In addition to the methods described in Fitzpatrick et al. (2005), we examined 56 video clips of Pileated Woodpeckers in flight, recorded by our field crews in Arkansas during the 2004-2005 field season. We searched for video clips that presented views of the dorsal and/or ventral wing surface, especially those presenting angles similar to the woodpecker in the Luneau video. We deinterlaced video stills from selected video clips, as we did earlier with the Luneau video. If Pileated Woodpeckers were larger in the frame than in the Luneau video, then we reduced the resolution to that in the Luneau video. If video stills were in better focus than in the Luneau video, then we added blur using Photoshop 6.0 until the bird and vegetation appeared similarly out of focus as in the Luneau video.
We measured wingbeat frequency of a known Ivory-billed Woodpecker flying away from its nest tree by analyzing a spectrogram of a sound recording made by A. A. Allen and P. P. Kellogg in 1935. We measured wingbeat rates of the woodpecker in the Luneau video and of Pileated Woodpeckers by counting the beats among video-frame sequences shot at 29.97 frames per second (=59.94 deinterlaced video fields per second).
On 15 March 2005, we filmed a reenactment of the Luneau video episode using life-sized models (right) of Pileated Woodpecker and Ivory-billed Woodpecker painted to resemble the respective species in terms of black-and-white patterns.
The Ivory-billed Woodpecker model had a typical narrow central black wing stripe on the underwing. The Pileated Woodpecker had a white underwing patch that was as large as the largest underwing patch of museum specimens of this species, to make the test as conservative as possible. In order to recreate features of the flying bird in the Luneau video, the wings of the models could be flapped rapidly by a system of strings and elastic bands to yield reasonably authentic angles and motion blur.
During the reenactment, all conditions were set to match as precisely as possible the conditions under which the original Luneau video was shot. The canoe with camera was moored at the exact location where it was when the bird took off from its perch. All sequences were filmed using the same camera (Canon model GL2), under similarly clouded skies, and slightly out of focus. We shot numerous video takes with different shutter speeds as low as 1/8 sec, in a specific effort to produce images containing motion blur and white-bleeding comparable to those in the Luneau video.