Yesterday, I posted a first draft of a “focus merge:” a series of 48 photos, each taken at a slightly different position, to create a montage with most of the image in focus.
Although it was interesting to have as much as possible in focus, it is also possible to look for alternatives using subsets of the images. This leaves some portions of the image out of focus, but there is more to it than just that.
There is a type of artifact in a focus merge that is hard to avoid if you have a lot of images: haloes. These halos occur because items that are closer to (or further from) the midpoint of travel blur much more than other portions. It is hard for the software to remove these for the simple reason that the out of focus haloes literally block the stuff behind them.
In the example below, taken from yesterday’s image, the blur occurs in two ways: in the background, and around the edge of the leaf.
The top image doesn’t include as many images, it doesn’t try to go as deep, and the foreground object stands out very very clearly from the blurry object in the background.
There are other artifacts in a deep focus merge with many images, but I carefully edited them out. I drew the line, so to speak, and more complex blurs like the one above. As it was, it took about half an hour to do the ‘easier’ clean-up.
For the image at the top, I selected the image range I included with a specific idea in mind: I wanted the white pistils to stand out against the background petal. To achieve this, I looked for an image that had the pistils very sharp, and the petal behind them at least somewhat out of focus.
I had a range of choice; there were three images that had sharp pistils. I chose one that left the deeper part of the petal unfocused, but the edges focused. This was a total of 13 images, a little more than half.
Today’s version also used images that were 100% of the originals for the focus merge. It took a great deal longer to generate the merge, even though I used fewer images. Yesterday’s image was made from images that were 50% of the original.
Here is a detail from the above photo showing the pistil’s stigma and the anthers on the stamens.
I also lightened this image to show the coloring on the anthers more clearly. The original images were about 75% of full life size.