Focus stacking

kae1

All-Pro
Location
West Yorkshire
Name
Ken
Had my first attempt at using focus stacking with my Panasonic G90 this morning. Whilst on a small screen the image IQ appears "alright", on closer inspection it is quite mushy and there's some interesting effects on what should be straight lines. I'm putting this down to not using a tripod, then I'm wondering whether its because it is an Olympus lens on a Panasonic camera or is that stacking introduces some "added extras"?. Anyone got any experience and tips?

P1540407decrop.jpg
Join to see EXIF info for this image (if available)
 
Anyone got any experience and tips?
Looks pretty much fine to me at forum resolution. Is this a 100% crop?

You didn't mention what you're using to do the stacking but, in general, in camera and handheld stacking aren't the best for avoiding artifacts. However, no stacker is going to improve upon the in focus regions of the stack's input frames. If the problems you're seeing aren't in the input frames it may be worth stacking in Helicon or Zerene, if you haven't already, to check the G90's frame to frame alignment performance.

Also, no stacker currently attempts to reconstruct portions of objects which are obscured. Usually this occurs where out of focus blur expands the parts of the image occupied by foreground objects, washing out parts of the background. Another mechanism for this highlight trailing—since stacking algorithms prefer high contrast they'll preferentially take somewhat out of focus highlights over background pixel values—which occurs to a minor extent here along the left side of the lamp and possibly also in the top glass. Normally you'd retouch this.

Additionally, if this is an in camera stack from 4k post focus keep in mind it's working with 8.3 MP 4k frames which are maybe jpeg quality 70 due to the video encoding. So the output won't pixel peep like a 20 MP stills frame because the input didn't.
 
Looks pretty much fine to me at forum resolution.
Thank you so much for your detailed response which has given me a much better insight into how the process works and the potential limitations.

The picture was an uncropped incamera image generated from an automatic merge of all the separate frames and TBH was a spur of the moment unplanned attempt. I think if I try again with a bit more effort, i.e. the camera on a tripod then doing a review of each frame before merging, it might produce a better image.🤞

Thanks again.
 
Back
Top