Long VS Short Exposures in Astrophotography - Broadband

hey there guys how's it going during a recent live stream a question popped up where someone was wondering is it better to take lots and lots of very short exposures or fewer but much longer exposures and my advice during the live stream was it's probably better to take longer exposures as long as you're not saturating things on your camera as that generally seems to me at least in my experience to be the best way to go about things but it did make me think why not do a little bit of an experiment on this and share the results with you guys so here we are the very next clear night and I've set up just such an experiment now I'm going to take you through the parameters that I've been using for this experiment so you're completely on board with exactly what we're doing and then the second part of this is going to be taking a look at the data side by side from both Stacks in pix in site at various stages of processing where hopefully we'll be able to pick out some differences between these and you'll be able to decide for yourself and I will too with any look as to what I might want to do in the future going forward so this is a Broadband test that we're doing with just uvir cut filtration there's actually a half moon up in the sky right now as well so it's not exactly perfect conditions and my home Sky conditions normally here a bottal seven quite a low bottle seven but still thereabouts now the telescope I'm using for all this test is a very very fast telescope indeed it's Celestron rasa 11 at f2.2 paired up with this I've got a onot color camera from player one the Poseidon c pro it's an IMX 571 sensor uh tremendous sensor and the two different settings I've used for this so for my 10 minute sub exposures like the one that you're looking at on the screen right here I've used gain zero and offset 100 and then for the one minute exposures which we're going have taken a look at and I'll explain how I've gone about into leaving these in a moment they're taking a gain 210 in order to engage High conversion gain mode and drop the read noise and offset 200 so uh everything's tried to be equalized as much as I possibly can between these as you'll note if you can see just down below the mean background value if you like of a 10 minute sub uh at gain zero offset 100 is about 9,000 or so and the background value on a 1 minute sub at gain 210 is about 10,000 so it's as close as I can reasonably get well experiment so all through the night now I've been into leaving a long sequence effectively where I've been uh taking one 10-minute exposure allowing it to dither and then taking 10 one minute exposures with the settings that I laid out before and after this is done we're going to stack up one hour of data from each different data pool if you like we're going to take the best data we can um and allow software to select that kind of thing I'll throw out any major outliers and hopefully we should have two very fair to compare Stacks I think so uh yeah that's basically the outline of this experiment I hope that you're going to find this one interesting and we'll jump over onto pix sight in just a moment but I do want to say if you enjoy this kind of thing please do leave a like on the video for me it really really helps push these videos out to the right people which is all I really want to do I want to reach as many people as possible on my channel and if you have any other suggestions by the way on other experiments you'd like to see so perhaps the same thing but repeated with a slower telescope or maybe with the same telescope but a narrow band filter that kind of thing there's lots of kind of variables that we could apply to the experiment uh and find out different things together I think it could be some fun but anyway on with the next part all right then guys so I'm finished now I've stacked up the two sets of data so the 61 minute subs and the six 10 minute subs and we are ready to do some comparisons now I've hooked up four separate comparisons for you and we're going to take a look at those right now so here in this first case if I just zoom out a little bit we are looking at the data just as it left the stacking software with only an stf screen stretch applied and um a slight change in terms of fast rotation and a horizontal mirror to make them the right way around because we're using aasa and it flips the image so hopefully as you can see when looking quite zoomed out there isn't really much to observe in terms of visual uh differences between these two the faint extensions of you know dust if you like is I know it's Stars really around these galaxies really nothing to be uh separated between them I would say it's just as strong on the left as it is on the right on really the faintest of regions on this galaxy so in terms of signal to noise ratio on the Galaxy itself and the fan extensions at this point in processing there's nothing in it now there is a slight difference when we get close up in terms of color at this point the colors are very slightly richer on the right hand image but again this is before any processing we can indeed change all that now there is one final thing I do want to point out on this image before we move on to perhaps a more interesting step of uh processing is that as you can see if I just zoom in one more uh step to a three to one view we ended up with very slight elongation of these Stars so some degree of eccentricity brought in I imagine due to the slightly Gusty conditions tonight pushing around just a little bit some of those 10 minute Subs so we gave something up there a little bit so in this case 10 minutes was a Touch Too Far on the night I know that my Mount is capable of doing these 10 minute Subs but he needs the night to match you know what I mean so I asked a little bit too much of it on a nightlight tonight um in terms of overall resolved detail though I think there is a very slight Advantage now we're extremely close in on the left hand shot in this Dusty region right up here if we just compare that view hopefully this is coming through for you on YouTube I can see a little bit more detail right here than there is on the right hand side overall the Stars just looked that little bit smaller too uh but I think that's down to the tracking accuracy as I mentioned now uh one other thing I've just noted that I want to bring up as well as I moved the view there the kind of background quality of the noise is definitely smoother on the left hand side where the actual stack of data itself is simply just more robust so the noise has had a chance to really even itself out a little bit more even over the exact the same total integration time because there simply 60 samples of noise versus six so slight advantage to the left hand side I would say right there now if we move on to a more interesting step of processing here we are looking at the same spot but this time we've applied a crop to get rid of stacking artifacts with applied gradient expert to both images to get rid of the residual gradients that were left behind that weren't corrected by flats and things like that and also have applied spectr photometric color calibration in order to try and bring the color Cali cration into these images and nothing else has been done so that is the phase at which you're looking at these two images now so uh over on the right the six 10 minute shots it's exhibiting very slightly stronger color as we'd observed before on the completely unprocessed stack I think it's worth noting that it is there but it is very very very slight um once again that background noise uh does look slightly cleaner on the left but there really isn't much uh between it so I'm not going to focus on it overly much and when I was taking a look around before hopefully you can really see now if I just equalize these views a moment over on the right because we have such a small stack of data um we've got a lot of Unsolved satellite Trails going through the field of view uh now if I'd taken a longer experiment at this and maybe gotten up to about 20 10 minute Subs something like that uh I would have had a view that looks looks a bit more like the left hand side where median Capa Sigma clipping uh stacking algorithms has gotten rid of all of those little transient events that's gone off in those individual Subs uh for us so if you are just going to do a short night's Imaging I would definitely advise you to take short enough Subs that you can end up with at least 20 30 maybe even more frames to actually stack up so that you can get rid of things like that so you're not having to deal with big ugly Trails going through your uh your images you know what I mean you quite simply don't have to but here they are for the sake of this comparison all the same now one more thing I uh want to point out at this point is that down here in the background there is a distant background Galaxy which on the left hand side the 61 minute shots I feel I don't know what you guys think put in the comments down below if you want I I think I can see an S shape already being resolved in there so there is more resolved detail in the left hand shot versus the right now is this down to again uh it being more perhaps statistically robust on the left I don't know is it down to the slight increase in tracking accuracy because the the wind gusts and things like that have been averaged out you know uh due to simp simply more samples to provide the stacking software with I can't really say um but I think the image on the left overall takes it takes the win for me on this so shorter Subs in this case do look a little bit better even though the signal is very slightly stronger on the Galaxy itself over on those 10 minutes Subs there really isn't much in it now the next step of processing was to apply uh noise exterminator and blur exterminator in the opposite way around actually so it was blur Terminator first so I'll just undo both of those so you can see again the data left as it was a moment ago we'll go ahead and zoom in just a touch more now we're going to apply on the left hand side BL exterminator first you can see those stars immediately shrink right down and the finer details in the car of this Galaxy start to become cleanly resolved these are default settings by the way I haven't gone and done anything crazy to this and over on the right so this eccentricity again gone so it really does take care of any small you know tracking artifacts and things like that it does a really wonderful job even when it's provided with uh not very much to work with you know this is six samples of this galaxy six 10-minute shots and uh the deconvolution that it's applying is working really well I have to say these these pieces of software are just incredible I I've said it many times before I wouldn't want to process without them now they quite simply make the game just more enjoyable to me um as always I will mention I do have affiliate links for these and if you guys want to use those it would provide me with a lot of support for my channel for which I'd be extremely grateful to you and I am indeed already grateful to everybody who's used those links as well you guys helped me out an absolute ton so hand on heart thank you very much indeed for your support through those um wonderful software though wonderful wonderful software and I think let's move on just a little bit so that's with blur exterminator applied now if we go ahead and apply noise exterminator this was at 75 strength detail slider left alone in both cases all that background kind of salt and peppery color morly noise gone we're now left with quite a smooth background in both cases however I would say because the noise looked a little bit better going in on the left it looks a little bit better coming out on the left so you know you get what you put in effectively I think and it's true with Astro to you know if you're providing extremely noisy day you know it's only so far that it can be polished reasonably speaking but now that both those tools have been applied I would say one other thing has become apparent on these images and that is that there's a bit of a contrast advantage on the right hand side so those longer Subs seem to have brought out more contrasty detail within the spiral arms of this galaxy so these dust Lanes in particular look a little bit darker even though the exact same processing has been applied to both of these images at this point um I think it's also worth noting the fine detail around these small nebulous regions here you can see on both both images give up something to the other there is no clear winner in this there's parts that I like more about the Right image there's also parts that I like more about the left so for me it's a bit of a draw on this one I think overall probably want to continue processing the left hand image more than the right there isn't really much in it and without comparing them directly side by side like this you know who'd know who'd know so it's good to see there's a wide acceptable range of exposures that you can go with now the next final step that I applied to these was to apply the same processing steps to take them just a few steps further so now this is stretched data I used just a very gentle General hyp uh generalized hyperbolic stretch on both of these I made the data match in that case and then I applied saturation to both of the images in the exact same amount in both cases in order to see as you can see uh that saturation difference that natively these images started with so the right hand side definitely takes the lead it looks like there is more color visible on that image I think um it seems to be just a little bit more color accurate than the left but I really do want to point out the ease with which I could turn the left image into the right can't be overstated you know it just it would be so easy to do but I didn't for the sake of this comparison I wanted you to see that there is you know a bit of a differ in the underlying data and that application of saturation has helped us to see that um but it isn't much I think it's very fair to say there really isn't much between them um there is a bit of a color advantage to longer Subs perhaps maybe a different test on another night would give a different result but I think you'll agree that in both cases for one hour of day or on a very very moonlit night not bad uh overall now if I want to take another look down at that interesting little background Galaxy we pointed out earlier hopefully you can see that s that we talked about very cleanly resolved on the left not really there on the right I would say maybe a hint of it is present but uh it looks to me like there is perhaps more resolved detail on the left hand image just overall generally speaking um so for me probably on balance shter so in this case have worked that little bit better uh than the longer ones tonight now on a night of perfect tracking all night long maybe it would be different I don't know but how often do those nights come for me not very not very often at all if I'm if I'm being honest so I think this has been an interesting experiment for me because I have often wondered you know am I under selling himself by going you know very long on exposures sometimes and sometimes very short it looks like not really it seems to be a very wide range of acceptable exposures As Long as You Follow one of the basic rules and that's make sure that your histogram is separated from the left hand side which it was in both cases it was comfortably matched effectively between these two there if you recall earlier in the video mean values of about 9,000 on the left and about 10,000 on the right so both we know way away from those zero values on the left hand side as they always should be I guess if your offset is set correctly now that is about it from me guys I would say so I want to say as always thank you very much indeed for your support I do appreciate you all and uh you know the fact that you make this doable for me so thank you thank you very much indeed uh one final shout out for those affiliate links again if you do want to use that kind of software and you wouldn't mind supp to me it would be extremely appreciated and thanks again to anybody who has use those uh huge welcome to all the new channel members and thanks again people who popped into my recent live stream which was extremely extremely busy and uh a very interesting night so yeah thank you all again guys anyway look forward to seeing you in future videos live streams and things like that make sure to hit like if you enjoyed this one and uh subscribe if you haven't already look after yourselves and I'll catch you in the next one cl guys

Channel: lukomatico

Views: 20,114

Keywords: astrophotography, astronomy, telescope, deep sky astronomy, planetary, pixinsight, blurxterminator, noisexterminator, long vs short exposures

Id: T0JDvllCaV4

Length: 18min 0sec (1080 seconds)

Published: Fri Apr 19 2024