M83

[peter_4059]

Had a great imaging session last night - the first good night for many months. Here's a quick process of M83 taken with the DSI II pro through my 10" f5 newt. 50x1min subs.

I also produced a 1 hr animation of Lulin but have no idea how to post the resulting 6.5MB GIF :D

[Matty P]

Excellent image Peter. Lots of detail within the core and spiral arms.

Very well done. :thumbsup:

P.S. You should ask Mike about posting the GIF of Lulin. ;)

[AlexN]

Peter - Well done indeed mate! Plenty of detail in that shot!

As with your galaxy images a few months ago, this one seems to lack a bit of punch.. maybe a quick tweak of the curves will give it a bit more oomph.. Otherwise a stunning up close look at beautiful target!

Alex..

ps. good to see you made good use of the beautiful night too! :)

[Kevnool]

Great image Peter with loads of detail.
Cheers Kev.

[peter_4059]

Thanks for the feedback Matty, Alex and Kev. I've given it a another shot with curves to boost the contrast.

[Tilt]

Great looking detail in this image Peter, well done.

Michael

[AlexN]

like the repro a lot Peter! Fantastic!

Alex.

[troypiggo]

Nice work, Peter. Kills mine.

[peter_4059]

Thanks Michael/Alex/Troy. I was thinking about staying up and getting the colour channels last night as the conditions were fantastic but just couldn't keep my eyes open any longer. This galaxy is definately on my list to get an LRGB image this year.

Looking forward to seeing the Leyburn results Troy!

Peter

[Peter Ward]

Pete...to be blunt, the dynamic range and noise... suck.

But...wow..the resolution is simply amazing! . Beautifully guided and superbly resolved stars abound....the gear and the seeing obviously all were working superbly well.

Stick with it :thumbsup:

[AlexN]

Agreed, the resolution is incredible.. Theres a close double on the top right of the galaxy that are clearly separated, they came together in my shot from last night.. We did have some stunning seeing last night in brisbane... very steady.. And Peters gear obviously took full advantage of the situation.

[Bassnut]

Yes, the repro is impressive for such short subs (tad noisy), lots of detail. Some overtime in PS would be worth it.

[peter_4059]

Thanks Pete,

Not sure what to do about the noise apart from getting more/longer subs or a cooled camera but that's what I get from 50 x 1min subs from the DSI II operating at 26 degC. What do you do about improving dynamic range?

I had an issue with the guide camera nosepiece - I didn't have an allen key small enough to tighten it up but the electrical tape must have done the trick ;)

The conditions sure were a big improvement on the past three months.

Peter

[allan gould]

Peter
What an absolutely brilliant photo. Looks great to me and is a superb example of what can be done with good amateur equipment and a lot of care, patience and ability.
Someone must have had an avgas leak into their oxygen cylinder to suggest that the dynamic range and noise suck for an image with a $600 camera.
I would be as pleased as chips the have done half as well.

[Peter Ward]

Pete...longer subs is the key. Try at least 5 mins so you'll be sky limited in noise, rather than camera limited.

I'm still blown away from the resolution you have achieved.....the double at 1:30 o'clock is rarely split in images with gear costing 10x as much. You've done really well there.

As for the noise....does the camera not self cool? If not, you may get a significant improvement via an ice pack or dry ice pack strapped to the camera.

[peter_4059]

Thanks for the comments.

I've kept the subs to 1 min for two reasons - firstly because 50x5 mins would make for a long session with meridian flips and I don't know how to deal with these from a processing perspective or multiple sessions and all the framing issues associated with my non-permanent setup (and because I'm impatient). The second reason has to do with the histogram while imaging. With 1 min subs I have the histogram almost full so have been worried about clipping at the white end. How is this handled if you want to take longer subs? I can turn the gain down but the Meade forum advise is to not do this.

Great to hear the resolution is good. I went to a fair bit of trouble up front checking the colimation and focus (love that Bahtinov mask).

The DSI is not cooled - just has a heat sink so you typically get Ambient + 3 degC. I've heard of people strapping dry ice to the back but haven't been game to try this yet. :P

Peter

[allan gould]

Hmmmmm! Dry ice - must try that.

[Ric]

Hi Peter, that's a wonderful image that you've captured. Lovely detail right throughout and it really shows what the little DSI's are capable of producing.

Cheers

[kinetic]

Firstly Peter, beautiful result! I'm amazed that your capture is so clear
and noise free with 26°C ....My DSI seems to work well when ambient
is below 20.

Secondly, and at the risk of hijacking the thread, I think Peter Ward's
comments are a bit harsh.
The dynamic range of this image is pretty awesome. You are viewing it on
the forum probably as a scaled down JPEG.
As a vendor of high end gear you would know that the lowly DSI Pro II
is still 16bit, so capable of 65,000 shades of grey.
Now Peter's image probably had most of the data down
the lower end of the histogram and so a lot of the DR wasn't even used.
But take a step back and look at exactly what he has captured from the
very centre of the core , out to the faint halo......
More DR isn't going to add a lot more to this except for maybe more
halo. And the resolution is pretty awesome considering he was PF
on a 10" scope with 1 minute subs!

I know you qualified your comments in a later thread Peter (W) and I
also respect that brutal honesty is also helpful.
But Pete's image here is pretty damned good all the same.

Well done Peter.

Steve

[peter_4059]

Steve,

Here's a typical screenshot of an un-processed frame from envisage showing the histogram peaking at 59770 out of the range 65535. While a lot of the histogram is bunched up on the LHS, if I take longer subs the stars are clipped. What do I do about this?

Peter

[kinetic]

Pete,

I will PM my processing routine.
It might not work, I'm no expert, but you may want to give it a try?

Steve

[AlexN]

I dont know for sure, but I think that you're a long way from clipping any real data.. cant hurt to try I suppose... I'd be taking a few images that bring the main bunch of the histogram out a bit from the black point, and take a few shorter subs to mask the stars back in should they clip out badly ?

[peter_4059]

So are you keeping the histogram below 65k while imaging or setting the exposure time and then taking multiple times and masking to avoid clipping?

[Peter Ward]

You are entitled to that opinion. I don't comment often , but am definitely *not PC* (and make no apologies for that) and simply call it as I see it.

I had hoped to spur Pete on to try *much* longer subs for a truly great result.



No it's not... either lowly or 16 bit. A 16 bit D/A does not a mean a 16 bit image. The fact is: many CCD images are simply not 16 bit despite their D/A resolution...which includes many CCD's in the SBIG/FLI/Apogee range.

IMHO the image looks flat. Beautifully resolved (and hence really worth further effort) but flat just the same. Longer subs will bring out the really faint stuff....not seen thus far (ie low DR) ...but easily obtainable given the excellent result thus far.

[renormalised]

I was going to suggest something similar myself....to get all the good bits it would pay to go longer on the subs. Just as long as his tracking and alignment are right.

[peter_4059]

Did a complete reprocess from scratch. Still a bit of noise. I think there is another close double at 2:30 in this image.

[jase]

Spectacular work Peter. Keep them coming. :thumbsup:

[peter_4059]

Thanks for the comments and advise Steve, Alex, Peter, Carl and Jase. Still trying to understand the merits of fewer long exposures or many short ones and how to manage clipping with longer exposures but will give some 5 min subs a try next time.

Peter.

[AlexN]

Some stars might saturate. but thats not that much of a big deal, although when they saturate they may bloat a little, costing you some of that magnificent resolution... However the longer your exposure, the more faint details you'll get... 100x1min images does not create an image as detailed as 10x10minute images... 10x10min will get more of the outer arms, a lot more dynamic range in your image, and a lot more background galaxies/stars etc.. 100x1min images may be smoother noise wise, because there is 100 subs to smooth it out... You may need to run some shorter subs to control the core of the galaxy though, as M83 is fairly bright in the central region..

Cant wait to see some 5 min subs!

Alex.