To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/cuivlazygeek . You’ll also get 20% off an annual premium subscription. My Patreon: www.patreon.com/cuivlazygeek My Merch Store: cuiv.myspreadshop.com/ Link to spreadsheet: docs.google.com/spreadsheets/d/17vDrXmXeTLB6d3Lc5y7oF8C996lHGkrT_IcD2VJ_3b4/edit?usp=sharing Three different people/teams reached out telling they had built a program to replace my own clunky spreadsheet! Here are some info and the links - Online calculator by Astrophilos: astrophilos.com/calculation/ (instagram.com/_astrophilos_/) - Online calculator by Orphi: astronomytools.orphi.eu/fbsc.html (contact available on the webpage) - The astronomical students team from the Polytechnic University of Valencia, instagram.com/astroupv/ - unfortunately their program is an .exe so I can't really share it, but they actually use the Integral method which is amazing!! Amazon affiliate: amzn.to/49XTx01 Agena affiliate: bit.ly/3Om0hNG High Point Scientific affiliate: bit.ly/3lReu8R First Light Optics affiliate: tinyurl.com/yxd2jkr2 All-Star Telescope affiliate: bit.ly/3SCgVbV Astroshop eu Affiliate: tinyurl.com/2vafkax8
Cuiv, this has to be one of the best videos you have produced to date. Appreciate the amount of research this video must have required. Whilst not the most accessible for beginners, it certainly will help avoid disappointment for people choosing the right filter for their setup. It will also help determine value for money between different filters. As we all know filters aren’t cheap and your modelling shows that there may be circumstances where the conventional wisdom of getting narrower band pass, more expensive filters to minimise light pollution is actually counterproductive for a particular telescope setup.
Hey, CUIV! Your latest videos on filters is very helpful, since light pollution is a serious problem for in-city imagers. By the way, I understand the reasons for ZWO products: controller, cameras, focuser are very good, with little if any issues with use. Other items a astrophotographer might try to assemble with their OTA probably will have intermittent problems during use. I enjoy your videos and tutorials. Thanks very much.
Interesting to see it laid out like that. I've thought a bit about this problem. In microscopy there are so called infity corrected objective lenses. That means, they dont produce an image on their own but instead give you a collimated beam that you can stuff all kinds of optical devices (filters, beam splitters etc.) into without any problems. Only after that a secondary lens is focuses the image onto the camera sensor. I'm wondering if that's also possible with (fast) telescopes...
@@CuivTheLazyGeek Yes, as far as I understand it. I heard about it when considering to do macro photography during the plentiful cloudy nights we get where I live. I think microscopeworld has a nice short article entitled "Infinity Corrected Optics" explaining it. It also helps with parfocality apparently. No more worrying about backfocus shift due to the change of optical path length between different filter thickness / diffractive index. It seems so tempting. I'm really wondering, why it's not used on telescopes (that I know about).
@@Triforian Trying to add a collimated space into the optical path would only complicate things greatly for us amateur AP folks. It is often done in large telescopes where they need to pipe the light around to different instruments like spectrographs, although that can be done these days with fiber optics.
@@richardshagam8608 Interesting. Thank you for the insight into what's used on the large telescopes. Can you give a short explanation what gets more complicated or a hint where I could find some more info?
Great video as usual Cuiv, and a very useful tool you've created. Thanks! I am glad my refractive index measurements are being put to good use. Cheers, Jim T.
Incredibly useful video thanks Cuiv! I admit I am about average at Maths but I was easily able to understand everything, because you did a very good job at explaining it! This video is essential viewing for us amateurs astro nuts impo! Many Thanks and clear skies! Wes, Liverpool, England.
There are some tweaks to the spreadsheet that can make the rest of the table update automatically based on the aperture and central obstruction. I'll try to send them to Cuiv if he wants to post them. I also tried using Wolfram Alpha to do the integral of arctan(stuff) * normal distribution equation and (assuming I had my equation set up correctly) it couldn't do it. I tried an even simpler version and it could only give a Taylor series. I modified the spreadsheet to have a parameter called "slice thickness" and made the doughnut min/max radii calculate themselves from it. Changing the slice thickness from 1mm to 0.1mm changed the final result by less than 1% (at least in my initial test case). Bottom line - while there are other methods that are more accurate in principle, the difference looks insignificant. Thanks so much Cuiv! I am just about to get a Hyperstar and have been trying to figure out what to do for filters.
Over the past couple of years I have gathered an array of narrowband and LRGB filters. Some are definitely better than others. Thanks for sharing your results.
I am looking to change my ZWO 1.25" HSOLRGB filters because I have some nasty reflections and I also want to upgrade to a RC and 1.25 will not be compatible anymore. I am looking for 36mm HLRGB (no S and O for the time being) and I was wondering what would you recommend. The Ha definitely needs to be in 3nm wavelength. I honestly don't know what to pick..
what a GREAT tool! Just cleared an upcoming question for me - if the L-Extreme ist "OK" for my planned SW 200P Quattro :) It is! Thank you very very much for that easy to use thingi!
Thanks! How would you go about that? 🧐 If there's a filter + a scope, your bandpass shift is fixed - you can't really compare with anything else without completely changing other variables
Thank you so much Cuiv for this very interesting video !! I did yet adapted the form for the Origin (I have not yet, but ...) as to be able to test my own filters with it. With the Sol'Ex, it's possible to do the same as your spectrometer, and so, to test filters like you. Amzing stuff !!!!!!! Merci Cuiv.
QUESTION: How about fast lenses like Samyang/Rokinon 135mm at f2? What filter to use? How about filters that you could put in front of the lenses? Any thoughts?
If the filter is at the back of the lens, then use the calculator spreadsheet! The aperture of that lens should be 67.5mm so radius 33.75! But if you put a filter on the front of the lens all the light rays are square to the filter (for the center of the FOV) and so there is no bandpass shift!
As far as I understand for a front filter the field of view is important not the f-ratio. But at 135 mm even a normal 7 nm filter might be on the edge of what works. It also might be expensive and difficult to find.
@@Triforian Correct! And even if the FOV were something like 10 degrees per 10 degrees, the largest angle of light rays to the filter would be 5 degrees, which isn't a big deal in terms of bandpass shift!
What does this mean exactly? I aim to use the Sv220 7nm duoband filter in front of my camera lenses of 50mm, 135mm and 300mm (only tried a little so far). Should i be expecting som issues?
nice, very useful! just an idea about google sheets, instead of using fixed slice of 1 mm - use (aperture - central obstruction) divided by number or steps (can be 100 or 1000, etc). So it will be possible to avoid hardcode, just update telescope parameters and everything can be calculated automatically
Excellent video as always. Would be a great resource (like Astronomy Tool) where you put in your scope and what filter you want. Whats interesting about your results is not just the loss of light that you want, but the increase of light that you dont want. For example, if i use my LP filter on a super fast scope, not only will i get less good light but purposefully get lots more unwanted light. That would be a real kick in the pants.
That's a great remark! It doesn't let in more light, but it's true that the outside of the aperture would let in LP (for the bandpass of the filter, so not the end of the world) while not getting any signal! In theory it's then better to close down the aperture! You can actually see this with the slices breakdown in the spreadsheet!
Interesting video Cuiv. Last month I performed a test with my SW 150P quattro comparing the exposures obtained with Optolong l-extreme with those get with l-ultimate and came to a conclusion close to your but based on SNR. It would be useful and interesting to explain, maybe with a new video 😊, how filter transmittance, at different wavelengths, affects SNR. Well done and thanks again!
Amazing video! If you were really lazy, you'd just have asked the Starizona support what they think about your setup, I have to admit that's what I did 😂
Thanks for an interesting video. It would be interesting to fold the Halpha spectrum with the filter response as well. If I understand your explanation, for the calculation you perform, the assumption is the Ha light entering the optics is a delta function, whereas it is a spike and has a width of its own (gross approximation of a boxcar function). Thus the total light entering the sensor should be a bit higher than estimated as even with the shift of the transmission wavelength of the off axis rays, there will be light at these wavelengths from the nebula.
@@CuivTheLazyGeek With a filter in front of the objective lens ALL the light rays, i.e. the ENTIRE FOV, are perpendicular to the filter and this problem goes away. Indeed only feasible for small apertures. For clarity, this would remove the problem completely for ANY aperture size if they made filters arbitrarily large.
@@Mutrino Not so! The light rays hitting the objective lens (and in this case the filter at the front) are parallel to one another, yes, but are only perfectly square to the filter and objective lens on axis. For off axis stuff (e.g. anything that's not at the center of the FOV) the incoming light rays are still parallel to one another but hit the objective lens/filter at a slight angle. The max angle will be FOV/2. This is why a dew shield can cause vignetting if it is too long, and its maximum recommended length is proportional to tan(fov/2)!
@@CuivTheLazyGeek Assuming the filter is the first object in the imaging train, If the incoming rays are parallel to each other, which you can assume for an astronomical object, then *by definition* the angle of incidence is 90 degrees over the *entire* FOV and the problem does, in principle, not exist. The only errors in that regard are caused by imperfections in the filter. Vignetting caused by a dew shield that is too long is an entirely different issue.
@@Mutrino No man, sorry, that is incorrect and a common misconception! Say you have star A at the center of the field of view. Star A is effectively at infinity. The light rays from it are parallel and are 90 degrees to the objective lens. Makes sense, and we are in agreement here. Now you have Star B, say 30 arcminutes from Star A. It gets focused on the sensor (part of the FOV). Star B is also effectively at infinity and its parallel light rays hit the objective lens. Necessarily, Star B's light rays come from a slightly different direction than Star A's light rays (at a larger scale, I think it's obvious that the light rays from Vega and Deneb from instance come from wildly different directions, yet those stars can be on the same FOV of a picture). In Star B's case, its parallel light rays will hit the objective lens at an 89.5 degrees angle. If the light rays from Star B hit the objective lens also at a 90 degrees angle, it would just get focused at the center of the field of view and would be conflagrated with Star A in the FOV! A good summary is this: "Our telescopes' view of the sky is a cone, not a cylinder. Light rays from a single object do arrive parallel to one another-- light rays from two separate objects do not arrive parallel to one another.". The source of that quote, and a good illustration diagram, is here: www.cloudynights.com/topic/89911-understanding-the-optics-of-telescopes/#entry1185053 Good illustration (using optics simulation software) also here: th-cam.com/video/VfMq1AmNXxw/w-d-xo.html So in terms of bandpass shift, if you have you filter mounted right in front of the objective lens, and you have a 10x10 degrees field of view, the filter will experience at worst light rays hitting it at an 83 degree angle (for the corner of the FOV, using Pythagoras).
As always, great work. Just one question: does this calculation only apply to the center of the image, and if so, how bad are edges of fast telescopes with narrow-band filters?
Technically it only applies to the center of the image - but it should be a perfectly good approximation for the rest of FOV, since off-axis, you'll have half the rays at an angle steeper than on-axis, but the other half will be shallower, thus averaging itself out!
Nice, Cuiv. Thanks to you, I'm using my ultra precise pharmaceutical lab diode array spectrophotometer to measure my narrowband filter bandwidths. Don't know if that is good or what, but hey, at least I already happen to know spectrophotometry, so it's pretty exciting 😂. Its weird to switch between astrophotography filters and drug solutions though 😅
super weel done, love it. Now, we know that narrowband filters cut the light input by their own nature. Thus, by boosting the exposure or gain, or both, is the low effectiveness of these filters overcome?
I think it'd be really interesting to see the overall transmission plotted against the target wavelength. This would make it very easy to see where peak transmission is for a given telescope/filter combo. Ideally this would update whenever you change any of the input parameters, but this would be difficult to implement in a spreadsheet. I'd much rather do it in Python.
@@CuivTheLazyGeek I definitely have the ability to code up either, I just don't know if I have the time. I've got multiple side projects in progress for astro stuff already, from trying to design an Arduino based eaf to working on a data management tool to organize, auto-archive, and auto-stack (using siril, because my custom stacking script is also a wip) all of my subs.
@16:00 - but that total peak transmission assumes a perfect filter with 100% bandpass at the center. And we know real world filters don’t do that. You’re lucky to get 90% bandpass at the peak. So, that also needs to be modeled in.
Nice! Assuming they're up to specs of course (the ones I got a while back were horribly out of spec, but I've heard Baader has gotten better since I called them out in a video)
@@CuivTheLazyGeek they reengineered them a while ago and published a thorough explanation on their blog. This is the document with the theory and spectral graphs of each filter: [EDIT: my comment got automagically deleted - twice😅, presuming it contained the direct link to the document] Look up the pdf document "Narrow-band Filters on Astronomical Telescopes" on the Baader - Planetarium com (whitepaper_narrowband_filters_on_astronomical_telescopes pdf)
Great article, but the question for me is whether the Schmidt corrector or Maksutov corrector were also included in the calculation in this context? This property also changes when using the Starizona Hyperstar lens system and any other correctors. Great video 👍
The only thing that matters is the angle of the light ray when it hits the filter - what this does is map a slice of aperture to an angle at the filter, assuming it is sitting straight in front of the center. Overall this should work for optics with Schmidt correctors (including Hyperstar) and Mak designs as well
@@CuivTheLazyGeek Thanks for your answer,.... That's exactly the issue because it changes depending on the focal length. The only option would be a fixed focal length or that the filter sits in front of the corrector or the optics. I only noticed this when converting a Mak-Cassegrain to Rasa, which was included in the calculation by mirror designers. The angles are different before and after the secondary mirror. Best regards
Thx for a exelent video. Is there any difference in filterchoise, when to choose a filter, when the sky is only slightly polluted? I live in a border 4 area.
Hi Cuiv. Loved the video. However, I have a question. When using a reducer/flatener (in my case a 0.63x reducer for my C8), do I use the reduced focal length in the spreadsheet? I do, right?
You do if the filter is placed between the reducer and the camera (which it would be in your case), you don't if the filter is placed in front of the reducer!
Some fast refractors and lenses (Rokinon 135mm f2 for instance) are definitely quite affected! Slightly less due to their lack of central obstruction, but still;
I was wondering whether it makes a difference if I screw the filter onto the front (near the mirror) or the end (near the camera) of my somewhat long coma corrector in my 8'' Newtonian. At the end nearer the mirror (farther from the camera), there could be more light rays with larger incident angles than at the camera end, right?
Hi Cuiv,Many thanks for the video. A very naïve question, why not putting the filter where the light rays are parallel, i.e. in front of the telescope (at least for refractors)? Cost, I guess
You got it right! Cost! Making a large narrowband filter is expensive - and you will still have to contend with incident light ray angles of up to FOV/2!
Hey Cuiv! Amazing video! I guess I need to calculate my l-Ultimate for my RC8. Maybe the old l-Enhance may work better for H-Alpha with that telescope.
Another great video! I'm taking advantage of this great video on filters to ask you a question. I have an ASI 2600 MC Pro camera and I have an Optolong L eXtrem filter for the HaO3. I would like to acquire an SII filter to make real SHO. On the other hand, I'm hesitating between the 4nm S2O3 filter from Altair Astro, or the 5nm S2Hb filter from Antlia which, even if it's wider, allows you to have the Hb band. What would you advise me?
Hb is a lower intensity copy of Ha - I personally don't really see the point, and would go with and SIIOIII filter to get more OIII! Make sure your scope focal ratio supports it!
Thank you very much for your quick response! This confirms what I thought. It is better to strengthen the O3. My aperture is 7 so I wouldn't have a problem with that.
That’s really great! 👍🏻 But, you only computed for the center of the sensor. What about the sensor corners where some of the light rays can have even more incident angle? The math may get significantly more complex, I fear. But who doesn’t want the nebulosity to show in the corners of their image?
The math isn't actually that much more complex - the light rays hitting the objective lens or primary mirror are parallel but at an angle (at most FOV/2 angle) and that means that for a given point on the sensor we will have two angles to consider for a given slice of aperture. But one will be sharper, the other will be shallower - so it more or less averages itself out!
Ignorant question here as mathematics is not my strong point. It seems to me that the calculation treats the telescope as a straight tube with the light converging directly from say the 150mm objective to the sensor. In the case of a newtonian or schmidt-cassegrain the sensor is effectively imaging the much smaller secondary mirror so the angle of convergence should be measured from the secondary mirror and therefore be much less than from the 150mm aperture. I don't think the convergence of the optical paths from the primary to the secondary counts. Just a conjecture on my part. I don't have the knowledge to test this idea.
@@givemespace2742 Well, I don't quite understand, since we are effectively doing that? The angle of convergence in the end has to be what the focal length and the aperture dictate, there's just no way around it. The secondary in a newt is just a relay, it does nothing to the angle of convergence. The secondary on an SCT actually changes the angle of convergence *to* the value we use (effectively folding the focal length into a smaller distance). Regardless of how the focal length is achieved, it (together with the aperture) uniquely determines the angle of the light rays that hit the sensor. The secondary mirrors and other optical elements are just the tools that allow us to get that focal length and have the light rays reach the sensor.
Almost all your videos attract people to go tomorrow to buy expensive Ha/SII/OIII filters. The tricky thing is that you always mention Tokyo as a super light-polluted city, so the superNB filter is a must. But don't you mind always adding a disclaimer that if you are in Bortle 1-4 or even 5, you generally do not need a super narrowed filter with all these shift issues, or not that expensive, as the effectiveness of these is dependent on the typical sky conditions in a given region? And let's take into account that the _absolute majority_ of hobbyists have an average budget, an average equipment, are not pedantic and have a non-business-related interest in astrophoto. No question, your channel is super, and the info is super valuable; just please tell the audience that if you do X, you don't need Y at all.
Weeeellllll I'm always clear about the purpose of the filters, and I can't really make a general statement like that... The fact of the matter is, even in B4 you will benefit from a tighter bandpass in terms of SNR, assuming of course it has good overall transmission for your optics. All I can do is give people the tools to figure out what works for them.... I trust my viewers' intelligence...
@@CuivTheLazyGeek When you are in a hardware shop, if you are in doubt, a good salesman before they suggest a proper form and weight hammer, will ask first what kind of nails you are going to drive, when and where. However, if the one gives the audience a range of rubber, titanium, steel and golden hammers starting from 50g up to 10kg and says I trust in your intelligence, so now go and select one that fits you better, we just confuse them all as we did not tell them important aspects of significant effect, and manipulate with the effectiveness micro deltas just encourage e buy the most expensive hammer that works 5000m underwater or in supersonic conditions as the ordinary consumer assumes the most expensive and nicely wrapper branded hammer, presented by the influencer, will solve all their issues, regardless of whether a consumer is a deep diver or jet pilot.:)) Anyway, a huge thanks for your efforts!
I hear you. Every dollar hurts. Cuiv’s calculator will also show you where cheaper filters may also provide better value/performance for your own particular setup. I also have the luxury of imaging in low bortle skies, so often just use no filter or basic broadband filters, but I also like the fact that my low budget narrowband filters allow me to image throughout the month, not just around the new moon.
Hi Cuiv! These calculations made me wonder about the angles of the light rays as they reach the filter or sensor. It's straight forward to calculate them with a simulated single lens refractor or a "plain" reflector w/o any corrective elements such as a coma corrector. Is it safe to assume that the angle of such light rays depend only on the f-ratio of the optics? I'm thinking, for instance, in a petzval refractor: is the angle the same as it comes from the last glass element as it is if it would come from a single lens element in the front?
Thanks for this! Amazing. My little 40mm aperture refractor is golden with my 3nm filters. Assuming that quality control was paying attention the day my filters were tested. 😂
Great video - how confident are you in those calculations though? When I type in the data for the Redcat 51 (which is f/4.9) into your excel sheet, according to your calculations I should buy the preshifted "high speed" Antlia ALP-T filter because it should give me about 20% more transmittance. Antlias website states that for scopes slower than f/3.6 the standard version of their filter is highly recommended though. Hmm...
Quite confident - it's likely you made a mistake. The most common is to create the slices from the diameter of the aperture rather than the radius. For the Redcat, your FL should be 250mm and your slices from 25.5mm down to 0!
@@CuivTheLazyGeek Okay, not only are you a wizard in astrophotography, you are a wizard in remote debugging as well. This was precisely my mistake and now it all makes sense. You sir, are a gentleman and a scholar. I will join your patreon at once.
@@CuivTheLazyGeek You can define a slice by radius or diameter, that is a matter of personal preference, it has no impact on the end result. You actually use the radius in your spreadsheet.
Thank you for this amazing video and the sheet, Cuiv. I really appreciate it. But I wonder: I am using a svbony sv220 duo filter with 7nm band pass. In their product description it explicitly says the filter cannot be used with scopes faster than f5. Now I am considering speeding up my f5 Newtonian with a Nexus reducer, bringing it to f3.75. According to your sheet, transmission would decrease by 4.3 percent points, which seems acceptable to me. (Of course I dunno the physical specifics of the filter, but I assume it's worst case). Now I wonder why they would limit use of the filter in their advertising like that. However, I conclude that using the reducer with my filter should be ok 🙂
I all ways try to put my 2 " filter as close to the camera sensor as possible. And still keep my spacing. I assumed this my happen with angles and going into the camera sensor.
Hello Cuiv! Thank you very much for this very interesting rational approach. However, this assumes that the optics do not have at their end an optical system to parallel the light beam so that it impacts the sensor at a 90° angle. This is used on many good quality long focal telephoto lenses. Are there any astro optics that use such a correction system?
Erm, that doesn't make sense to me, if all the light rays impacted the sensor at a 90 degrees angle (i.e. all the light rays are parallel to one another at the sensor) you'd just get an unfocused mess - I think you're referring to an area in the light path where the light rays are parallelized before being focused again (before hitting the sensor, obviously). You would need to have a way to slot your filter in exactly in that place .I've never seen that in astro optics, you'd need corrective optics in front of the filter (to parallelize the light rays again), and then between the filter and the sensor! (To unparallelize them).
With a rasa telescope, why is it that you do the calculation with the full aperture rather than the diameter of the secondary mirror that reflects the light through the back of the telescope & filter there. It seems like the angle that light rays intersect the filter should be measured from the secondary mirror rather than telescope aperture
There's no secondary mirror in a RASA? That's where the camera is... Even in a non-RASA SCT you would need to use the primary mirror/objective lens since this is where the light is collected (and the collected light does make it to the final image)
@CuivTheLazyGeek Probably (in)effectiveness of the filter also strongly depends on the filter distance to the camera sensor vs a secondary mirror or lens, and on the size of the sensor.
Great video, too much technical but useful. It will be good if you put the comparing filters by brand and transmission in the end, for comparison. Thank you!
@@CuivTheLazyGeek I actually bouhgt IDAS NB1 because I switch between F5.6 telescope and Rokinon 135 F2...and I read, that Optolong filters have huge degradation bellow F5...if you go all the way to F2 as with the Rokinon, you already lost like 50% of the signal. And thats not the case with IDAS, they preserve their transmission all the way trough
@@CuivTheLazyGeek Ok. Thanks We had once “conversation” about this. I still remain on the same viewpoint. 5nm bandpass Chroma Oiii works better than specialized Astranomik bandpass 6nm MAX-Fr filter. I measured more simple way. Just took 20 frames each alternating filters. Chroma generated higher SNR. This is what matters most.
Are you sure that the filters are designed for an AOI of 0°, or rather for a small angle to be in center of the angular distribution of an F/5 or F/6 scope?
Only in the sense that it determines the effective FOV, therefore the angle of the parallel light rays at the aperture (and to the sensor). However, in the case of the filter at prime focus in front of the sensor, the sharper angles also come with shallower angles, so they effectively average out! That said, if you put the filter in front of the objective lens, then the FOV uniquely determines the angle of the light rays hitting the filter, and it is then more important!
Thanks!! In theory with your filter at the front of the telescope (as it should be with the Seestar) the light rays are square to the filter (for the center of the FOV) so you shouldn't have bandpass shift issues...
What the problem your experiencing? I have the same filter and have noticed horrible colored halos on ALL stars. I use it in front of my camera lenses. Might only be affecting my 135mm that way though, so further testong is needed for me...
@mistaskate8715 Almost impossible to focus so I used Cuiv's recommendations of focusing first with Seestar LP filter only then adding the SV220 filter. Stars are bloated and unusable which isn't too big of a deal. Also, getting about half subs as using LP filter only.
Always look for your videos, Cuiv... another good one. Qestion, please sir: 80mm triplet, ZWO533MC PRO, Bortle 8/9 metroplex skies.... One filter for every DSO? I'M using a Svbony 220 right now, and it's OK. Svbony 240 is said to work for nebula and galaxies both but produces some star bloom. I hear that Altair has an answer...what's your choice that's not crazy expensive? Thanks, Michael
Depends on what you want to do! Check my review of the IDAS GNB and also check Lukomatico's since the SV240 is a clone of that - but unless you have something specific in mind you probably don't need it?
Wait, so, bandpass shift is a function of the refractive index of the filter substrate?! Hmmm... That is news to me. I might have skipped your video a bit too fast, but I don't remember you explaining why that is, so maybe I am missing something. My understanding is that narrowband filters use the magic of interferences to only pass a small sliver of the spectrum. Therefore, what matters is the thickness of each of the tens (or maybe even hundreds) of layers that are deposited onto the substrate, and their respective refractive index. I do not see why the substrate itself would matter all that much. That being said, I have the ability to precisely measure (down to 0.03nm) the exact bandpass shift for a given angle, so I might just do a video on that, JUST TO PROVE YOU WRONG! lol. Just kidding of course. Great video as always! And very interesting tool! More on this soon, I hope. またね。
Just to be clear, it doesn't depend on the refractive index of the substrate, but on the effective refractive index of the filter overall! This is exactly the result of the combination of all of the layers with their own thickness and refractive index (e.g. exactly what you say - I hope I didn't confuse people in the video, it's always hard to have the right balance in terms of details!) See the following (which I only found after I had filmed the video, sigh. Still, it confirmed what I was saying, which is good): cdn.shopify.com/s/files/1/0267/1172/0013/files/whitepaper_narrowband_filters_on_astronomical_telescopes.pdf?v=1656667062
@@CuivTheLazyGeek Cool. Thanks for sharing the document and confirming. Even taking into account the overall refractive index, it was not at all obvious to me that there necessarily is a direct relationship between that and the amount of bandpass shift. The white paper states that the formula is an approximation, but it probably is "good enough", and I assume that the Baader engineers experimentally verified that the formula matches reality. Anyway, fascinating stuff!
fun experiment would be showing this math paired with gear lottery, would be fun to see how my filters from optolong match up since mine act a bit strange hehe (bad and good gear lottery on 1 sho package ) '
@CuivTheLazyGeek No problem, I have taken Comet A3 subs with my Seestar, and just wanted to process them. I have seen there are other tutorials, but nothing like you do! Your tutorials helped me a lot. Will you be able to do a PixInsight tutorial on Star Clusters?
@@CuivTheLazyGeekI had the same thing but hadn't realised your sheet makes you manually enter the slices instead of calculating them from the aperture. Not lazy enough!
I haven't used calculus or differential equations or any of the other such high-level math in nearly 25 years... so I'm _not_ the guy to create such a program for you :).
Yes, I don't give edit rights for obvious reasons - you need to go to file -> make a copy to create your own editable version on your Google Drive, or download it locally
I don't think your math needs to get any more complicated. Numerical integration like you performed is good enough. You could improve it a bit by performing a trapezoidal summation, but there's really no good reason to do that, either. Consider how accurate the filter curve measurements are, anyway. Frankly, you shoudn't even display your 'overall transmission' result (or any other values on the spreadsheet) to any more than three significant figures--think slide rule accuracy. Anything more than that is meaningless (and we're all going to die, to quote Dylan O'Donnell).
To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/cuivlazygeek . You’ll also get 20% off an annual premium subscription.
My Patreon: www.patreon.com/cuivlazygeek
My Merch Store: cuiv.myspreadshop.com/
Link to spreadsheet: docs.google.com/spreadsheets/d/17vDrXmXeTLB6d3Lc5y7oF8C996lHGkrT_IcD2VJ_3b4/edit?usp=sharing
Three different people/teams reached out telling they had built a program to replace my own clunky spreadsheet! Here are some info and the links
- Online calculator by Astrophilos: astrophilos.com/calculation/ (instagram.com/_astrophilos_/)
- Online calculator by Orphi: astronomytools.orphi.eu/fbsc.html (contact available on the webpage)
- The astronomical students team from the Polytechnic University of Valencia, instagram.com/astroupv/ - unfortunately their program is an .exe so I can't really share it, but they actually use the Integral method which is amazing!!
Amazon affiliate: amzn.to/49XTx01
Agena affiliate: bit.ly/3Om0hNG
High Point Scientific affiliate: bit.ly/3lReu8R
First Light Optics affiliate: tinyurl.com/yxd2jkr2
All-Star Telescope affiliate: bit.ly/3SCgVbV
Astroshop eu Affiliate: tinyurl.com/2vafkax8
Cuiv, this has to be one of the best videos you have produced to date. Appreciate the amount of research this video must have required. Whilst not the most accessible for beginners, it certainly will help avoid disappointment for people choosing the right filter for their setup. It will also help determine value for money between different filters. As we all know filters aren’t cheap and your modelling shows that there may be circumstances where the conventional wisdom of getting narrower band pass, more expensive filters to minimise light pollution is actually counterproductive for a particular telescope setup.
Perfect timing! I’m considering the purchase of a hyperstar and totally forgot that I will have to budget for new filters as well!
Yep, you really need to take that into account! Cheers!
Hey, CUIV! Your latest videos on filters is very helpful, since light pollution is a serious problem for in-city imagers. By the way, I understand the reasons for ZWO products: controller, cameras, focuser are very good, with little if any issues with use. Other items a astrophotographer might try to assemble with their OTA probably will have intermittent problems during use. I enjoy your videos and tutorials. Thanks very much.
Thank you! Glad this is helpful!!
Interesting to see it laid out like that. I've thought a bit about this problem. In microscopy there are so called infity corrected objective lenses. That means, they dont produce an image on their own but instead give you a collimated beam that you can stuff all kinds of optical devices (filters, beam splitters etc.) into without any problems. Only after that a secondary lens is focuses the image onto the camera sensor. I'm wondering if that's also possible with (fast) telescopes...
Huh, basically an area of parallel (and perpendicular to the lenses) light rays?
@@CuivTheLazyGeek Yes, as far as I understand it. I heard about it when considering to do macro photography during the plentiful cloudy nights we get where I live. I think microscopeworld has a nice short article entitled "Infinity Corrected Optics" explaining it. It also helps with parfocality apparently. No more worrying about backfocus shift due to the change of optical path length between different filter thickness / diffractive index. It seems so tempting. I'm really wondering, why it's not used on telescopes (that I know about).
@@Triforian Trying to add a collimated space into the optical path would only complicate things greatly for us amateur AP folks. It is often done in large telescopes where they need to pipe the light around to different instruments like spectrographs, although that can be done these days with fiber optics.
@@richardshagam8608 Interesting. Thank you for the insight into what's used on the large telescopes. Can you give a short explanation what gets more complicated or a hint where I could find some more info?
amazing tool for such an important spec which is often overlooked!
Thanks!! Hopefully someone will make it into a simple website page!
Great video as usual Cuiv, and a very useful tool you've created. Thanks! I am glad my refractive index measurements are being put to good use. Cheers, Jim T.
Woohoo, Jim! Thanks for stopping by, and thank you so much for your work, letting us figure out a refractive index range for those filters :)
Incredibly useful video thanks Cuiv! I admit I am about average at Maths but I was easily able to understand everything, because you did a very good job at explaining it! This video is essential viewing for us amateurs astro nuts impo!
Many Thanks and clear skies!
Wes, Liverpool, England.
Thanks so much Wes, glad it was helpful!
Great video Quiv, thanks for all the effort you put into this ! Very interesting, but most of us have to rely on what the manufacturer says it is
Thank you very much for this. Very valuable
Glad this is helpful!
There are some tweaks to the spreadsheet that can make the rest of the table update automatically based on the aperture and central obstruction. I'll try to send them to Cuiv if he wants to post them. I also tried using Wolfram Alpha to do the integral of arctan(stuff) * normal distribution equation and (assuming I had my equation set up correctly) it couldn't do it. I tried an even simpler version and it could only give a Taylor series. I modified the spreadsheet to have a parameter called "slice thickness" and made the doughnut min/max radii calculate themselves from it. Changing the slice thickness from 1mm to 0.1mm changed the final result by less than 1% (at least in my initial test case). Bottom line - while there are other methods that are more accurate in principle, the difference looks insignificant. Thanks so much Cuiv! I am just about to get a Hyperstar and have been trying to figure out what to do for filters.
Over the past couple of years I have gathered an array of narrowband and LRGB filters. Some are definitely better than others. Thanks for sharing your results.
Absolutely! Some respect their published specs,... And many dont!
I am looking to change my ZWO 1.25" HSOLRGB filters because I have some nasty reflections and I also want to upgrade to a RC and 1.25 will not be compatible anymore. I am looking for 36mm HLRGB (no S and O for the time being) and I was wondering what would you recommend. The Ha definitely needs to be in 3nm wavelength. I honestly don't know what to pick..
@@GoldenJackalTutorial if your budget is infinite, go with Chroma. Otherwise Antlia. Otherwise Askar. (In general)
@@CuivTheLazyGeek Appreciate it, mate!
what a GREAT tool! Just cleared an upcoming question for me - if the L-Extreme ist "OK" for my planned SW 200P Quattro :) It is! Thank you very very much for that easy to use thingi!
Excellent, cheers!
Thanks Cuiv this was really helpful! Appreciate your work on this!
Glad it was helpful! It was quite a bit of work!
Nice video... would like to see actual pics that show the difference. Really like your channel !
Thanks! How would you go about that? 🧐 If there's a filter + a scope, your bandpass shift is fixed - you can't really compare with anything else without completely changing other variables
Thank you so much Cuiv for this very interesting video !! I did yet adapted the form for the Origin (I have not yet, but ...) as to be able to test my own filters with it. With the Sol'Ex, it's possible to do the same as your spectrometer, and so, to test filters like you. Amzing stuff !!!!!!! Merci Cuiv.
That's ideal! Measure your filters via spectrometer, then use exactly your measurement to apply to the calculator :)
Once again, an amazing video explaining important concepts that can have a huge impact on our purchasing dollars.
Thanks Marvin!
WOW, Very interesting, Thank you.
QUESTION: How about fast lenses like Samyang/Rokinon 135mm at f2? What filter to use? How about filters that you could put in front of the lenses? Any thoughts?
That is precisely the way you do it in practice :)
If the filter is at the back of the lens, then use the calculator spreadsheet! The aperture of that lens should be 67.5mm so radius 33.75! But if you put a filter on the front of the lens all the light rays are square to the filter (for the center of the FOV) and so there is no bandpass shift!
As far as I understand for a front filter the field of view is important not the f-ratio. But at 135 mm even a normal 7 nm filter might be on the edge of what works. It also might be expensive and difficult to find.
@@Triforian Correct! And even if the FOV were something like 10 degrees per 10 degrees, the largest angle of light rays to the filter would be 5 degrees, which isn't a big deal in terms of bandpass shift!
What does this mean exactly? I aim to use the Sv220 7nm duoband filter in front of my camera lenses of 50mm, 135mm and 300mm (only tried a little so far). Should i be expecting som issues?
nice, very useful! just an idea about google sheets, instead of using fixed slice of 1 mm - use (aperture - central obstruction) divided by number or steps (can be 100 or 1000, etc). So it will be possible to avoid hardcode, just update telescope parameters and everything can be calculated automatically
Excellent video as always. Would be a great resource (like Astronomy Tool) where you put in your scope and what filter you want.
Whats interesting about your results is not just the loss of light that you want, but the increase of light that you dont want. For example, if i use my LP filter on a super fast scope, not only will i get less good light but purposefully get lots more unwanted light. That would be a real kick in the pants.
That's a great remark! It doesn't let in more light, but it's true that the outside of the aperture would let in LP (for the bandpass of the filter, so not the end of the world) while not getting any signal! In theory it's then better to close down the aperture! You can actually see this with the slices breakdown in the spreadsheet!
Interesting video Cuiv. Last month I performed a test with my SW 150P quattro comparing the exposures obtained with Optolong l-extreme with those get with l-ultimate and came to a conclusion close to your but based on SNR. It would be useful and interesting to explain, maybe with a new video 😊, how filter transmittance, at different wavelengths, affects SNR. Well done and thanks again!
Thank you I really appreciate your channel.
Thank you so much!
Amazing video! If you were really lazy, you'd just have asked the Starizona support what they think about your setup, I have to admit that's what I did 😂
Great video!
Thanks a lot Cuiv!..
Thanks for watching!
Thanks for an interesting video.
It would be interesting to fold the Halpha spectrum with the filter response as well. If I understand your explanation, for the calculation you perform, the assumption is the Ha light entering the optics is a delta function, whereas it is a spike and has a width of its own (gross approximation of a boxcar function). Thus the total light entering the sensor should be a bit higher than estimated as even with the shift of the transmission wavelength of the off axis rays, there will be light at these wavelengths from the nebula.
It will perform much better with a filter IN FRONT of the objective lens. At this (tiny) aperture is doable.
Yes, for tiny apertures that works, all the light rays are then square to the filter (for the center of the FOV)
@@CuivTheLazyGeek With a filter in front of the objective lens ALL the light rays, i.e. the ENTIRE FOV, are perpendicular to the filter and this problem goes away. Indeed only feasible for small apertures. For clarity, this would remove the problem completely for ANY aperture size if they made filters arbitrarily large.
@@Mutrino Not so! The light rays hitting the objective lens (and in this case the filter at the front) are parallel to one another, yes, but are only perfectly square to the filter and objective lens on axis. For off axis stuff (e.g. anything that's not at the center of the FOV) the incoming light rays are still parallel to one another but hit the objective lens/filter at a slight angle. The max angle will be FOV/2. This is why a dew shield can cause vignetting if it is too long, and its maximum recommended length is proportional to tan(fov/2)!
@@CuivTheLazyGeek Assuming the filter is the first object in the imaging train, If the incoming rays are parallel to each other, which you can assume for an astronomical object, then *by definition* the angle of incidence is 90 degrees over the *entire* FOV and the problem does, in principle, not exist. The only errors in that regard are caused by imperfections in the filter. Vignetting caused by a dew shield that is too long is an entirely different issue.
@@Mutrino No man, sorry, that is incorrect and a common misconception!
Say you have star A at the center of the field of view. Star A is effectively at infinity. The light rays from it are parallel and are 90 degrees to the objective lens. Makes sense, and we are in agreement here.
Now you have Star B, say 30 arcminutes from Star A. It gets focused on the sensor (part of the FOV). Star B is also effectively at infinity and its parallel light rays hit the objective lens. Necessarily, Star B's light rays come from a slightly different direction than Star A's light rays (at a larger scale, I think it's obvious that the light rays from Vega and Deneb from instance come from wildly different directions, yet those stars can be on the same FOV of a picture).
In Star B's case, its parallel light rays will hit the objective lens at an 89.5 degrees angle. If the light rays from Star B hit the objective lens also at a 90 degrees angle, it would just get focused at the center of the field of view and would be conflagrated with Star A in the FOV!
A good summary is this: "Our telescopes' view of the sky is a cone, not a cylinder. Light rays from a single object do arrive parallel to one another-- light rays from two separate objects do not arrive parallel to one another.". The source of that quote, and a good illustration diagram, is here: www.cloudynights.com/topic/89911-understanding-the-optics-of-telescopes/#entry1185053
Good illustration (using optics simulation software) also here: th-cam.com/video/VfMq1AmNXxw/w-d-xo.html
So in terms of bandpass shift, if you have you filter mounted right in front of the objective lens, and you have a 10x10 degrees field of view, the filter will experience at worst light rays hitting it at an 83 degree angle (for the corner of the FOV, using Pythagoras).
As always, great work. Just one question: does this calculation only apply to the center of the image, and if so, how bad are edges of fast telescopes with narrow-band filters?
Technically it only applies to the center of the image - but it should be a perfectly good approximation for the rest of FOV, since off-axis, you'll have half the rays at an angle steeper than on-axis, but the other half will be shallower, thus averaging itself out!
Nice, Cuiv. Thanks to you, I'm using my ultra precise pharmaceutical lab diode array spectrophotometer to measure my narrowband filter bandwidths. Don't know if that is good or what, but hey, at least I already happen to know spectrophotometry, so it's pretty exciting 😂. Its weird to switch between astrophotography filters and drug solutions though 😅
super weel done, love it. Now, we know that narrowband filters cut the light input by their own nature. Thus, by boosting the exposure or gain, or both, is the low effectiveness of these filters overcome?
I think it'd be really interesting to see the overall transmission plotted against the target wavelength. This would make it very easy to see where peak transmission is for a given telescope/filter combo. Ideally this would update whenever you change any of the input parameters, but this would be difficult to implement in a spreadsheet. I'd much rather do it in Python.
Yep, would need to provide it in a Python script, or better, embedded in a webpage!
@@CuivTheLazyGeek I definitely have the ability to code up either, I just don't know if I have the time. I've got multiple side projects in progress for astro stuff already, from trying to design an Arduino based eaf to working on a data management tool to organize, auto-archive, and auto-stack (using siril, because my custom stacking script is also a wip) all of my subs.
@16:00 - but that total peak transmission assumes a perfect filter with 100% bandpass at the center. And we know real world filters don’t do that. You’re lucky to get 90% bandpass at the peak. So, that also needs to be modeled in.
Absolutely! And I show how to do that in the video :)
This video explains exactly why I love my blueshifted 3.5nm Baaders with over 95% transmittance at the emission line 😍
Nice! Assuming they're up to specs of course (the ones I got a while back were horribly out of spec, but I've heard Baader has gotten better since I called them out in a video)
@@CuivTheLazyGeek they reengineered them a while ago and published a thorough explanation on their blog. This is the document with the theory and spectral graphs of each filter: [EDIT: my comment got automagically deleted - twice😅, presuming it contained the direct link to the document]
Look up the pdf document "Narrow-band Filters on Astronomical Telescopes" on the Baader - Planetarium com (whitepaper_narrowband_filters_on_astronomical_telescopes pdf)
Great article, but the question for me is whether the Schmidt corrector or Maksutov corrector were also included in the calculation in this context? This property also changes when using the Starizona Hyperstar lens system and any other correctors.
Great video 👍
The only thing that matters is the angle of the light ray when it hits the filter - what this does is map a slice of aperture to an angle at the filter, assuming it is sitting straight in front of the center. Overall this should work for optics with Schmidt correctors (including Hyperstar) and Mak designs as well
@@CuivTheLazyGeek
Thanks for your answer,....
That's exactly the issue because it changes depending on the focal length. The only option would be a fixed focal length or that the filter sits in front of the corrector or the optics. I only noticed this when converting a Mak-Cassegrain to Rasa, which was included in the calculation by mirror designers. The angles are different before and after the secondary mirror.
Best regards
Thx for a exelent video. Is there any difference in filterchoise, when to choose a filter, when the sky is only slightly polluted? I live in a border 4 area.
Hi Cuiv. Loved the video. However, I have a question. When using a reducer/flatener (in my case a 0.63x reducer for my C8), do I use the reduced focal length in the spreadsheet? I do, right?
You do if the filter is placed between the reducer and the camera (which it would be in your case), you don't if the filter is placed in front of the reducer!
Great video Cuiv, as usual. many thanks. am i right to understand that this issue is more crucial for reflectors than it is for refractors? thanks
Some fast refractors and lenses (Rokinon 135mm f2 for instance) are definitely quite affected! Slightly less due to their lack of central obstruction, but still;
I was wondering whether it makes a difference if I screw the filter onto the front (near the mirror) or the end (near the camera) of my somewhat long coma corrector in my 8'' Newtonian. At the end nearer the mirror (farther from the camera), there could be more light rays with larger incident angles than at the camera end, right?
Hi Cuiv,Many thanks for the video. A very naïve question, why not putting the filter where the light rays are parallel, i.e. in front of the telescope (at least for refractors)? Cost, I guess
You got it right! Cost! Making a large narrowband filter is expensive - and you will still have to contend with incident light ray angles of up to FOV/2!
Hey Cuiv! Amazing video! I guess I need to calculate my l-Ultimate for my RC8. Maybe the old l-Enhance may work better for H-Alpha with that telescope.
Another great video! I'm taking advantage of this great video on filters to ask you a question. I have an ASI 2600 MC Pro camera and I have an Optolong L eXtrem filter for the HaO3. I would like to acquire an SII filter to make real SHO. On the other hand, I'm hesitating between the 4nm S2O3 filter from Altair Astro, or the 5nm S2Hb filter from Antlia which, even if it's wider, allows you to have the Hb band. What would you advise me?
Hb is a lower intensity copy of Ha - I personally don't really see the point, and would go with and SIIOIII filter to get more OIII! Make sure your scope focal ratio supports it!
Thank you very much for your quick response! This confirms what I thought. It is better to strengthen the O3. My aperture is 7 so I wouldn't have a problem with that.
Great video! BTW, have you done a calculation for the new Celestron Origin which I understand is a RASA 6.
No - it should be very similar to the C6, feel free to use the spreadsheet to make your own computation!
That’s really great! 👍🏻 But, you only computed for the center of the sensor. What about the sensor corners where some of the light rays can have even more incident angle? The math may get significantly more complex, I fear. But who doesn’t want the nebulosity to show in the corners of their image?
The math isn't actually that much more complex - the light rays hitting the objective lens or primary mirror are parallel but at an angle (at most FOV/2 angle) and that means that for a given point on the sensor we will have two angles to consider for a given slice of aperture. But one will be sharper, the other will be shallower - so it more or less averages itself out!
Ignorant question here as mathematics is not my strong point. It seems to me that the calculation treats the telescope as a straight tube with the light converging directly from say the 150mm objective to the sensor. In the case of a newtonian or schmidt-cassegrain the sensor is effectively imaging the much smaller secondary mirror so the angle of convergence should be measured from the secondary mirror and therefore be much less than from the 150mm aperture. I don't think the convergence of the optical paths from the primary to the secondary counts. Just a conjecture on my part. I don't have the knowledge to test this idea.
@@givemespace2742 Well, I don't quite understand, since we are effectively doing that? The angle of convergence in the end has to be what the focal length and the aperture dictate, there's just no way around it. The secondary in a newt is just a relay, it does nothing to the angle of convergence. The secondary on an SCT actually changes the angle of convergence *to* the value we use (effectively folding the focal length into a smaller distance). Regardless of how the focal length is achieved, it (together with the aperture) uniquely determines the angle of the light rays that hit the sensor. The secondary mirrors and other optical elements are just the tools that allow us to get that focal length and have the light rays reach the sensor.
Almost all your videos attract people to go tomorrow to buy expensive Ha/SII/OIII filters. The tricky thing is that you always mention Tokyo as a super light-polluted city, so the superNB filter is a must. But don't you mind always adding a disclaimer that if you are in Bortle 1-4 or even 5, you generally do not need a super narrowed filter with all these shift issues, or not that expensive, as the effectiveness of these is dependent on the typical sky conditions in a given region? And let's take into account that the _absolute majority_ of hobbyists have an average budget, an average equipment, are
not pedantic and have a non-business-related interest in astrophoto. No question, your channel is super, and the info is super valuable; just please tell the audience that if you do X, you don't need Y at all.
Weeeellllll I'm always clear about the purpose of the filters, and I can't really make a general statement like that... The fact of the matter is, even in B4 you will benefit from a tighter bandpass in terms of SNR, assuming of course it has good overall transmission for your optics.
All I can do is give people the tools to figure out what works for them.... I trust my viewers' intelligence...
@@CuivTheLazyGeek When you are in a hardware shop, if you are in doubt, a good salesman before they suggest a proper form and weight hammer, will ask first what kind of nails you are going to drive, when and where. However, if the one gives the audience a range of rubber, titanium, steel and golden hammers starting from 50g up to 10kg and says I trust in your intelligence, so now go and select one that fits you better, we just confuse them all as we did not tell them important aspects of significant effect, and manipulate with the effectiveness micro deltas just encourage e buy the most expensive hammer that works 5000m underwater or in supersonic conditions as the ordinary consumer assumes the most expensive and nicely wrapper branded hammer, presented by the influencer, will solve all their issues, regardless of whether a consumer is a deep diver or jet pilot.:)) Anyway, a huge thanks for your efforts!
I hear you. Every dollar hurts. Cuiv’s calculator will also show you where cheaper filters may also provide better value/performance for your own particular setup. I also have the luxury of imaging in low bortle skies, so often just use no filter or basic broadband filters, but I also like the fact that my low budget narrowband filters allow me to image throughout the month, not just around the new moon.
Very interesting indeed…👍🏻
Thank you!
Hi Cuiv!
These calculations made me wonder about the angles of the light rays as they reach the filter or sensor. It's straight forward to calculate them with a simulated single lens refractor or a "plain" reflector w/o any corrective elements such as a coma corrector.
Is it safe to assume that the angle of such light rays depend only on the f-ratio of the optics? I'm thinking, for instance, in a petzval refractor: is the angle the same as it comes from the last glass element as it is if it would come from a single lens element in the front?
Another Great Video!
Thanks!
Thanks for this! Amazing. My little 40mm aperture refractor is golden with my 3nm filters. Assuming that quality control was paying attention the day my filters were tested. 😂
Based on the insane images you bless us with, I'd say your filter specs are good 😊
@@CuivTheLazyGeekawwwww shucks!
Haha, I just today found videos from Brent Mantooth on the same topic. He also mentioned you, Cuiv.
I didn't even know of these videos, I'll have to check them out!
Great video - how confident are you in those calculations though?
When I type in the data for the Redcat 51 (which is f/4.9) into your excel sheet, according to your calculations I should buy the preshifted "high speed" Antlia ALP-T filter because it should give me about 20% more transmittance. Antlias website states that for scopes slower than f/3.6 the standard version of their filter is highly recommended though.
Hmm...
Quite confident - it's likely you made a mistake. The most common is to create the slices from the diameter of the aperture rather than the radius. For the Redcat, your FL should be 250mm and your slices from 25.5mm down to 0!
@@CuivTheLazyGeek Okay, not only are you a wizard in astrophotography, you are a wizard in remote debugging as well. This was precisely my mistake and now it all makes sense.
You sir, are a gentleman and a scholar. I will join your patreon at once.
@@CuivTheLazyGeek You can define a slice by radius or diameter, that is a matter of personal preference, it has no impact on the end result. You actually use the radius in your spreadsheet.
@@Mutrino obviously yes! :)
Thank you for this amazing video and the sheet, Cuiv. I really appreciate it.
But I wonder: I am using a svbony sv220 duo filter with 7nm band pass. In their product description it explicitly says the filter cannot be used with scopes faster than f5. Now I am considering speeding up my f5 Newtonian with a Nexus reducer, bringing it to f3.75. According to your sheet, transmission would decrease by 4.3 percent points, which seems acceptable to me. (Of course I dunno the physical specifics of the filter, but I assume it's worst case). Now I wonder why they would limit use of the filter in their advertising like that.
However, I conclude that using the reducer with my filter should be ok 🙂
Maybe they're taking into account manufacturing tolerances with somewhat variable centering of the band passes!
I all ways try to put my 2 " filter as close to the camera sensor as possible. And still keep my spacing. I assumed this my happen with angles and going into the camera sensor.
Hi cuiv, how can i give you the program i made? i just finished programming it.. after 2hrs !
That's awesome! Can you send it to cuivlazygeek at gmail dot com?
@@CuivTheLazyGeek yes, tomorrow (Italy time) I will send you everything for email 🥰
@@CuivTheLazyGeek i just sent you an email!
I’m 65 now and remember when (late 1970s) SLR filters only crewed to the outside of your lens.
A lot of DSLR lenses still have those external threads (in other words, you're still young! ;))
Hello Cuiv! Thank you very much for this very interesting rational approach. However, this assumes that the optics do not have at their end an optical system to parallel the light beam so that it impacts the sensor at a 90° angle. This is used on many good quality long focal telephoto lenses. Are there any astro optics that use such a correction system?
Erm, that doesn't make sense to me, if all the light rays impacted the sensor at a 90 degrees angle (i.e. all the light rays are parallel to one another at the sensor) you'd just get an unfocused mess - I think you're referring to an area in the light path where the light rays are parallelized before being focused again (before hitting the sensor, obviously). You would need to have a way to slot your filter in exactly in that place .I've never seen that in astro optics, you'd need corrective optics in front of the filter (to parallelize the light rays again), and then between the filter and the sensor! (To unparallelize them).
@@CuivTheLazyGeek Yes Cuiv, thanks for correcting me, you are certainly right. I probably misunderstood some statements from lens manufacturers...
great info, thanks
but its not the same for uhc or cls, right
less effects since its wider I guess
It's actually exactly the same effect at play, but because the band passes are so wide, the impact is indeed minimal or null!
With a rasa telescope, why is it that you do the calculation with the full aperture rather than the diameter of the secondary mirror that reflects the light through the back of the telescope & filter there. It seems like the angle that light rays intersect the filter should be measured from the secondary mirror rather than telescope aperture
There's no secondary mirror in a RASA? That's where the camera is... Even in a non-RASA SCT you would need to use the primary mirror/objective lens since this is where the light is collected (and the collected light does make it to the final image)
@CuivTheLazyGeek Probably (in)effectiveness of the filter also strongly depends on the filter distance to the camera sensor vs a secondary mirror or lens, and on the size of the sensor.
Great video, too much technical but useful. It will be good if you put the comparing filters by brand and transmission in the end, for comparison. Thank you!
I can't since it depends on the telescope! So it's about the combination of scope and filter!
@@CuivTheLazyGeek I actually bouhgt IDAS NB1 because I switch between F5.6 telescope and Rokinon 135 F2...and I read, that Optolong filters have huge degradation bellow F5...if you go all the way to F2 as with the Rokinon, you already lost like 50% of the signal. And thats not the case with IDAS, they preserve their transmission all the way trough
Cuiv, I have coded your calculations into a vba script if you want it.
Nice video! What about f3.3 Epsilon? I measured for my filters. Interesting what will you get.
Feel free to make a copy of the spreadsheet and compute for your Epsilon!
@@CuivTheLazyGeek Ok. Thanks
We had once “conversation” about this. I still remain on the same viewpoint. 5nm bandpass Chroma Oiii works better than specialized Astranomik bandpass 6nm MAX-Fr filter. I measured more simple way. Just took 20 frames each alternating filters. Chroma generated higher SNR. This is what matters most.
Are you sure that the filters are designed for an AOI of 0°, or rather for a small angle to be in center of the angular distribution of an F/5 or F/6 scope?
Yes, since I measure them with an AOI of 0!
@@CuivTheLazyGeek Interesting. I would have expected that the filters are design with a small pre-offset.
Doesn’t sensor size also play a role?
Only in the sense that it determines the effective FOV, therefore the angle of the parallel light rays at the aperture (and to the sensor). However, in the case of the filter at prime focus in front of the sensor, the sharper angles also come with shallower angles, so they effectively average out!
That said, if you put the filter in front of the objective lens, then the FOV uniquely determines the angle of the light rays hitting the filter, and it is then more important!
Wow, good stuff. Hopefully, I can now figure out why my SV220 doesn't want to work well with my Seestar...
Thanks!! In theory with your filter at the front of the telescope (as it should be with the Seestar) the light rays are square to the filter (for the center of the FOV) so you shouldn't have bandpass shift issues...
What the problem your experiencing? I have the same filter and have noticed horrible colored halos on ALL stars. I use it in front of my camera lenses. Might only be affecting my 135mm that way though, so further testong is needed for me...
@@CuivTheLazyGeek Ahhh, this makes sense. Thank you for the clarification.
@mistaskate8715 Almost impossible to focus so I used Cuiv's recommendations of focusing first with Seestar LP filter only then adding the SV220 filter. Stars are bloated and unusable which isn't too big of a deal. Also, getting about half subs as using LP filter only.
Always look for your videos, Cuiv... another good one. Qestion, please sir: 80mm triplet, ZWO533MC PRO, Bortle 8/9 metroplex skies.... One filter for every DSO? I'M using a Svbony 220 right now, and it's OK. Svbony 240 is said to work for nebula and galaxies both but produces some star bloom. I hear that Altair has an answer...what's your choice that's not crazy expensive? Thanks, Michael
Depends on what you want to do! Check my review of the IDAS GNB and also check Lukomatico's since the SV240 is a clone of that - but unless you have something specific in mind you probably don't need it?
@@CuivTheLazyGeek Thanks, I'll do that.
Wait, so, bandpass shift is a function of the refractive index of the filter substrate?! Hmmm... That is news to me. I might have skipped your video a bit too fast, but I don't remember you explaining why that is, so maybe I am missing something. My understanding is that narrowband filters use the magic of interferences to only pass a small sliver of the spectrum. Therefore, what matters is the thickness of each of the tens (or maybe even hundreds) of layers that are deposited onto the substrate, and their respective refractive index. I do not see why the substrate itself would matter all that much. That being said, I have the ability to precisely measure (down to 0.03nm) the exact bandpass shift for a given angle, so I might just do a video on that, JUST TO PROVE YOU WRONG! lol. Just kidding of course. Great video as always! And very interesting tool! More on this soon, I hope. またね。
Just to be clear, it doesn't depend on the refractive index of the substrate, but on the effective refractive index of the filter overall! This is exactly the result of the combination of all of the layers with their own thickness and refractive index (e.g. exactly what you say - I hope I didn't confuse people in the video, it's always hard to have the right balance in terms of details!)
See the following (which I only found after I had filmed the video, sigh. Still, it confirmed what I was saying, which is good): cdn.shopify.com/s/files/1/0267/1172/0013/files/whitepaper_narrowband_filters_on_astronomical_telescopes.pdf?v=1656667062
@@CuivTheLazyGeek Cool. Thanks for sharing the document and confirming. Even taking into account the overall refractive index, it was not at all obvious to me that there necessarily is a direct relationship between that and the amount of bandpass shift. The white paper states that the formula is an approximation, but it probably is "good enough", and I assume that the Baader engineers experimentally verified that the formula matches reality. Anyway, fascinating stuff!
Idk if its on purpose, but at 1:52 ish (a bit before) the video is cut and you repeat the same thing twice
Thanks! Yep, small editing error, it happens
fun experiment would be showing this math paired with gear lottery, would be fun to see how my filters from optolong match up since mine act a bit strange hehe
(bad and good gear lottery on 1 sho package )
'
Please do a comet processing tutorial in PixInsight
I'm not qualified for such a tutorial, I simply don't have the knowledge!
@CuivTheLazyGeek No problem, I have taken Comet A3 subs with my Seestar, and just wanted to process them. I have seen there are other tutorials, but nothing like you do! Your tutorials helped me a lot.
Will you be able to do a PixInsight tutorial on Star Clusters?
Nice calculator,But why I'am getting insanely low (0.000000001%)with my Rokinon 135?And anything below 400mm have this kind of issue.
You're likely making a mistake - I've done computations for the Rokinon 135, and definitely don't see that
@@CuivTheLazyGeekI had the same thing but hadn't realised your sheet makes you manually enter the slices instead of calculating them from the aperture. Not lazy enough!
I haven't used calculus or differential equations or any of the other such high-level math in nearly 25 years... so I'm _not_ the guy to create such a program for you :).
Hahaha that's fine! To be honest the slices method is good enough :)
The spreadsheet requires a premission.
Yes, I don't give edit rights for obvious reasons - you need to go to file -> make a copy to create your own editable version on your Google Drive, or download it locally
@@CuivTheLazyGeek Thanks. Got it.
I don't think your math needs to get any more complicated. Numerical integration like you performed is good enough. You could improve it a bit by performing a trapezoidal summation, but there's really no good reason to do that, either. Consider how accurate the filter curve measurements are, anyway. Frankly, you shoudn't even display your 'overall transmission' result (or any other values on the spreadsheet) to any more than three significant figures--think slide rule accuracy. Anything more than that is meaningless (and we're all going to die, to quote Dylan O'Donnell).
Yep absolutely agree :) the 1mm slices are definitely more than precise enough... But integrals and limits are so beautiful... 😂
Brilliant. Ha Ha!
Mwahaha :)
Excellent Video, Thank you