For a very good explanation of Stimulated Raman Scattering (the primary nonlinear process in this video) please check out: Your Favourite TA's video on Stimulated Raman Scattering here: ua-cam.com/video/sZEGMDZKyPQ/v-deo.html
Excellent work. It's a shame these videos have so few views, mainly because most people haven't got the slightest idea of what's going on or why the underlying phenomenon is so truly weird and extremely unusual to actually observe.
@Les' Lab no, keep them as is. Some of the unfamiliar terms in your videos prompt research into them on my part, such as soliton fission. I honestly have no idea what that means but I am headed to the internet to find out.
Depends on where you want the channel to go I guess. To be honest I think Styropyro pretty much has the fire and explosions and burning random shit with lasers for 14 year old viewership angle covered on youtube ("not that there's anything wrong with that"), but that's what the masses like if you really want the megaviews! I'm content with the deep, real science and moderate views, for my part.
The peaks at 16:52 arise due to a process called Stimulated Raman Scattering (SRS). To understand this mechanism, consider a simple, isolated molecule like H2. The molecular bond can vibrate at a set of distinct frequencies due to quantum mechanics. If we were to shine a laser onto the molecule, a photon can "donate" some energy to the molecule by exciting vibrations. Since the photon has lost energy, its frequency has decreased by an amount corresponding to the vibrational frequency of the bond. Since the frequency has decreased, the wavelength must have increased. For an amorphous material like silica, the molecular bonds are extremely irregular compared to a simple H2 molecule. Therefore, the possible vibrational frequencies form a continuous "band" with a peak at ca. 13.2 THz and a width of around 5 THz as shown in Figure 2 in this link: www.rp-photonics.com/raman_gain.html Therefore, when shining a highly intense, coherent laser pulse with a certain carrier frequency into a silica fiber, coherent molecular vibrations will be excited. This transfers power from the carrier frequency to one that is approximately 13.2 THz lower; this is SRS! Furthermore, this lower frequency will keep "stealing" power from the carrier until it gets so intense that the process can repeat for a frequency that is an additional 13.2THz below that one (and so on in a process called "cascaded SRS"). This explains the observed "comb" structure. Considering the peaks at 16:52 labelled 497.7nm and 519.4nm (with an unlabeled one in between), we can see that the peak spacing is 10.85nm. Converting this difference in wavelength (BW_lambda) at 497.7nm (lambda_0) to a difference in optical frequency (BW_f) using BW_f = c*BW_lambda/lambda_0^2, we get 13.13 THz, which is very close to the peak frequency of the Raman gain spectrum presented in the link above. If you are interested in some of the theory behind nonlinear phenomena in optical fibers, feel free to check out my channel, where I both demonstrate these effects experimentally and simulate them numerically using custom python code: Experimental Self-Phase modulation : ua-cam.com/video/wZyaTVTmmBk/v-deo.html Experimental supercontinuum: ua-cam.com/video/ZsZXqxwJBcw/v-deo.html Numerical simulation: ua-cam.com/video/xIdozUy9Nas/v-deo.html I plan on upgrading my own code to handle Raman effects in the future, but I know that others have created packages that allow for simulating SRS in optical fibers: github.com/WUST-FOG/gnlse-python This was a really awesome demonstration, great job :)
In addition, though it's not an effect I have worked with myself, I don't think soliton fission is taking place in your experiment. The sources I have been able to find (see links below) state that soliton fission takes place for pulses on the scale of tens to hundreds of femtoseconds, which is much shorter than the 2ns pulses used in the present experiment. The basic reason seems to be that a pulse with a duration of around 100fs has a spectral width, which is close to the 13.2 THz peak of the Raman gain spectrum (see previous comment). Therefore, the spectral tails of the pulse itself end up stealing power from the carrier frequency, which is different from the SRS case, where the lower frequency photons are not initially present but "created" from a mechanical vibration excited by the pulse. Combined with other nonlinear effects (Self-Phase-Modulation, Cross-Phase-Modulation and Four-Wave-Mixing) as well as dispersion (different frequencies propagating at different speeds leading to broadening in the time domain), the spectral output is usually much more "messy" than the one shown at 16:52. Again, this effect is not my main area of expertise, so if anyone suspects that soliton fission is present after all, I would be interested to see the explanation. www.rp-photonics.com/supercontinuum_generation.html en.wikipedia.org/wiki/Supercontinuum#Soliton_fission_regime Chapter 12 & 13 of "Nonlinear Fiber Optics" by Govind P. Agrawal (www.elsevier.com/books/nonlinear-fiber-optics/agrawal/978-0-12-817042-7) opg.optica.org/oe/fulltext.cfm?uri=oe-14-21-9854&id=116387 opg.optica.org/directpdfaccess/b9dfcc88-eb69-4928-991b84b0a0c94b38_81189/oe-12-19-4614.pdf?da=1&id=81189&seq=0&mobile=no
Excellent thanks! This is really clearing things up! The original paper describes CSRS and SPM as the primary mechanisms for the Supercontinuum, however, when I first replicated with work with 9um fiber, I saw a very smooth continuum and concluded what I was seeing in 50um fiber was a different process, hence my best guess of Soliton fission. Given the regularity of the peaks, and how well what you have described fits with what I see, my guess is clearly erroneous. Now I see all the pieces it looks like SPM must be a more prominent process in 9um fiber and smooths out those peaks so they are no longer visible, which would explain the stark differences. See: ua-cam.com/video/w1wSHizmbYg/v-deo.html if you haven't already :-) Thanks for your awesome contributions to my understanding of what is going on here, it is much appreciated!
@@LesLaboratory Hello Les I have posted a video explaining Stimulated Raman Scattering on my own channel. Please feel free to check it out when you have the time. Best, YFTA
@@yourfavouriteta Thanks! I have and it's great! For anyone else following these threads, check this video out here (Highly recommended) ua-cam.com/video/sZEGMDZKyPQ/v-deo.html
This is something that I would like to have a go at building. The only thing I don`t have to hand is the lens (to focus uv onto the dye) but you said in another one of your videos that you could use a lens from a laser printer, bit low on transmission but usable. When I get some time away from work, I will be giving this a try. This is one of better channels on UA-cam, you are vastly under subscribed!
Thanks! Yes, in Laser printer assemblies, there is normally a Fast-Axis collimator lens in front of the Laser diode (It's square in section) It robs a bit of power from N2 Lasers, perhaps as much as 25% but it's not like these things are short on peak power.
@@LesLaboratory I love your work, keep it up! Can you get less separation in the peaks? I would love to couple this to my monochromator and have a really intense tunable light source. I considered a doubled Ti:Sapphire laser but the cost astronomical. Also, that spectrometer you used for your rpi spec, it might be capable of hyperspectral imaging, the pushbroom kind. I was talking to the engineers that made the larger version and from what I have learned, the optics are suitable. Just need to put an imaging lens at the slit and scan it horizontally as you make the datacube.
hey les, I have a really interesting thing to tell you about. Ive had this pen for a while in its current state and only recently thought to myself that it could be a diode that hit super continuum. Whats going on was I was starting a 488 diode in its first light, underdriven. I did this to attempt to get 480nm (I think) out of it which has actually worked 2/2 tries including this one until the end. Anyways, It was so underdriven that by the time I got to its threshold driving current it stayed at 1mw. My technique was to press the button for a second and drive it up slightly and press again. Once it turned it too high still waiting for it to light up bright it suddenly turned green. I thought oh no and turned it back down and the green went away. Keep in mind each time I see light from it I press the button for a second. Sometimes a few times when the wavelength dropped low. When I put a lens on the laser as it was staying at 1mw now forever, the dot had bright green shading on the outside of the dot. Its easily 525nm from the naked eye. The dot in the middle looks 490nm. At only around 1 foot away its possible to see 460nm or something.... Im not sure it could be 450 honestly, its dark blue but only a small part of it, noticably. I took the laser outside one night to play with it and I stood almost directly in its path about 50 feet away and in my eye I could see orange to yellow color as I let the beam very close to my eye. I never saw any red at all and orange is barely there, I swear, I saw orange. Mostly yellow which was noticeable. I can only see it when I point it near my eye from 50ft away. Not on the dot. Im actually wondering if you want to check this laser out for yourself. In any capacity, like some pictures I can send you, I have none of the yellow color yet. If you want to, Id mail it to you in agreement of sending it back to me when youre done with it, if you wanna play around and/or analyze it. I would love to hear about what you think no matter what you get out of it, as a curiosity and your opinion. Or best outcome, make a video out of it :^D The laser dot has a strange footprint now too. I wanna see what your spectrometer has to say too. I just thought that I live in america idk where your stationed.
Interesting video.. a few comments as I'm watching: -1GW/cm^2 pretty nice. Almost enough to drive optical parametric conversion - Beam size from your fiber: the fiber tester is extreme Multimode and the fiber piece you're using is very short, so even the biggest modes that experience high loses in the multimode fiber make it through that short bit. In the long fiber however those high orders have enough time to die out. If you can, try the tester on the long fiber and compare. Your Dye laser beam is also probably of much lower order than the fiber tester to begin with. You can also always check if reducing peak power with attenuation changes the beam size coming out of the fiber if you suspect non linear processes at work. The comb like structure is interesting. Did you try to influence it? (Push against the coupling mirror, stress the fiber...) To see if it changes in periodicity? 20% efficiency is pretty nice as well. Do you still get that with the 100m fiber? If you feel like experimenting you could take a few meters of the multimode fiber, get the good coupling efficiency and then with a torch heat a small section in the center and stretch it. It'll reduce the fiber diameter in a relatively controlled fashion. You might be able to push your efficiency and continuum bandwidth without the need for very long pieces of fiber.
YT channel: Your Favourite TA has suggested a robust experiment to substitute my fiber tester for something more suitable for a more scientific comparison, to try and narrow this one down. The comb it turns out is Stimulated Raman Scattering, and is dependent on the properties of the silica fiber. In previous experiments (9um fiber) this was not visible because self phase matching smoothed out the peaks. The coupling efficiency could be better, but yeah, 20% isn't terrible. I have not made further measurements with 100m yet, this whole thing is a live experiment in progress. That could be really interesting, like creating a tapered fiber, cool! I will look into this and see if it achievable practically.
I will do. In the mean time, you will find that, and some other very interesting things on this channel: Your Favourite TA: ua-cam.com/channels/QA1HNbn6PloeM2_YZXN1uQ.html
Spectacular work! very inspiring! would it work to shoot the nitrogen laser directly down the fibre? the collimation many not be great but the efficiency may be higher...
Thanks! There are a couple of problems that might stand in the way: You would have to find a way of coupling it into this small fiber. The fiber may be strongly absorbing in the UV. That said, it is on my mind to at least try it... 🙂
Very awesome !!! You said Pink Floyd album cover and I looked up at my gold disc album picture and it's about the closest thing I've seen yet ! Have around five or six he/ne lasers, mostly Phillips laserdisc player units and a nice Chinese disco laser with two open tubes but now I really want a nitrogen laser and various dyes please 😎
Thanks! Yeah, it really is beautiful. Good old He-Ne Lasers is where I started, lovely beam quality. N2 laser are about the easiest and least expensive of all the Lasers to build, and well worth the journey.
@@LesLaboratory wow, thanks for the reply les ! I've been all over the country collecting these old philips laserdisc players (4 so far) to make a working unit as all the optics are gummed up !!! As you say, the beam is pin point from those, one tube has even been modified with an external mirror at some point ? Are built N2 units easy to come by and the cuvettes and dyes as I really love the UV stuff !
@@johnwalker194 I have an external mirror He-Ne, but sadly the gas is contaminated. N2 Lasers (or parts) show up on eBay from time to time, but are a bit of a rarity. I built most of the ones I have. Cuvettes are pretty easy to come by on Aliexpress and eBay but they can be quite pricey.
Great observation. As per my other comment on this video, the supercontinuum arises due to cascaded Stimulated Raman Scattering. Basically, an optical frequency 13THz below the carrier frequency will steal power from the carrier. When this "thief" frequency gets sufficiently strong, another frequency 13THz below it will begin to steal its power and so on. Therefore, the lowest frequency order (highest wavelength going from blue through green to red) will be extremely sensitive to variations in the initial optical power of the incident pulse. Thus, I suspect that the strong variation in the power of the green part of the spectrum is a "magnified" version of much more minute power fluctuations in the incident pulse from the dye laser.
If this were down in the radio spectrum I would suggest harmonic resonance for the frequency comb. Not sure what the equivalent mechanism is for the optical spectrum.
Could it be some sort of etaloning effect happening in the fiber? I know that usually happens in a CCD detector, but it might be worth investigation whether etaloning can occur in long (+50-100 m) fibers also? Cool video anyways - thanks!
It's definitely Stimulated Raman Scattering. A spacing of 12nm at 500nm corresponds to 13THz, which is exactly the location of the Raman gain peak i silica. It's a very interesting and useful effect! Check out my other comment for more details :)
That, is something I want an answer too as well, but I want to see the difference in pulse shape across the spectrum as well. I have some parts on the way from Aliexpress to build a small monochromator to measure it :-)
Multimode fibers can demonstrate some complex behaviour. Of course I understand you wouldn't show all the testing you did in the final edited video, but If I saw a difference in mode coupling like that, the first thing I would do is play with the incoupling alignment to see if I could get the same affect for both lasers with the same length. The essential property of a multi-mode fiber is in its name, multiple modes are stable. however, higher-order modes have higher propagatino loss, which means they are attenuated more strongly, the longer the fiber. Do you see the same behavior if you connect the fault locator on the longer MMF?
An interesting suggestion that I will look into. It might also be a good idea to set several wavelengths as well, I just did not expect to see such a tight output beam I will give it a go.
There are a couple of things responsible here: The red output is significantly lower than the blue, so the color balance of the 'white' beam is off. Blue light scatters forward much more then red. Try it with a 405nm Laser pointer and a 650nm pointer of similar powers, and you will see the 405 is much more visible. Camera color rendering is shocking, even in decent ones :-S
Would it be possible to use this to create a full color hologram? The thought emporium used three colored lasers and a combining cube to get a "white" laser for creating color holograms and it worked okay, but this seems like it could capture the actual entire spectrum if it could be used.
I suspect not.The problem is that Supercontinuum lasers, they have good spatial coherence (i.e. you can make a very tight beam) are not temporally coherent (huge spread of frequencies) and so you are unlikely to be able to produce the required interference pattern on the film. Litiholo plates are excellent for Continuous wave lasers, but I have yet to get it to react sensibly to short pulse, microjoule lasers like these. All that said, a faded memory of mine reckons that probably buried deep in a journal somewhere, somebody somewhere will have generated something that works with white light...
I suspect so, but my scope was not fast enough to detect it. The fiber itself is long enough to broaded pulses, even without a nonlinear component. I'm going to resurrect this project at some point, as I have new kit to play with...
Off topic question. Do red laser pointers like the ones you see used as cat toys emit only in the 630-670nm range or can the emit in the UV-blue as well. I know some of them even emit in Infrared. I’m having a issue with a photographic emulsion that I want to be IR sensitive. I know it is UV-Blue 300-520nm sensitive. After sensitizing With some dyes I found when I used a red laser on the emulsion with a 720nm filter. The emulsion reacted to the light. So can red laser pointers emit uv light or did I succeed in the Near infrared/red light spectrum?
Red Laser diodes emit no light in the UV or blue portion of the spectrum. Light sensitivity of photographic emulsion does not have the expected sharp cut-off at each end of its quoted range, but instead will tail off towards the UV and IR at each end. The manufacturers of the emulsion may not show this on their graphs or datasheets, since they are quoting a usable range for regular circumstances. A Laser beam, with a power of just a few milliwatts has extraordinarily high brightness when compared with other light sources, and can easily expose silver halide emulsions, regardess of their quoted sensitivities.
@@LesLaboratory I made the emulsion. I’m testing for sensitivity. So from what you say here that means i in fact made an emulsion that is near infrared sensitive. I filtered the laser through a 720nm and 760nm filter. Only the 720nm showed results. Thank you for the help. This is great news. I need to test more dyes to find some 800nm+ fluorescing dyes from cheap everyday things. So far if I understand right I found my 720 dye.
@@morningstarsci Not currently. The lowest wavelength is ~420nm (the pump wavelength) and the longest is determined by the length of the fiber and the amount of energy you can couple into it. For multi octave spanning super-continuum from UV to IR you would require Photonic Crystal Fiber, and a suitable laser to pump it. This is very costly, with the fiber alone costing hundreds of dollars per metre! There 'may' be other methods and it is something I will be looking in to.
It'd be an interesting project to make an interface to sync the N2 laser to the camera's shutter. I wonder if the HDMI outputs of modern cameras are synced to the shutter? Back with analog composite it would've been pretty easy to extract the sync pulse, phase shift if necessary and trigger the laser at 25 or 30Hz
Hmmm, that is an interesting proposition. I think you would need an FPGA to extract the equivalent of vsync. I would bet it can be done with a Pi-Camera....
Hi Les, I have sent you an email that will probably end up in your `junk` folder as it contains an Ebay link for something I found while looking for supplies. You might find it useful. To help you find my mail, the word quatermass is part of the sender ID.
Since manufacturers abroad stopped caring about color coding. It threw me as well, it is 50/125. These days manufacturers are letting people pick jacket colors as well, which will just cause confusion.
@@LesLaboratory great :/. i guess.. for short runs, it wouldnt be the end of the world if one accidentally grabbed a yellow fiber expecting it to be a sm, as ive (in a pinch), used mm with a sm signal. but that def wouldnt work the other way around.. id be like trying to shove a grapefruit through a drinking straw...
I have not tried it, but from what I have read and experienced, I would say the probability is high that you would see some SRS from grin fiber. I will see if I can get hold of some at reasonable price.
also.. maybe try to reach out to @corningcablesystems with some of your questions about whats going on with some of your experiments... if anyone knows, the optical engineers a corning should =]
Cool, thanks for the tip! @My Favourite TA has cleared up the comb issue and suggested and experiment to test why the mode is so clean (see comments below). This comments section is awesome, there is a real sense of community in here!
PhD student in matsci here…these videos are a joy to watch! thank you so much for all the effort in development and clear presentation! subscribed (and shared with lab mates:))
What a coincidence! I made the exact same comment about the quality of the experimentation, it’s superb, I know of many supervising tutors who would not measure up to this level!
Excellent work building something like this. Not an easy task at all. The things required to make supercontinuum emission are high peak power, a very wide transverse mode spacing of several nanometers if possible, and a lot of fiber to allow mode mixing and stimulated raman scattering. Essentially the power from the light creates a strong electric field undulator that the longer wavelength mix with to create parametric mixing. The energy is so high it works a lot like a free electron laser. Why you are ending up with discrete bands of wavelengts is because the fiber is only allowing gain at those frequencies, but each color bunch should have the same mode spacing as your dye laser excitation beam in absolute value ie its 4nm across. The math describing it will make your brain bleed though. 😅😲🤓❤
Can you reduce the appearance of speckle by mechanically vibrating the coil of fibre? If you watch a recent Mike's Electric Stuff video you'll notice he does a teardown on a biology lab instrument using what I assume is a solid state green laser and it's run through a coil that's mounted to a motor with an eccentric weight. The resulting motion creates a rapidly varying set of path lengths through the fibre and reduces speckle especially on longer exposure readings.
@@LesLaboratory Oh yes, the individual fat globules, I expect? The reason I ask is that I work in film and TV where we are very interested in continuous-spectrum light sources for the purposes of good colour rendering. LEDs are the current go-to but the spectrum of white LEDs, while usable, is often significantly less than ideal. There are other issues around shutter timing which really mean it ought to be continuous wave, or at least very high frequency. Useful power levels are on the order of hundreds of watts and the entire purpose is to beat the efficiency of tungsten halogen sources, which might remove supercontinuum lasers from consideration. Still, multiple ganged emitters would also reduce speckle!
I was not surprised to see the cleanliness of the beam exiting the fibre. When you couple a laser beam into the fibre it "just' travels through the fibre and gets scattered. In this case the fibre become the 'resonator' of sorts, so the actual lazing happens in the fibre for what you get at the output, hence the decoupled beam is the result of the laser cavity, eg. fibre. With destructive and constructive interference the beam 'cleans' it self up, or just lazes what is permitted or supported by the cavity.
This is absolutely amazing! I wonder if the same effect could be made with a shorter length of fiber but with the output coupled back to it's input? or perhaps just reflected backwards and using the fiber as cavity of sorts?
@@great__success I agree that it must be Stimulated Raman Scattering. Figure 3 in the paper you mention matches the spectrum in the video exactly! Also, the spacing between the lines of around 12nm corresponds to ca. 13THz when operating near 500nm. This is exactly the frequency where the Raman gain spectrum of silica peaks as per my previous comment.
@@yourfavouriteta I fully agree! A couple of years ago my eyes almost fell out when I entcountered this the first time. I coupled a laser with 532nm, 10kHz rep rate, ~1ns pulsewidth, ~5kW peak power into a HP780 fiber (~20m length) to send it across the canary islands. When I checked the transmitter because the guys on the other island did not see anything on their tracking cam (they had a filter which was transmitting only 532nm) I found out that I was sending actually orange light instead of green 😂. It took a while until i figured out what was going on... Cheers Johnny PS.: Congratz to your cool channel Les! I am always smiling when a new video comes out! I wish you much more subscribers! You are doing very cool stuff!
@Your Favourite TA, I was hoping you would comment! So the comb then is just a product of SRS, awesome! Any thoughts as to why the output from a raw fiber has such low divergence in this case?
Les, try to set the dye laser up as a ring dye laser optical setup, and if you have an AO, modulator set it up in the cavity as well with a variable frequency controller to the AO, and I have an Idea of using two piezoelectric disk with small first surface mirrors glued onto them , in the cavity as a beam stabilizer, inside the ring configuration, as well put a spacial beam filter in the cavity of the ring lasr then sync the pizo, frequency with the AO, then pump it into the Fiber,. Im going to try this after tge first of 2024, I have all the gear to do the experiment. However I think you might find thus very intriguing and interesting. Perhaps make the fiber bundle into a ring laser as well with a polarizing plate and a 1/2 Wave plate cavity dump. Use pizo setup in its ring cavity as well. Brain storming I sm. I am Dale Robertson
Very nice work! A few points regarding the mode profile: 1. What's the mode profile of the Dye laser beam? How does it look like after propagating through the fiber? 2. What is the bend radius of the multimode fiber? As you may know, overly tight bend radii induce loss in higher order modes. Is it possible to change the bend radius? 3. How does the mode profile look after various distances of propagation? 4. Do you know how to extract the delay from spectral interferogramm? It would be nice to see the Fourier transform of those spectra, as the spectra do look like the output of a white light interferometer. 5. You may be able to do FROG or IFRAC with your pulses in an reasonably sized interferometer since solitions maybe compressed compared to the initial pulse.
Thanks! 1) It is poor, see: ua-cam.com/video/w1wSHizmbYg/v-deo.html at 09:00, once it has propagated through the fiber it appears Gaussian, but this has not been measured. 2) I have not tried it but I have considered experimenting with this. 3) Again, this is something I want to try, I was speaking to Sam (laserfaq) and am considering sacrificing a cable just to see. 4 & 5) might be problematic. I have considered FROG at least however on a budget it might be difficult to attain meaningful precision, but I like to try stuff ....
@@LesLaboratory 1) Interesting to see that the even the original beam is seemingly single mode after coupling to the fiber. Does that single mode-ness change with power of the beam? If you truly see the non-linear self cleaning, it should be power-dependend and the effect should be easy to see even for a small attenuation. It would be cool to see some interferometry. 5) There is spatial and spectral resolution to concern about as well as dynamic range of your spectrometer. The spectral resolution of your spectrometer should be easily measured by just looking at a few narrow line width sources like a HeNe. As for dynamic range of your spectrometer, i.e. how small of a pulse on a underground you can see, that should be in the spec sheet of your sensor. As for spatial/temporal resolution for setting the delay, you get 2*300nm/fs for a single bounce in an Michelson interferometer, so 10 um positioning resolution like from a stepper motor turned lead screw means 66 fs resolution. Enough to see some pulses, I'd say.
Hey Les and followers. I'm making good progress with my own SC system! I built and calibrated a Pi Spectrometer from your earlier video. It works amazingly well. 'I can't believe how accurate it is. I bought some coumarin-1 dye (mixed with acetone) and managed to make a dye laser and shoot it into 30 meters of 50/125 fiber (FC/UPC ends) from a used Model VSL-337ND-S Spectra Physics laser. From Thorlabs I bought an adjustable collimating lens for launching the dye laser into the fiber. I'm not sure this is ok: CFC2-A Adj. FC/PC Collimator, f = 2.0 mm, ARC: 350-700 nm. I'm a little concerned that the UPC ends can affect my coupling and also not sure if the 2mm f is ok. My question is this - I'm not really getting amazing SC yet. I get a huge peak at 446nm but not much spreading, and the light still looks blue. Any thing you might recommend?
Glad you are following the work! I have sent you e-mail . It could be a few things, maybe the solvent, maybe the Dye Laser setup, maybe the fiber, may be alignment. It is quite a touchy setup and requires persistence. A couple of other though have successfully replicated this, so it is in reach!
So from comments in this thread, it is confirmed that is is solely Stimulated Raman Scattering that is responsible for the comb. In single mode fiber, I have not observed a comb (yet), but it looks like other nonlinear processes are smoothing the comb out into a continuous spectrum in small diameter fiber.
What kind of safety precautions are necessary for working super-continuum lasers? I guess regular safety glasses are of little use. Or are the powers low enough that it doesn't matter?
@Curix, the peak powers are such that it will happily damage your eyes, and you are right regular glasses are no use. I have Glasses for the pump Lasers, (Nitrogen and Dye) but for the supercontinuum, you have to make good use of beam stops (no un-terminated reflections of beam) and cameras
Sure, I should probably do a short video on that. The process is very easy, you just need the right tools for the job, and they can be had inexpensively.
If you're interested, I have a video on my channel demonstrating how to cut, strip and splice fibers. Check it out here: ua-cam.com/video/Q1gcAD0gkw4/v-deo.html
Just found your channel. Very interesting stuff! It's always a treat listening to someone who really knows what he's doing. BUT the coolest thing is the absolutely perfect T-shirt for doing LASER experiments!
They are made from a stack of glass microscope slides. Soda glass will absorb a fair bit of UV, but slides are thin enough and flat enough to make reasonable attenuators inexpensively.
@@LesLaboratory i dont think you mentioned any of this or shown them ,what is the point of using them then removing them and why is the spectrum then more visible when you do this?
I cannot think of a better visualization of the dirac com from my signals/systems class. But this one gets created by some weird nonlinear effect, not by a dirac comb in time. Fantastic!
This has a shorter fiber, so it doesn't quite reach those wavelengths. The longer the fiber, the greater the span of the Supercontinuum (up until a point!)
What a great experimentalist, I love your work and the passion with which you deliver your results deserves being a part of every Physics & Physical Chemistry undergraduate’s learning material. Some is well into postgrad PhD quality. Keep up the good work. I wish I had you as a neighbour! 👏👏👏👏👏👏
I was waiting for this for whole week :) great work ! I can't think of a way to measure but probably there are micro pulses within your ns long pulses. Probably distance between these micro pulses are at the order of femtoseconds.
It's something I intend to investigate. The original paper used a monochromator a PMT and an oscilloscope to scan the spectrum, and measure the pulse width at each wavelength, and it looks like there is something to see there. I have just acquired a scope (the one int he video), I won't get down to femtoseconds with it, but after the hack, it will be a 1GHz scope, so subnanosecond resolution.
@@LesLaboratory That would be another epic content. I used PMT's in the past but it was configured to measure few photons . The ones we used was old, nevertheless had few ns rise time. I never used them in the presence of considerable amount light and I am curious. :))
Hi, your videos are great, but you should work on volume levels - your speech volume level is way to quiet compared to other videos, and intro at 0:11 is way louder than your speech, you need a gain of about 20dB.
My guess is that the modes, or "striations" you see after the reflecting diffraction grading are a result of multimoding occuring within the fiber. I believe that these discrete modes are occuring because the MM fiber allows for numerous, discrete TEM modes to propagate, whereas the single-mode fiber you used previously, did not. This is why we see the modes occur after the grating, because all of the power coupled into the fiber, exists propagating within these discrete modes in the fiber. I'd be curious to see if applying stress to the fiber causes these modes to shift, widen, vary in amplitude, transfer power between modes, etc...
Interesting suggestion, but the consistent spacing of ca. 12nm (corresponding to ca. 13THz when operating in the visible range) matches the spacing expected from Stimulated Raman Scattering as per my other comment. This is a well-known effect in optical fibers and is even exploited for amplifying the optical telecom signals that were most likely used to transmit this video to your device! It's a very cool effect :)
Can something like this be done for regular bright light? I am in need of such a bright continuous spectrum light source such as this fir mineral identifications as early experiments have demonstrated fluorescence at specific frequency combinations. Something which cod easy be achieved with this setup and proper filters however this appears anything but portable. Picture a flashlight which causes specific minerals to stand out in the dark similar to how UV reactive ones appear under blacklight. That is what i am pretty sure can be created but proper bright configurable is anything but simple.
For a very good explanation of Stimulated Raman Scattering (the primary nonlinear process in this video) please check out: Your Favourite TA's video on Stimulated Raman Scattering here: ua-cam.com/video/sZEGMDZKyPQ/v-deo.html
Excellent work. It's a shame these videos have so few views, mainly because most people haven't got the slightest idea of what's going on or why the underlying phenomenon is so truly weird and extremely unusual to actually observe.
Thanks! Yeah, I'm stuck with that one, should I do a simpler series of videos for the masses do you think?
@Les' Lab no, keep them as is. Some of the unfamiliar terms in your videos prompt research into them on my part, such as soliton fission. I honestly have no idea what that means but I am headed to the internet to find out.
Keep your style, Les 🙌 There's already enough BS on YTube. Yours is Gold !
Yeah, it's more about people deserving to watch this than about the creator trying to get lots of views 🙂
Depends on where you want the channel to go I guess. To be honest I think Styropyro pretty much has the fire and explosions and burning random shit with lasers for 14 year old viewership angle covered on youtube ("not that there's anything wrong with that"), but that's what the masses like if you really want the megaviews! I'm content with the deep, real science and moderate views, for my part.
The peaks at 16:52 arise due to a process called Stimulated Raman Scattering (SRS). To understand this mechanism, consider a simple, isolated molecule like H2. The molecular bond can vibrate at a set of distinct frequencies due to quantum mechanics. If we were to shine a laser onto the molecule, a photon can "donate" some energy to the molecule by exciting vibrations. Since the photon has lost energy, its frequency has decreased by an amount corresponding to the vibrational frequency of the bond. Since the frequency has decreased, the wavelength must have increased.
For an amorphous material like silica, the molecular bonds are extremely irregular compared to a simple H2 molecule. Therefore, the possible vibrational frequencies form a continuous "band" with a peak at ca. 13.2 THz and a width of around 5 THz as shown in Figure 2 in this link:
www.rp-photonics.com/raman_gain.html
Therefore, when shining a highly intense, coherent laser pulse with a certain carrier frequency into a silica fiber, coherent molecular vibrations will be excited. This transfers power from the carrier frequency to one that is approximately 13.2 THz lower; this is SRS! Furthermore, this lower frequency will keep "stealing" power from the carrier until it gets so intense that the process can repeat for a frequency that is an additional 13.2THz below that one (and so on in a process called "cascaded SRS"). This explains the observed "comb" structure.
Considering the peaks at 16:52 labelled 497.7nm and 519.4nm (with an unlabeled one in between), we can see that the peak spacing is 10.85nm. Converting this difference in wavelength (BW_lambda) at 497.7nm (lambda_0) to a difference in optical frequency (BW_f) using
BW_f = c*BW_lambda/lambda_0^2,
we get 13.13 THz, which is very close to the peak frequency of the Raman gain spectrum presented in the link above.
If you are interested in some of the theory behind nonlinear phenomena in optical fibers, feel free to check out my channel, where I both demonstrate these effects experimentally and simulate them numerically using custom python code:
Experimental Self-Phase modulation : ua-cam.com/video/wZyaTVTmmBk/v-deo.html
Experimental supercontinuum: ua-cam.com/video/ZsZXqxwJBcw/v-deo.html
Numerical simulation: ua-cam.com/video/xIdozUy9Nas/v-deo.html
I plan on upgrading my own code to handle Raman effects in the future, but I know that others have created packages that allow for simulating SRS in optical fibers:
github.com/WUST-FOG/gnlse-python
This was a really awesome demonstration, great job :)
In addition, though it's not an effect I have worked with myself, I don't think soliton fission is taking place in your experiment. The sources I have been able to find (see links below) state that soliton fission takes place for pulses on the scale of tens to hundreds of femtoseconds, which is much shorter than the 2ns pulses used in the present experiment.
The basic reason seems to be that a pulse with a duration of around 100fs has a spectral width, which is close to the 13.2 THz peak of the Raman gain spectrum (see previous comment). Therefore, the spectral tails of the pulse itself end up stealing power from the carrier frequency, which is different from the SRS case, where the lower frequency photons are not initially present but "created" from a mechanical vibration excited by the pulse. Combined with other nonlinear effects (Self-Phase-Modulation, Cross-Phase-Modulation and Four-Wave-Mixing) as well as dispersion (different frequencies propagating at different speeds leading to broadening in the time domain), the spectral output is usually much more "messy" than the one shown at 16:52.
Again, this effect is not my main area of expertise, so if anyone suspects that soliton fission is present after all, I would be interested to see the explanation.
www.rp-photonics.com/supercontinuum_generation.html
en.wikipedia.org/wiki/Supercontinuum#Soliton_fission_regime
Chapter 12 & 13 of "Nonlinear Fiber Optics" by Govind P. Agrawal (www.elsevier.com/books/nonlinear-fiber-optics/agrawal/978-0-12-817042-7)
opg.optica.org/oe/fulltext.cfm?uri=oe-14-21-9854&id=116387
opg.optica.org/directpdfaccess/b9dfcc88-eb69-4928-991b84b0a0c94b38_81189/oe-12-19-4614.pdf?da=1&id=81189&seq=0&mobile=no
Excellent thanks! This is really clearing things up!
The original paper describes CSRS and SPM as the primary mechanisms for the Supercontinuum, however, when I first replicated with work with 9um fiber, I saw a very smooth continuum and concluded what I was seeing in 50um fiber was a different process, hence my best guess of Soliton fission.
Given the regularity of the peaks, and how well what you have described fits with what I see, my guess is clearly erroneous.
Now I see all the pieces it looks like SPM must be a more prominent process in 9um fiber and smooths out those peaks so they are no longer visible, which would explain the stark differences. See: ua-cam.com/video/w1wSHizmbYg/v-deo.html if you haven't already :-)
Thanks for your awesome contributions to my understanding of what is going on here, it is much appreciated!
@@LesLaboratory Hello Les
I have posted a video explaining Stimulated Raman Scattering on my own channel. Please feel free to check it out when you have the time.
Best,
YFTA
@@yourfavouriteta Thanks! I have and it's great! For anyone else following these threads, check this video out here (Highly recommended) ua-cam.com/video/sZEGMDZKyPQ/v-deo.html
@@LesLaboratory Awesome! Don't forget to subscribe ;)
I love everything about it ❤
Major props. Your videos always brighten my day!
Thank you for making these videos. ❤
Thanks! I am glad they make your day! ☺️
This is something that I would like to have a go at building. The only thing I don`t have to hand is the lens (to focus uv onto the dye) but you said in another one of your videos that you could use a lens from a laser printer, bit low on transmission but usable. When I get some time away from work, I will be giving this a try. This is one of better channels on UA-cam, you are vastly under subscribed!
Thanks! Yes, in Laser printer assemblies, there is normally a Fast-Axis collimator lens in front of the Laser diode (It's square in section) It robs a bit of power from N2 Lasers, perhaps as much as 25% but it's not like these things are short on peak power.
Loving the real genius reference. Now remember-I want 5MW by mid may.
Ha! I hoped somebody would would see it!
@@LesLaboratory I love your work, keep it up! Can you get less separation in the peaks? I would love to couple this to my monochromator and have a really intense tunable light source. I considered a doubled Ti:Sapphire laser but the cost astronomical.
Also, that spectrometer you used for your rpi spec, it might be capable of hyperspectral imaging, the pushbroom kind. I was talking to the engineers that made the larger version and from what I have learned, the optics are suitable. Just need to put an imaging lens at the slit and scan it horizontally as you make the datacube.
hey les, I have a really interesting thing to tell you about. Ive had this pen for a while in its current state and only recently thought to myself that it could be a diode that hit super continuum. Whats going on was I was starting a 488 diode in its first light, underdriven. I did this to attempt to get 480nm (I think) out of it which has actually worked 2/2 tries including this one until the end. Anyways, It was so underdriven that by the time I got to its threshold driving current it stayed at 1mw. My technique was to press the button for a second and drive it up slightly and press again. Once it turned it too high still waiting for it to light up bright it suddenly turned green. I thought oh no and turned it back down and the green went away. Keep in mind each time I see light from it I press the button for a second. Sometimes a few times when the wavelength dropped low. When I put a lens on the laser as it was staying at 1mw now forever, the dot had bright green shading on the outside of the dot. Its easily 525nm from the naked eye. The dot in the middle looks 490nm. At only around 1 foot away its possible to see 460nm or something.... Im not sure it could be 450 honestly, its dark blue but only a small part of it, noticably.
I took the laser outside one night to play with it and I stood almost directly in its path about 50 feet away and in my eye I could see orange to yellow color as I let the beam very close to my eye. I never saw any red at all and orange is barely there, I swear, I saw orange. Mostly yellow which was noticeable. I can only see it when I point it near my eye from 50ft away. Not on the dot. Im actually wondering if you want to check this laser out for yourself. In any capacity, like some pictures I can send you, I have none of the yellow color yet. If you want to, Id mail it to you in agreement of sending it back to me when youre done with it, if you wanna play around and/or analyze it. I would love to hear about what you think no matter what you get out of it, as a curiosity and your opinion. Or best outcome, make a video out of it :^D
The laser dot has a strange footprint now too. I wanna see what your spectrometer has to say too.
I just thought that I live in america idk where your stationed.
Interesting video.. a few comments as I'm watching:
-1GW/cm^2 pretty nice. Almost enough to drive optical parametric conversion
- Beam size from your fiber: the fiber tester is extreme Multimode and the fiber piece you're using is very short, so even the biggest modes that experience high loses in the multimode fiber make it through that short bit. In the long fiber however those high orders have enough time to die out. If you can, try the tester on the long fiber and compare. Your Dye laser beam is also probably of much lower order than the fiber tester to begin with. You can also always check if reducing peak power with attenuation changes the beam size coming out of the fiber if you suspect non linear processes at work.
The comb like structure is interesting. Did you try to influence it? (Push against the coupling mirror, stress the fiber...) To see if it changes in periodicity?
20% efficiency is pretty nice as well. Do you still get that with the 100m fiber?
If you feel like experimenting you could take a few meters of the multimode fiber, get the good coupling efficiency and then with a torch heat a small section in the center and stretch it. It'll reduce the fiber diameter in a relatively controlled fashion. You might be able to push your efficiency and continuum bandwidth without the need for very long pieces of fiber.
YT channel: Your Favourite TA has suggested a robust experiment to substitute my fiber tester for something more suitable for a more scientific comparison, to try and narrow this one down.
The comb it turns out is Stimulated Raman Scattering, and is dependent on the properties of the silica fiber. In previous experiments (9um fiber) this was not visible because self phase matching smoothed out the peaks.
The coupling efficiency could be better, but yeah, 20% isn't terrible. I have not made further measurements with 100m yet, this whole thing is a live experiment in progress.
That could be really interesting, like creating a tapered fiber, cool! I will look into this and see if it achievable practically.
Great content!
A follow up on fiber cleaver is due I guess? Amazing work..
I will do. In the mean time, you will find that, and some other very interesting things on this channel:
Your Favourite TA:
ua-cam.com/channels/QA1HNbn6PloeM2_YZXN1uQ.html
Blown away.
This is worth gold
Thanks!
fantastic work
Thanks! :-)
Spectacular work! very inspiring! would it work to shoot the nitrogen laser directly down the fibre? the collimation many not be great but the efficiency may be higher...
Thanks! There are a couple of problems that might stand in the way:
You would have to find a way of coupling it into this small fiber.
The fiber may be strongly absorbing in the UV.
That said, it is on my mind to at least try it... 🙂
@@LesLaboratory if one person can do it, I am sure it's you! I would love to know the result either way!
😮 beautiful
Thanks!
Very awesome !!! You said Pink Floyd album cover and I looked up at my gold disc album picture and it's about the closest thing I've seen yet ! Have around five or six he/ne lasers, mostly Phillips laserdisc player units and a nice Chinese disco laser with two open tubes but now I really want a nitrogen laser and various dyes please 😎
Thanks! Yeah, it really is beautiful. Good old He-Ne Lasers is where I started, lovely beam quality. N2 laser are about the easiest and least expensive of all the Lasers to build, and well worth the journey.
@@LesLaboratory wow, thanks for the reply les ! I've been all over the country collecting these old philips laserdisc players (4 so far) to make a working unit as all the optics are gummed up !!! As you say, the beam is pin point from those, one tube has even been modified with an external mirror at some point ? Are built N2 units easy to come by and the cuvettes and dyes as I really love the UV stuff !
@@johnwalker194 I have an external mirror He-Ne, but sadly the gas is contaminated. N2 Lasers (or parts) show up on eBay from time to time, but are a bit of a rarity. I built most of the ones I have. Cuvettes are pretty easy to come by on Aliexpress and eBay but they can be quite pricey.
I wonder why the green flickers so much.
It would be interesting to see if it's sensitive to fiber temperature or magnetic fields or something.
Great observation.
As per my other comment on this video, the supercontinuum arises due to cascaded Stimulated Raman Scattering. Basically, an optical frequency 13THz below the carrier frequency will steal power from the carrier. When this "thief" frequency gets sufficiently strong, another frequency 13THz below it will begin to steal its power and so on. Therefore, the lowest frequency order (highest wavelength going from blue through green to red) will be extremely sensitive to variations in the initial optical power of the incident pulse.
Thus, I suspect that the strong variation in the power of the green part of the spectrum is a "magnified" version of much more minute power fluctuations in the incident pulse from the dye laser.
If this were down in the radio spectrum I would suggest harmonic resonance for the frequency comb. Not sure what the equivalent mechanism is for the optical spectrum.
Wow!! Beautiful!!!
Thanks! :-)
Could it be some sort of etaloning effect happening in the fiber? I know that usually happens in a CCD detector, but it might be worth investigation whether etaloning can occur in long (+50-100 m) fibers also? Cool video anyways - thanks!
It's definitely Stimulated Raman Scattering. A spacing of 12nm at 500nm corresponds to 13THz, which is exactly the location of the Raman gain peak i silica. It's a very interesting and useful effect! Check out my other comment for more details :)
@@yourfavouriteta this sounds right!
@@yourfavouriteta wait, does that mean one could generate a 13thz signal microwave signal using silicon?
What does the output of the fibre look like in the time domain? Do the pulses get spread temporally as well as spectrally?
ooh i'm very curious about this too
That, is something I want an answer too as well, but I want to see the difference in pulse shape across the spectrum as well. I have some parts on the way from Aliexpress to build a small monochromator to measure it :-)
I would assume that combing of the spectrum would vary with the length of the fibre.
Yes, the effect is depended on fiber length, the longer the fiber the more SRS you will see (up to a limit).
Subscribed!!
Awesome!
Multimode fibers can demonstrate some complex behaviour. Of course I understand you wouldn't show all the testing you did in the final edited video, but If I saw a difference in mode coupling like that, the first thing I would do is play with the incoupling alignment to see if I could get the same affect for both lasers with the same length.
The essential property of a multi-mode fiber is in its name, multiple modes are stable. however, higher-order modes have higher propagatino loss, which means they are attenuated more strongly, the longer the fiber.
Do you see the same behavior if you connect the fault locator on the longer MMF?
An interesting suggestion that I will look into. It might also be a good idea to set several wavelengths as well, I just did not expect to see such a tight output beam I will give it a go.
@17:30 the light appears more blueish even though it has red wavelength on it! 🤔
There are a couple of things responsible here:
The red output is significantly lower than the blue, so the color balance of the 'white' beam is off.
Blue light scatters forward much more then red. Try it with a 405nm Laser pointer and a 650nm pointer of similar powers, and you will see the 405 is much more visible.
Camera color rendering is shocking, even in decent ones :-S
Would it be possible to use this to create a full color hologram? The thought emporium used three colored lasers and a combining cube to get a "white" laser for creating color holograms and it worked okay, but this seems like it could capture the actual entire spectrum if it could be used.
I suspect not.The problem is that Supercontinuum lasers, they have good spatial coherence (i.e. you can make a very tight beam) are not temporally coherent (huge spread of frequencies) and so you are unlikely to be able to produce the required interference pattern on the film.
Litiholo plates are excellent for Continuous wave lasers, but I have yet to get it to react sensibly to short pulse, microjoule lasers like these.
All that said, a faded memory of mine reckons that probably buried deep in a journal somewhere, somebody somewhere will have generated something that works with white light...
does the pulse length change as part of the frequency broadening?
I suspect so, but my scope was not fast enough to detect it. The fiber itself is long enough to broaded pulses, even without a nonlinear component. I'm going to resurrect this project at some point, as I have new kit to play with...
Try buildings an color laser interleaved tv unit
Off topic question. Do red laser pointers like the ones you see used as cat toys emit only in the 630-670nm range or can the emit in the UV-blue as well. I know some of them even emit in Infrared. I’m having a issue with a photographic emulsion that I want to be IR sensitive. I know it is UV-Blue 300-520nm sensitive. After sensitizing
With some dyes I found when I used a red laser on the emulsion with a 720nm filter. The emulsion reacted to the light. So can red laser pointers emit uv light or did I succeed in the Near infrared/red light spectrum?
Red Laser diodes emit no light in the UV or blue portion of the spectrum.
Light sensitivity of photographic emulsion does not have the expected sharp cut-off at each end of its quoted range, but instead will tail off towards the UV and IR at each end. The manufacturers of the emulsion may not show this on their graphs or datasheets, since they are quoting a usable range for regular circumstances. A Laser beam, with a power of just a few milliwatts has extraordinarily high brightness when compared with other light sources, and can easily expose silver halide emulsions, regardess of their quoted sensitivities.
@@LesLaboratory I made the emulsion. I’m testing for sensitivity. So from what you say here that means i in fact made an emulsion that is near infrared sensitive. I filtered the laser through a 720nm and 760nm filter. Only the 720nm showed results. Thank you for the help. This is great news. I need to test more dyes to find some 800nm+ fluorescing dyes from cheap everyday things. So far if I understand right I found my 720 dye.
Where did the red color in some of these shots?
It flickers in and out. If the fiber were longer, the red end would be more stable.
@@LesLaboratory Does any produce any UV or NIR/IR wavelengths?
@@morningstarsci Not currently. The lowest wavelength is ~420nm (the pump wavelength) and the longest is determined by the length of the fiber and the amount of energy you can couple into it. For multi octave spanning super-continuum from UV to IR you would require Photonic Crystal Fiber, and a suitable laser to pump it. This is very costly, with the fiber alone costing hundreds of dollars per metre! There 'may' be other methods and it is something I will be looking in to.
is that the shirt worn by chris knight in real genius?
Yep! It was a moral imperative!
@@LesLaboratory lol indeed
where is styropyro
does changing the orientation of the coiled fibre change the output?
I have not tried this, but there are a number of things I want to investigate along these lines.
It'd be an interesting project to make an interface to sync the N2 laser to the camera's shutter. I wonder if the HDMI outputs of modern cameras are synced to the shutter? Back with analog composite it would've been pretty easy to extract the sync pulse, phase shift if necessary and trigger the laser at 25 or 30Hz
Hmmm, that is an interesting proposition. I think you would need an FPGA to extract the equivalent of vsync. I would bet it can be done with a Pi-Camera....
The hq pi camera has an external sync in, to force when a frame begins
@@cleverca22 Wow, that opens up a lot of cool options, I'll look into it myself, cheers!
Hi Les, I have sent you an email that will probably end up in your `junk` folder as it contains an Ebay link for something I found while looking for supplies. You might find it useful. To help you find my mail, the word quatermass is part of the sender ID.
since when does MM fiber come in yellow?
Since manufacturers abroad stopped caring about color coding. It threw me as well, it is 50/125. These days manufacturers are letting people pick jacket colors as well, which will just cause confusion.
@@LesLaboratory great :/. i guess.. for short runs, it wouldnt be the end of the world if one accidentally grabbed a yellow fiber expecting it to be a sm, as ive (in a pinch), used mm with a sm signal. but that def wouldnt work the other way around.. id be like trying to shove a grapefruit through a drinking straw...
Not sure if this is a stupid question, would grade Index Fiber do this trick?
I have not tried it, but from what I have read and experienced, I would say the probability is high that you would see some SRS from grin fiber. I will see if I can get hold of some at reasonable price.
also.. maybe try to reach out to @corningcablesystems with some of your questions about whats going on with some of your experiments... if anyone knows, the optical engineers a corning should =]
Cool, thanks for the tip! @My Favourite TA has cleared up the comb issue and suggested and experiment to test why the mode is so clean (see comments below). This comments section is awesome, there is a real sense of community in here!
PhD student in matsci here…these videos are a joy to watch! thank you so much for all the effort in development and clear presentation! subscribed (and shared with lab mates:))
Awesome! Glad you all like them!
What a coincidence! I made the exact same comment about the quality of the experimentation, it’s superb, I know of many supervising tutors who would not measure up to this level!
Fascinating stuff. BTW your audio level is a bit on the low side.
Thanks! Ah ok, it thought it was just me, I will pull up the levels next time.
Excellent work building something like this. Not an easy task at all. The things required to make supercontinuum emission are high peak power, a very wide transverse mode spacing of several nanometers if possible, and a lot of fiber to allow mode mixing and stimulated raman scattering. Essentially the power from the light creates a strong electric field undulator that the longer wavelength mix with to create parametric mixing. The energy is so high it works a lot like a free electron laser. Why you are ending up with discrete bands of wavelengts is because the fiber is only allowing gain at those frequencies, but each color bunch should have the same mode spacing as your dye laser excitation beam in absolute value ie its 4nm across. The math describing it will make your brain bleed though. 😅😲🤓❤
Thanks I'm sure it would! I have been reading up on the subject, and yeah, complex stuff. If only there was more time in my day!
It's like a smoke ring. You push a wave through the middle of another wave of course you're gonna get self-focusing
Absolutely stunning!
Thanks! :-)
Is it possible to pump the fiber with more than one wavelength at once to get more output in wavelengths above 570nm?
Can you reduce the appearance of speckle by mechanically vibrating the coil of fibre? If you watch a recent Mike's Electric Stuff video you'll notice he does a teardown on a biology lab instrument using what I assume is a solid state green laser and it's run through a coil that's mounted to a motor with an eccentric weight. The resulting motion creates a rapidly varying set of path lengths through the fibre and reduces speckle especially on longer exposure readings.
@@hfuy8005 Yep, it is possible to reduce speckle this way. If you want truly absent speckle liquids can be used as well. Milk works very well indeed!
@@LesLaboratory Oh yes, the individual fat globules, I expect? The reason I ask is that I work in film and TV where we are very interested in continuous-spectrum light sources for the purposes of good colour rendering. LEDs are the current go-to but the spectrum of white LEDs, while usable, is often significantly less than ideal. There are other issues around shutter timing which really mean it ought to be continuous wave, or at least very high frequency. Useful power levels are on the order of hundreds of watts and the entire purpose is to beat the efficiency of tungsten halogen sources, which might remove supercontinuum lasers from consideration. Still, multiple ganged emitters would also reduce speckle!
I was not surprised to see the cleanliness of the beam exiting the fibre. When you couple a laser beam into the fibre it "just' travels through the fibre and gets scattered. In this case the fibre become the 'resonator' of sorts, so the actual lazing happens in the fibre for what you get at the output, hence the decoupled beam is the result of the laser cavity, eg. fibre. With destructive and constructive interference the beam 'cleans' it self up, or just lazes what is permitted or supported by the cavity.
This is absolutely amazing!
I wonder if the same effect could be made with a shorter length of fiber but with the output coupled back to it's input? or perhaps just reflected backwards and using the fiber as cavity of sorts?
Please make the explanation video if you find out why it produces such peaks in the spectrum)
This paper: DOI 10.1364/OPEX.12.004366 states, that it is Raman scattering
@@great__success I agree that it must be Stimulated Raman Scattering. Figure 3 in the paper you mention matches the spectrum in the video exactly! Also, the spacing between the lines of around 12nm corresponds to ca. 13THz when operating near 500nm. This is exactly the frequency where the Raman gain spectrum of silica peaks as per my previous comment.
Thanks!
@@yourfavouriteta I fully agree! A couple of years ago my eyes almost fell out when I entcountered this the first time. I coupled a laser with 532nm, 10kHz rep rate, ~1ns pulsewidth, ~5kW peak power into a HP780 fiber (~20m length) to send it across the canary islands. When I checked the transmitter because the guys on the other island did not see anything on their tracking cam (they had a filter which was transmitting only 532nm) I found out that I was sending actually orange light instead of green 😂. It took a while until i figured out what was going on...
Cheers
Johnny
PS.: Congratz to your cool channel Les! I am always smiling when a new video comes out! I wish you much more subscribers! You are doing very cool stuff!
@Your Favourite TA, I was hoping you would comment! So the comb then is just a product of SRS, awesome! Any thoughts as to why the output from a raw fiber has such low divergence in this case?
Les, try to set the dye laser up as a ring dye laser optical setup, and if you have an AO, modulator set it up in the cavity as well with a variable frequency controller to the AO, and I have an Idea of using two piezoelectric disk with small first surface mirrors glued onto them , in the cavity as a beam stabilizer, inside the ring configuration, as well put a spacial beam filter in the cavity of the ring lasr then sync the pizo, frequency with the AO, then pump it into the Fiber,. Im going to try this after tge first of 2024, I have all the gear to do the experiment. However I think you might find thus very intriguing and interesting. Perhaps make the fiber bundle into a ring laser as well with a polarizing plate and a 1/2 Wave plate cavity dump. Use pizo setup in its ring cavity as well. Brain storming I sm. I am Dale Robertson
0:56 Yooooo! 😂
and @Les' Lab that shirt is so smart. Nice GaAs reference, among other things
Very nice work!
A few points regarding the mode profile:
1. What's the mode profile of the Dye laser beam? How does it look like after propagating through the fiber?
2. What is the bend radius of the multimode fiber? As you may know, overly tight bend radii induce loss in higher order modes. Is it possible to change the bend radius?
3. How does the mode profile look after various distances of propagation?
4. Do you know how to extract the delay from spectral interferogramm? It would be nice to see the Fourier transform of those spectra, as the spectra do look like the output of a white light interferometer.
5. You may be able to do FROG or IFRAC with your pulses in an reasonably sized interferometer since solitions maybe compressed compared to the initial pulse.
Thanks!
1) It is poor, see: ua-cam.com/video/w1wSHizmbYg/v-deo.html at 09:00, once it has propagated through the fiber it appears Gaussian, but this has not been measured.
2) I have not tried it but I have considered experimenting with this.
3) Again, this is something I want to try, I was speaking to Sam (laserfaq) and am considering sacrificing a cable just to see.
4 & 5) might be problematic. I have considered FROG at least however on a budget it might be difficult to attain meaningful precision, but I like to try stuff ....
@@LesLaboratory 1) Interesting to see that the even the original beam is seemingly single mode after coupling to the fiber. Does that single mode-ness change with power of the beam? If you truly see the non-linear self cleaning, it should be power-dependend and the effect should be easy to see even for a small attenuation.
It would be cool to see some interferometry.
5) There is spatial and spectral resolution to concern about as well as dynamic range of your spectrometer.
The spectral resolution of your spectrometer should be easily measured by just looking at a few narrow line width sources like a HeNe.
As for dynamic range of your spectrometer, i.e. how small of a pulse on a underground you can see, that should be in the spec sheet of your sensor.
As for spatial/temporal resolution for setting the delay, you get 2*300nm/fs for a single bounce in an Michelson interferometer, so 10 um positioning resolution like from a stepper motor turned lead screw means 66 fs resolution. Enough to see some pulses, I'd say.
Does it look white in real life or does it still look strongly blue as shown on the camera?
It looks white, but a cold white. The camera makes it look bluer. If the fiber was longer, more red would be produced, creating a warmer white.
Very pretty! Never get tired of beam shows, and you're producing one without a set of galvos! :)
Thanks! It sure is beautiful! I should probably get some galvos, just, because...
Hey Les and followers. I'm making good progress with my own SC system! I built and calibrated a Pi Spectrometer from your earlier video. It works amazingly well. 'I can't believe how accurate it is. I bought some coumarin-1 dye (mixed with acetone) and managed to make a dye laser and shoot it into 30 meters of 50/125 fiber (FC/UPC ends) from a used Model VSL-337ND-S Spectra Physics laser. From Thorlabs I bought an adjustable collimating lens for launching the dye laser into the fiber. I'm not sure this is ok: CFC2-A Adj. FC/PC Collimator, f = 2.0 mm, ARC: 350-700 nm. I'm a little concerned that the UPC ends can affect my coupling and also not sure if the 2mm f is ok.
My question is this - I'm not really getting amazing SC yet. I get a huge peak at 446nm but not much spreading, and the light still looks blue. Any thing you might recommend?
Glad you are following the work! I have sent you e-mail . It could be a few things, maybe the solvent, maybe the Dye Laser setup, maybe the fiber, may be alignment. It is quite a touchy setup and requires persistence. A couple of other though have successfully replicated this, so it is in reach!
Liked and subscribed
Was the comb effect present at all in the single mode fiber? I wonder if the comb spacing is related to fiber width.
So from comments in this thread, it is confirmed that is is solely Stimulated Raman Scattering that is responsible for the comb. In single mode fiber, I have not observed a comb (yet), but it looks like other nonlinear processes are smoothing the comb out into a continuous spectrum in small diameter fiber.
What kind of safety precautions are necessary for working super-continuum lasers? I guess regular safety glasses are of little use. Or are the powers low enough that it doesn't matter?
@Curix, the peak powers are such that it will happily damage your eyes, and you are right regular glasses are no use. I have Glasses for the pump Lasers, (Nitrogen and Dye) but for the supercontinuum, you have to make good use of beam stops (no un-terminated reflections of beam) and cameras
Fantastic ! Can you tell us more about how you stripped the finer cleanly?
Sure, I should probably do a short video on that. The process is very easy, you just need the right tools for the job, and they can be had inexpensively.
@@LesLaboratory That'd be ultra helpful. Last week I struggled with it for hours!
If you're interested, I have a video on my channel demonstrating how to cut, strip and splice fibers. Check it out here:
ua-cam.com/video/Q1gcAD0gkw4/v-deo.html
We want a video about laser gyroscopes , please !
It is on my very long list of stuff!
Just found your channel. Very interesting stuff! It's always a treat listening to someone who really knows what he's doing.
BUT the coolest thing is the absolutely perfect T-shirt for doing LASER experiments!
Thanks! Yeah, a little Easter egg for those that know 😉
@@LesLaboratoryI love it! Watching this movie is a moral imperative for every nerd out there! :)
0:00 _"I
Reference to the 80's film "Real Genius" :-)
Awesome results Les; would love to see this in person, any chance you could join a crowd of like-minded individuals in Lincolnshire in March?
I would love to, trouble, is work is always manic this time of year, couldn't make it last year. what is the date of the meet this year?
@@LesLaboratory It's 31st March to 3rd April.
what are these attenuators you are refering to?
They are made from a stack of glass microscope slides. Soda glass will absorb a fair bit of UV, but slides are thin enough and flat enough to make reasonable attenuators inexpensively.
@@LesLaboratory i dont think you mentioned any of this or shown them ,what is the point of using them then removing them and why is the spectrum then more visible when you do this?
Wow that dirac comb was dope! btw whats the pulse frequency and duration of individual pulses?
I have a question. What optics are used at the input and at the output. Lens size and focal length.🤔
10mm FL lenses, these ones specifically www.aliexpress.com/item/33057917525.html
Thanks for the answer. I like watching your videos. You are great.😉👍
I cannot think of a better visualization of the dirac com from my signals/systems class. But this one gets created by some weird nonlinear effect, not by a dirac comb in time. Fantastic!
i dont see the red and yellow orange full spectrum like i seen the first video you done?
so the longer the cable to more the spectrum? i figured the lower the power intensity too no?
This has a shorter fiber, so it doesn't quite reach those wavelengths. The longer the fiber, the greater the span of the Supercontinuum (up until a point!)
There will be undoubtedly losses with a longer fiber, but this is still orders of magnitude brighter than SM fiber.
@@LesLaboratory ok
@@LesLaboratory got it thx
Is that just a 3.5" floppy disc on your scope? It seems somehow an odd size in the video.
Yes, its just a dodgy camera angle! I might replace it with a floppy emulator, it takes an age to write data to it!
What a great experimentalist, I love your work and the passion with which you deliver your results deserves being a part of every Physics & Physical Chemistry undergraduate’s learning material. Some is well into postgrad PhD quality. Keep up the good work. I wish I had you as a neighbour! 👏👏👏👏👏👏
Thanks so much! There is so much to explore with this stuff, so there will be plenty of content to come!
Absolutely Awesome ! and yeah would look good on the Floyd cover, can't help with the solution but can hardly wait for more....cheers.
Thanks! Solutions have been offered in the comments, which are turning out to be real quality, so it looks like more experiments to be done.
Nearly 10k subscribers, nice going brother!
Thanks! I am there! :-D
Great work! This is super interesting!
Thanks!
I was waiting for this for whole week :) great work ! I can't think of a way to measure but probably there are micro pulses within your ns long pulses. Probably distance between these micro pulses are at the order of femtoseconds.
It's something I intend to investigate. The original paper used a monochromator a PMT and an oscilloscope to scan the spectrum, and measure the pulse width at each wavelength, and it looks like there is something to see there. I have just acquired a scope (the one int he video), I won't get down to femtoseconds with it, but after the hack, it will be a 1GHz scope, so subnanosecond resolution.
@@LesLaboratory That would be another epic content. I used PMT's in the past but it was configured to measure few photons . The ones we used was old, nevertheless had few ns rise time. I never used them in the presence of considerable amount light and I am curious. :))
Hi, your videos are great, but you should work on volume levels - your speech volume level is way to quiet compared to other videos, and intro at 0:11 is way louder than your speech, you need a gain of about 20dB.
Thanks! Yes, I have had audio issues lately. I need to see if I can get the editing software to auto level the cuts for me.
My guess is that the modes, or "striations" you see after the reflecting diffraction grading are a result of multimoding occuring within the fiber. I believe that these discrete modes are occuring because the MM fiber allows for numerous, discrete TEM modes to propagate, whereas the single-mode fiber you used previously, did not.
This is why we see the modes occur after the grating, because all of the power coupled into the fiber, exists propagating within these discrete modes in the fiber.
I'd be curious to see if applying stress to the fiber causes these modes to shift, widen, vary in amplitude, transfer power between modes, etc...
Interesting suggestion, but the consistent spacing of ca. 12nm (corresponding to ca. 13THz when operating in the visible range) matches the spacing expected from Stimulated Raman Scattering as per my other comment. This is a well-known effect in optical fibers and is even exploited for amplifying the optical telecom signals that were most likely used to transmit this video to your device! It's a very cool effect :)
Can something like this be done for regular bright light? I am in need of such a bright continuous spectrum light source such as this fir mineral identifications as early experiments have demonstrated fluorescence at specific frequency combinations. Something which cod easy be achieved with this setup and proper filters however this appears anything but portable. Picture a flashlight which causes specific minerals to stand out in the dark similar to how UV reactive ones appear under blacklight. That is what i am pretty sure can be created but proper bright configurable is anything but simple.
Try gradient index Fiber
I have considered this, do you know of any cheap sources?