M81 and M82, can't get out the details

[Jannis]

Well, as the topic sais, i'm having a bit of problems getting the details out of this, but not sure what i messed up in shooting or stacking?

Stacked mostly of 30 sec pics of both iso 1600 and iso 3200, total exposure with 143 frames is 1 hous 16 min and 31 sec. Also added around 20-30 dark and around 20 flat.

Was it a bad idea to stack both iso 1600 and 3200 at the same time in dss?

Here's what i was able to come up with after DSS and PS, and a single sub of each ISO.Also attached a dark frame at iso3200, gained up to see it. It looks like i have issues with the dark frames?

I'll try to add a link to the TIFF a bit later, but need to get my webserver up and running again first. probably in a few hours.

[Jannis]

Well, untill i can upload teh full res TIF, here's a crop of M81, if that's of any help...

[Bizibilder]

Did you have darks and flats at both ISO settings? You would need to stack all the ISO1600's and 3200's together seperately. You say "mostly" 30 sec subs? You must also only stack subs/darks/flats of the same exposure time together. Hope this helps. Btw for 1 hr and a bit you have a good amount of data there - I recently ran 1h 42mins for a similar result.

[roundycat]

I would say you are on the right track. I would stick to one ISO setting, I doubt there is much to be gained by shooting at 3200 other than severly restricting the available dynamic range. I adjusted your M81 slightly, there is more there than you might think.

Dennis

[Jannis]

Thanks for the reply guys. yes, have flats and darks of each iso at same exposure. there was a coupple of different exposures in there also if i remember correct on what DSS accepted to stack. but that's only like 2-4 pics. would i need longer single exposures, of many many more 30 sec exposures to get out more details?

[Jannis]

A TIF file is awailable here, if anyone want to give it a try. this is after stacking, setting brightnes and a little saturation in DSS. and levels and curves in PS.

http://jannis.no-ip.info/astro/M81.tif

note that it's 108MB, and i have a very slow internet connection, so download may take a while.

[Cloudwatcher]

I was having a go at extracting detail from the cropped view of M81,not with great success,when,reducing the size,it struck me that more could be seen on a smaller scale.

Anyway,I isolated and made a copy,transferring it to the widefield version as a separate layer,darkened the background (losing a few stars in the process )and sharpened M82 adding some colour for the fun of it................

[Jannis]

Quite a bit improvement in both colors and detail extraction there. I still have a long way to go on processing..

[Jannis]

Ok, time to update this thread again...

I've given it a 2nd try, but can't seem to have any improovement from first try, even though i have 4x as much data...

This is with, well, as you probably guess, 17x 15 min frames stacked in DSS.

http://jannis.no-ip.info/astro/workinprogress/m81-17x15min.jpg

And this is 17x 15 min frames, plus a 10 min frame and an 8 min frame. Looked much the same, but tried to drag more details out of it and get rid of that sky glow, with little success. also cropped a bit as edges didn't look very nice.

http://jannis.no-ip.info/astro/workinprogress/m814hours33min50sec.jpg

Compared to the first attempt i had with only like 70 min total exposure:

http://jannis.no-ip.info/astro/full/M81%20M82%20143%20frames%20at%2030%20sec.jpg

I have not applied darks, as they only messed up my pic... bias applied though.

I know i'm not good at processing, but would expect to see at least a slight improvement to go from 70 min exposure with 143x 30sec frames to 255 min exposure with 17x 900sec frames.

[Jannis]

If anyone want to have a go at the TIFs from DSS, feel free. I'm more or less stuck.

They're a bit slow to download though. 26MB each:

17x 15min + 1x 10min + 1x 8 min with no darks applied

17x 15min + 1x 10min + 1x 8 min with darks applied

Only the 17x 15min frames and bias, no darks

Any advice on improving it any more would be great!

[RikM]

I think it will be a great image when you get it figured out.

I had a quick play with it, but couldn't do any better I am afraid. Someone will put you right I am sure.

I made an artificial flat duplicate layer and then clone stamped out the galaxies and applied it as 'difference' 75% in Photoshop elements 7. A bit crude but sometimes works for me.

[Jannis]

Thanks. It sure helped on the sky glow. Managed to get a similar resulth as yours also, but unable to drag any mroe details out of the galaxy's fainter area. Or if i Did, the M82 would be almost completly washed out. Maybe these two needs to be processed separatly?

Managed to get this, after also trying to make it a bit less blue, as i asume i got way too much blue colour in it?

A tiny bit of progress, but, babysteps...

To reduse the noise and increase the details in the fainter outer part of the galaxy, what do i need to go?

Longer exposures, more exposures, getting my darks to work, flats, or just more processing skills in PS?

Also struggeling a bit to get the correct colours out of the stars in the galaxy, and the galaxy itself.

[Jannis]

Well, getting closer on M81 i guess. More normal colours at least. But it made M82 look rather overprocessed...

Also applied a bit of noise reduction here.

Is it normal not to get more details out of a 4.5 hour exposure pic? I mean, i've seen pics here of M81 wich are Far better, in both details, colours, and noise.

Some are taken with CCD, some with filters, some in an optimal area with no light pollution, but also some with nothing but a DSLR like i am.

And seeing what you guys can drag out of a seemingly rubbish pic, i'm sure there Must be some way to drag out more details from this too?

Hmm...

...

>_>

...

o.o

free hugs and cookies for everyone who comes with advises/kicks me in teh right direction...

And thanks for all the help so far everyone, i'm certanly getting there, just with baby steps at the time..

[Jannis]

And just for comparison, here's a single frame, completly unprocessed from the camera, only cropped to roughly same size.

Can clearly see that M82 is very overprocessed (no, it's not supposed to be a UFO, it's a galaxy, lol) compared to what details are actually in the light frames... >_<

Maybe better to start this from scratch again in DSS, lol.

[Jannis]

I think i shuold say i'm happy for now. after all this work, when comparing teh two, i can clearly see a big difference.

Not as much as i wanted, but i guess not bad for the stuff i used at least.

I mean, using a rather cheap DSLR on a 1000mm F/5 scope, in a light polluted area, no filters, no modded camera filter, no coma correctors or anything, and with dark frames working rather bad...

So, final pic for now, here's 1.2 hours vs 4.5 hours:

[Trull]

Hi Jannis, I'd crack open a Tou at this result, considering the humble beginnings its a nice one. What do you fancy next?

[ollypenrice]

I'm at a loss over this 'flats at the same exposure' business! If you are taking darks for flats, then yes the darks for flats must match the flats.

But don't take darks for flats, it is a waste of time. Make a master bias and use this as the dark for your flats.

It does not look to me as if those flats are working. You have a huge vignetting effect in the earlier images and in the later ones, though I may be wrong, the vignetting has only been removed by badly clipping the black point.

Certainly process each galaxy separately. Also, on the spiral, do a main stretch for the whole thing and forget about the core burning out. Ignore that. Then do a second stretch just for the core, ignoring the arms. Paste the good core onto the good arms, put a big feathered lasso around the good core, select inverse and erase the arms from the shorter stretch. That's how I did this one, though it is not finished by a long way since I want the very faint stuff as well; http://ollypenrice.smugmug.com/Other/Galaxies/HaLRGBM81M82FIN4/1213684521_9XMon-X2.jpg

For anything faint, more subs and longer subs. No free lunch!!!

Olly

[Jannis]

Thanks you very much for the reply olly!

Also, it turned on a lamp up in my head, as i realiced i've written a big misstake here. I never added flats, like i said, only darks and bias. That explains the vignetting... ;;

so this: "total exposure with 143 frames is 1 hous 16 min and 31 sec. Also added around 20-30 dark and around 20 flat."

should actually be: "total exposure with 143 frames is 1 hous 16 min and 31 sec. Also added around 20-30 dark and around 20 bias.".

Only 1 small word differense, but a veeeeery big differense in final result...

But, about flats, does it matter Where the scope is for taking these, or can i for example take them now indoors against a bright wall, make a master flat, and use this as a permanent flat for the rest of my pics?

I mean, the vignetting isn't going to change anyway unless i change scope, right?

And about "For anything faint, more subs and longer subs. No free lunch!!!", is 15 min at still not ling enough? I coudl try longer i guess, but i think i'm pushing the luimits for the DSLR as it's allready getting lots of hot pixels at 15 min. I Did take a 31 min exposure with the integrated dark frame noise reduction on, and didn't see any signs of pixel error though, so i guess it Will work with 30 min subs also as long as i get my darks working as well.

I'll give it a try though, but unfortunally probably not before in the fall again.

[ollypenrice]

As long as you don't alter the imaging train at all you can take the flats where you like. The camera has to stay in the scope and the focus must not change or the size of the dust bunnies will change. If I can't get my flats done at the end of the session I do them inside because I can't get the temp down outside in the daytime or keep the bright sunlight out. (Hey, life is hard in the south of France!!)

You cannot rotate the camera, though, between flats and lights. You might think the vignetting would be symmetrical but it never is. Look at yours and you can see that it isn't.

Flats are quite hard to get right but are utterly essential. It took me a while to learn this. An EL panel is the easiest way. You can't go wrong with a Gerd Neumann Aurora panel. Gerd Neumann jr.-Aurora Flatfield Panels-

Well, that is not quite true because in imaging anything and everything can go wrong and frequently does!!!

Have a care with bias frames. Don't subtract them as well as subtracting darks because the bias is present within the darks. I use bias frames only as darks for my flats, along with many other imagers.

Olly