M31 - More subs, worse image

[swag72]

Hi Folks - Well I am truely p'eed off. DSS has spent 8 hours today stacking a little ove 3 hours of M31 data. I previously had a little over 1 hour and stacked it and it was OK. Since adding the other data - I have lost ALL detail in the core area - It's now just a fuzz of blur.

I haven't got the example - That's now found the bin and my eyes are square from looking at a screen for hours on end. I guess my question is this.

I thought that the more subs I added the better the image would become - Clearly not. Would light pollution cause a problem? There was more light in my garden as the neighbours reported some street lights out. Have I used TOO many subs?

I know I've not got a picture to show you, because I am about ready to cry with utter frustration. But some general pointers would be good.

Thank you to those who listen to my incessant moans and questions.

[Starblazer]

It could well have been the result of one or two light polluted images. Go through your images one by one and choose the one's that look the clearest.

It's a long project but worth it for the results.

[swag72]

I will go through the images and weed some out. I have been really wracking my brains, and remember that I used the Endrophy (HDR) method, whereas before I used the Kappa Sigma - I think I'll restack overnight.

[Beyond_Vision]

M31 has a high dynamic range and it can get blown out in the core when trying to stretch the image to reveal the outer areas. DSS will automatically apply a stretch to the data and it may blow it out when it show the final result of stacking. You either can change the levels and curves function in DSS then save with changes applied or you can save with changes embedded but not applied and do levels and curves adjustment in something like GIMP or photoshop.

Another way is to layer and feather a less stretched version of the core on to image in photoshop or an equivalent package.

Regards

Kevin

[swag72]

Thanks Kevin - Is it better to start off from a file with changes embedded but not applied or with changes applied? I have always started off from the changes applied as the embedded image seems really hard to manipulate.

[Beyond_Vision]

Some people will turn in there grave if you attempt to use the tools within DSS but I find them ok for an initial stretch. In all honesty I have tried both and can end up with much the same results.

When the image popped up after stacking did you change any of the settings before saving it with changes applied?

[swag72]

No, I didn't change any of the settings.

DSS is having a churn again - I've discarded what looked like the most light polluted images, 15 minutes in total - Interestingly, the discarded images tended to have the higher scores attached to them by DSS, so the programme thought they were great!!

[Beyond_Vision]

That could be why the image would be blown out as the data may be overstretched and the channels out of balance.

First option

You can adjust the RGB sliders to align the peaks of the histogram and make the peaks sit at where the curve transitions. You can also change the curve type appled on the left for each colour and you can alter the shape. There is also a check box to apply changes to RGB at the same time or individually. You will also need to adjust the saturation to around 17-18%. Then save with adjustments applied.

Second Option

Save with adjustments embedded but not applied and do a levels and curves adjustment with photoshop or equivalent.

Regards

Kevin

[swag72]

What is the point of having the adjustments embedded but not applied?

[Beyond_Vision]

I think you can use them as adjustment layers in photoshop which are not actually applied to the image layer until a latter stage when the layers are combined.

In reality what most people will do is open the image in photoshop and not use the embedded information. In effect ignoring any stretches etc that have been applied by DSS. What you will most likely see when the image is openned is a dim image that needs to have it's levels adjusted and curves appled to brighten the image.

[James4]

Hey Sara, I also found DSS assigning good scores to bad frames - to be fair I haven't read how it assigns scores. Since then I now manually rename the bad ones to 'moved-image5' etc. I also found on a couple of runs DSS didn't come up with a good result (on M81 I remember). I trialed a copy of Nebulosity and it produced a way nicer image on the same stack. I don't let Neb pick my frames either ...

I think both programs are great, but its good to have a second option.

I would try running a smaller stack through with any tweaked settings before running your full set through again. Good luck! James