Another Helix

[rustigsmed]

hello,

Plenty of fantastic Helix's around the place, here is my attempt with 2 hours 35min data, mostly 7min subs and some 5 min subs at iso 800. unfortunately a some steep gradients to deal with.

have managed to reveal part of the brighter shell but could not get a whiff of the other side. Maybe I need longer subs :shrug: or darker skies, or some filters.
I plan on getting 6-8 hours weather permitting.

Hi res below
https://www.flickr.com/photos/80336656@N07/14951927477/in/photostream/lightbox/

any tips appreciated - cheers

Rusty

[Retrograde]

That looks pretty good Rusty.

If you look at some of the deep images of the Helix it does seem that one side of the outer shell is brighter than the other so maybe you just need more data to bring it out?

[raymo]

Very nice; knocks my efforts into a cocked hat, as the old
saying goes.
raymo

[rustigsmed]

Thanks retrograde - yes I had a quick search of DSLR helix images and found Richard's superb 20 hour pic. he did 10 min subs at iso 1600 and just managed some faint colour on the otherside. Im guessing perhaps a filter and longer exposure still might be needed. so that's ok i'll work on smoothing out what I have (and doing something about the stars - Richards were way smaller :eyepop:)



cheers raymo, don't be so hard on yourself i doubt there would be any better versions out there with the same amount of data collected and with 60 second subs! this is probably my fifth go at processing the helix - the other versions were promptly deleted after many frustrating hours at the laptop!

also I have been reading a few photoshop tutorials on gradient removal which makes a huge difference - the helix can benefit greatly from doing this. i had tried a few methods which removed the gradient but black clipped it to buggery. i'll see if i can find the most useful one i found (and used for this version) and link it for you.

cheers

rusty

[raymo]

Thanks very much Rusty, very kind.
raymo

[Rex]

That's a great image Rusty, well done. The other lobe is very faint, I have a 10hrs worth of data on this object from last new moon that I have only very roughly stacked and stretched to see what I had to work with and it is still very very faint. Hopefully will find some time soon to process and post it.

[LightningNZ]

Great pic Rusty.

I reckon with an H-alpha filter you might start to get some of it, it'll really knock the background down in a big way.

If you're going for 10 minute subs in the future you might want to check how much dark signal you are getting, to see if it's putting a low-end "fog" into your subs. That'll reduce the sensor's sensitivity to faint stuff unless you can knock it back with cooling.

Good luck!

[BruceG]

Thats a mighty fine Helix you have there Rusty. Really nice going. The scale of the object you guys with the reflectors get is making me think that I now need at least 8" of fast Newtonian. My little 90mm semi APO is fine for things like Carina and Orion, but really doesn't cut it for the smaller objects. I have captured a couple of galaxies recently (for the first time) and they are so small compared to the sensor, it is ridiculous.
Keep up the fine work - i am jealous!
Bruce.

[rmuhlack]

that's a good start Rusty. The other shell is indeed very dim, and I found that even after 20 hours at f6.4 it was only just showing. I think that a Ha filter is really required to improve the contrast to get that shell to show above the background sky signal. That said you are using a faster scope than me, so with more data you should improve detail and reduce noise "across the board"

In terms of sub lengths and ISO settings, what you want for the really dim stuff is to maximise the signal to noise ratio. With an uncooled camera you are limited in your sub times before dark noise starts to dominate your signal. I have done various experiments with my previous uncooled Canons, and found that SNR was maximised at ISO1600 and (depending on the ambient temps) sub times of 3-4 mins (see first graph attached). Refer here for details of the test procedure: http://www.iceinspace.com.au/forum/showthread.php?t=106775

Now that my DSLRs are cold finger peltier cooled I can go for longer subs, and am currently using 10min subs on my 1000D (on a f6.4 scope) and 8mins on my 450D (on a f5 scope), See second graph attached showing active cooling SNR data for my 1000D

[rustigsmed]

Cheers Rex, looking forward to your rendition - have been enjoying your images of late :thumbsup:



Thanks Cam, hmmm I think you could be right!



Cheers, Bruce! No harm adding another scope to the arsenal ;) an 8" should do the trick no worries :thumbsup: this is cropped due to some odd stacking artefacts around the perimeter.



Thanks for the detailed response Richard. Some interesting points you've added there, perhaps I should look at iso 1600 and go for around 3 minute subs :question: rather than the 7min at iso 800. if I were to grab a filter (Ha or cls) then I guess longer would be better.

the reason I went for iso 800 was that I remember reading somewhere that iso 793 was the canon 600d's 'native iso' I guess that doesn't necessarily translate to the best signal to noise ratio though.

as I've only completed one other image with my new setup I'm still trying to find my feet in terms of how long I should image a target for considering my local conditions/optics. I was wondering if I went too far with this and struggled to get rid of the grey/white fog/gradient (its removal I think has made the coloured areas appear 'noisier').

Have you had any dew issues with the finger cooled Richard? At the moment i'm contemplating modding my own, going central cds (new), or a mono QHY9 when their new 7x36mm filter wheel arrives. That of course is another learning curve and expense.

Cheers,

Russell

[rmuhlack]

I also previously used ISO800 having read that it corresponded to "unity gain" with my camera. The thing is however that (as Roger Clark puts it) "unity gain is a flawed concept". See here: http://www.clarkvision.com/articles/digital.sensor.performance.summary/#unity_gain The SNR procedure I used is also from Roger Clark's website, who observes a similar reduction in SNR at longer sub times, and an increase in SNR with increasing ISO. See here: http://www.clarkvision.com/astro/canon-10d-signal-to-noise/ The results from applying this test (with my uncooled cameras at least) are clear - ISO1600 results in a higher SNR at all sub times, and a sub time of 3-4 minutes has a higher SNR than at 5 or 10 mins.

For an uncooled DSLR, I don't think the use of a Ha filter changes the optimal sub/ISO setting. Roger's procedure calculates a relative SNR value. For my example of the uncooled 1000D, the relative SNR of a Ha filtered DSLR image at 3-4mins and ISO1600 will still be greater than a Ha filtered image at 10mins and ISO800.


In terms of dew with my cold finger mod, I was experiencing severe dew issues when using a setpoint of -5C or 0C, and intermittent dew issues with a setpoint of 5C (this is despite using a nichrome heater around the front of the sensor). Increasing the setpoint to 10C and the dew issues go away, with only a minor loss of relative SNR. The setup is regulated which means that I can consistently apply matched darks, which I think (together with the much lower sensor temperature) has significantly improved my images.

[rustigsmed]

Thanks for the info again Richard greatly appreciated :thumbsup:
I'll have a go at shortening my sub length till I can nab a new (cooled) camera. Not sure i'm game (or handy enough) to go with finger peltier cooling myself.

Cheers
Rusty

[SimmoW]

Great result there Rusty, esp with an SLR. Hope to image it one day myself, with my tiny 430mm f/l scope (but bright).

I was researching similar slr noise issues last night, and came across this great Craig Stark article

http://www.stark-labs.com/craig/resources/Articles-&-Reviews/CanonLinearity.pdf

In it he concludes that ISO400 is the sweetspot. I've always (in my very short career!)used 800 but after this article, will experiment with 400.

I've only recently been able to get to 10 min subs, again I think we need to experiment on a single object to see just how much more, if any, detail we are getting compared to say, 5 mins. I'm acutely aware that dslrs are ultimately a compromise. All good fun!