 In the morning, use the hashtag C3T, or write us an email to hello at 3tlingew.org. It's a laser physicist, it's a science-slammer, you know, it's a very big picture. It's just, it's about television, Microsoft. It's about picture, how you get them and how they're related to each other. And how our image is informed by them. Thank you very much. Yes, good morning. Good morning everybody. Let's get started. First of all, me, why am I standing here? I'm a physicist, which changed to biophysics to build a microscope. I told that about a comic during and explained to her like half an hour. And she drew that picture. She's a very good artist and follow her on Twitter and I want to thank you again for this great image. I'm a physicist and somehow I got my boss to biochemistry and started building microscopes. The talk's name is about the four things in life. So let's first have a look at the four things which I'm talking about. Here you can see a one-set piece. You can see a little piece of glass on it. There are cells on it. You can't see that in this camera. You have to use your first microscope for this. And that's a face-contrast microscope. And the first picture you get, that looks like that. You see a little bit on the right-hand side, there are some little structures. On the left-hand side, you can see the shapes from the sand piece. So a feeling for the size. If I zoom in to 20, you can see a little bit more of the cells. At 40, you can see. You see the cores of the cells and also the cell periphery. If you want to see more of the details, you can't use a face. You have to use flour since microscope. So you color and you put ionates and you register the slide. And then you can see the structures in the cell. So we still use the 40 times of magnification as we use it here. And change to fluorescence. And here, you can see we get very beautiful images you can see here. And that's why I really love the field of microscopy. If we zoom in in one of these cells, on the left-hand side, you see the whole picture on the right-hand side. It's just the part of the cell. There you see there are a lot of things in there. And if you zoom in even more. So you want to see the internal structures of the cell. And then you can see it gets blurry. It looks like Lord Spaghetti. So before I tell you why this is happening. So what are these kind of things? You can see there, these are microtubules. That's very important part of the cytoskeleton of the cell. So they can keep shape. It's formed from proteins, which form such rings, fall from 13 parts, which form like a cell case. The diameter is 25 nanometers. On such of these macroscopic pictures, you can see 25 nanometers. They don't appear to size. So if you look at the scale bar on the top right corner, this doesn't fit. These tubes are way bigger. And that's not because of magnification. But here, this is the key reason. Then physics. Some of you have already heard about the fraction limits. And Abe had come up with this. The most important part for Christens is this person over here. That's Rayleigh. He came up with the Rayleigh Criterium. And this Criterium tells you how can you, if you have two point shaped sources, when can you still defect them. So for Christens' macroscopy, all the styles we use, which gave us these very nice pictures, are just form sources. This is 250 nanometers. If you go smaller, you can't tell them apart anymore. At least not with normal macroscopy. You can go around to use a refresh limit. There are various points. But you have other problems if you use these other techniques. I have a few of these. The first is structure dilumination. Sim stimulates emission depletion. And the last one is single molecule localization. 240 for stats, the Nobel Prize was awarded in Chemistry. How you use the techniques to go around this fraction? For Sim, you use free transformations. Stat goes into laser technology. So you directly use the light you're going to use. And the last one, storm, you use a fitting of signals. Stat is a completely interesting technique. But it's a little bit boring in this context here. Sim and storm need a lot of calculations and a lot of programs. That's what I would like to talk about today. I'd like to start about Sim. Sim works like a firing. We use the Mojev pattern effect. On the left hand side, you have two grids. If you move them back to each other, there are funny structures. You maybe notice when you're going out to barn with the car and you see how the pattern also happens on bridges. On the right hand side, you can't see much. So start shaking your hand. Can you see something? I hope you do. You should see a person in a red hoodie. This also is based on this more fringes effect. If you don't shake your head, you just don't see it. Okay, someone will see it. So it works. So using this Mojev fringes, you can see something. That's why we're using it in microscopy. We use a grid and we illuminate it. We project this grid into our sample and shift it, turn it around and again. That's how you make this Fourier transformation from this Rodeida. That's how they look like. If you put them on top of each other, you get this flower-like shape. Fourier transformation, what? That's just a transformation from the image to the frequency space. So on the outside, you have the small differences and essentially, you have the big differences. So what you get out of it are these funny shapes in the circle in hand. That's like a normal microscopy picture. But sim, these rotations, etc., it allows us to get more information from the corner parts. And using this technology, you can get such a picture. And you can see here, we have much more information in one of these images. Just because we have rotated and shifted. So we have taken several pictures. How does it work? We take the Rode data, the shifts, the rotations and put it into a software. For example, here, FairSim. FairSim is an open-source software. If you want to just look at GitHub and have a look at it. And someone I've suited with at the University of Bielefeld. No reaction. It's early in the morning. The Rode is software. And for sim, usually there are only the software from these microscope developers. And then it's just a black box. And I just wanted to start a free software, open-source software. And in February, there will be the next paper about it, which will be also an open-access paper where they start co-optation of GPUs. They will beat every commercial software there. And commercial software needs two minutes for one image and this one 30 milliseconds. And that's, yeah, right. And the nice thing about this technology is you can use the Rode data and perhaps someone thinks, oh, we could do it even better. Maybe this grid isn't a perfect sign and we use another one and we have different software and we could put it into another software and then our picture will get even better. The Rode data is just fantastic. If you take away something from this talk, then Rode data is great. And so I have another picture here. It's a sim image. And it's an image that's really great. So the Rode data is a super-solution for globalization, macroscope from liver cells. And it was found that in diagnostics. So the applications where it has to be fast, there was really an advantage from super-resolution for certain diseases because the pores in the liver cells have the perfect size to use super-resolution paper and open access. You can look at it online. Please do that. And this is from my dear colleague, Fjola Mönchmüller. And now we go to localization microscopy. That was my PhD thesis. And the technology is called Direct Stochastic Optical Reconstruction, or DSTORM for short. And this is about blinking for the chemicals and dyes on it with chemistry. And the dyes don't want to do that anymore. And one out of 2,000 is working. And the others are off. And if you look at that in a microscope, that's what it looks like in the beginning. I turn on the laser. And then I really need a laser for that. And then by and by, they all turn off. And then they start flashing. And these signals are really signals from the individual molecules. They are point-sized. And if I record these signals, 10,000 or 20,000 pictures as raw data, and then from such a blurred picture, we can get this one. If I put it into the right software, it's called RapidStorm or ThunderStorm, very creative naming. Also free software. You can download it in just Google for localization software, ThunderStorm, and localization software. And with ThunderStorm alone, you get different results. But you can, if you reconstruct it, like here, you can see that, aha, it's not just one microtubule, but there are two. And you can really distinguish the two and which made the biologist in me very happy. It's very clear that this is just an improvement in resolution and how it works. Thank you to Ricardo Enricas, who made this video. If you can see, if you record the blinking of the Eiffel Tower, then at first you search for the signals and the same way you do it in the cell. And then you reconstruct the picture dot for dot. And the signals we found are also diffraction-limited and blurred. But if you do it point by point, you know that every point is a single dot. And the resolution is only dependent on the number of photons and not the limits of the optics. And that's why the resolution is increased substantially. If you want to do more colors, there are some problems. So we stay in France with the colors. So chromatic aberration causes this blur. And if you use a white light and from red, white blue, it's smeared. And it's chromatic aberration. Normally, you have to correct that with registration. And you take the two pictures and try to make them match. But such a registration is never perfect and there are bound to be errors. In my PhD thesis, I developed a technique that is specter de-storm. And we make color from raw data. And you can look at the software, how the software works. Raw data are great. And here's some more examples of what it looks like. And we can just make two colors. We can also make 3D reconstructions. All this looks a bit pixelized. But we are really at such a high resolution that you can hardly believe it. And because you hardly believe it, I will give you a little comparison of scale. And so we went down from the beginning. So this little sphere on the right side has a diameter of 150 nanometers, which really isn't a lot. And such a cent coin has 15 millimeters diameter. And if we imagine that one of these physicals is a pinhead of 3 millimeters, then for comparison, the cent piece would be 300 meters of three soccer fields. So to remember the picture from the beginning, these pinheads of physicals are these little green dots that you can see on the normal fluorescence microscopy. So we are really, really well into it, getting stuff out of the data. So it's probably a good idea to build your own microscope because there's a spectral demixing de-storm. There was nothing like that. So here's a time lapse how we built our microscope. So we just move it into another room. And some people say, especially in the life sciences, DIY bio, no, you don't do that. Don't you have enough money? So you can just order one. And I say tinkering is science. And science is tinkering. And that's part of it. Otherwise, there's no progress. Otherwise, we wouldn't have this wonderful technology. Many of these things were developed in 2006 to 2008 because there were some crazy physicists who teamed up with biologists and thought, how can we increase resolution? I don't understand that. Everyone who thinks that should think about it and listen to me for some more. And building this microscope works because there is open software. And it's micromanager based on an imageJ. An imageJ is used by biologists for image processing. But micromanager is wonderful. And the modern microscopes can be controlled by that, our STD storm can control all our different microscopes that we have. They were detected. And anyone has a new device without a driver, he just rides a driver, puts it into micromanager community, and everyone can use it. It's wonderful. And what's also cool is this picture here. This is a picture from my 12 euro microscope, which I have here. So the software which can control the most advanced microscope can also control this 12 euro microscope, which you have from online order. And Arduino also works if you build your own microscope table. So look at this, it works. So some people thought, hey, building ourself, wonderful, we can. So there's a blueprint for cost effective to just have microscopy, so such microscopes like these you can buy for 800,000 microscopes and can do two colors and 25 nanometers resolution. But you can build it for 20,000 euros with two colors, 40 nanometers. And it's important that we do these, think about these things. Because our universities say we can offer to students things from research. And what that means for third world countries where there's a university who wants to research, I don't think I have to emphasize it anymore if you look at the numbers. The same is for SIM. As fair SIM, I talked about it before, fast SIM is an idea how to do structured illumination microscopes cheaply. And if you look at the commercial offer, it's about a million for a commercial SIM microscope can do four colors, needs time minutes for the image reconstruction. Or you just use the open free software and buy the parts with not quite quite as beautiful, really neat surfaces. A lot cheaper, 25 kilo euros, three colors and 30 milliseconds for fast SIM. And it's just wonderful. The team, yeah. So that was really the applause for all the people who work in this field. And I'll meet them again in two or three weeks. Maybe they're watching. Hi. So why do so few people play with microscopes if you're working in the life sciences to research first and foremost, talk about, think about their experiment and the cell and the little bit of life that they have. And so making all the experiments with it and that's you have a head full with all the things. And sometimes you have completely different ideas doesn't so stick a banana in your head or something silly. All this works a bit slower. So thinking about open data, open source and making the software available. But more and more, I realize that it really can't work differently. And what I want to tell you is it's not so much about pictures but about data. And it's about making everything open and accessible. And we need good people who can write better software as I can do. I'm just a physicist. So programming is just with a hammer. And it's just a sculpture, although it should be a painting. But it's a stupid comparison. So there's the conclusion. I talked about the diffraction barrier and how it makes problems. So I talked about better computers, better software culture, modern cameras, and all in all, better technology so we can have super resolution such as SIM and D-Storm. And then we can make cheap DIY things. But it's mostly about data. So microscopy is data collection and not necessarily making pictures. So raw data are great. So I hope you remember that. And we scientists are a bit slow. So many people have opened their software and published on the open access and published data on the internet. And it has to be more, so be patient with us. But it'll be more and more open science and everyone can play. And I invite everyone who has love for imagery construction and programming. And so look at the things I told you about today. And maybe have other cool ideas that we didn't have in our labs. And so I say thank you very much for listening. I'll be in the assemblies, hacking and communications near the Zendertzentrum by the wardrobe. And yes, thank you very much for your attention. Thank you very much for the extra applause for open science. We have time for about four questions. Signal Angel, is there something from the internet? Unfortunately, no. Here are the microphones. I will start with you. Question. You can start immediately with your question, but I want to say something. I have micro-etchor running here on this crabby 12 euro microscope, I put your ascent. And even with a 12 euro microscope, you can see the magnification here. It's pretty impressive, though. What is the star? There, you can see the edge of the star. Softer, which runs very well. OK, that's about the question. Easy question. You said you made several pictures, and then you put them together, and then you have the whole picture. So what is happening when the things start to move on the microscope? Answer, that's an important question. If you want to look at dynamics and biology, you have to be very fast. Using Sibon outside in the beginning, if you want to high resolution, you have way different problems. If you have to structure the nation, you have 15 pictures, five pictures, rotation, movement, rotation. If you manage to take these 15 pictures fast enough, let's say in 20 milliseconds, and you have enough emission, if you have a fast reconstruction, you have the possibility to observe these dynamics in biology. If you have to take 20,000 pictures, then I won't be able to observe very fast processes. But with modern CMOS cameras, you have the first approaches of this localization microscopy. In one brush, you take 10,000 pictures, reconstruct them, and then in perspective, you have four pictures. Then you have high resolution microscopy. That's a start for slow dynamics. In principle, all this microscopy technology, such as concentration microscopy, you exchange resolution with time. So what you see in space, you lose in measuring time. So for one of these two color images, it takes you about 10 minutes. OK. A question? If you use self-built microscopes, how are the data reproducible questions answered? In fact, even the commercial build is not like a standard. It was self-constructed. So that's your numbers, maybe three. You have optics that are standardized in a strong way. Yes, OK. If you add something yourself. A person asking asks, are there more standardized optics? So why is this? By the things I just showed, you're still in 20,000 euros. You still need a good objective. And you have to use a standard objective. And that's the point. That's the point where it stands in force. And camera and all the optics. You just standardize material. And once you've built this microscope, you have to characterize, you have to test measurements. And then it works. Then it's reproducible. But still, they're looking very, very good results. And they think, yeah, that's good. But still, it looks like a real microscope. But yeah, it is a real microscope. You must be afraid to build your things yourself. You have to do good test measurements. Then you have reproducible measurements. But the same for commercial measurements. You have to prove they're stable. Otherwise, hm. Question? There are different types of techniques, such as conversion microscopes, to go through these limits. For example, using different wavelengths. Was this, in fact, a green, or was it colored? And are there other approaches using maybe UV, or X-UV, or maybe emulsion microscopy? So you can, beyond this diffraction limits, you don't know where this 150 nanometers are coming from. So is it going in this direction? Answer? Wavelength is not really helping. In conventional way, this is not very helpful. If you want to tune a wavelength to get a better resolution, why would you want to do this? So you're not losing time. So you want to observe life forms. If you go into your hard UV and you look at living cells, it's only a very short time you can do this. If you have a cell which is able to move fast enough, it will move away if it's able to. You can hunt them, basically. We also have to have fun in the lab, okay? So wavelength, it's not very helpful. It's in maybe a little bit. You maybe use 405 nanometers. That's the best resolution where you can change the grid, but you can't go into your UV. In standard microscopy, there are also other approaches such as convolution or kind of focal microscopes where you can't just take out and calculate something else. But if you really want to go to this resolution of 150 nanometers, you can still see a structure. There are still structures smaller, which you can see. And we use red light. So it's 643 nanometers. So the wavelength is not that important. The point is, as many of us as possible have to come out. That's what gives our resolution. Last question from the Internet. The Internet asks, what's the size of such a microscope? I didn't know what are the relations between size and complexity of such a microscope itself. Also, there are two sides of the thing. If you have a commercial microscope, they're not allowed to sell it. If you are able to look into this ocular and you can switch on these lasers, it makes sense that you're not allowed to do this. Because otherwise, you only have two eyes and you're able to get lasers into them. So all the thing is bigger because it should be stable in temperature. And in the end, you want to have the standard size boxes. In your own boss, microscopes, you have big advantage. Why do I need oculars? I just want to do a higher resolution. I don't want to look through them. So my formation is perfectly aligned with my camera. I'm just looking at the monitor. And later safety is perfect. Because I seem to be tempted to look into this. So I just see it on the camera. So the philosophy of how to do this is just completely different. You can get a point in the middle, but the point you really save money is these big boxes, heavy things, stable status. Whereas possibilities to save something, you can put in a filter, you can soft them, the software knows what it's doing. You can click something. In a cell phone set up, you have to screw a little bit and then you have to put this in there. And you have to look, is everything still set? It may be a little bit annoying. It's not that nice and user-friendly, but still there's a solution. In its high cost, you have also maintenance and support. And so you have a guarantee which you don't have on your own building. It doesn't change anything on the price. Okay, thank you very much, André. Thanks again. Everyone who is interested in physics can stay here and talk about certain afterwards. So let's thank the speaker again. And we are thanking you for listening to your translation. This has been Pink Dispatch and Frau Blauwal. Please give us feedback.