Pulling out the faint stuff with Topaz Labs Adjust

[Xiga]

Hi guys

So i've been mucking around in PS, playing with the Topaz Labs 'Adjust' package. I was just experimenting tbh, but i may have stumbled upon something that others might find useful. Basically, with the right settings it seems like it's doing a pretty good job of pulling out very faint nebulosity, and with hardly any work required at all. I'll start first by explaining the settings that i used.

I had been selecting Shadow Recovery each time and changing various settings. But i think you can actually save a New Preset, which would speed things up a bit. Here are the settings i was using:

As you can see, it's basically a combination of Exposure adjustments and Noise Reduction. 

The 1st two exposure settings i just left on their defaults, everything else was turned off. And in the Noise section i left the 1st setting on the default, but dropped the NR strength way down, as the default of 0.58 was way way too high. 

That's it! The only other thing i did, which is actually very important, was i applied a Layer Mask which was simply an inverted copy of the image itself. The mask had a basic Levels adjustment, just to ensure the background was as close to white as possible, with the stars bloated somewhat to give them some protection, as well as the high signal areas too of course. I then proceeded to run the filter no less than 7 (!) times, just to see how far i could take things. 

A few notes on the images below:

1. This is the only image of my own which contains faint nebulosity, hence why i have used it (not because it's particularly deep or anything). It is based on 16 x 600s subs with a Nikon D5300 and an IDAS-D1 filter, taken from a Green zone (Mag 20.5) and with streetlights nearby. 

2. The 'Before' image below was my LRGB stack from back when i shot this last January. I created a synthetic Lum and combined it with a Colour layer. Only basic processing has been done on this (and no noise reduction or sharpening). 

3. We're not comparing final images here, remember. I'm mainly just concerned with the areas of faint IFN in the background. 

'Before':

'After':

As you can see, the filter has really brought out the IFN.
Now, i'm far from an expert on PS, but as far as i know i think IFN is meant to be quite hard to capture and process, is that right? And does it not usually also require a LOT of exposure, and from very dark skies? (at 2 hrs 40 mins from a Green zone the image above fulfills nether of these). The signal is there alright, but it's pretty faint. I'm sure in the hands of an experienced image processor, they would no doubt be able to manually pull out the IFN, but i'm guessing that would be very hard to do indeed, whereas the above method is pretty fool proof. 

It's by no means perfect of course. In the above example, the Noise Reduction is a bit too OTT, but perhaps with more tweaked settings a better balance could be found. I'm only posting this, not because i think it is necessarily great, but more in the hope that others will also experiment with it too, and maybe together we can all find out a bit more about how best to use this for certain images (or even just parts of an image). 

Edit - i just realised that in the 'After' image above, i accidentally hadn't reset the settings in the 'Detail' section. So there will be some slight sharpening going on in it too (hopefully not much, given the mask that was being applied). In any case, it shouldn't really have effected the IFN in the background, which was the main objective. 

[Demonperformer]

There are certainly some powerful plugins for PS out there and this seems to be one of them. My version of PS is too early for it, but I'm sure it will be of interest to others.

[wimvb]

On 16/10/2018 at 02:18, Xiga said:

As you can see, the filter has really brought out the IFN.

AfaIk, this is not ifn (see link below). But it's an interesting technique nonetheless. Thanks for sharing.

https://www.astrobin.com/forum/c/astrobin/generic-discussions/misuse-of-the-term-ifn/

[ollypenrice]

As Wim says, no IFN here. The IFN is much harder to find and even harder to find in colour.  However, the dusty features you've brought out are in good agreement with those of my very long exposure M45 so we can't both be inventing them!  (I'm always wary of 'creating' features in the faint stuff so comparing one's results with others is a good idea.)

Olly

[RichLD]

I'm a big fan of Topaz plugins - Adjust, Clarity, Detail and Remask get a lot of use in all of my photographic exploits - highly recommended!

[Xiga]

15 hours ago, wimvb said:

AfaIk, this is not ifn (see link below). But it's an interesting technique nonetheless. Thanks for sharing.

https://www.astrobin.com/forum/c/astrobin/generic-discussions/misuse-of-the-term-ifn/

 

14 hours ago, ollypenrice said:

As Wim says, no IFN here. The IFN is much harder to find and even harder to find in colour.  However, the dusty features you've brought out are in good agreement with those of my very long exposure M45 so we can't both be inventing them!  (I'm always wary of 'creating' features in the faint stuff so comparing one's results with others is a good idea.)

Olly

Thanks Wim and Olly for pointing this out. I had no idea, i had just assumed it was IFN. Always good to learn something new ? 

Thanks for confirming the faint stuff agrees with you long exposure of M45 Olly. That's good to know! I might keep this technique in the back of my mind in case i'm ever struggling to bring out any faint stuff in future. 

[Xiga]

4 hours ago, RichLD said:

I'm a big fan of Topaz plugins - Adjust, Clarity, Detail and Remask get a lot of use in all of my photographic exploits - highly recommended!

I agree Rich. I haven't used Clarity or Remask myself, but i can say that i have now moved to solely using Topaz Detail for all of my sharpening. It's very impressive i have to say, the control you have over Small, Medium and Large scale structures actually reminds me in a way of using the different Layers in Registax. Also, the InFocus one is probably the closest thing a DSLR user can get to a Deconvolution-type tool. I've only tried it once or twice, and it did tighten things up a smidge for me, but i need to use it more to really know for sure. 

I'm starting to think they have some very smart mathematicians working for them over at Topaz Labs.