Cuiv, I am rather a user of cooled astrocams, but I wonder, how can average thermal signal (which is not noise btw, noise is the just the square root of the average level, which U ofc know) swamp signal? Signal grows linearly with time, and thermal noise grows square rooty with time. No matter what multiplicative constant, the signal will asymptotically win. How can that problem be real? I suspect someone has misconceptions in misconceptions and they really mean "amp glow" and related effects (which are signal! and grow linearly) not averaged thermal noise. (btw, just thinking out loud, why using multiband filters with OSC sensors and DSLR's, when it is pretty affordable nowadays to use cooled mono+NB. the first option is cutting tons of light and resolution, and doesn't have any flexibility)
Hehehe I followed the exact same reasoning while preparing the video - we think alike. You are absolutely correct of course - in the end my understanding is absolutely about how DSLRs react to high temperatures in the sensor (easily 40-50 celsius in summer) and/or high ISO. In that case the dark current noise itself is of course an issue, but as it grows as a sqrt, that can be easily overcome. In examples I saw it showed horrible fixed patterns that simply didn't want to be fixed by calibration, dithering, and sigma clipping, the worse the longer you exposed (or the higher you got the ISO). Which yes, is not noise but something similar to amp glow (but difficult to calibrate out). I always think I need to be clearer in my videos, but then I end up making super long videos are doing millions of takes, so I have to compromise :) Thanks for pointing this out though!
@@CuivTheLazyGeek Your vids are great! I am positive you know many maths/physics/software details (I am a theoretical physicist working part as a programmer). You are one of the few astro-youtubers with really scientifically accurate vids. I only try to enhance what you've said. Ppl (not u) fall for the "termal noise" trap, when it is really the average they observe. Also that weird pattern noise should be visible on darks but the pattern probably travels as the temperature changes and different elements in the electronics heat up. Considering this are 10 minute subs, calibrating them is extremely unpractical. This is why cooling was invented :) In cams like ASI 2400, 2600, 6200 at
Just to comment on your thoughts, most people who use DSLR or Mirrorless cameras for astrophotography already have them and want to fully utilise whatever equipment they have at hand.
@@rafserrano2164 Just to comment on your thoughts: if you already have a dishwasher , but want a garden sprinkler, you can try to use the dishwasher's components to setup the garden sprinkler to fully utilize the equipment you already have :)
@@przemekmajewski1 I'd rather use the garden hose, thank you ;) didnt know physicists tend to make things more complicated, must be a very stressful job 😆
It has to be the most common rookie mistake regarding exposure time. I think it comes when someone imaged the Orion nebula and it shows up on nearly all cameras nicely with/without any filter. So moving on to a fainter object, the rookie WANTS to see the nebula on each sub, although many fainter nebulae will not show up like that unless the image is severely over exposed and the stars are over saturating and bloating like a big flashlight. So, moving on to a light pollution filter or narrow band, and the "ahhh, finally, I can shoot 10 minute subs so now I will be able to see the nebula in each of my subs" . Meanwhile the nebula emit the same amount of light , the narrow band filter letting through the same amount of light too, with the exception of the wavelength that are blocked. You are still receiving the same amount of light at the sensor. Some blocked, others lights allowed through, the nebulae or the stars aren't emitting less or more. People asked me "why are you only taking 180second subs with the narrow band? " I'm like, because I tried 300 seconds and my stars were bloating. Plus if I got an airplane or cloud ruining my sub, I don't have to throw away a whole 5 minute exposure, only 3 minute. It adds up!
I use a ZWO 533MM camera with LRGB, Ha, O-III and S-II filters. I run the cooler at -10C. I find longer exposures to be advantageous to a point, but there are factors that limit that advantage even with cooled cameras. 300-second subs in hydrogen alpha generally have the same noise levels as 180-second subs. I have yet to expose O-III and S-II subs for less than 300 seconds. What limits the benefits of going 300 seconds or longer for me are the hordes of satellites and aircraft passing overhead, and my mount's ability to track accurately. Light pollution is another limiting factor. Poor seeing also messes with your auto guiding too. What seems to work best for me so far is exposing through the LRGB filters subs for 30 to 120 secconds depending on the object and filter, and 180 to 300 seconds through the 7nm narrow band filters depending on the object. In this way a passing satellite or glitch in the auto guiding results in less lost integration time. I am using an 80mm F/6 apochromatic refractor with a .8X focal reducer/flattener.
This is a great video and was something I too have struggled to understand in the past. The clarification of the two sentences "I need to expose longer/I can expose longer" is gold.
Great info Cuiv. I wish I had viewed something like this four years ago - Would have saved a lot of disappointing imaging sessions with my DSLR until I finally figured it out by trial and error.
I worked as a radiometrologist (retired, physicist). We would use a multi-channel analyser (MCA) to examine the spectrum of say, X-rays. By setting a region of interest (ROI), your narrowband filter's window, we would exclude the unwanted data either side, and its noise too. That may mean you can expose for longer, because you have less noise and can count for longer because you have less signal too. Can does not mean you necessarily want to. That depends on your signal to noise ratio with a camera (& counting statistics). A smaller camera sensor will produce less heat, which is proportional to its area (& design a bit too), and that is noise to be removed somehow. You still need to determine your optimum ISO and exposure time, tracking may be a factor with the latter. But you may not need to increase your exposure time because you are still getting the same count in your ROI, you have simply got rid of the rubbish on either side.
Thank you for tackling a complex subject there are dozens of folks who are addresing the entry level problems and only a handful willing to go to the next step. Thanks and please keep up the good work.
Great video pointing out a common misconception. However if I heard correctly, at one point you imply that increasing exposure time will increase thermal noise. The opposite is true, sensor thermal noise will decrease in each frame as the square root of exposure time. For the stacked image it is the total exposure time that matters. For the same total exposure time the sensor thermal noise component will be the same. P.S. Kn addition to increasing image contrast the NB filter will also decrease the shot noise in the final image that comes from the sky glow. Keep up the good work. I always enjoy your videos!
Well explained Cuiv!! You've probably seen this, but for those that haven't, a talk by Roger Glover (of SHARPCAP fame) giving an in-depth talk on the 'dynamics' of CMOS astroimaging.
Awesome info bro! Extremely helpful in understanding the narrow band filters. I almost got tricked too into thinking I need to expose longer or raise the ISO but it makes perfect sense what are you saying.
Thanks for the explanation! I was really falling for that misconception "If I get an L-Extreme I will have to expose longer and my camera is already at it's limit!". It's well explained, that's a new sub for you :D
L-Extreme does however like longer exposures. According to the exposure calculator my saturation for Luminance is 6 seconds while for the L-Extreme it's 2 minutes.
@@licho52 Ideally it indeed prefers longer exposures as it cuts off the skyglow and thus the shot noise (and thus read noise becomes more considerable compared to shot noise). But it doesn't mean that you absolutely have to expose to optimal exposure time (especially if thermal noise ends up overwhelming read noise quickly), and even if read noise is not overwhelmed by other sources of noise because of other constraints on exposure time (whatever those constraints may be), it's going to be fine (and with modern low read noise sensors better than suffering from considerably more skyglow shot noise with broader band passes)
Depends on how tall the peak is, on the spikes in the graphs y-axis. Moving from no filter, to a filter, you're getting less light because some of it is lost on the y-axis but the aim is to cut light out on the x-axis. Moving from one filter to another, you're probably not losing any of the light on the y-axis (which will depend on the filters you choose to use). So it makes sense that you don't need to increase exposure times. You're still getting (pretty much) the same amount of light in the wavelengths you want light from, unless those peaks are a lot lower. However, increasing exposure time means you're gathering more light and less of it is light pollution.
Thanks Cuiv! Great info. I'm using a DSLR right now but plan to move to a dedicated astro cam. This is great info, thanks again. ps.. I lived in Japan for 2 years. Loved it. It's not all raw fish ;-)
That is true only if the nebula emission is limited to that narrowband. If there is a certain emission close but out of that band you will lost data there. I do prefere to capture with and without and blend them
Glad you’re back and doing better. I would love to see a video on how do the initial setup in NINA with an auto focuser. I bought and installed a Senso quite a while back and although I was able to get it installed and see that the motor would drive my focuser I’m completely lost on what to do next to actually make AF work. Every video I’ve seen always assumes certain knowledge and skips over steps making it impossible to figure out. A short simple tutorial showing all the steps with NINA would be very helpful. I’ve been using NINA for a while and have all the other basic functionality working but AF is the final piece. Anyway, keep up the great content!
@@CuivTheLazyGeek someone nds good. I actually remember looking at one of them when I first got the AF but still couldn’t really figure it out. I’ll probably relook at them as well
I think using these on a unmodded DSLR is kind of a waste- The blocking filter on the camera blocks out most of what you're paying for the filter to pass. Great video and great to see you feeling a little better. Welcome back!
Yes, you can't deceive the laws of physics, but knowing them, you can better understand how to use them to achieve your goal. My first Canon 550D, I myself am a lover of the Sony system, I turned into a Frankenstein) He could sing and dance and even cool down at a delta of -26 °. The reason was simple, there was no money for an astro camera. However, it was then that I realized all these relationships when using narrow band filters. Now my astro-Canon is in my "museum", I do not sell it, as a memory of where it all began. And comparing his raw footage today with what I'm getting now from Toptek 571, I don't always see a huge difference. Cooling is a big advantage for digital cameras. Not counting, of course, high-aperture optics.
Excellent video!! I’m a beginner (& starting to get some decent results with my DSLR) and this was easy to understand, along with being very informative. Thanks!
Hello Cuiv, I l Really love your video's. Have seen a lot of them and learnt a lot. In the meantime, Optolong has the L-Ultimate out. What about this one for OSC camera's ?
In theory it's excellent! I've only measured a couple in my spectrometer, one was good (up to spec) the other terrible though... So in theory excellent, in practice lottery
Cuiv, I know you are on a hiatus right now and will be back in a few months and all, however, perhaps it would be funny to do a very short video in french when you are back! Just short 1 minute stuff. Just an idea
Well, not completely the same. There's about a 3% loss to the filtering material and another 5% loss to the glass plate it is on... so you ARE losing about 9% of the narrowbandpass you want as opposed to not using a filter at all, but in the end its usually less total loss than using a broad band-block like the pollution filters. Additionally if you're going for deep galaxies the 5 to 7nm narrow filters are a better option. Using an actual astrograph instead of an "astrophoto telescope" will make a much greater difference in your light gain overall but not many people are willing to do the tiling and knitting that RCs need for broad objects.
This is the perfect explanation of what I asked a few days ago. Thanks so much, Cuiv! One more thing: what is the real relationship between band pass and telescope speed, i.e. focal ratio? Why are there filters designed for f/2 telescopes for example?
The issue is with the angle of the incoming light. Narrowband filters are interference filters, which work be causing light other than the band you want to pass to be removed by interference. These are very sensitive to the angle of the incoming light, as you change the angle you shift the wavelength which gets transmitted. Faster focal ratios have light coming in to the sensor at a wide angle, this means that more of the light sees the filter as shifted and so the light you do want ends up getting mostly blocked.
What Ben said - at fast focal ratios, the filters will start becoming opaque wrt angled incoming light rays. Effectively this has the effect of diminishing the telescope aperture (and thus increasing the focal ratio artificially) for the useful signal. You'll still get images (and good images at that), but effectively you're not using the full potential of the fast scope.
@@CuivTheLazyGeek I have to now, you just single handedly invalidated the very first video of my astrophtography CZcams channel 😅😅😅😃. But really though, this is great information and I am glad you can help steer me, and others in the right direction. I will test this theory soon enough. Once again, great video! 👍👍
Generally the exposure time graph shows 3 minutes is optimal with diminished returns after. Canon 6d has such low read noise, my sub tropical climate shows its fine even in the heat.
Newbeone here. Still learning, no HW yet. Is exposure time to capture nebulae based on equipment you are using, trial and error, or are there some basic guidelines to start out with? My current HW plan is edgehd, AM5 and a 2600mc pro(maybe the duo).
I never dial up the ISO. It heats the camera. It does not increase the number of photons collected. I leave my ISO at the values where read noise, DR are optimized. I can increase the gain in post.
Hi! This might be a dumb question after what you said, but because of the response i've gotten from asking store owners about filters i still wanted to ask. I use a star tracker with an HA-modded Nikon dslr and lenses between 50mm to 300mm and f1,8 to f8. I do this unguided and get between 1-5min exposure time. I want a filter to be able to do astrophotography i full moonlight and in summertime when it doesn't get as dark at night. Because what you say makes sense to me i believe that i could go with a narrow bandpass filter like, for example, the L-extreme for this, but several store owners have said it's not a good idea. Do you think i should go for 7nm, or would you suggest a wider bandpass in my case? I should add that i want to mount the filter in front of my lenses as that makes most sense to me with camera lenses. If you know a filter that makes that possible with normal camera step down rings i would love suggestions as this is not an easy topic to get clear answers on either!
Excellent, clear explanations as always ! In the video you mentioned that narrowband filters are not a good idea for OSC cameras - why is this ? Thanks
Let's say single band filters are not super efficient with an OSC camera. So an Ha filter would only make the red pixels react, so only 1/4th of the surface area. Still, it can work fine if combined with RGB data from the same camera.
Nice video, great explanation. Question or tips. I have a 450D canon modified myself (remove IR\UV filter). I am using 12nm halpha filter. I notice that I have a lot of bandings (horizontal) in my raw images @ 400mm (refractor). Compare to my camera lens 50mm I don't see bandings. Is there a way to avoid the bandings? Why do I see them @ 400mm but not @ 50mm? Maybe it's still there ? Thanks!
Actually the person who teaches me astrophotography is wrong. I was wondering why I have to exposed longer for Ha. It totally didn’t make sense for me. Thanks for the clarification
He's likely thinking about swamping the read noise, which is indeed desirable - but if it's going to make things worse, then it's fine to exposure less long!
I think that this video, even beeing more suitable for DSLR users, adresses a little my question about the L-Extreme and the ASI533MC. I've used that filter on M31 and took about 50 120s frames (out of 75 due to a wall on my line of sight 😭) and I'll now combine those with a bunch of 120s or 180s frames to be taken with the Baader Neodymium (I live under a Bortle 7/8).
It’s usually not recommended to use any filters for galaxies. Especially not narrowband unless you actually want to capture the faint H-alpha signals from that galaxy. This is because most galaxies emit their light in broadband and any filter will reduce the signal that your camera receives from the galaxies
@@ripwords I'm aware of that. This was an experiment since I'm starting in the hobbie, I've done the NGC7000 with this filter (you can see the Picture @astrobin) and wanted to see the result on galaxies and M31 was the obviuos target 😁😇). I'm at this time taking 120s frames with the Baader Neodymium light pollution filter since I'm imaging from a Bortle 7/8 zone and later I'll combine with the previous taken with the L-EXTREME and see the result
Using the L-eXtreme as a booster of nebulae within M31 will likely work - after all there are multiple methods to combine Ha and LRGB for galaxies, to stress out their nebulae. I've never done that myself though!
Thanks for the video. I'm new to narrowband filters so found the video interesting but left me a bit confused. Perhaps it would be helpful to share specific examples of solutions on what settings and exposure times would be good? In my case I'm getting an h alpha modded canon EOS R and a drop in narrowband L extreme filter.
I am using these techniques with my DSLR. Hopefully, Santa Clause will bring me a dedicated cooled camera for next summer! Until then I’m just trying to improve my skills.
Thanks Cuiv, great video! I was working my way to your conclusion but you got me over the line! I'm a DSLR user and just starting out - I wonder if there is an issue using the very narrow bandpass filters with manual focusing with a Bahtinov mask?
You really want a bright star to use a Bahtinov mask with a very narrowband filter and a DSLR - but it should still then be visible in live view. Otherwise you'll need to take short exposures and adjust in between!
Hello - when I use my DSLR , I was under the impression that I need to have my histogram 1/3 of the way for the peak value - otherwise , the image is underexposed . I have been following this rule of thumb When I use my multi band pass filter , The histogram is too far to the left , so I increase my exposure time to get it back to 1/3 . So are you saying that I don’t need to increase exposure time ? If the histogram peak is all the way to left , that is ok ? I feel like it will be too dark . A 30 second exposure with and without a filter is not the same on the histogram Can you please elaborate on this ? Thank you - and thank you for all the great content ! Regards
In deep sky astrophotography, you want to expose to the left rather than to the right. The only thing you need is to have the histogram "spike" separated from the left edge of the histogram. That should guarantee you are swamping the read noise, which is what we need. Since after stacking, you can stretch the image, you don't need to worry that it appears too dark with individual sub frames. You can expose longer of course but then you're likely saturating the stars.
@@CuivTheLazyGeek thank you for your reply . So is it better to have the peak on the histogram as close to the left as possible ? As long as the peak is to the right of the left edge ?
Hi Cuiv, I'm a lazier amateur astrophotographer here and following your channel since your AZ-GTI tuning video :) , what are your thoughts of using narrowband and an unmodified regular camera? I am planning to use the Optolong L-enhance with my 61mp Sony Camera, am I to expect terrible results with a regular camera and narrowband filters?
Mmmmmh, well you would get a result, and you'd likely get a better result than without in a light polluted zone. H-Alpha is so strong on so many targets that it could gather enough SNR, even when cut down by 80% by the sensor filter. I'd say give it a try - but be prepared to want a modded camera after you try! (a lot of cheap, old Canon modded cameras floating around)
@@CuivTheLazyGeek thanks for sharing your thoughts on this, I didn't know that there is that much reduction from the sensor filter. I have an Astronomik clip in CLS filter for my camera and there is a lot of colour shift to it even if I set the grey point in white balance that is why I'm thinking of trying narrow band filters, I guess its better to move to a less light polluted area in my case 😆
Dithering changes the framing vey slightly in between images. This helps get much better images. See details here: czcams.com/video/Da33jfKJl1M/video.html
@@CuivTheLazyGeek Thank you!!! I now completely understand.. you've taught me quite I bit with your videos.. I must say, I am starting new again in AP after 16 years of downtime and now having to learn new programs and having to replace my 300D and Toucam Pro II with a 450D and zwo426 for guiding and planetary imaging.. (still waiting on the zwo, tho).. Thank you again!
What about with the HCG-mode of the 2600, 533 cameras with ZWO saying we should either shoot at 0 gain or 100 gain, 100 gain being the HCG-mode. Would really love to pick your brain on something Cuiv maybe I will try an DM you on discord
From what I can tell, there is basically no drawback to using 100 gain/HCG mode - you get much better read noise while basically retaining dynamic range
The OPT affiliate link takes me to spammy stuff. Is there a typo? If i copy the text of the link it works all right, but if i click on it i get redirected to stuff that wants to show me notifications and install software, and never arrive at OPT.
There seems to be an extra u+200b character (zero-width space) at the end of the link. This causes a redirect to a malicious site. Not clear to me why incognito prevents this, but i see the extra Unicode character in the output from wget, so it's present even if a browser is not involved, so not the result of any extension.
That's super weird, thanks for letting me know! Yeah there seems to be an invisible character at the end of the link for some reason.... I've removed it for this video and will check on others too. Thank you!
I have a full spectrum modified 60D and find when I shoot everything is very red. I know I can process much of this out… but… I think I am missing a filter… what clip in filter do you recommend for someone just starting out? I will be attaching my camera to a FLT91 (when ever it arrives) 😊
I think Optolong has most of their filters available as clip in. Depending on your targets I'd recommend the L-Pro or the L-eXtreme, but it is really up to you! Also I'd probably go with a filter drawer rather than clip-in simply because it's then easier to switch! That's a nice scope you have on the way :)
Interesting topic! I have the f/2 3.5nm Baader narrowband pass filters for Rasa. I am wondering if they can be used for normal scopes > f/4 and if they are better than using my older 12nm narrowband filters.
They won't work on normal scopes - their bandpass is pre-shifted and transmission will be close to 0 on a F4+ scope. That said, I've just had my OIII f2 4nm Baader NB filter (CMOS-Optimized) measured via spectrometers. It is pre-shifted too much, and barely passes any OIII at f2.... so yeah, I'm not having a good experience with Baader.
@@CuivTheLazyGeek Wow. Thanks for the information. I know the previous Baader f2 6.5nm ones have halos, so they should have already fixed the 4nm filters in all departments. This could be controversial. Will you be releasing any video on this soon?
Hi Cuiv. A question regarding NB imaging. I'm starting working with monocrom from a light polluted area (Bortle 7/8) and I'll try the Horse Head Nebula as HaRGB. The question is shall I use short exposures in RGB and larger ones in Ha or can I go to longer ones in RGB. From your experience with light pollution what will be the best approach?
So the best approach is to swamp the read noise, while not saturating the sensor. For the RGB exposures, the latter will happen very quickly. In a Bortle 7/8 I don't think I'd go beyond 60s exposures for RGB (assuming F4-F5 telescope or so), and for Ha I would probably go for 180s (7nm filter) or 360s (3nm filter). So long exposures for Ha, short exposures for RGB!
@@CuivTheLazyGeek Thanks for the help. I'll try it as soon as the clouds get off my way. And yes the imaging train will be an Evostar72ED F5.8 with the ZWO 7nm Ha along with the ASI1600MM
Maybe I need to come back with a fresher head but I didn't really pick up a whole lot of what was said. Just went in through one ear and out the other. Maybe someone can explain it to me in an even more simple way.
To sum it up: Narrower bandpass filters will let the same amount of light of the target wavelength through, it will only decrease the collected light of the rest of the spectrum (which is what we want in the first place). Or, to use a simily: if one (the light) walks on a specific line through an open door, it doesn't matter if it is a regular door or a barn door, one will make it through just as fine (unless it's yo' momma, 'cause she's so ... ), but the barn door will let through all kinds of stuff you don't want (possibly even yo' momma). ;)
@@oscpe262 No, I know how narrowband works it was a lot of what was send in the video that I didn't understand. I can't remember what it was that I didn't understand right now as it was almost a whole day ago.
Cuiv, I am rather a user of cooled astrocams, but I wonder, how can average thermal signal (which is not noise btw, noise is the just the square root of the average level, which U ofc know) swamp signal? Signal grows linearly with time, and thermal noise grows square rooty with time. No matter what multiplicative constant, the signal will asymptotically win. How can that problem be real? I suspect someone has misconceptions in misconceptions and they really mean "amp glow" and related effects (which are signal! and grow linearly) not averaged thermal noise.
(btw, just thinking out loud, why using multiband filters with OSC sensors and DSLR's, when it is pretty affordable nowadays to use cooled mono+NB. the first option is cutting tons of light and resolution, and doesn't have any flexibility)
Hehehe I followed the exact same reasoning while preparing the video - we think alike. You are absolutely correct of course - in the end my understanding is absolutely about how DSLRs react to high temperatures in the sensor (easily 40-50 celsius in summer) and/or high ISO. In that case the dark current noise itself is of course an issue, but as it grows as a sqrt, that can be easily overcome. In examples I saw it showed horrible fixed patterns that simply didn't want to be fixed by calibration, dithering, and sigma clipping, the worse the longer you exposed (or the higher you got the ISO). Which yes, is not noise but something similar to amp glow (but difficult to calibrate out). I always think I need to be clearer in my videos, but then I end up making super long videos are doing millions of takes, so I have to compromise :) Thanks for pointing this out though!
@@CuivTheLazyGeek Your vids are great! I am positive you know many maths/physics/software details (I am a theoretical physicist working part as a programmer). You are one of the few astro-youtubers with really scientifically accurate vids. I only try to enhance what you've said.
Ppl (not u) fall for the "termal noise" trap, when it is really the average they observe. Also that weird pattern noise should be visible on darks but the pattern probably travels as the temperature changes and different elements in the electronics heat up. Considering this are 10 minute subs, calibrating them is extremely unpractical. This is why cooling was invented :) In cams like ASI 2400, 2600, 6200 at
Just to comment on your thoughts, most people who use DSLR or Mirrorless cameras for astrophotography already have them and want to fully utilise whatever equipment they have at hand.
@@rafserrano2164 Just to comment on your thoughts: if you already have a dishwasher , but want a garden sprinkler, you can try to use the dishwasher's components to setup the garden sprinkler to fully utilize the equipment you already have :)
@@przemekmajewski1 I'd rather use the garden hose, thank you ;) didnt know physicists tend to make things more complicated, must be a very stressful job 😆
It has to be the most common rookie mistake regarding exposure time. I think it comes when someone imaged the Orion nebula and it shows up on nearly all cameras nicely with/without any filter. So moving on to a fainter object, the rookie WANTS to see the nebula on each sub, although many fainter nebulae will not show up like that unless the image is severely over exposed and the stars are over saturating and bloating like a big flashlight.
So, moving on to a light pollution filter or narrow band, and the "ahhh, finally, I can shoot 10 minute subs so now I will be able to see the nebula in each of my subs" .
Meanwhile the nebula emit the same amount of light , the narrow band filter letting through the same amount of light too, with the exception of the wavelength that are blocked. You are still receiving the same amount of light at the sensor. Some blocked, others lights allowed through, the nebulae or the stars aren't emitting less or more. People asked me "why are you only taking 180second subs with the narrow band? " I'm like, because I tried 300 seconds and my stars were bloating. Plus if I got an airplane or cloud ruining my sub, I don't have to throw away a whole 5 minute exposure, only 3 minute. It adds up!
I use a ZWO 533MM camera with LRGB, Ha, O-III and S-II filters. I run the cooler at -10C. I find longer exposures to be advantageous to a point, but there are factors that limit that advantage even with cooled cameras. 300-second subs in hydrogen alpha generally have the same noise levels as 180-second subs. I have yet to expose O-III and S-II subs for less than 300 seconds. What limits the benefits of going 300 seconds or longer for me are the hordes of satellites and aircraft passing overhead, and my mount's ability to track accurately. Light pollution is another limiting factor. Poor seeing also messes with your auto guiding too. What seems to work best for me so far is exposing through the LRGB filters subs for 30 to 120 secconds depending on the object and filter, and 180 to 300 seconds through the 7nm narrow band filters depending on the object. In this way a passing satellite or glitch in the auto guiding results in less lost integration time. I am using an 80mm F/6 apochromatic refractor with a .8X focal reducer/flattener.
This is a great video and was something I too have struggled to understand in the past.
The clarification of the two sentences "I need to expose longer/I can expose longer" is gold.
My God Cuiv, you are such the Geek! Thank you for that!!!
Great info Cuiv. I wish I had viewed something like this four years ago - Would have saved a lot of disappointing imaging sessions with my DSLR until I finally figured it out by trial and error.
It makes totally sense what you say. I do use a DSLR and will move to a much more polluted area, so your advice is highly appreciated, thanks!
I worked as a radiometrologist (retired, physicist). We would use a multi-channel analyser (MCA) to examine the spectrum of say, X-rays. By setting a region of interest (ROI), your narrowband filter's window, we would exclude the unwanted data either side, and its noise too.
That may mean you can expose for longer, because you have less noise and can count for longer because you have less signal too. Can does not mean you necessarily want to. That depends on your signal to noise ratio with a camera (& counting statistics). A smaller camera sensor will produce less heat, which is proportional to its area (& design a bit too), and that is noise to be removed somehow. You still need to determine your optimum ISO and exposure time, tracking may be a factor with the latter.
But you may not need to increase your exposure time because you are still getting the same count in your ROI, you have simply got rid of the rubbish on either side.
Thank you for tackling a complex subject there are dozens of folks who are addresing the entry level problems and only a handful willing to go to the next step. Thanks and please keep up the good work.
My pleasure - it's the complex stuff that's the most fun! :)
Great video pointing out a common misconception. However if I heard correctly, at one point you imply that increasing exposure time will increase thermal noise. The opposite is true, sensor thermal noise will decrease in each frame as the square root of exposure time. For the stacked image it is the total exposure time that matters. For the same total exposure time the sensor thermal noise component will be the same. P.S. Kn addition to increasing image contrast the NB filter will also decrease the shot noise in the final image that comes from the sky glow. Keep up the good work. I always enjoy your videos!
Well explained Cuiv!! You've probably seen this, but for those that haven't, a talk by Roger Glover (of SHARPCAP fame) giving an in-depth talk on the 'dynamics' of CMOS astroimaging.
Yep, it's a great talk! Thank you!
Awesome info bro! Extremely helpful in understanding the narrow band filters. I almost got tricked too into thinking I need to expose longer or raise the ISO but it makes perfect sense what are you saying.
Good to see you cranking the info.
Thanks for the explanation! I was really falling for that misconception "If I get an L-Extreme I will have to expose longer and my camera is already at it's limit!". It's well explained, that's a new sub for you :D
My pleasure! Glad it was helpful!
L-Extreme does however like longer exposures. According to the exposure calculator my saturation for Luminance is 6 seconds while for the L-Extreme it's 2 minutes.
@@licho52 Ideally it indeed prefers longer exposures as it cuts off the skyglow and thus the shot noise (and thus read noise becomes more considerable compared to shot noise). But it doesn't mean that you absolutely have to expose to optimal exposure time (especially if thermal noise ends up overwhelming read noise quickly), and even if read noise is not overwhelmed by other sources of noise because of other constraints on exposure time (whatever those constraints may be), it's going to be fine (and with modern low read noise sensors better than suffering from considerably more skyglow shot noise with broader band passes)
Depends on how tall the peak is, on the spikes in the graphs y-axis. Moving from no filter, to a filter, you're getting less light because some of it is lost on the y-axis but the aim is to cut light out on the x-axis. Moving from one filter to another, you're probably not losing any of the light on the y-axis (which will depend on the filters you choose to use).
So it makes sense that you don't need to increase exposure times. You're still getting (pretty much) the same amount of light in the wavelengths you want light from, unless those peaks are a lot lower.
However, increasing exposure time means you're gathering more light and less of it is light pollution.
Thanks Cuiv! Great info. I'm using a DSLR right now but plan to move to a dedicated astro cam. This is great info, thanks again. ps.. I lived in Japan for 2 years. Loved it. It's not all raw fish ;-)
That is true only if the nebula emission is limited to that narrowband. If there is a certain emission close but out of that band you will lost data there. I do prefere to capture with and without and blend them
Glad you’re back and doing better. I would love to see a video on how do the initial setup in NINA with an auto focuser. I bought and installed a Senso quite a while back and although I was able to get it installed and see that the motor would drive my focuser I’m completely lost on what to do next to actually make AF work. Every video I’ve seen always assumes certain knowledge and skips over steps making it impossible to figure out. A short simple tutorial showing all the steps with NINA would be very helpful. I’ve been using NINA for a while and have all the other basic functionality working but AF is the final piece. Anyway, keep up the great content!
I have some older videos on this but I am planning on doing a refresher.
@@CuivTheLazyGeek someone nds good. I actually remember looking at one of them when I first got the AF but still couldn’t really figure it out. I’ll probably relook at them as well
Had to abandon my modified DSLR and jumped to a cooled osc due to the thermal noise. I now use the L-Extreme. What a difference! Thank you, sir.
Great to hear! My pleasure, as always :)
@@CuivTheLazyGeek As time flies now using the L-Ultimate. Be aware readers that this too struggles with halos on bright stars.
I think using these on a unmodded DSLR is kind of a waste- The blocking filter on the camera blocks out most of what you're paying for the filter to pass. Great video and great to see you feeling a little better. Welcome back!
That is of course very true - I think for narrowband, most understand that the camera needs to be modded (or something like the EOS Ra)
Yes, you can't deceive the laws of physics, but knowing them, you can better understand how to use them to achieve your goal. My first Canon 550D, I myself am a lover of the Sony system, I turned into a Frankenstein) He could sing and dance and even cool down at a delta of -26 °. The reason was simple, there was no money for an astro camera. However, it was then that I realized all these relationships when using narrow band filters. Now my astro-Canon is in my "museum", I do not sell it, as a memory of where it all began. And comparing his raw footage today with what I'm getting now from Toptek 571, I don't always see a huge difference. Cooling is a big advantage for digital cameras. Not counting, of course, high-aperture optics.
Clear explanation, it makes sense. I believe I am better prepared to shoot Astro photo.
CTLG nails it again. Many thanks Pal
Awesome content and approach Cuiv! Great videos - keep up the great work!
This has saved me lots of effort. Thank you!
You're welcome!
Thank you - hugely interesting explanation.
Glad it was helpful!
Great explanation Cuiv, the exposure time is always an issue
Another great video with useful info. Content is key and you just opened a few doors for me. Thanks again Cuiv. Keep them coming. HJ
My pleasure HJ!
Excellent explanation Cuiv! 👍👍
Cuiv going crazy with the videos :D
I've been in a recording and editing frenzy, there's more coming down the pipeline :)
Excellent video!! I’m a beginner (& starting to get some decent results with my DSLR) and this was easy to understand, along with being very informative. Thanks!
Excellent to hear it was helpful! Thanks for the feedback!
Hello Cuiv,
I l
Really love your video's. Have seen a lot of them and learnt a lot.
In the meantime, Optolong has the L-Ultimate out. What about this one for OSC camera's ?
In theory it's excellent! I've only measured a couple in my spectrometer, one was good (up to spec) the other terrible though... So in theory excellent, in practice lottery
Cuiv, I know you are on a hiatus right now and will be back in a few months and all, however, perhaps it would be funny to do a very short video in french when you are back! Just short 1 minute stuff. Just an idea
Well, not completely the same. There's about a 3% loss to the filtering material and another 5% loss to the glass plate it is on... so you ARE losing about 9% of the narrowbandpass you want as opposed to not using a filter at all, but in the end its usually less total loss than using a broad band-block like the pollution filters. Additionally if you're going for deep galaxies the 5 to 7nm narrow filters are a better option.
Using an actual astrograph instead of an "astrophoto telescope" will make a much greater difference in your light gain overall but not many people are willing to do the tiling and knitting that RCs need for broad objects.
This is the perfect explanation of what I asked a few days ago. Thanks so much, Cuiv! One more thing: what is the real relationship between band pass and telescope speed, i.e. focal ratio? Why are there filters designed for f/2 telescopes for example?
The issue is with the angle of the incoming light. Narrowband filters are interference filters, which work be causing light other than the band you want to pass to be removed by interference. These are very sensitive to the angle of the incoming light, as you change the angle you shift the wavelength which gets transmitted. Faster focal ratios have light coming in to the sensor at a wide angle, this means that more of the light sees the filter as shifted and so the light you do want ends up getting mostly blocked.
What Ben said - at fast focal ratios, the filters will start becoming opaque wrt angled incoming light rays. Effectively this has the effect of diminishing the telescope aperture (and thus increasing the focal ratio artificially) for the useful signal. You'll still get images (and good images at that), but effectively you're not using the full potential of the fast scope.
@@CuivTheLazyGeek and Ben, thanks! Clear skies!!
Excellent lesson! I've been schooled. :)
Thanks for making this video! This was explained very well! Appreciate it! 😃
Glad you enjoyed it! Hopefully you can test the theory at some point :)
@@CuivTheLazyGeek I have to now, you just single handedly invalidated the very first video of my astrophtography CZcams channel 😅😅😅😃.
But really though, this is great information and I am glad you can help steer me, and others in the right direction. I will test this theory soon enough.
Once again, great video! 👍👍
This interesting video is focused on DSLRs, right? It doesn't apply to dedicated cooled astronomical cameras, correct? Thanks!
Glad you're back!
Generally the exposure time graph shows 3 minutes is optimal with diminished returns after. Canon 6d has such low read noise, my sub tropical climate shows its fine even in the heat.
Thank you!!
Makes perfect sense! Good advice, thanks
Very interesting. Thanks.
Exactly so and we’ll put! Thanks!
Thank you!
Newbeone here. Still learning, no HW yet. Is exposure time to capture nebulae based on equipment you are using, trial and error, or are there some basic guidelines to start out with? My current HW plan is edgehd, AM5 and a 2600mc pro(maybe the duo).
I never dial up the ISO. It heats the camera. It does not increase the number of photons collected. I leave my ISO at the values where read noise, DR are optimized. I can increase the gain in post.
Thank you for a great explanation. It makes sense and is very useful info.
My pleasure, glad it helped!
Hi! This might be a dumb question after what you said, but because of the response i've gotten from asking store owners about filters i still wanted to ask.
I use a star tracker with an HA-modded Nikon dslr and lenses between 50mm to 300mm and f1,8 to f8. I do this unguided and get between 1-5min exposure time. I want a filter to be able to do astrophotography i full moonlight and in summertime when it doesn't get as dark at night. Because what you say makes sense to me i believe that i could go with a narrow bandpass filter like, for example, the L-extreme for this, but several store owners have said it's not a good idea. Do you think i should go for 7nm, or would you suggest a wider bandpass in my case?
I should add that i want to mount the filter in front of my lenses as that makes most sense to me with camera lenses. If you know a filter that makes that possible with normal camera step down rings i would love suggestions as this is not an easy topic to get clear answers on either!
Top video…👍💪pls make a video with Seestar s50 and Lenses… what’s there the difference? Ect…😊 thx
This video is gold! 🥇
Thank you!
Excellent, clear explanations as always ! In the video you mentioned that narrowband filters are not a good idea for OSC cameras - why is this ? Thanks
Let's say single band filters are not super efficient with an OSC camera. So an Ha filter would only make the red pixels react, so only 1/4th of the surface area. Still, it can work fine if combined with RGB data from the same camera.
Nice video, great explanation. Question or tips. I have a 450D canon modified myself (remove IR\UV filter). I am using 12nm halpha filter. I notice that I have a lot of bandings (horizontal) in my raw images @ 400mm (refractor). Compare to my camera lens 50mm I don't see bandings. Is there a way to avoid the bandings? Why do I see them @ 400mm but not @ 50mm? Maybe it's still there ? Thanks!
Actually the person who teaches me astrophotography is wrong. I was wondering why I have to exposed longer for Ha. It totally didn’t make sense for me. Thanks for the clarification
He's likely thinking about swamping the read noise, which is indeed desirable - but if it's going to make things worse, then it's fine to exposure less long!
I think that this video, even beeing more suitable for DSLR users, adresses a little my question about the L-Extreme and the ASI533MC. I've used that filter on M31 and took about 50 120s frames (out of 75 due to a wall on my line of sight 😭) and I'll now combine those with a bunch of 120s or 180s frames to be taken with the Baader Neodymium (I live under a Bortle 7/8).
It’s usually not recommended to use any filters for galaxies. Especially not narrowband unless you actually want to capture the faint H-alpha signals from that galaxy. This is because most galaxies emit their light in broadband and any filter will reduce the signal that your camera receives from the galaxies
@@ripwords I'm aware of that. This was an experiment since I'm starting in the hobbie, I've done the NGC7000 with this filter (you can see the Picture @astrobin) and wanted to see the result on galaxies and M31 was the obviuos target 😁😇). I'm at this time taking 120s frames with the Baader Neodymium light pollution filter since I'm imaging from a Bortle 7/8 zone and later I'll combine with the previous taken with the L-EXTREME and see the result
Using the L-eXtreme as a booster of nebulae within M31 will likely work - after all there are multiple methods to combine Ha and LRGB for galaxies, to stress out their nebulae. I've never done that myself though!
Thank you, Prof. Cuiv ;-) Very Interesting 👍
My pleasure - this will be in the finals though so make sure to learn it well :p
Thanks for the video. I'm new to narrowband filters so found the video interesting but left me a bit confused. Perhaps it would be helpful to share specific examples of solutions on what settings and exposure times would be good? In my case I'm getting an h alpha modded canon EOS R and a drop in narrowband L extreme filter.
I am using these techniques with my DSLR. Hopefully, Santa Clause will bring me a dedicated cooled camera for next summer! Until then I’m just trying to improve my skills.
Let's all root for Santa to get you that cooled camera :)
Thanks Cuiv, great video! I was working my way to your conclusion but you got me over the line! I'm a DSLR user and just starting out - I wonder if there is an issue using the very narrow bandpass filters with manual focusing with a Bahtinov mask?
You really want a bright star to use a Bahtinov mask with a very narrowband filter and a DSLR - but it should still then be visible in live view. Otherwise you'll need to take short exposures and adjust in between!
Nice, thx for this explication!
My pleasure!
Hello - when I use my DSLR , I was under the impression that I need to have my histogram 1/3 of the way for the peak value - otherwise , the image is underexposed . I have been following this rule of thumb
When I use my multi band pass filter , The histogram is too far to the left , so I increase my exposure time to get it back to 1/3 .
So are you saying that I don’t need to increase exposure time ? If the histogram peak is all the way to left , that is ok ?
I feel like it will be too dark .
A 30 second exposure with and without a filter is not the same on the histogram
Can you please elaborate on this ?
Thank you - and thank you for all the great content !
Regards
In deep sky astrophotography, you want to expose to the left rather than to the right. The only thing you need is to have the histogram "spike" separated from the left edge of the histogram. That should guarantee you are swamping the read noise, which is what we need. Since after stacking, you can stretch the image, you don't need to worry that it appears too dark with individual sub frames.
You can expose longer of course but then you're likely saturating the stars.
@@CuivTheLazyGeek thank you for your reply .
So is it better to have the peak on the histogram as close to the left as possible ? As long as the peak is to the right of the left edge ?
could you review some of the best astrophotography cameras for low budget and the enthusiast? - thanks
I'd like to try - but I'd need to get my hands on those, that's a big project. I have something like that on the back burner
Hi Cuiv, I'm a lazier amateur astrophotographer here and following your channel since your AZ-GTI tuning video :) , what are your thoughts of using narrowband and an unmodified regular camera? I am planning to use the Optolong L-enhance with my 61mp Sony Camera, am I to expect terrible results with a regular camera and narrowband filters?
Mmmmmh, well you would get a result, and you'd likely get a better result than without in a light polluted zone. H-Alpha is so strong on so many targets that it could gather enough SNR, even when cut down by 80% by the sensor filter. I'd say give it a try - but be prepared to want a modded camera after you try! (a lot of cheap, old Canon modded cameras floating around)
@@CuivTheLazyGeek thanks for sharing your thoughts on this, I didn't know that there is that much reduction from the sensor filter. I have an Astronomik clip in CLS filter for my camera and there is a lot of colour shift to it even if I set the grey point in white balance that is why I'm thinking of trying narrow band filters, I guess its better to move to a less light polluted area in my case 😆
Can we assume that the 18nm vs the 12nm vs the 3nm filters all have the same transmission value?
Correct, this is an assumption made, which is usually fairly reasonable.
Thanks for the explanation.. but what exactly does "dithering" do?
Dithering changes the framing vey slightly in between images. This helps get much better images. See details here: czcams.com/video/Da33jfKJl1M/video.html
@@CuivTheLazyGeek Thank you!!! I now completely understand.. you've taught me quite I bit with your videos..
I must say, I am starting new again in AP after 16 years of downtime and now having to learn new programs and having to replace my 300D and Toucam Pro II with a 450D and zwo426 for guiding and planetary imaging.. (still waiting on the zwo, tho).. Thank you again!
What about with the HCG-mode of the 2600, 533 cameras with ZWO saying we should either shoot at 0 gain or 100 gain, 100 gain being the HCG-mode. Would really love to pick your brain on something Cuiv maybe I will try an DM you on discord
From what I can tell, there is basically no drawback to using 100 gain/HCG mode - you get much better read noise while basically retaining dynamic range
The OPT affiliate link takes me to spammy stuff. Is there a typo? If i copy the text of the link it works all right, but if i click on it i get redirected to stuff that wants to show me notifications and install software, and never arrive at OPT.
It may be that a Chrome extension is interfering, as this does not happen in incognito mode. Still trying to track it down.
There seems to be an extra u+200b character (zero-width space) at the end of the link. This causes a redirect to a malicious site. Not clear to me why incognito prevents this, but i see the extra Unicode character in the output from wget, so it's present even if a browser is not involved, so not the result of any extension.
That's super weird, thanks for letting me know! Yeah there seems to be an invisible character at the end of the link for some reason.... I've removed it for this video and will check on others too. Thank you!
@@CuivTheLazyGeek Thanks! With the character removed i am no longer seeing the bad behavior.
@@TheAntibozo Thanks again for letting me know!
I have a hyperstar for my C6 scope. What do you recommend me for my monochrome camera, filters 2'" vs 1.25". .?
Mmmmh it depends on the camera sensor. But at F2, if larger than M43 sensor, then 2 inch would likely be preferable to avoid vignetting
I have a full spectrum modified 60D and find when I shoot everything is very red. I know I can process much of this out… but… I think I am missing a filter… what clip in filter do you recommend for someone just starting out? I will be attaching my camera to a FLT91 (when ever it arrives) 😊
I think Optolong has most of their filters available as clip in. Depending on your targets I'd recommend the L-Pro or the L-eXtreme, but it is really up to you! Also I'd probably go with a filter drawer rather than clip-in simply because it's then easier to switch! That's a nice scope you have on the way :)
@@CuivTheLazyGeek Is stacking filters a thing? Is it something I should account for when purchasing filters?
Interesting topic! I have the f/2 3.5nm Baader narrowband pass filters for Rasa. I am wondering if they can be used for normal scopes > f/4 and if they are better than using my older 12nm narrowband filters.
They won't work on normal scopes - their bandpass is pre-shifted and transmission will be close to 0 on a F4+ scope.
That said, I've just had my OIII f2 4nm Baader NB filter (CMOS-Optimized) measured via spectrometers. It is pre-shifted too much, and barely passes any OIII at f2.... so yeah, I'm not having a good experience with Baader.
@@CuivTheLazyGeek Wow. Thanks for the information. I know the previous Baader f2 6.5nm ones have halos, so they should have already fixed the 4nm filters in all departments. This could be controversial. Will you be releasing any video on this soon?
Hi Cuiv. A question regarding NB imaging. I'm starting working with monocrom from a light polluted area (Bortle 7/8) and I'll try the Horse Head Nebula as HaRGB. The question is shall I use short exposures in RGB and larger ones in Ha or can I go to longer ones in RGB. From your experience with light pollution what will be the best approach?
So the best approach is to swamp the read noise, while not saturating the sensor. For the RGB exposures, the latter will happen very quickly. In a Bortle 7/8 I don't think I'd go beyond 60s exposures for RGB (assuming F4-F5 telescope or so), and for Ha I would probably go for 180s (7nm filter) or 360s (3nm filter). So long exposures for Ha, short exposures for RGB!
@@CuivTheLazyGeek Thanks for the help. I'll try it as soon as the clouds get off my way. And yes the imaging train will be an Evostar72ED F5.8 with the ZWO 7nm Ha along with the ASI1600MM
👍🏻
Maybe I need to come back with a fresher head but I didn't really pick up a whole lot of what was said. Just went in through one ear and out the other. Maybe someone can explain it to me in an even more simple way.
To sum it up: Narrower bandpass filters will let the same amount of light of the target wavelength through, it will only decrease the collected light of the rest of the spectrum (which is what we want in the first place).
Or, to use a simily: if one (the light) walks on a specific line through an open door, it doesn't matter if it is a regular door or a barn door, one will make it through just as fine (unless it's yo' momma, 'cause she's so ... ), but the barn door will let through all kinds of stuff you don't want (possibly even yo' momma). ;)
Peson summarized it pretty well there - hopefully this helped!
@@oscpe262 No, I know how narrowband works it was a lot of what was send in the video that I didn't understand. I can't remember what it was that I didn't understand right now as it was almost a whole day ago.
@@Kyle_Hubbardsounds like a you problem.