Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

Filtering background grain


Recommended Posts

Hello all,

All of my pictures to date have suffered from some background noise or grain that comes through after extensive processing (more processing, more grain). It's not so much a gradient (I'm trialling GradientXTerminator which does on ok job on the colour gradients, but the 'noisiness' still remains).

I have tried cutting out the galaxies to seperate Photoshop layers, then blurring the background inbetween the stars however when I paste the DSO's back in, they have a brighter background and look like they've been stuck back on with glue.

Even blurring the edges between the fixed background and the galaxies isn't working very well.

Does anyone have any tips for sorting out the background (I'm aiming for the background to be dark and homogeneous) whilst still preserving detail in the DSOs?

post-18683-133877568206_thumb.jpg

Stacked .tiff: http://dl.dropbox.com/u/3717210/Leo%20Doublet%20%28M65%20%26%20M66%29%20-%202011-04-24%20Stacked%202x%20drizzle.TIF

Thank you kindly for any advice offered.

Mike

Link to comment
Share on other sites

First recommendation; Get more signal.

Second; Use Neat Image.

Third; Convert to Lab Mode in Photoshop and apply a small blur to the a and b channels. About 0.3-0.5 should do it. Different blur to each channel. Go back to RGB.

Dennis

Link to comment
Share on other sites

Mike,

I have had a play in Pixinsight.

I agree it needs more signal. The Colour Calibration, DBE and noise reduction routines in PI are excellent and can really sort out a problematic background.

5648640721_92e822fbcf_z.jpg

I uploaded to my flickr page too to display here, hope you don't mind.

Link to comment
Share on other sites

Thanks chaps,

John - I don't mind at all (in fact a lot of the kit used was yours once upon a time :)).

In terms of more signal, what would you both recommend? Longer subs or just more of them? This is an hour and 40 mins worth of five minute subs with ~5 deg C of matched darks.

This was taken from a dark site so I can now only add data from home - I am not sure if adding more frames (but less clear) would be a good thing for this image.

I'll try PixInsight although I only have the LE edition (which I only use for DBE) and the Photoshop lab workaround seems intriguing. However I expect the labs colour adjustment would negatively impact the galaxies too?

Cheers,

Mike

Link to comment
Share on other sites

Dennis's Lab colour won't affect the galaxies because the resolution is retained in the Lightness channel.

I agree that PixInsight is THE tool for background sky. Background Neutralization works superbly.

I get some slight noise reduction by using colour select to collect all the background pixels then I drop the saturation a little and also reduce contrast, the only time I ever touch that one.

I agree that Neat Image is good but in any of these tricks the thing is to do it as a layer so you can keep the sharp, high signal areas unadulterated by noise reduction.

I don't know DSLRs but I suspect that beyond five minutes without cooling may actually add noise. If this is true than more subs beat longer subs. If I'm aiming to present an image a full size then I would never expect a result in less than four hours with a cooled CCD at f5 to 7. More like 6 hours minimum.

One of my regulars uses physio cold packs around his DSLR. He has a couple or three in the freezer so he can rotate them.

Olly

Link to comment
Share on other sites

Hi Olly,

Thanks for the advice and the tips.

Beyond five minutes the noise starts to ramp up to unacceptable levels and instead of being more easily removed in dark frames, it starts to encroach across the entire histogram and sometimes adds nasty banding effects.

Physio cold packs sound like an ace solution. I'm going to keep my eye out for some small cold packs and some form of mechanism to conveniently attach them (perhaps flush to the underside of the camera).

As for adding data - I have an extra 2 hours of data. The trouble is that it's from a light polluted location. I've tried adding these into Deep Sky Stacker but I get worse results than I did with less (but from a darker site) data.

Best,

Mike

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.