Hi The Thought Emporium, I'm Rafael de la Fuente, the creator of diffractsim; I'm very glad you used diffractsim in this nice video! The discovery you have made is very interesting. There are also more specialized methods to make amplitude holograms with better quality. Do you want to have a talk, guide you, and we try to make them work? 🙂
Oh man, this was my phd thesis! - the simplest method to create a hologram is to just use 2D FFT. In far field, light amplitude and phase is just the Fourier transform of the aperture! - keeping the phase of the FFT result should get you a hologram already, but you need more optimization to get really good phase only hologram. - the conversion to black and white is basically binarizing your phase plate + adding DC value. The binarization creates phase ambiguity, that’s why you get two copies. - if you etch the pattern on glass via plasma etching, you get much brighter hologram. - the binarization is non-linear, so when I was working on this area the best way was to ‘error diffuse’ the phase using simulated annealing. It should be super fast these days to do that on GPU. I had a matlab code for that if you’re really interested. More than happy to share the phd thesis if you’re interested. I got super deep into this lol.
Is your thesis "Free-space optical communications with retro-reflecting acquisition and turbulence compensation"? Searching your username found that on your scribd as "Thesis 11", and from there I found the PDF on Oxford's site via Google Scholar.
Check out 10.1007/BF02935980 I was gonna include it, but the actual holograms worked better so it was cut for time. but it's pure dots and gratings. literally can be calculated in an excell spread sheet and drawn by hand if you were so inclined.
Holograms are wild. As far as I know, this is the best way of making holograms that aren't just rapidly spinning fans, and I bet this would look SUPER cool. Congrats!
@@collin6526 yeah, and that's why I don't really consider them as such. It's just that it's the widely accepted method for creating "high-tech" stuff but the rapid LED spinning doesn't look that great.
LitiHolo made (and sold for a while) holographic film and a printer. They got flooded so are "film only" for a little while. The cool thing about their film is that it's self-developing and self-fixing. The bad thing is that it's... pricey.
@@thorjohnson5237 I love litiholo, good to see them start producing again. The worst thing about the film being pricey is I tend to resist using it up quickly, and then its other challenge, a short shelf life) comes into play :(
I am currently writing together my thesis on direct laser writing of phase change material based metasurfaces and I have to say you essentially did the same things I did just on a MUCH simpler and more affordable basis :D Of course the holograms I produced have better uniformity and such .. but I guess that is to be expected when comparing this to cutting edge research. It's increadible what you can achieve with this setup. The projects your are doing just absolutely AMAZING, especially considering the breadth of topics. Please never stop
Quick suggestion (computer generated holography is pretty much my job). To go from a phase hologram to a intensity hologram, instead of linearly converting phase to grayscale, you should take the square of the cosine of the phase, it should work better.
Also, there are ways to make the mirror image less noticeable (unfortunately it never goes away with intensity holograms). Too long for a comment, but feel free to contact me in case!
@@atmel9077 Yes, sorry, i made a mistake. It's actually (cos(phi)+1)^2, else you double the spatial frequency, which is not that terrible, it just changes the size of the projected hologram. Anyway, the cosine of the phase plus one is the electric field of the interference with a reference wavefront (which you would be using to record an actual hologram on holographic paper, and should be equivalent to the light you shine on the printed one, so a flat wave), and then you take the square, which represents the intensity of light at the recorded hologram plane. The trick to make the ghost image less apparent would be to use something more complex than a flat wavefront as a reference, such as a converging or diverging one, and then add an appropriate lens in front of the laser you shine on the printed hologram.
@@bitsurfer0101 Well, only if you REALLY lower your expectations. At the moment "holographic screens" are only about 1 cm wide, and have a viewing angle of only 20 degrees at best. Also, if you were to make a holographic videogame, you would measure the frame rate in frames per hour. This all makes them unusable as proper screens. They have a lot of cool applications, though, you can use them to manufacture stuff with lasers, and do some pretty cool tricks in microscopy experiments.
The coolest hologram I've ever seen was in the museum of illusions in pittsburgh; it was a hologram of a microscope, and you could walk up to it and actually look through the lens.
I saw that microscope hologram once at the Oregon Museum of Science and Industry (OMSI). It was really cool. Each time I went, I wanted to buy the "Astronaut Ice Cream", but it was too expensive. Finally I saved up enough money and bought some. It was a freeze dried slab of Neapolitan Ice cream, it tasted like dried foam, not much flavour, not very sweet, and I would not buy again. Real ice cream tastes so much better.
I have seen an illusion with a telescope in wich you could look throuh when i was akid on a trip . i never succeeded in replicating it .Would be super interested by some materials on it
This is actually a trick! in the first famous example, the hologram is generated using two different holograms merged during the process. So you are not actually looking down the microscope. But theoretically, it might be possible to make such a hologram with a magnified real image.
6:55 What's shown is not an Argoscope. It's an eclipse maker. The pedals around the edge are designed to *inhibit* any such Argo effect and not throw any light into the center of the image!
That's pretty neat. Do they bend the light away from the center geometrically or do they use that same bending to destructively interfere with the bent light that creates the argo point in the first place? (Or neither)
@@juhtahel7454 Yes, they use destructive interference. To minimise diffraction you really want a gaussian opacity circlur shade that fades to see through at the edges, leading to a smooth non-peaky gaussian roll off for the diffraction pattern rather than the bright rings of the airy disc. Using just 100% opaque material this is approximated with petals, and the design of the petals has been optimised to cancel out with destructive interference at different scales as smoothly as possible.
also btw this sunshield is to block the incoming light of stars so the telescope can search for exoplants orbiting the blocked star and study those planets
I REALLY like where this series is going :) 1. The diffraction pattern and the image are related via a Fractional Fourier Transform, so you can get a better starting pattern using an invFrFT, though tuning the fraction can be a pain. I've never used RF's defractsim though - it looks amazing, so that's my weekend gone! 2. If/when you get to use a photon sieve transform on these, note that the oversizing on the holes effects the noise pattern in the whole image and so becomes another useful tunable parameter when trying to make a visually good image. (but is really annoying if you're trying to optimize for radiometry ) 3. When you get to thinking about the telescope, consider a Push Broom approach (with a modified mount) and the fun things that'll let you do if you happen to also have a diffraction grating around. again loving these vids!
I just wanted to stop by and let you know that your videos encouraged me to go into synthetic biology. I used to watch you a lot during my undergrad studies and never thought I would have a chance to actually do similar work but landed a spot in a synbio lab for my honours and will be finishing my project up this year before going on to possibly do my PhD in the same lab. We are engineering fungi to produce antimicrobial peptides and it's some of the most rewarding shit, I never thought I would end up doing such awesome work and I wanted to thank you for keeping me inspired to do things that I thought would be impossible.
You have an awesome way of explaining things. You condensed 4 years of holography, wave propagation, and digital image processing courses into 20 minutes of practical applications.
6:50 I actually wrote a paper on this topic. Using a pair of axicons to transform the input beam into a very thin annular ring beam (with a diameter slightly larger than the diameter of the obstacle) will produce a higher intensity Arago spot than the traditional method of using a Gaussian beam. We showed it should also be possible to generate an Arago spot with an intensity higher than the intensity of the input source using an axicon and focal lens (under ideal conditions with a very precise setup). I can't speak about resolving an image using our setup, but it was an idea we had thought about trying.
Epilepsy warning for these time ranges 1:02 - 1:09 2:30 - 2:44 3:42 - 4:36 4:49 - 4:54 6:18 - 6:23 7:03 - 7:12 7:39 - 7:51 even at 1080p Enhanced Bitrate the center of the detraction waves are flickering well above 3hz which is the typical baseline that justifies providing an epilepsy warning. I realize there's probably not much you can do now, but it would be extremely cool of you to provide in the future.
That's considerate of you. It is very rare for flashing lights to trigger a seizure, only about 3% of people with epilepsy show any increase in seizure activity due to flashing lights. So that means 97% are not affected. Still, for those 3%, I'm sure it is helpful.
I appreciate it. I've got photosensitive epilepsy. This video wasn't a problem for me but, I'd really love it if this kind of warning was more common on anything with flashing lights. I don't know why youtube doesn't just scan the video and put a label on it. it would be a simple test I would think
@@jameshughes3014 It should be almost trivial to implement as it only has to be done once per video, each video gets trans-coded and scanned and what not a number of times as it gets saved into the archives and advert breaks are detrmined. I believe it is a specific range of frequencies that is the most triggering and this could be reduced to a number or even better the scenes could be played back at a different speed for those that have checked the "photosensitive epilepsy" checkbox on their profile.
With JWST, the main part of the diffraction spikes is from the fact that the primary mirror is a hexagon made of smaller hexagons. Two of the spike pairs from the secondary mounting arms overlap with those from the primary mirror shape, while the third, from the "vertical" arm, creates the fainter, shorter pair of horizontal spikes.
When the first JWST images were released, some politicians and other people were complaining that for the billions of dollars it cost, there were these weird spikes ruining the images lmao
yes, I had the same confusion at the time because of the central smaller spike, I had no idea where it came from. after a while and talking with people that had the same confusion I finally figured out.
"I have major resolution issues" meanwhile "There's no point in using the microfilm people suggested that's about 8x more linear resolution than Pan F is!" ... sigh. Keep in mind that you cannot get the resolution out of microfilm though with closed down apertures, you need like F/2 and just a lens that's quality enough to be pretty sharp at f/2
This! The CMS20 film should be able to get a resolution of about 500 line pairs per millimeter. I've used it for pictorial photography myself (not the intended use case but still awesome), it's absolutely amazing. I used Rodinal as a developer. The Adotech developer is meant for greyscale reproduction. If you don't get the hard black and white you want (the extreme contrast is another feature of this film!) you can try developing it in your other developer.
@@mfbfreak Yeah you actively don't want to use adotech here, since it's for making it usable for normal purposes. You want like the harshest developer possible. Possibly even something like dektol paper developer with constant agitation. A super fast development will decrease dynamic range just the opposite of stand development increasing it, giving more pure black and white binary. (Dektol might be too much just because of blotchiness from the speed, but whatever the fastest strongest stock solution you can find is that takes just long enough to not show liquid swishing patterns)
What you call an "Arago spot" I knew as a "Poisson spot". Poisson predicted them and thought that they were so ridiculous, it disproved Fresnel's wave theory of light. Arago then decided to perform the experiment and sure enough, they exist, to Poisson's chagrin.
I love how this guy can just go from completely chill talking about some kind of mind-bending science, to whatever you call that in about half a second
It's probably been said but the bat signal is symmetrical so you could turn it 45 degrees and have the center line on the axis it's mirrored around, for a clean single bat signal
Hi, are you familiar with coherent diffraction imaging? What you're doing here is kind of a weird back-asswards version of that. If you let yourself stick to the specific case of Frauenhofer diffraction (far field), then the image:diffraction pattern relation is a Fourier transform (amplitudes not intensities), and there's *tonnes* of algorithms which can "solve" that for you, hybrid input output (Fienup 1986) is probably what would suit you best (likely the "fourier phase retrieval" you see in that python code uses a similar algorithm). Can send more information/advice/maybe whip up code that'll do what you want based off of my own if you'd like or don't want to dive into that particular kettle of fish, though it's really not super complicated when you get into it!
Yet another example of how useful fourier transformations are. Kind of like the science version of when you buy a new car, suddenly you notice that car everywhere you go. You should run with this idea to apply a fast fourier transformation but have the output constrained to a modified photon sieve rather than the fresnal lense approach. I'd love to see that actually.
Another interesting point is that this is the technology behind EOTECH optics. When you look into the box you’re seeing a hologram projected onto the target, which means even if you’re looking off-axis, if the holodot is on the target its gonna hit.
I’ve been into holograms and learning about how they’re made since the late 80s in high school but now, thanks to your video, I finally have an understanding of how photo film based holograph generation works! Your channel was the quickest I’ve ever subscribed to!
This gives me a flashback 41 years to the past where I made holograms for my final high school project (I was 17). At that time in Belgium, you needed to present sort of an validation exam before doing university studies, and as I intended to study physics... I made holograms myself on glass plates of various objects, and also made double exposures to look at slight object deformations due to stress or temperature changes. Just before it was my turn to present my project to the jury, and my classmates, I wanted to show my glass plates to the girl sitting next to me, and... they dlid out of the folder and crashed on the ground😮. Most plates were in too small pieces to be useable and I was in shock... all the work that took weeks were scattered. I still had a few big glass pieces left, and it allowed me to present most of my prepared slides and more: The fact that each piece of a broken hologram contains the full image of the object. In the end, I got the best evaluation, started studying physics and became a lecturer. The lady that caused my hormones to break the plates also became a professor and I'm still in contact with her.
There's a team at MIT using a spatial light modulator & lasers to project a 3D hologram into a gelatinous cube of UV curable resin in order to create a full 3D resin print simultaneously with amazing detail resolution with no layer by layer printing- it does require post processing in that the beams converge creating a hardened shell of the object, & would then be cured to make the inside solid. I was thinking would also be very possible to drain / suction it out of the center if you just need something as a reference or something ornamental etc. You can also buy 3D printer bed plates usually coated with PEI that will allow whatever side is printed on the surface of the bed to leave an etched hologram into the surface. Your refined technique could be perfect for making custom ones, like for instance a hologram of a company logo, or The Thought Emporium...
12:40, dude I watched his video on this phenomenon and I was enthralled. It's one of the most mind blowing things I've ever seen. I thought I understood diffraction, having taken college physics courses, but the lensing he demonstrated in that video was nothing short of astonishing. Like it makes total sense that it works, but it feels like magic. You make the special pattern and light bends to your whims like some kind of arcane spell.
Awesome work. One minor nitpick: 17:45 That’s not just mirroring, it’s 180 degree rotation… or double-mirroring, across both the X and y axis, which is functionally identical to 180 degree rotation, in that it produces images which can be overlaid and match up without flipping them over.
Before changing the pattern to greyscale try keeping only one color channel. The color version may be accounting for the chromatic aberration when drawing the pattern. My theory is a monochrome version doesn't need all of the color channels.
Flipping a battery backwards in the middle of a big project is the perfect description of why jobs of an engineer or scientist often involve project time estimates that are reasonable, but every few projects, the estimate is wrong by 3-5x.
Nope, this is a perfect example of what you shouldnt do, when you go climbing watermelon trees . You dont want to find that perfect big juicy ripe melon at the top of the tree . Only to Drop it because your camera battery was inserted backwards 🤷♂️🤦♂️
On Canadian currency, shine a laser (or pin point source of light, cell phone flashlight at a distance works) through the Maple Leaf, and it will project an image of the denomination.
fun fact: you can create a similar lensing effect with your fingers by creating a small pinhole with the tips and looking through it. it's easiest to see when you start really close and pull away. moving it around also shows it once you find a good distance
This is awesome! It's so cool how "just" using film lets you make features that would normally be so hard to manufacture. Though I was expecting a shoutout for gratings that show the denomination through our money.
Wow! That contraption making each pixel output a different wavelength each time is awesome. That means you can make more natural colour images than with only RGB.
JWST has 8 spikes. The 6 brighter main spikes come from the hexagonal edges of the mirrors, and two of the angles also happen to line up with the two lower angled secondary supports. The weaker pair of horizontal spikes you ignored is from the vertical secondary support that doesn't line up with any mirror edges.
The video has just started, I but I saw your want to transfer holograms onto chocolate. I'm an artist researching the use of lasers to create structural colouration on sculpture surfaces and I'm currently looking at fiber lasers and seeing if they can possibly colour engrave silicone and then if it's possible to use the silicone as a mold to transfer the iridescent pattern. I feel like this fiber laser - silicone avenue may be of interest to you? Regardless, thank you for the endlessly fascinating and informative videos. You've been a major source of inspiration over the years. Anywho, back to watching!
oh my god, amazing, i am currently working on my end school project where i want to show as many oldschool photography i can lear, actually i am only film shooter and i am trying older technique, but this? this is something else and perfect example about another art look
I designed grating holograms while doing my graduate thesis, and I can tell that there is nothing less intuitive that wave mechanics... if you think you got it, you are wrong. A few helps: 1) with your method you could do better than a frenel plate: if you got a lense and you cut it in "2pi-mod slices" you will got a flat lens that do makes image at focal poit, but if you change the curvature of the lens' slices by straight lines, you got a frenel lense, which makes focal point but cannot make image since has different focal points, now even worst, if you just make a binary pattern of those slices, you will have a frenel plate... since you could indeed do make grayscale patterns, for the telescope better try the first curved-slices such you could make a real flat lense (just a frenel plate with sinusoidal instead of binary pattern would work better, FP its the worst way) 2) also, you could design from the very beginning an aspherical lens instead of a classic parabolic or spherical one, it would reduce a bit the chromatic aberration (but the 2pi-mod is color dependent, so it will still be there) 3) looks like those holograms designed dot patterns instead of making "real images", but that could be also be done: the Fraunhoffer approximation tells you that the far field image is the fourier transform of the holographic mask... you could, in principle, to make modulated gratings (so you get twings images away from the central beam), modulated with you desired pattern on its phase (its too long to put eqns here, but you could find them in internet)
I actually manufacture the master stamps and actually a lot of the time just cut a diffractive grating into lenses and parts through diamond turning. It is very expensive and time intensive but the results are unmatched
Buy a bunch of CMS II 20 now, EPM (the company that contracted Agfa to make it) went belly up about 2 years ago, and Addox has said they are running out of the film. I did send a message regarding where you can get other microfilm in bulk rolls if you need it, I hope that found you well.
Ooh... could you pass that along to me? I haven't been playing with film for a while, but... I can't afford a spatializer. Half of me wonders if I can can draw a hi-rez arbitrary pattern on a CDR...
You can get an amplitude SLM from LCD projectors. Yet you'll also end up with a pattern convoluted to a grid (you may be willing to mask the others points thus). I already did two of those for teaching purpose, it's quite cool (since you can drive it with slides displayed on your laptop), yet you don't get phase control. If you want to try this, don't hesitate to ask questions!
Try using a laser that pulses at high rate, AND rotate the laser (IE spin it). Pi is transcendent number, you're rotation won't be 100% centered and so the particles in laser beam will change angle + trajectory of how they're hitting the film, causing 2+ sets of defractive angles. This should help with resolution AND looking at the hologram from different angles (besides just straight on.
The space "flowers" you showed are not for using diffraction as a telescope, they actually do the opposite. They are designed to reduce diffraction to a minimum to effectively block out starlight from a star to make exoplanets around it visible. It's called a star shade.
the starshade purpose is to reduce glare of a star, so that the extrasolar planets can be seen, with star-glare reduced. starshades are placed further away from a telescope, and it is not in the diiffreaction-focus.
One of the best holograms I've seen was that of a microscope, and you could actually look through the holographic microscope and see an image of what was on the microscope slide.
just a correction, and something I was also confused about at the time. the 6 diffraction spikes from JWST don't come from the arms, they come from the shape of the mirror. the diffraction spike from the arm is actually the smaller central 7th spike, the other 2 spikes from the other two arms are actually aligned with the shape of the mirror, so they overlap.
I was looking into this back in the early 2000s for a particular project and there was a single white paper and demo program that did something similar, and like your method, the idea was to print the pattern and then reduce it onto film. I'd long given up on the idea but I may revisit it for a project I'm working on.
Amazing that youve explained the theories behind holograms. I love this video so much, it has theory and practical aspects to it. I learned a new thing today 🙏
You can use lcd projector as spatial light modulator, but with half of the possible light phase shift compared to commercial SLMs. I done it and it works
@ 4:34, that is why the beam from space, that destroyed Halina Hawaii, didn't strike Blue stuff. That reflection from a pool would destroy the emitter.
10:26 - it's not about resolution ) Adox CMS is not a unique magical thing. it's just a technical film made for photocopy process (exactly what you are doing). such films have two features that make them work great as a mask: 1. their base are clearer than regular film and not scatter light too much. 2. their image are denser than image on regular negative. so when with regular "black and white" negative you actually get it more like "light gray and somewhat darker gray" with photocopy film instead yo will get something more like clear base and crisp black mask. also you can try Rollei RPX25 or Silberra U200
Nice and kind of answers a question. Building a tunable metamaterial for slits and the orbital energy levels for part of the tunable range. Thank you for sharing. Keep up the good work.
I think one simple change that may help is taking the grayscale image and using a threshold filter to make it pure black and white. as I would assume the gray parts of the image are causing the laser beam to scatter randomly.
Thinking about why you don’t get good picture resolution at 19:23 is likely because of: - your grating resolution is not high enough. Ie you need more dense encoding in the grating to have higher fidelity images. - with higher resolution, you might want to add more offset to the grating so that you have less area near DC (or the optical axis). - the laser beam phase uniformity probably comes into play is not simulated in the software. The complex amplitude of your laser beam will convolve with the desired output, creating specking and broadening each pixel.
To turn an amplitude hologram to a phase hologram you need to bleach it with a potassium dichromate solution or other holographic bleach. You get better results if the film is overexposed. I played with computational holograms back when I had to use a PC XT for calculating (it took about 12 hours per point, so my holograms only had a few points). The results were printed on a LaserJet and photographed from across the room. Fun times! Looking forward to future updates.
One thing I did with a lazer pointer I once had (well, the diffraction disks that came with it) I held one of them up to my eye and looked at a point source of light (basically a street light outside), and I could see the image without needing to shine a lazer though it.
I think just the fourier transform of the image should work, as long as you pass through a lens afterwards. This is a standard technique of fourier optics. Although I think the lack of phase control necessitates an image doubling, since (in fourier optics speak), the one pulse at the center in x-space need a infinite spike pattern in k-space (so infinite diffraction grating gives a point), but translation by x of the point results in multiplication by a phase factor e^{ikx} (k is the wavelength in fourier space). The way to counteract this would be to have a second copy that is translated by -x, so the total phase factor is (e^(ikx) + e^{-ikx} = 2cos(kx)), which is real. Another option is to have the second copy by out of phase, so you can sin(kx) or any e^{i theta} cos(kx + theta/2) in general. This is ok because the phase is not varying, but a second copy is necessary. Maybe phase control can be achieved with some sort uv cure resin, which can be partially cured in places and then washed away? Then the varying heights of the resin (with it's different refractive index), can give static phase control, and this can overlaid on the mask which gives amplitude control.
You could rip open an old lcd monitor, you can control it from any laptop. They also have a bunch of diffusers and polariser sheets which I'm sure you could be creative with. I mean you could use the sheets to diffuse the lasor beam, and then parallel it again. making your projections larger. or use it to concentrate the white light so you don't get such large rainbow diffusions.?
I really wanted to make holograms in the 80s when I was a teen. I had access to a couple of HeNe lasers, but it was so hard finding books about it and even harder finding the equipment for a kid in small town NZ.
When I was young and I did not have glasses yet, I used 4 fingers to make a tiny hole to look thru.. It worked like a lens to see sharp in the distance.. Still use it when I loose my glasses..
Wow that's so interesting! Have you considered doing an episode on Upside-Down goggles? The image appears upside down whether or not the user wears the goggles upside down.
I wonder if you could use direct digital film exposure to get better results. For example, the imagesetters used to make films for prepress. They do 2540dpi, and the film is extremely high-density, so black is exceptionally black. And since these things are usually tabloid+ size, you could potentially make enormous diffraction gratings with them.
For anyone interested in observing an Arago spot, I've had great success using a DSLR camera without its lens, a laser pointer, reading glasses and a ball from a ball pen smashed between two transparent panes.
It feels a little like I'm having a fever dream, because most of it has been scrapped from the internet and I can't find information regarding it. Some 12-15 years ago there was a company going by the name "zebra imaging" that could print holographic plates using 3D models, that were about the size of an A3 sheet of paper, full color, White light viewable, and with a very significant angle of view. I recall they had one hologram of a building that, on a different set of viewing angles would hide the concrete and show only the hydraulic network. It feels to me like this is lost tech! I can't find either into on how they did it, or other companies that did it. And honestly, this should be put in the open. Don't know if a normal DIY would be able to get close, but considering we do have people building electron microscopes and ICs in their garage... Food for thought, this might be a direction you would enjoy taking.
In the first place, there is no light interference at all. Light speed remains same regardless of media or type of light, if you still have doubt, please refer to " Science Alert, current research in Physics".
Instead of using a binarized (black/white) fresnel lens, consider using sin and cos like fresnel lens. "Hard" edges are inducing higher frequencies. So a grayscale image instead of a black and white print will give you an even better focus spot
I used to watch hologram made by the sun. During cloudy days with a bit of the sun shining. The roof of the house was corrugated iron, it was riddled with holes. At the right angle, you could see the clouds in the house going in a different direction. The hologram was bigger, yet not focused enough 😊
Imagine making a hologram of an entire galaxy with everything perfectly sharp, you could zoom in to the grains of sand on Earth from hundreds of light years away.
Since there seems to be a lot of people that know stuff about this here, i have a question to y'all. I've seen a hologram (of the type he made the church with) of a microscope, and you could see through the eyepiece to a magnified sample. It got me wondering, if you could just plop a camera lens on a table, make a hologram out of that, and use the hologram as a lens? I get that you couldn't adjust the focus on it, but would it work? A paper thin telephoto lens would go hard
You can print these directly to a transparency with an ordinary laser printer... I did back in Ye Olden Days. The spatial resolution sucks, but that 'only' limits the angle of the diffraction. OK, the images were very, very narrow. But cool. Luckily our school logo was symmetrical around the origin, so It did not suffer from the negative diffraction orders. And naturally, you need to expand the laser in order to get any kind of resolution from such a coarse mask.
Hey, there's a film stock that was used to print sound on film. It's basically very high contrast, with very dark black. Might be a good stock to try for this purpose.
17:48 This is also a property of the 2D FFT. Every point corresponds to a sine wave and if you look at a 2D sine wave, it will look the same if you flip it around its axis. Mathematically speaking, the spectrum is always symmetric if the pattern is real and not complex, which is true, since you're just changing the amplitude and not the phase
Fresnell lens are weird. Oldschool projectors used them, and if you looked at the lens directly with the projector off, there was an effect as if the reflection of outside light curved out of plane.