Note: Bias/dark flat frames can help reduce the extra noise added from short sub length exposures, I will be exploring this in another video.

This topic is not talked about enough. Great that you are tackling it. Many will benefit.

This may be the best and simplest explanation of this issue I have come across. Excellent job!!

One of the best explanation I have come across on the net. Period!

I think, the less the time you expose, the more frames you acquire AND the most well-guided images will be taken. At the same time, you ensure your guiding will continue to be as desired, and the focus, as well, so lees frames to be thrown away. Flat fields are an absolute must (fortunately with NINA it seems to be relatively easy, as it calculates the ideal time per filter depending on the minimum and maximum time per filter you set).

Absolutely fantastically explained video. You didn't leave anything out, something that's been missing from other videos I've seen. Loved the formulas and charts, and also liked the fact that you did a real test between two methods. The level of detail in this video made me sub instantly. Keep up the awesome content!

All people confused total exposure time with integration time. A 30 second exposure remains 30 seconds even with 100 30 second frames. 3000 seconds is just integration, not exposure. Exposure collects light, integration dusts away noise, improving signal-to-noise ratio, but does not increase exposure. I normally write like this: "photo x, exposure time 120 seconds, total integration time 3 hours". The longer an exposure of a single light is, the more light enters the sensor, as long as light pollution allows it. Integration lets you discover wonders that you didn't believe that your sensor was able to collect with such a short exposure. But you don't collect more light by integrating. This concept is so important as it is greatly misunderstood.

Short sub minimum requirements = $300 uncooled camera, $400 budget mount, $400 small APO $100 flattener Long sub minimum requirement = $900 cooled camera ($1400 mono), $1200 mount, $400 small APO $100 flattener Unless you are VERY dedicated to the hobby or are willing to spend 3x as much for slightly better images,go for the short sub exposures. EAA is FUN and easy, astrophotography is challenging

Might be a good point to also include the difference in file size, 30 500s subs is not nearly as expensive as 500 30s subs when each frame is 50 to 100Mb each… All of the sudden you need 50 terabytes of storage, 256 gig of ram and 4 $2000 processors… And your image processing machine set you back $37,000 lol…

Excellent video. 2 years old but still the easiest explanation I was able to find. I see a lot of people talking about gain/ISO, and that that should be part of the explanation. I disagree. First you figure out your exposure time, and then pick a gain to make sure you aren't clipping or over saturated. I come from the planetary imaging world where everything was super simple. Make your exposure time as short as you can (because that is frame rate in planetary imaging). We are trying to beat seeing, so ms level images. Then adjust your gain so you can see the image. That's why planetary cameras have their optimal gain up at 180-200. Hot pixcels be damned. We don't care!

Great video. Giving us something to work with is awesome. I have the 294 mm and mc. And going to a star party in Northern California end of June 2022. This is one of those videos I’ll watch a few times.

With faint objects i have run into dynamic range problems using short subs. If your Nebula just fills the first 100 buckets of your camera sensor well you will have to live with these 100 brightness values across the entire nebula. When stretching aggressively you will see banding. Long exposures avoid that by using more of the dynamic range of the sensor.

To be honest, When I imaged the orion nebula, a 60 second exposure was way more vibrant than a 120 second exposure although the 120 secs was exposed for 30 mins, and the 60 mins for just 10 mins. For Bright Objects, short exposures are perfect. But when I went for the sculptor galaxy, a 60 second exposure absolutely failed to bring out much detail except for a bit at the centre. A 120 sec was much better and brighter, but still could not detail the outer edges of the galaxy. That is when you realise a 300 -600 sec exposure is perfect for dim objects.

Excellent video and thank you, was glad you mentioned Dr. Robin Glover. He has a great video that is very specific to exposure times. I have had good success with exposing at AVG ADU 1200 above the bias level on my 294MCPro. The exception is when I use my L-extreme filter as I am not able to achieve that level at a camera GAIN 125. I'm in a Bortel 5/6 zone and do not go over 3 min subs.

Well my exposures are usually 15-30 minutes, depending on object brightness, shortest i have done was 30 seconds on Cat eye nebula to get details in the middle :) Longest i went was 60 minutes for distant galaxy cluster :)

Very insightful and useful advice, many thanks! If I may please be pedantic with one small detail regarding the noise (R) factor in the minimum sub length formula (C times R squared over P): "[...] get the lowest noise that you can, because, as it gets higher, the squaring of it will make it exponentially worse, and you will want this to be as low as possible." I think you meant to say "quadratically worse", instead of "exponentially worse" ;-)

The big thing here is that many mounts will give a high yield of 30-second exposures even if they can't track a 3-minute exposure perfectly at all. I strongly recommend using flats and flat darks. (Flats cannot be applied correctly without flat darks, required for the calculation.)

I’d rather use a lower exposure and get a more accurate detailed shot, than potentially blowing out my shot with an exposure I had no business taking. The hobby isn’t going anywhere. Take your time and enjoy your time under the stars.

All in all, exposures depend on your bortel zone, your target, and most importantly your scopes aperture and your cameras pixel size, crop factor, and so on. There is no perfect exposure time as there is no perfect rig for DSO, planetary and so on. Astronomy like other fields has many rabit holes to go down. Clear skies

I live in a Bortle 7/8 area and use a Optolong L-Ultimate 2" Light Pollution Dual Band Filter. How would the color camera (ASI2600MC) exposure numbers change? Thank you and thanks for sharing this video.

Interesting comparison, something I just started thinking about as I just got a tracking mount. But I think you got something wrong about why the long exposure image is less noisy: In general MORE IMAGES = LESS NOISE! Because the stacking software doesn't ADD, it AVERAGES the images. So the random noise in every sub exposure will be averaged in the stacked image and the signal to noise ratio improves. Otherwise there wouldn't be really a benefit of stacking images, except the clouds/shake problem you talked about. In addition, the noise in the short exposure time subs to me looks more like dark current/fixed pattern noise, what should be cleaned out by using darks (but you did)... maybe something went wrong there? I don't know the settings of both exposure times, but I would assume, that differences in photon noise had a great impact on your images. For people who don't know: There are several types of noise in your image, for example "read noise" (a.k.a. downstream noise, caused by the electronics of the camera after the gain/iso is applied to the signal before it reaches the ADC) and photon noise (caused by the effect that not every pixel was hit by the same amount of photons, so the pixel voltages varies to each other) and some more. Read Noise: The higher gain/iso you choose, the more the signal gets amplified before the read noise is applied -> less noise in the image. Photon Noise: The longer the exposure time, the less impact has a varying photon count of each pixel -> less noise in the image. So if your gain setting was the same on both exposure times I would guess it is mainly photon noise and bad SNR (more visible read noise) you see there. If gain was adjusted for the different exposure times, may be your darks calibration has something to do with that...(?) So for some reason your longer exposed subs are cleaner. I am not saying your test result is wrong/bad but your conclusion why the test result is like it is might be off. BUT for sure it is NOT less noise because you stacked less images! It's the opposite! Thanks for your video though! Cheers

This has explained so much, thanks very much 👍

Yep, i guessed right but it was easy, you can clearly see you have pulled out more of the fainter details with the longer exposure.

Very inter and good for beginners like me

I live in a Bortle 4 area. 180sec subs is my usual go to (with a OSC and pretty accurately tracking NEQ6 mount). Might reduce that to 120sec if I'm shooting something like Andromeda or Orion, or if the moon is above 50%. Might increase it to up to 300sec if I'm shooting a distance nebula with the L-Enhance filter. But I'm also a lazy astrophotographer and don't shoot bias or flats and also tend to borrow darks I've shot on previous nights under the same conditions.

I greatly appreciate your explanation on this topics. Now I understand why my shorter exposures tend to be better than longer ones.

Getting into mono and I've been taking notes using your videos. Super super helpful info! Thanks!

Great explanation! Really helpful! Thank you so much.

Theoretically lucky imaging is the way to go if your telescope is sharp enough, even if it takes longer to image the same thing. On long exposures the best result you can achieve is the average atmospheric conditions which usually are way under the telescope’s potential. By taking loads of say 2/5 seconds exposures you can just take the perfectly sharp ones and avoid the ones blurred by instabilities in the atmosphere.

I miss the times I use to chill film emulsions with dry ice to reduce the effects of reciprocity failure during hour long exposures. Then all of the time spent in a dark room processing the plates.; and praying that my manual guiding was good enough for a solid image.

Ioptron Skyguider Pro. Guidestar61 + field flattener. Stock Canon T7. =2 minutes subs. My best night of astrophotography ever!

Thank you for your very interesting talk, like the in depht explanation of exposure times etc. Wim Nelissen, The Netherlands

Really useful video for someone of limited knowledge like myself..good stuff man 👍🏻

The shutter speed is increasingly dependent not on the illumination of the sky, but on the number of Starlinks))😆

Overexposing subs is also a consideration!

That is super helpful RoRo. I am just getting going with Deep space astro. I have an ESPIRT 100ED/EQ6 mount +1600MM with all the bells and whistles, controlled with a stellarmate on rasbery pi. Got it all working for the first time the other week (after some very frustrating nights of getting nowhere) and got a shot of M42. Luminence turned out well (1 minute subs, 1 hour luminence/1 hour RGB) although it was a bit off centre and the bright core was a bit saturated(gain was a bit too high I think). My RGB channels were out of focus (I thought they were supposed to be par focal). This is the next stop for me, to optimise the exposure settings! Ive seen Robin's talk on youtube. It really cuts into the science of CMOS astro imaging. I come from a science and engineering background and even I will admit that it's not an easy topic. I think tutorials like yours are super helpful. You need a primer to begin with and then depending on personal interest, people can dig into the detail..... I am gona be compiling notes based on tutorials like yours and the work Robin and others have done ( whatever I can get my hands on ).... As is the case with astro, it an be days or weeks between imaging sessions ans so its easy to forget some of the fubdamental principles.

Thanks for this very interresting video! It is very clear.

Absolutely amazing video! So well explained and easy to understand and follow.

Very well made video, this cleared up a lot of my questions regarding this topic. Thank you for the info!

Great Video! Only one I could find that cleared up all my questions. Good job!

I just switched over from DSLR to TEC imaging and this video presents another piece of information for me to assimilate into my approach. With DSLRs I was pretty happy if I got 180s subs from my Bortle 8 observatory. With my ZWO ASI533MC Pro, 3 minute subs seem to leave me with bloated stars that are tougher to manage in PS. After viewing this video I will be working with much shorter subs on my 80mm f/8 refractor and LP filter, More like 60s to 90s or so. Great info to have for consideration. Thanks!

According to the rule of 500, my 254mm reflector on a Dobsonian with a focal length of 1270 and an f5 has a maximum shutter speed of .39 seconds, before star trailing becomes and issue. Sad.

This is great info. But for me, IMO LOL, the right exp time depends on how long my mount will allow me to keep the shutter open before my stars turn into eggs. Provided of course my histogram is not clipping or I am not blowing out the stars. MORE DATA! MORE!!!!!!!!

You also need to factor in overheads such as download time, at very short exposures these can be considerable

Very helpful; you have a really great channel!

Short exposures are good, just need more subs to increase signal. Also depending on wethor or not you are using an un-cooled camera, Long exposures past 5 minutes don't look to great because of the sensors heat, yes dark frames can help, but temperature isn't precise and connont be regulated easily with dslr's and so on.

So the 15 seconds is like lucky imaging then? It seems like it’s longer than like the lucky planetary imaging.

I think you have missed the key on sub exposure, which is brightness of the object. You definitely won’t expose the same time for m42 and sh2-240.

I can find planets as well as many other deep space objects through various viewing lenses. When I take out the viewing lens and insert the camera I am clueless as to how to magnify the image so I can see it on the computer and phone APP's. The moon is easy enough to see and photo, but I can never see Jupiter or get it to come into focus. How is this done using cameras similar to the ASI 120MC-S model? I am very new at this!!

best explanation I've ever seen

Great info. It helped to resolve a number of questions that I had. I just started with this hobby about 9 to 12 months ago, and I've struggle with determining my ISO and exposure times. Maybe a short comparison might be an interesting topic, or not.

If I take shorter exposures the histogram looks very low and not the normal third across the screen. Would this hurt?

I suppose one can just begin exposing 30 seconds, 60, 90, 120.. and observe when the background greyness peak begins to detach noticeably from the black end of histogram..

Spot on all those details. Especially failure rates. I often vary exposures on a single object from 1 to 3 minutes (DSLR) during an imaging session (clouds, unexpected lights, etc). DSS stacks those without a problem but I'm unsure how that works in other stackers?

Thanks for the well presented, balanced and objective information. In evaluating these types of videos, I honestly believe most of these crazy long exposure times, and, crazy amounts of hundreds or even thousands of exposures, is unnecessary overkill. And I think it is also at times the idea of a "look what I can do", attitude. So, I think I am more of a shorter exposure guy! ha! ha! But, each to their own. So, if 10 minute exposures makes someone happy, or if 1000 exposures is your bag, by all means, go for it! 😁

Great video, I have been trying to wrap my head around calculating ideal exposure times for my setup and until now I've been eyeballing it. I've been severely undershooting narrowband at 3-5 minutes and overshooting my LRGB at 2-3 minutes. With my location and my 183MM Pro @ F4.5 I supposedly should shoot for 14+ minute narrowband subs, 45 second RGB, and 15 second luminance. I think I'll shoot a bit longer LRGB than that though, while cutting back the narrowband. In any case, now I have some ballpark numbers to work with and hopefully this should improve my images!

G'day mate, nice video with good information. There is one thing missing however, something important. It was good of you to disclose that you have based this whole video on somebody else's hard work, but you really should mention Dr. Robin Glover by name & give him the credit he is due. And it's not too late to pin it up in a comment.

Interesting, the light pollution map says my estimated SQM is 21.52, which equals to bortle 4 on that site. But on sharpcap tools it equals to a bortle 2.4? That can't be quite right?

Excellent explanation.

Excellent presentation and explanation! Much appreciated. I just became your subscriber.

Thank you so much for making this confusing topic (for beginners) more understandable. Much appreciated!!!! Wondering how a light pollution, broad ban filter would alter the cheat-sheet exposure times (Bortle 7-8)?

Thanks so much for the vid. I'm still in the DSLR arena and the principals of gear, conditions & experience are equally applicable. What changes perhaps is the recommended exposure times .... yes? Do you have a vid or other source you can point me to regarding this? Cheers :)

Ok, this is a great video and shedded some light on my process. Here is what I have: RedCat 51 ASI 533mc 30mm guidescope/ASI 290mm mini EQ6 Pro L-Extreme/L-Pro Filters Bortle 7 skies I have been shooting at a gain 100 x 300 second exposures. Are my exposures too long? Should i be following this 180 second threshold in my suburban area?

I think that is quite a helpful video - many thanks. I have multiple scopes in a small 2.3 metre astro dome I control remotely from my house. I most commonly image nebular using a Williams Optics FLT 110 pared to a ZWO ASI071MC-cool Pro camera and a 235mm Celestron Carbon Fibre F10 SCT connected to a ZWO ASI1600MM-c. I mount this and a Williams optics 98mm Doublet (the 235mm and 110 side by side, with the 98mm piggybacked on the 110) on an old Vixen Atlux mount controlled by a Vixen SkySensor2000-pc. I guide off axis from the SCT or on axis from the 98mm using a ZWO 120MM-s or QHY 5IL cameras respectively. With guiding I can go out to 30 minute subs and still get round stars - (but super saturated frames) So I know the tracking and guiding is really decent. I didn't know there was a light pollution map (big thanks you for that) that tells me I am in Bortle class 6 skies. Typically since moving from modded Canon DSLRs to ZWO cameras I have shortened my subs to 2-6 minutes - depending on how faint my target is. This has been all trial and error - your video gave me some science to tune into my experimentation - another big thank you. Lastly I haven't settle on single processing software yet. I typically transfer my images from my astrolab PC to a 8 physical core CPU with 64 GB RAM and 2 NVME drive and 4 SSDs to to stacking and processing. But I have only played with Deep Sky Stacker and an old Photoshop CS4 software to process my images. From your video I am sensing I should look at one of the big three astro image processing software suites - are you totally happy with PixelInsight or do you think go simpler? Many thanks Matthew

Excellent many thanks

That was effing brilliant! This is using math to substantially reduce the impact of light pollution. I'd love a QHY600, but I think reality will constrain me to a 268! Add in quality NB filters (i.e.

i have been wondering this for so long, but nobody has ever really done it in this much detail, i dont have any experience with photographing stars, but i really want to try it, although the only DSLR i actually have is an old pentax K-5, which is i think a decade old now, or maybe even older

That means i dont need star tracker if i stack multiple short exposure images?

Hello. So 12:03 let's pretend that there are no unwanted artifacts in that single exposure photo. Would it be fair to assume it would yield the best result?

0:13 Why would I take 60 or 240 photos with the same exposure time? That is stupid. If I didn't know what was the best, I would take one half hour photo, one 15 minute photo, one 7 minute photo, one 3 minute photo etc, until I have used up one hour. Then one of them will probably be good.

Well done! Thank you!

Very good video on this topic! :) I would just add that the lense f/ratio will also have an influence on the optimal sub exposure time.

Id love if you'd do a video about how to find the best gain to set your camera at in some detail like you did for this video

Thanks for some very helpful information. I usually go for 5 minute subs on clear nights, mostly to reduce disk storage requirements and computational power for stacking. However, when there are some fast moving clouds going through the sky, going to a shorter exposure subs would allow some data collection. My experience has been that at 5 minutes almost every exposure catches a cloud. It's good to know what my options are, and that I could go much lower and still have very acceptable results.

It would be really helpful if you could provide sub lengths for different setups representing beginner to pro. Starting with Sony A1 + 200-600 lens under a borle 4 sky.

Hi. I would love to hear your thoughts on varying the gain rather than exposure time when capturing RGB stars for a narrow band image. I have an ASI2600mm and am just getting a feel for it. I have clearly been overexposing my stars and the colours have suffered but I wonder if a lower gain would improve the colour quality more than shortening the exposure.

SImple - Tracking and Histogram, to start with, ISO, then light pollution... iteratively; that's how I see it as a process. Thank you.

Great! Best wishes from Norway

How about taking milkyway images?

Recently came across your channel, subbed! Thanks for the detailed explanation. When these issues crop up, though, there is always something that is not entirely clear. When you give the sub times for narrow band filters, I am assuming you are basing that on a mono sensor. However, there are many of us who shoot OSC with duo band filters (take the L-eXtreme, for example, which is 7nm in both Ha and OIII). If the calculated subs are for a mono sensor with a 7 nm filter, shooting the same 7nm filter with an OSC adds the extra filtration of the Bayer matrix, so how does that affect exposure calculation? Multiply (exposure time) by what factor? 3? Thanks!!

Good explanation

Well that was thorough. Thanks.

Buy 3nm narrowband filters and you limit is only how long you can stay in focus. For me 3,600 seconds. MUST GET YOUR POLAR ALIGNMENT WITHIN 30".

This is an easy question. 30 seconds or less because I have a Sony a6000 and bulb mode eats stars for lunch.

very clear. thx

What about DSLR photography?

Excellent explanation - thank you VERY much! Couple of quick questions... 1. You mention this is based on dedicated AP camera, are the same values for exposure time by sky conditions the same for astro-modified DSLR? 2. Is the brightness level of the target a factor in any way?

Yes, - but how do I turn: "Sky Electron Rate 1.31 e/pixel/s" Into seconds of exposure??

Thank you for your informations !

Hello Rowan, just landed here from nowhere and this was the explanation video I've been looking for in the last 2 months! Absolutely clearly explained, thank you so much. I think there is a last point it could be interesting to deepen. Given the pro and contra of 60 15 seconds and 5 3 minutes exposures, in order to achieve the 5 3 minutes SNR, how many additional 15 seconds exposures would be required?

Do we also need to consider dynamic range when thinking about exposure length? Presumably we would want to have the brightest parts of each sun frame to be almost filling the full pixel depth, so maximising the dynamic range? This would then count against shorter exposures

Great vid

What about 1 trillion of 1/4000s sub exposure? Can you imagine how crispy details

How can you manage dithering with short exposure lengths? Wouldn't you be loosing a lot of time in settling between the subs compared to longer exposure times?

Really interesting video! I saw that you recently got the 294mm pro and I also saw a red cat 51 on the mount. I am really interested in the deep sky performance of this camera with a wide field refractor, like the redcat, and especially how the 1x1 binning mode compares to 2x2 binning mode for deep sky imaging since there are some caveats with the 1x1 binning mode.

Excellent video!! Congratulations! One conclusion about the video is that if we move from a zone of bortle 9 to one of bortle 2, the minimum subexp is prolonged. Especially if narrow band filters are used. In my case (ASI 294MC-Pro), using a Ha or SII filter in bortle 2 gives me a calculation of more than 600 seconds of minimum exposure time, making it very difficult to guide to reach those times. My question is: Is it correct to increase the gain (without losing so much dynamic range) to reduce the minimum exposure time in very little polluted skies?

I think there's something wrong with the very last calculation. It does not make sense that a Luminance frame has a longer minimum exposure time than RGB where only ~1/3 of the light is coming through the filter.

Why is there no effect on exposure time to compensate for the f/stop of the lens/telescope, seems like a f/2.8 lens lets in 8 time more light then f/8 lens and would effect exposure time like it does on a DSLR?

So, sharpcap has two tools, sensor analysis, and the brain feature found in the histogram tool. How do you correlate your formula with the results in this tool. I’m not as far as I would like with using it because along the process I discovered that my newt has light leaks. And winter has moved in with wind, rain, snow. But I’ll hopefully get back to it soon.

I live close to an airport and I have gone from 30 minutes down to 15, down to 10 and now mostly 5 minutes exposures when doing narrowband imaging using my mono CCD's due to the loss of frames thanks to planes.. also nowadays I just set up behind my house (bortle 8) instead of driving 20 minutes for bortle 6 (lower takes too long to drive) and 300sec seems to be fine for me.. my mount handled the 30 minutes just fine, probably more, but I don't see too much difference between three 5 minutes vs one 15 minutes exposures.. too short exposures like 15 to 30 sec. would waste too much time for download between shots, as I am relatively limited how long i can image a night (trees and my house in the way)