Are longer Exposures really better for Astrophotography?
Vložit
- čas přidán 5. 09. 2024
- Welcome to the new video!
Short Info: I do know that the exposure length depends on a lot of factors like read noise, sky brightness, ADU and FWC/focal ratio, but that is beyond this video. I'm also aware that there are a couple of exceptions to this like short exposure lucky imaging...But this is just about the impact longer exposures have on SNR! :)
Thank you for watching and clear skies!
==============================================
You want to support my work? Then please, go check out my Patreon!
/ crazedconceptions
If you wish to contact me, you can always reach me on Instagram!
/ crazed_conceptions
Or alternatively, send me an E-Mail!
crazedconceptions@gmail.com
Intro Music: Away by Patrick Patrikios
==============================================
I've done a lot of experimenting with exposure times and I can honestly say that I have found that the benefits of shorter expsosures massively outweigh longer exposures.
Of course - obvioulsy - if we're talking about *one* exposure, a longer exposure is better than a shorter one.
But if we're talking about whether it's better to take 60 x 1 minute exposures or 15 x 4 minute exposures - it's a no brainer. Shorter exposures and more of them are always better.
By doing that, any photobombing due to planes, satellites, UAPs, etc are all much lessened. Wind issues are alos minimised. At 60 seconds, most mounts can track _unguided_ . Of course, we are fed the line that everyone needs a guide camera, but that's because it's big $$$ to sell them. They are honestly not needed, and neither are exposures of more than a minute or even 45 seconds. People get sucked into believing that guiding and longer exposures are necessary but the truth is that they're not.
Wrong. I don't understand why people keep repeating this nonsense. I mean, it would make sense in a perfect world(perfect uniform background signal with no LP, or other sources of noise), with a perfect camera(perfect quantum efficiency and no noise), but neither the world nor our equipment is, nor will ever be - perfect.
You want faint stuff? You go long exposure. Not even a billion subs will reveal the faint stuff in a stack if the individual sub doesn't get enough photons for whatever you wanna resolve to get above the combined noise floor, and the camera is not the only source of it.
And even in a straight apples to apples comparison - short sub stack loses in SNR department. Longer subs increase SNR linearly, while increasing the total exposure length merely doubles it, whenever you QUADRUPLE the exposure time. So, for example: a 4x240s will have twice the SNR of a 16x60s stack.
You made good points and your production was great also.
Don't forget, always look at the histogram. When you see ANY signal appearing close to the right side, you will find bloated stars because pixels are spilling over to their neighbors. That is telling you back off the exposure. Clear skies.
Thank you Michael!
I agree with you. But: I personally am fine with sacrificing a few stars (as in, clipped white) to achieve better SNR in my images. But that is a highly individual choice 👍
@@CrazedConceptions You could get the best by shooting so the target is spread out across the histogram and a lower exposure to capture the brighter stars. PixInsight's WBPP script would process the different exposure levels. Then mask and blend to get the best exposure for those bright stars. M42 is a target which begs for this approach.
@@mesample1705 yeah exactly! That's the plan! :)
Have you tried to compare for example 1x240s sub versus a stack of 4x60s subs? It would be interesting as a follow-up video.
I haven't yet, but that's definitely on my list! :)
In the case of 1x 240s vs 4x 60s, the shorter subs should easily win, because of the reduction in noise.
4x240s vs 16x 60s on the other hand would be interesting indeed!
@@CrazedConceptions The 60 second stack should have smaller stars and slightly better detail, but the 240 second would have better dynamic range. Dynamic range depends on the percentage the target's signal spreads across the histogram in the subs, and the number of bits of the light sub files. The higher the ISO setting, the lower the bits.
@@CrazedConceptions Yes, my main idea is that it would be good to compare equal exposure times. I think the noise gets even out by the simple fact of having more exposure time, regardless if it's from one or multiple subs.
@@CrazedConceptions that is not true!
1x240s will always be better snr than 4x60s subs.
You will need to learn how gain works!
If you expose at the same gain, the 60s shot will have less noise, and stacking the 4 shots will also have less noise, however you are gathering less data! If you actually stretch them equally, you will see, that the 240s image is gonna have far less noise
Now lets say you increase the gain of the 60s shot, but keep the gain of the 240s shot lower. This will still not give you lower noise on the 60s shot, as stacking doesnt reduce noise linearly. 4x the images doesnt reduce noise 4x. Of course there might be gain settings which work optimally, but straight out saying, that shorter is less noisy is not the correct explanation!
My CCD camara requires me to take longer exposures to get to my calculated Mean ADU- That is what basically tells you how long of an exposures you can optimumly concidering sky limitations of course. I must say the long exposures are worth it. And that's why I invested in a CEM60EC mount. The encoder allows me to take longer exposures in the guide camara (like 5 seconds) without affecting the tracking. And therefor I can take 15-20 or even 30 minutes exposures.. That's why they say the most important part of the rig in imaging is the mount. That said now a days CMOS chips have low read noise and can reach the Mean ADU without needing such long exposures.
Great video though. Cheers!!!
You should compare 4 stacked 1 minute image to the 4 minute image. If the 1 minute image overwhelms your cameras read noise and exceed the recommended Min Sub exposure time for your sky background, you won’t see much (if any ) difference in a 4x60 vs 1x240 as far a S/N and detail goes
I thought you gave a great presentation with very helpful & practical tips. The only downside for me was that it was really hard - virtually impossible - to tell which image you were referring to at any one time IE whether you were referring to the 60 second image or 240 second image on the screen...... Maybe next time label your images clearly so we can tell which images you are discussing. Thanks again for an otherwise very helpful video.
Surely you should compare 4 minute with 4 one minute stacked?
You're experiences track with mine... especially with narrowband. Nice job!
Thank you!
I'm quite new and I can't understand if the two images have the same equivalent exposure so you played with gains or ISO and F stop to use a different shutter speed and keeping the same amount of light
Great video on a very controversial topic in astrophotography!
I too had gone through many many posts and articles online on this topic, and after some playing around with my gears, I came to a similar conclusion : it's much better to expose as long as possible for your conditions.
I image with an 8" classical cassegrain at F/12 FL2400mm and my main targets are small faint planetary nebulae. This system is at the opposite end of the spectrum compared to your RASA. It's very hard even with 40 minutes of single subexposure to bring out the faintest details in narrowband, but it certainly results in much better pictures than any shorter exposures such as 10 min with the same integration time albeit blown out bright stars, which can always be replaced with LRGB ones.
That sounds fascinating, and I totally agree! I would love to see some of your images, do you have Instagram? :)
This has been know for a very long time… I’m not sure why the 60 second sub myth keeps propagating… you will definitely capture more with 1 ten minute sub than ten 60 second subs the signal is just to close to the noise in short exposures
The improperly calibrated 60" images played a factor in the stacked result. Also, I wonder if those 60" subs were long enough to properly swamp the read noise when taken with the NB filters. I don't own a RASA, so 60" might be plenty long enough. I'd be interested to see the results of the properly calibrated 60" stack against the properly calibrated 240" stack, as well as a comparison between, say, a 240" stack and a 600" stack. Maybe something like, here's 60", 120", 240", 480"...
Don't get me wrong, I agree that in order to really pull the dim/faint details, your subs need to be long enough to separate those dim/faint details from the background. I just don't think a blanket statement of "long subs are always better" is accurate. You mention some of the cons right around 3:00 in. One you didn't mention was over-exposure. You'll start clipping highlights (blowing out star cores) as you overflow the wells of the pixels. The longer you keep that shutter open, the more areas you clip, which loses that data.
Hi Jonny, the improper calibration definitely made an impact, although i doubt its noticeable. I just had the wrong gain back when I took those darks, and never bothered to re-take them because I stopped shooting at that exposure length.
I agree with what you said, but keep in mind that I was only talking about the improvement in SNR. I know that we can go very technical here, but that was never the goal. I just tried to answer the question whether or not longer exposures result in better SNR, which they do.
Whether or not one wants to blow out star cores is an individual decision, I personally don't care about that. Its easy to take another set of short subs for the stars, its difficult to get good SNR in faint parts. But that is my choice...:)
Really interested by astro biscuits recent lucky imaging video. I’d be interested to see how much resolution you could get lucky imaging dsos with a RASA.
That's on my bucket list! But I'll probably wait until I have the 11", the additional 220mm focal length should make small targets more interesting ☺️
Yes indeed you are 100% correct. See the the Jan. 2024 S&T for more evidence.
Great video, but this comparison should have equal exposure times, because that is honestly what this question or test should answer. If you have a total 60 minutes of exposure time, do you get better results by doing 3 minute exposures or 30 seconds exposures. Without doing equal exposure times it's pretty easy to conclude that a three-minute total exposure time is going to give you more detail than just a one-minute total exposure time. Again great video and hopefully you will do another one that will test with total exposure time, I believe that will be a more beneficial test to answer this question.
I'm not sure why you'd think that these are different times! It's a 60 minute integration, each. Wouldn't be much of a comparison if that wasn't the case 😅
@@CrazedConceptions Sorry I missed the part where you said that this was a one hour integration for both photos. This is a fascinating topic for me I have seen several tests with mixed results.
@@brianlipsey9883 no worries! And yeah it definitely is a fascinating topic 👍
Just look into Roger Glover videos and find out optimal exposure depending on scope, camera, filter and importantly bortle of site. Darker site longer exposures are needed and much better pictures comes out.
Shooting hydrogen for 1min is just outright wrong even from bortle 9.
Bottom line, watch Glover’s video and stop discovering bike in your videos.