Editing the nadir

Carpebble-logo

Processing the photos part 2

In this series of web pages, I have documented the process which I currently use to create 360x180 degree panoramas. This page explains how I create a composite nadir image from two raw images. If you want to know how I capture the two source images, then look at my page on taking the photographs. During my processing, I use tonemapping to increase the detail visible in both bright and dark areas in my photographs, then I edit the nadir to create a composite image, before stitching using Hugin and creating a panosphere.

Sometimes, if I shoot carefully, it is possible to create a panorama in Hugin without first creating a composite nadir image. If things work out well, it is possible to shoot three nadir photographs, two through the tripod with the head rotated 180 degrees between shots and then once, hand held without the tripod. If these shots are imported into Hugin, sometimes Hugin will make a good job of merging these images. In order to remove people, shadows and the tripod, use masks in Hugin. Click here for a panorama created only in Hugin. The original files are in the same folder on PicasaWeb, if you want to try creating a panorama from them.

Otherwise, use the method described below to create a composite nadir before importing into Hugin.

Creating a composite nadir image

This procedure was suggested in outline by Bruce Torrence on Flickr, who recommended that I leave the camera on the tripod to take the nadir image. Previously I had only captured the nadir image hand-held, which was a nightmare to stitch correctly. Now, I take one image on the tripod and a second image as close to the same position as I can, hand-held. This is a problem if the shutter speed is longer than 1/40th second, as it is difficult to hold the camera still enough to take a sharp image.

After trying to load both nadir images into Hugin and use the mask option to cut out the tripod, I settled upon my current method, which is to create a composite nadir image in GIMP before reading the combined image into Hugin.

This method uses layers and layer masks in GIMP. I imagine a similar process can be achieved by a proficient Photoshop user, but using layer masks is not usually one of the first things that a new user would attempt in either package. Sorry, but this outline may mean relatively little to you unless you are already pretty familiar with the GIMP.

The procedure I use is this:

    • Open both nadir photos (with and without tripod) as layers in GIMP.
    • Put the version without the tripod on the top, make it partially transparent and rotate / stretch and skew until it closely matches the shape and position of the lower layer with the tripod.
    • Make both layers fully opaque and add a transparent (black) layer mask to the top layer.
    • Edit the layer mask, using a white brush to paint over the tripod, feet and shadows, to remove as much of them as possible.
    • Merge visible layers.
    • Save as a new image using minimal compression.

The next two pages deal with stitching using Hugin and creating a panosphere.

Edit nadir in GIMP