Oh M31 my Nemesis, is this vignetting?

[Jessun]

Hello all,

I have been relentless imaging M31 at every opportunity and I have now collected over 3 hours of the beauty.

After installing guide scope rings I can now easily frame and guide but the lack of accurate polar alignment (well, not accurate enough) shows up as a rotation of the image as the hours tick by, so every 30 I just reframed and continued.

Below is a jpeg of the core so far. This one still from a wobbly balcony in a city centre but S/N is king so it seems patience can help a long way extracting data. I chucked 154 2 min subs in the mix, 100% stack. I'll let the ol' laptop do another run later with the best 80%. I am curious to see if that makes a big difference. (I reckon you already have the answer to this )

The core shows the whole array of imaging errors, including black spot from something on the CCD for a quarter of the frames, egg stars, an erratic moving red object up in the right corner (helicopter anticollision beacon?), and the full book of optical errors haha BUT it's where I am at this early in the game and I am chuffed.

OK, subject of the post is however the second picture and it shows a massive colour shift from top to bottom and I am asking you if this is vignetting, and I wonder if it's caused by a non perpendicular CCD, and of course I wonder if this can be combatted in processing.

I'm determined to keep adding subs to this my first project!

All the best

/Jessun

[Jessun]

You guys leave me to learn the hard way

Spent some more time on this and saw Pixinsight and Noiseware being mentioned in other threads. So background somewhat corrected and noise also a bit reduced and I need an aspirine...

A mountain to climb this...

This is it so far then:

/Jessun

New crop SGL.tif

[JimStan]

I'm not sure where your color shift is coming from ( I will leave that to the experts ), but I can tell you that vignetting is a darkening of the field from the center to the edges of the image, as if you were looking at the FOV through a tube that was cutting off the light hitting the edges of your picture.

If you are actually showing us the whole frame, I would not say that vignetting is your problem.

You might be experiencing some sort of "amp glow", which is a result of the power being consumed by your camera, causing the sensor to heat up, and a coloration showing up in the image.

You might be able to eliminate that by taking flat frames. Look at that concept in the sticky notes concerning processing of stacked images. Good luck!

Jim S.

[johnrt]

If you have pixinsight use the SNCR tool set to green on the image - it will zap that colour shift for you.

[Jessun]

JimStan and johnrt thanks for your input!

I realize that the full picture would be better use to you to troubleshoot - I just like the "Shot-in-Panavision-look":)

I think the full picture will show exactly what you were talking about JimStan, the tube effect. The CMOS reported temp through APT was 27 degrees:cool:...

johnrt yeah I now have Pixinsight and Noiseware and below is a comparison of the original stack with some basic adjustments made to levels and curves in Canons own program and the latest edit which includes using Pixinsight's SNCR tool removing green, a BackgroundEnhance, a CanonBandingReduction and a DarkStuctureEnhance and finally also a Nightscene noisereduction in Noiseware.

It has taken the image leaps forward I think, but the tube effect is still present so I need to dig deeper in the menues of Pixinsight.

Finally I have attached a small crop from some red lines that I initially thought was something airborne until I found them in 4 more places, randomly spread it seems to me but perfectly identical pattern! What could these lines be?

Thanks again for your help!

/Jessun

PS uploaded images also suffer from jpeg artefacts, the TIFFs look nice.

[ollypenrice]

Did you take flats? These would cure vignetting.

If you are shooting in Lyon then the colour gradient will be massive LP. I'm south of you, near Sisteron, in total darkness (or pouring rain, to be strictly accurate right now!). PixInsight is the cure-all for that.

if your stacking won't cope with field rotation make sub stacks and align those by hand into full stacks.

Olly

[Jessun]

ollypenrice, no I didn't. I will get to grips with that very soon I think!

Of course that opens up a whole array of new questions... Every sub set has its own flats so for three nights you will end up with three complete stacks with all the essentials - three images basically and then you say you align the resulting images manually.

Ok, in the ideal world would you want to stack all subs in one go instead if it's possible? Would that even out noise better than aligning three separate stack results?

Would you adjust the sub results before aligning or adjust only the final image?

I am curious on how to add data day after day, as here downtown Lyon I only have about two hours to image before my target drops too low into the light and the smoke:p...

The slight field rotation seems controllable in software tho, just happy to have an easy guide star for my M31 conundrum.

I am mostly staring at Pixinsight's plethoria of functions. I am at the foot of that mountain too:p I am sure it can cure a lot of my errors!

Below is M31 Panavision style again. Looks cinematic I think and some artistic freedom is hopefully allowed:rolleyes:. This one has the background extracted that I then used as a layer in Paint.net and combined using 'colour burn' and it seemed to take out the funny colours. (PS trial ran out...)

Below also I have included two subs about an hour apart showing the red dot that made the star trails in the stack. Looking at it closely it's exactly the same spot pixel by pixel and I have concluded it's a frozen CCD pixel. It moves from frame to frame a pixel or two, BUT it's my field that moves! Heureka! Glad I solved that mystery...

Hope you're not flushed away entirely down there in the south! When things dry up I might just drive down there!

Thanks for your input

/Jessun

[JimStan]

I really like Paint.net. It does almost everything that PrintShop does, and it is FREE ! ( Free is good ! ! ! ) Nice improvement, and a really great picture ! You can be proud of that, as far as I am concerned!

Jim S.

[kookoo_gr]

the dot you show looks like a bad pixel. you can remove it with darks. shoot around 15-20 and create a master dark frame in order to use it with your frames

[dmahon]

Pixinsight is, for my mind's way of working, a real pig to use.

I do, however, like the DBE function (though every time I use it I have to watch the YouTube video first) which would be very useful to you here. I can't get on with the rest of it at all and am afraid that I would never buy something that I would end up tearing all of my hair out trying to use, especially at that price.

Fortunately PixInsight LE was freely available and offered the DBE function (although it is now very hard to find a copy legitimately, as it has been pulled by the publisher and should not be distributed on websites - though I am not sure whether or not it may be distributed in other ways).

[Jessun]

Thanks JimStan, appreciate it!

kookoo_gr yeah spot on there, but I was too lazy to do darks on two of these nights. Paying the price...

dmahon, I forked out for the full thing... I did the background extraction but could not get very much further from there. Doing I think substraction produced a reasonable result, but it was a simple job in paint.Net to flatten an inverted background with the main image. I am sure pixinsight can do black magic but for now what you're saying is true - it's difficult to use.

Got a flats panel coming this way btw. At this poing in my climb up the mountain I'd rate flats far more important than darks. Seems noise is mostly random and can be stacked to oblivion, whereas the other optical artefacts add up.

/Jessun