Something I'd see if they could do is just as you said, use some kinda image mapping to get a rough estimate where each picture is so it could smartly adapt them when combining them.

http://www.cambridgeincolour.com/tutorials/high…

I’d like to use it for some photography of the night sky where image stacking is often used to bring out faint images (see http://www.astrostack.com/). For image stacking you need to *add* the pixels in the images together rather than filter, such as selecting the median values (which I suspect is similar to what it currently does).

Another useful feature (essential for image stacking) is the ability to subtract a ‘dark frame’ from the photos before processing. A dark frame is a long exposure photo taken in pitch black, e.g. with the lens over on. Then bias in the individual CCD elements show up. Subtracting the dark from from the individual frames can improve long exposure photos greatly.

These features would be great for me – a complete mobile astrophotography lab in my pocket.

I’d like to use it for some photography of the night sky where image stacking is often used to bring out faint images (see http://www.astrostack.com/). For image stacking you need to *add* the pixels in the images together rather than filter, such as selecting the median values (which I suspect is similar to what it currently does).

Another useful feature (essential for image stacking) is the ability to subtract a ‘dark frame’ from the photos before processing. A dark frame is a long exposure photo taken in pitch black, e.g. with the lens over on. Then bias in the individual CCD elements show up. Subtracting the dark from from the individual frames can improve long exposure photos greatly.

These features would be great for me – a complete mobile astrophotography lab in my pocket.