Antennae

[Placidus (Mike and Trish)]

Originally Posted by SpaceNoob

Damn fine job with those fainter parts, I tried my best to capture that stuff but I just couldn't get them to rise much above the noise, not sure if it's because I need more aperture, tiny pixels or because of light pollution. 30minute subs showed an average background ADU of ~4500 pointing West (towards the Brindabellas). The fainter extensions were around 10-30 ADU above the background for me. Pretty much clutching at straws trying to push them out in processing lol.

That 20" really sucks down those photons, are you at a dark site?

Thanks kindly, Chris.

We're at a pretty dark site, midway between Orange and Wellington, 50 Km from either, and about 45 Km from Molong. The nearest house is 3km away. We're 660 m above sea level. There's nowhere around the Central West that is really high (like Mt Canobolas at 1300m) that isn't also a fog trap and cloud magnet.

A typical dark subtracted luminance 1 hour sub put the average for the entire length of the antennae at 3439 +/- 536 counts (mean+/-sd), and background at 3315 +/- 153 ADU counts. A difference of say 124 ADU counts or about 250 photons. The outer bits are of course proportionally fainter, but hard to tell where they end.

Best,
Mike

[NOMH (John)]

Great job with this target! I like the structure you captured and the color of the galaxies.

I didn't realize it was so dim and required as much exposure. This is on my to do list and I can barely reach this from my location. It may take too many nights at a dark site to get even close to what you have.

JB

[multiweb (Marc)]

An old favourite. Incredibly deep and detailed. One for the poolroom.

[Placidus (Mike and Trish)]

Originally Posted by NOMH

Great job with this target! I like the structure you captured and the color of the galaxies.

I didn't realize it was so dim and required as much exposure. This is on my to do list and I can barely reach this from my location. It may take too many nights at a dark site to get even close to what you have.

JB

Thanks John. The main body of the two colliding galaxies and the inner parts of the antennae are not so difficult, only the extreme outer ends.

Originally Posted by multiweb

An old favourite. Incredibly deep and detailed. One for the poolroom.

Thanks, Marc! In other news, sticking my head out the window as I type, I can see the pre-eclipse moon peeking between seriously heavy clouds. Not long to go.

[AlexSavoy (Alex)]

Truly look amazing. Beautiful colors and amazing quality. Work of art

[cometcatcher (Kevin)]

That's an awesome shot! I had fun imaging this a while back with the 8". (Through your finderscope )

[strongmanmike (Michael)]

Originally Posted by cometcatcher

That's an awesome shot! I had fun imaging this a while back with the 8". (Through your finderscope )

Ok Kevin...how many peoples finderscopes have you stollen..??

Mike

[cometcatcher (Kevin)]

Originally Posted by strongmanmike

Ok Kevin...how many peoples finderscopes have you stollen..??

Mike

This is going to sound dumb, but I would love to try my K-5 on that 20".

[Bassnut (Fred)]

I imaged this and found the antennae really hard to bring out, required masking out and seperate super stretching. Im guessing you didnt go that far, so this result is very pleasing indeed guys, well done!

[Placidus (Mike and Trish)]

Originally Posted by Bassnut

I imaged this and found the antennae really hard to bring out, required masking out and seperate super stretching. Im guessing you didnt go that far, so this result is very pleasing indeed guys, well done!

Thanks, Fred.

Grasshopper say: Easy question needs complicated answer. We used a mask, but one generated mathematically, not painted on by hand, whose only goal was to protect the galaxy core, not to "mask in" the antennae, which would be cheating. We produced two versions of the final image:

A: very strongly stretched after strong wavelet noise reduction, to show the antennae.

B: Deconvolved but with very little stretch, to optimally show the galaxies.

We then produced a mask M by low-pass filtering (blurring) image B with a filter constant of about 100 pixels. The final image was image A where the mask was dark, image B where the mask was bright, and pro-rata in between. The mask generation process did not involve optimizing the appearance of the antennae. That was just lots of exposure time.

Distinctly dodgy would be for example use PhotoShop to manually paint a mask on where you think the antennae should be, or where you think the galaxies should be. That way you just get what you were hoping to find, rather than what's actually there.

Legitimate is to adjust the black and white points on the mask, because that, plus the filter constant, are just 3 degrees of freedom, know nothing about the presumed shape of the antennae, and is unlikely to produce a pair of antennae out of nothing.

Hope that incomprehensible rave helps.

Best,
Mike