way upon the wall behind it. You could see caustics in a similar way that sunlight shining through a pool causes caustics upon the bottom and sides of the pool. It was truly mesmerizing, as it appeared holographic in nature, as the surface deformed with wave formation while the strong, single-point light source shone through it. I could rotate the jar and see the shapes changing as if a hologram was being rotated. It fascinated me. I can't imagine anyone else has had the same experience, because I have never seen any reference to this type of phenomenon since then. Maybe I'll try searching on RUclips. I love your curious videos, Steve. Thank you for dislodging this 41-year-old memory. Cheers.
I seem to recall there's a description of how to make these in "Dunninger's Complete Encyclopedia of Magic", Steve should definitely get a copy, it would keep him busy for weeks.
I appreciate the fact you are humble which requires honesty opposed to most other producers who act like they know it all
The traditional method of making these was polishing a piece of metal by hand on top of a stone with the image you wanted to project. Each one would take days or weeks to make. But I doubt that’s how they make these modern versions you used in the video.
Heavy embossing creates pretty much the same situation.
Of course, it seems like someone figured out how to recreate/improve the effect with etching, and it probably didn't take long for that to become the preferred method.
That's fascinating! Could you share where you got that info? I would love to learn more.
@@iankrom510 one of the crafters in the village used to make these.
a crossover with tim from grand illusions is the absolute last thing i was expecting, tim is a treasure
I just love these discussions. Even knowing how something works, it can still SEEM magical, knowing that people have been making these things for centuries before they knew exactly how they worked.
FINALLY!!! I've been waiting for years for someone to pick up controlled optical caustics. Super excited!
I believe it is - or was originally - done by having some design on both sides, with one side then being manually sanded and polished, stopping just short of erasing the image completely.
The progress could always be checked by holding a light to it, not only just when it's complete.
you are probably on the right track with the differential cooling.
If I had to take a guess the thinner parts that cool slower have a harder micro structure due to larger crystals.
When polished, this would result in the minuscule raise to surface compared to the material that cooled faster.
There is an article and a video in RUclips where a Japanese artisan explains how at least the traditional process works. The image is indeed produced from the casting on the back side. Then the mirror surface is just ground to exactly right thickness so that the small changes in the metals inner structure become visible.
Also Wikipedia article "Chinese magic mirror" has a reference to a book that gives a similar description of the process.
No shot you know Tim?! That man's an utter legend!
These play an important role in one book of Ken Liu’s fantasy series The House of Dandelions. It’s fun to see the science behind them.
I think that the way that they make it is with masked polishing.
you polish the whole mirror surface with the base curve preset, then you mask the image and polish just a litle bit more, the reflected image is then the controlled defects in the polished surface
That makes a lot of sense.
My guess is that after it is cast an image is stamped onto the back of the mirror, creating a pattern of differing density in the metal. Next, the back would be sanded with a flat, rigged abrasive to create a perfectly flat surface. If it is then polished with something with a bit of flex to it (not unlike a foam sanding block), the less dense areas will abrade slightly faster then the more dense areas. Also, the flex in the polisher would create a slightly concave surface, giving the mirror a slight convex surface.
Agreed. This seems the most logical to me. I belive an enchant would be less reliable, due to the very minor pollutants in the casting with the miniscule variation in the solution.
9:00 I love it when new insights and technologies allow for such beautifully intricate and advanced designs to be created onto otherwise rather mundane and simple materials.
Some places have souvenir "penny presses" that take a penny and use mechanical pressure to flatten and re-emboss it with a new image. Some of these are two-sided, but older ones left the back of the penny flat. However, even flattened, it's possible to see what that side of the penny originally looked like, and still read the minting date or text that was on that side of the coin.
Perhaps there's something similar happening here - An image is placed on the disc with a normal casting/minting process, then pressed into the elliptical-dish press to flatten out the image. If the disk is only pressed part of the way, the image from the original casting would still be present, just much flatter
I don’t think so. I’m pretty sure the penny effect is from the plating and/or oxidization. This means that parts of the metal have different colors. When you “smash” it you get patterns. This is somewhat similar to “Damascus” steel.
It made me so happy to hear that you got these from the grand illusion channel.
you cast it, you pregrind it, you press it tightly onto a stone with a leather lining, you polish it, done. magic. relying only on tension while casting wouldn't work because all metal acts inhomogenously in its crystalline structure referred to temperature induced tension, so your result would get wobbly
with a mismatched image it is pretty much the same precedure, but you reshape it afterwards for example. there may be a connection between mismatched mirrors and less "resolution" you are getting because of thicker metal and things. best thing about doing it like this is you can use your grinding stone to retain the convex shape.
the grinding stone could even be the casting shell. you make some kind of wooden stamp you hold into the molten metal to get the upper surface, let it cool off, then you literally lap your casting surface and your workpeace out on each other to equalize their surfaces, then you take that leather piece, hydraulic press on dead center, polishing, done. :) they could have used massive stones back then for example to get that pressure
By mentioning it's convex at the start you set it up perfectly for me to feel really clever for figuring out how it worked before you explained it
Earned a sub, very nice video with outstanding explanation of the thematic, pls don't stop doing stuff like this, we need more people like you in this world! :)
No idea why it makes me so happy when there are collaborations between fave RUclipsrs. It would be great to have Steve and Tim together.
Steve, we talked about this. If it doesn't guggle, whistle or have a 2D representation of it, I don't know what's going on!
(lol)
Made me laugh to imagine Steve trying to find a way to explain these mirrors with a 2d water contraption.
Didn't you see the 2d light ray diagram?
@@superfluidity ugh, you must be never invited to parties.
Oh look, in the sky, there goes the joke.......wayyyyyyy over your head.
Absent powerful calculations and clear understand of photon behavior, it is mind-bending to consider the effort, trial, error, and precision required to create the techniques required for these magic mirrors.
I wonder if there is any similarity to how Schmidt Cassegrain corrector plates are made? Apply vacuum to the embossed side, polish the mirror side flat, then release the vacuum. You can imagine that a suitable technique might result in some parts of the mirror being ground down more or less, resulting in a faceted mirror surface.
8:56
this is one of the coolest things i’ve ever seen, especially the rotation
Excellent video as always. Definitely looking forward to the video on the caustic patterns in glass!
I agree. Considering how hard it's to compute caustics even in 3D rendered images, it'll be interesting to learn how those glass pieces are actually created computationally.
I'm very satisfied that the real object deals with the interaction of a sheet of light hitting a plane, and the animations deal with a line of light hitting an edge, continuing your trend of making a 2D version of a 3D thing to more easily explain it
It reminds me somehow the method which it is used to see imperfections in lenses or transparent materials in industries.
I'm happy I solved this in my head before you explained it, solving a puzzle to start the day is a beautiful thing.
I'm still confused about how the light goes through the mirror. Your explanation looks like it just bounces back.
The instructions I've seen involved work hardening the bronze by use of a punch, then polishing flat. There would be tiny raised areas where the hardening has occurred.
Tim is such a cool guy. would be cool if you did deep dives on other tools and gizmos that Tim has. every item a mystery to be solved
10:20 Not only is there no smallest uninteresting integer, but that absence proves by induction that there are no uninteresting integers at all. If there were, there would be a smallest member of that set. The same proof you mention also works for "most uninteresting integer", as that property too would be very interesting.
Wow, it's incredible how those people 1000 years ago could predict that Einstein would be born and put his face on a transparent mirror
your probably memeing, but magic mirrors are made today as well, but they are made using different methods that make it easier and cheaper to make
From the THUMBNAIL I was thinking of Tim. I started the video, and then a minute later you can imagine my delight.
What if you had a concave metal "stamp" of the desired image that you pressed the soft brass into and then polish it so it's almost invisible.
Or maybe an image made of a clear lacquer that distorts the light is covered with one that doesn't so you can't see the difference?
Try a solvent on it!
I'm astonished how you pull up random things from all parts of the world and manage to baffle me more each video. This channel is just pure (technical) magic to me.
I absolutely lost it in the end with that clear card. Seriously, I was screaming "WHAT?!?!?!?!" at my screen - a kind of physical reaction that I rarely have when watching any video!
Thanks for bringing that kind of magic into my life, Steve! I'm sitting here, being just baffled as to how somebody can scratch all the itches in my brain so perfectly....
Once again I find your video topic overlapping with the world of ultrahigh power laser driven inertial confinement fusion reactors. That last transmissive version you showed is exactly what we do to the laser beams right before they fall onto the frozen hydrogen target. We don't use a lion image... as you may have guessed, but rather a random pattern of bright and dark areas across the whole area of the beam. This essentially destroys the phase coherence of the beam and DRAMATICALLY increases beam intensity uniformity or "smoothness" by suppressing laser speckle induced by sub-beam or intra-beam beamlet interference as it is focused from a third of a meter in diameter down to a tenth of a millimeter at the target capsule surface. This increase in beam intensity uniformity is incredibly important in suppressing the in-flight formation of Rayleigh-Taylor hydrodynamic instabilities of the 100 million Kelvin plasma as it collapses to a few microns in diameter. We call these optics "distributed phase plates" or DPPs, and without this technique the maximum fusion neutron yield on the machine would probably be something like an order of magnitude lower. Interested to see the next video.
This reminds me of the Cross Atlantic Communications Cable problem
The length of the cross Atlantic cable meant that, for traditional telegraphy, you needed very high voltage for your comms link, which burnt through the cable's insulation. This was solved by using a Mirror Galvanometer. The very small changes in the rotating mirror meant that much lower voltages could be used.
What if the moon we see is the result of a magic mirror effect?
I imagine the way these things are made are as follows: 1) The design is with nothing on the the mirror side. 2) The 'projection' is painted on the back half using clay (in much the same way that sword smiths in Japan have been 'staining' metal for centuries). 3) The device is cooked in the forge so that the clay stains and thus weakens a few microns of metal underneath. 4) The clay is then removed during the polishing process, taking stained/weakened metal with it. Not only would this likely achieve the desired affect, it would also serve as the perfect indicator for when to stop abrasive polishing ("as soon as the design disappears"), and it is perfectly doable with tools and techniques we've had for thousands of years. Acid etching is also a good theory, but I worry about how far back we're able to date the use/concentrations of acids.
What happened to the audio at 0:47?
You ended it on a cliffhanger??? That lion looked like it was a hologram. That's so cool.
i am a geologist specialising in trace fossils but also have a solid backgrownd in minerals and how they form the theory is on point but it will take alot of calibration to work out the right timing and thickness to produce variable crystal sizes that result in the correct level of deflection. Very impressive this is the first thing that really impressesme from a te4chnological point of veiw this is the same as the mechanism from the mediterian. it suggests that there is a lost knowledge or something like if you used this to make call signs, and you have a amplifier then you could send this 160 km on a clear day.
I remember hearing stories about forensics recovering VIN numbers from cars that had them sanded off in an attempt to anonymise them.
The stamping of the plate changed the properties in the metal making it much harder wearing.
The basic mirror shape, with the design on the back, was cast flat, and the convexity of the surface produced afterwards by elaborate scraping and scratching. The surface was then polished to become shiny. The stresses set up by these processes caused the thinner parts of the surface to bulge outwards and become more convex than the thicker portions. Finally, a mercury amalgam was laid over the surface; this created further stresses and preferential buckling. The result was that imperfections of the mirror surface matched the patterns on the back, although they were too minute to be seen by the eye. But when the mirror reflected bright sunlight against a wall, with the resultant magnification of the whole image, the effect was to reproduce the patterns as if they were passing through the solid bronze by way of light beams.
quoted from wikipedia
Enjoyed the video, and the exploration of what is behind a phenomenon I'd never seen before.
But, to be honest, the teaser for the future video about transparencies looks amazing! Definitely returning for that one.
The trick is this:
Cast the mirror in a mould which has the rough 3D image on its convex side. Using a press with a corresponding concave button and a flat punch on the top; press the casting's initially convex side until it is flat against the punch. This will in turn push the casting's initially flat side down into button. Now, the image side will be flat, and the blank side will be convex. Polish the convex side to a mirror finish, and voila!
The mould used for the casting was made the same way. A plug is crafted with the 3D image engraved into it in its expected flat appearance. Using that plug, a mold is formed. Then the mold is pressed into a convex shape, thus distorting the image such that; when a casting is made from it and later flattened, the 3D imprint is guaranteed to flatten back into the original artistic form.
The most interesting effect of these mirrors is how they lead the one to think that there is a causal relationship between the reflected image and the pattern on the reverse of the mirror.
No matter how many times one is told or even says, "looks can be deceiving," one often finds one's self fooled.
Its videos like this one that reminds me how much I love your content as well as your good friend Matt Parker. Found both your channels at the same time, thru number/computerphile many years ago.
I assume it was cast, stress relieved, and then cooled (or heated) for polishing.
I would assume the polishing is done chilled.
When it returns to room temperature, the thickness differences cause very minor distortions.
Perhaps it needs to be an original image on the polished surface side vs the distortion due to the thickness on the back side.
The thermal change before polishing should still translate after the thermal delta to room temp is "released"
But I think management of casting stresses would be quite important
^^^^ This
This phenomenon is basically the reflective equivalent of water caustics, the light patterns projected on the bottom of the swimming pool by the water whose surface is definitely not flat.
Speaking as someone familiar with metallurgy and historical metalworking techniques- I feel like this was probably done right in the casting process. The bottom of the mold would likely be stamped by a 'master' mirror, and then once the metal was poured a weighted cap with the design on the bottom surface would be placed on top of the still-molten metal. I feel like the molds and caps were likely made of a fine clay or pottery to ensure minimal surface defects. Think of how records are made- there's a master, and then copies are created based on that
This would explain the inconsistencies in the surface image vs. the image on the back, and would be a much less complex task than trying to etch it as others have suggested. Speaking from personal experience, it's hard to achieve the kind of smooth, sine-like curves required for this mirror using a chemical etchant. It would cause the etched surface to become a matte finish and require a lot of polishing to achieve the smooth transition between peaks and valleys, even with the miniscule variation in depth.
Perhaps that was how the master was created, but I feel like the majority of these had the images cast right in to them
when polishing and lapping diamonds for record cutting styli we found that the temperature at which we grind would affect the surface flatness ( fractions of a micron ) when its service temperature differed from its cutting temperature. In this way we could grind concave or convex surfaces on an essentially flat lapping skaife. As the surfaces we are grinding are less than 0.1 mm^2 laser reflection would reflect individual crystal lattice dislocations
If you support the structures on the back, seal around the perimeter, and pull a vacuum, the glass will deform where it is not supported while you polish the front. Once the vacuum is released, that beautifully polished front surface will deform as the glass springs back into shape.
It seems like the mirror is made before the embossing.
The embossing process will change the mirror surface very slightly.
nice, ive never heard another channel mention Grand Illusions.... nice that you worked together.
Hmm. I thought it might have something to do with those one-way dark mirrors that are hard to see through from one side, and then reflecting off the engravings on the back. Guess this makes more sense though
The easiest way is to make a cast, and then polish mirror side in high temperature of the piece. Cooling down will shrink material proportionally to the thickness (back shape) making small mirror surfaces deviations.
Now i finally know why they called that thing in Shenmue a "Mirror"! ..20 and some years later the mystery is lifted, thanks Steve :)
Very cool, am interested in the transparent glass projection technique as well.
Oh cool. You know Tim well enough to borrow stuff from him?
Grand Illusions is admittedly a comfort channel for me, so I wasn't expecting that.
Regarding the magic mirrors with mismatched images, it's possible that the piece was cast in a mould with an obverse (the mirror side) that has differences in thickness matching the buddha image. It's also possible that the mould has some sort of inlay of a material with higher/lower thermal conductivity and/or specific heat capacity (e.g. copper inlay over exterior portions of a steel mould). In this way, the mould itself could strongly influence the cooling time of the obverse, mirrored face of the piece, and thus the crystallographic properties.
polished while heated I'll bet. Then the contracting when cooled will give just what you need. Maybe?
Tim from grand illusions is such a cool guy :)
A hypothesis on how they're manufactured:
During the polishing process, the plates are heated/cooled such that thicker parts of the plate have a different temperature from thinner parts of the plate.
This creates thermal stress, so that some parts expanded more than others.
Once the thermal stress is relieved, the polished surface is deformed, revealing the thickness gradient.
Edit: this is very close to what's being described is the video, so :shrug:
Getting such sharp heat gradients on metal is not easy
The first thing I thought of was uneven tension in the metal, I know when people talk about casting (and glass making) that sharp edges and uneven cooling create high and low tension areas in the material.
I suspect thats why the castings are so deep, you could pour water or oil into the shallow parts to increase the temperature difference.
I'm still waiting for the video on the Rayform, and how to obtain one. That's really awesome.
According to Dunnigers Complete Encyclopedia of Magic most of these trick mirrors are made by etching the image desired, then polishing the surface.
To get the effect shown here, you could etch the entire design as opposed to just the outline.