10 min subs versus 20 min subs - empirical results...

[Tim]

...at least, as empirical as possible with a varying sky.

The OTA, camera, temperature and target position were all very similar, as was the seeing and transparency and lunar input. The flats, darks, bias and dark flats were identical masters.

I have processed them in a very simple curves and levels fashion, no noise reduction or anything else, just to aid a direct comparison.

What is more, I haven't compared these myself, as I am planning to combine the whole lot of data for my 'finished' image and dont want to get hung up on details, so I'm leaving it for you lot to make your/my mind up

Night one > 20 x 600 sec (10min) subs

Night two > 10 x 1200 sec (20min) subs.

What do you think?

TJ

[centroid]

'Chalk and Cheese' Tim

Dave

[ngc2403]

QED

[Psychobilly]

Indeed... and no ones doubting it provided that....

Your LP levels and or filtering allow it...

Your "camera" controls noise well enough to allow it...

Your tracking / guiding allow it.....

Your polar alignment is so good that you dont get field rotation....

As some of us have been saying each "person" needs to work out what works for their kit/location etc... and aim for the "longest" subs that "work" for them...

Peter....

[CELESCOPE]

As i expected Tim, and found , i could get away with 20 min subs no prob as thats what i used to do , but being in a stacking zone for aircraft, i lost so many subs .hence going back to 8 min sub s, but i see now i can get away with 12 mins , and get pretty good results ,its as i said really , whatever is best for your location etc etc .

But very impressive results there

Rog

[rich_t]

I thought I read somewhere, can't remember where, that the equivalent single exposure for a bunch of subs was equal to the sub length multiplied by the square root of the number of subs?

That's 45mins for your 10min subs and 63mins for the 20min subs?

Something like that anyway! I imagine there's a boat load of things that change that though.

Cheers,

Rich.

[acey]

Which galaxy is it?

[CELESCOPE]

Galaxy M101:)

[roundycat]

an interesting test, I wish I had the patience to do this properly. I'm a bit worried about comparing the two with the background being so obviously different. Were they both given identical stretches in PS?

Dennis

[MartinB]

Very interesting Tim. Throws up more questions than it answers. I'm guessing the top one is 10x1200 seconds given the brighter sky background. The 20x600 will need more stretching since the average levels will be lower. Ideally, to do the comparison, the brightness levels of the background should be identical.

I would be very interested in knowing the brightness of the background for the individual subs. Are these taken with your QHY8?

Lovely M101 btw!

[moocher]

The images speak for themselves Tim, so i take it your hanging on to the OO? Rog doesn`t Sigma Clip stacking get rid of plane and satellite trails?

Moocher:)

[Aza]

Nice images, will be great when you combine them all!!

I have heard people mention shooting the maximum your sky glow can allow.

How do you work this out? I sometimes worry I shoot my exposures too long, and would love to know how to get the optimal length!

[ngc2403]

i think Martin the top one is the short subs and the bottom one is the long judging by the amount of stars.

The thing about comparing them is that the longer subs have not been stretched as much as the short ones because they would of saturated badly.

[afm038]

By moving the mouse over the photos the top one says as part of the filename 10 min subs and the bottom one 20 min subs.

On first viewing I prefer the bottom one but are both good. I'm sure it'll look awesome when they are all combined.

Andy

[MartinB]

Ah yes, I was looking on my work monitor and the top one looked better!! Now looking on a decent montior the top one has clearly been pushed harder and the background noise has become intrusive.

[Darth Takahashi]

Hold on a moment, I would like to know some basic facts?

Whats the camera used and whats the optical train / system used?

Finally was each image processed the same? Simply raising the black point will already kill a lot of noise without damaging the image / signal.

It clear at least to me that an uncooled DSLR's limit is determined by the original noise produced by the chip and the amount of heat generate in the chip during the exposure. I have and can prove empirically the opposite point with a DSLR?

I have no problem tracking for more than the required 20mins, however, I would say that this is the top end of my sky limit, weather dependent of course.

I now do shorter sub's 5-10 mins and more of them to reduce the noise instead of going for longer subs.

Neil.

[Aza]

Neil are you sure that doing 5-10 minute subs on an DSLR is better then 20?

I dont know myself, but I routinely do 20 minute subs, and I wonder if I am wasting my time, and could in fact being doing double the amount.

When I do 20 minute subs, I dont seem to have a problem with the light pollution, and they seem to me to be much better in the noise department.

[Tim]

Still at work and its stocktaking night, so dont expect any sense out of me for a couple of days.

The images were stretched in an identical way by the same values.

I was hoping that some of the really faint stuff on the outer regions might have come through better on the 20 min subs, but I cant really see much evidence of that.

I suppose I could do an unstacked version, single 20 min against a single 10 min to compare the stuff that comes through. I think the limiting factor may be my skies though. The LP here is so bad that I can read in my garden just by skyglow, that's gotta hurt the faint stuff I would think.

Aza, I wouldnt even consider 20 min on a 12 bit dslr, I used to find 6 mins was about the maximum useful exposure for me, here.

Camera was QHY8 btw. Off axis guiding.

[Tim]

Also, there is no magic about the figure of 20 mins. it could be 3mins against 6, or 5 against 10.

As Peter was saying, usually, not in every case with every target under every sky, but usually, you might be best going for as long a sub as your equipment is capable of, rather than more of the shorter ones. DSLR users have a seperate issue of noise as Neil Hankey points out.

The real point, I suppose, is that if your kit is capable of a 10 min sub, but because you haven't taken the time to align properly you can only get 5mins, or you have flex issues, then it is well worth taking the extra time and effort to increase your sub length to its maximum potential.

Probably. I'm just glad to see a clear sky now and again, dont wanna get too hung up about mathsy type things

[Darth Takahashi]

Aza, the short answer is yes. I posted an article sometime back that proved this point. I think it was by Stark Labs, should be able to trace it back. Unless the DSLR is cooled its clear that the noise starts to build faster than the signal because of heat generated in the chip that cannot be removed.

Cooled camera don't have this issue and therefore benefit from longer exposures.

I'll look back for the article.

Neil.

[Aza]

Thanks TJ and Neil

If you can find that link it would be cool

So TJ says 6 minute was his max with an DSLR, what do you think is yours Neil?

And is there a way to find the optimum length for your skies?

Sorry if this is taking the thread away from your images TJ.

Perhaps you could PM me Neil

Cheers

[Sam]

Neil,

I read that article when you posted it and I recall that the summary I got from it was that a DSLR is good for about 5mins at ISO 400 and that gave the longest exp with the lowest noise, any longer than that andf noise started to increase too quickly to gain any benefit from the longer exposure - though if a hypermod was done then it may be a different story...... mmmmmm.

Sam

[Aza]

Wow I had better rethink my nights out! My images then could well be a lot worse than they should be

I could potentially have 4 times the number, and better, subs than I currently get....crazy!

[Psychobilly]

The latest Canon sensors (1000D and 450D) allow you to take very long subs with good noise control they are way better that teh ones used in the 300/350 's

Ampglow is non-existent even in 20 minute subs and even at room temp the noise levels are suprisingly good... editied to remove the 400D from the list based on Aza's comment...

I did post a thread back in Decmber last year whe I did my initial noise level tests for the new camera...

Peter....

[Aza]

SO is it possible some DSLR sensors can in fact go longer?

I have a 400D and have never had any ampglow with 20 minute subs.

Is there a way to test the noise levels in a 10 minute sub and 20 minute sub...to prove to me that the 20 minute subs are a waste of time?