home made twilight mode...

... for those suffering Nex envy... :smile:

Just give a try.

On "another" forum... this interesting Ctein's article (i'd forgot) was linked:

Now, a bit of an experiment based on that theory, Photoshop or equivalent needed:
- get 5 identical shots in a row (high iso!)
- have photoshop stack and align them, creating 5 layers
- number one (background) remains untouched
- the other four layers as mode=luminosity, opacity 25%

... done.
look at 100% and compare noise with one of the 5 originals.

I'm a bit busy to post examples right now, I've tried quickly with a series of B&W jpgs, and what happens to the noise is impressive... Hope it works with color too, let me know.

 

a quick test

Ok, here 2 100% crops, one single image and one stacked as reported:

They are in camera B&W jpgs taken for (other) testing purposes only. Five shots hand held, not very high iso (1600).
Not a micro 4/3 camera, but this makes no difference on the technique.
The entire point is that noise is random, different from shot to shot even if the shots are identical. So, more images you stack and more it will merge to a medium value, locally. I did not experiment with transparency percentages, I thought it was a good idea to have the stacked images sum up to 100%, but this is completely arbitrary and intuitive on my part.
BTW, it's how I think the "twilight mode" on the Nex is working.
Obviously, it's not a very useful technique, but perhaps there are circumstances to find a good application. Not applicable with moving subjects, or applicable but with severe limitations and requiring masking work.

 

@Narnian: yes, it's a power switch! :smile:
@mmzymxf: dunno. I am sure I didn't discover anything new, and remember a SW doing something similar. Yesterday I realized there's a simple way to do it with PS and thought to share here too. If I remember well, the super resolution stuff is based on the idea that micro movements will add detail when stacking images. In my uninformed opinion, it's more a pixel randomization matter (bayer sensors), and resolution is only apparently increasing cause of the refined tonal continuity. Which would be the reason the foveon sensor has an apparent better resolution...
But, apart from personal interpretations, this technique appears to work very well with noisy images! :smile: