Tuesday, August 15, 2006

More Luneau chicken scratches

The de-interlaced, 4x magnified digital video clip from which these animations are derived is copyright 2004 by David Luneau, and is available from Science Magazine Online, among other sources.

This is rough, some frames (like field 316) are real clunkers and need major reworking. But here is my rendition of the first 0.6 seconds of the Luneau video flight sequence. I show only the bird, always centered in the frame. I primarily focus just on the wing shape, with less attention paid to the body. The sketches are estimates/interpretations of the wing positions at about mid-exposure, without motion blur. Other penciled annotations include the field number as assigned in the original Science publication, and a little arrow beside the body showing its apparent direction of motion relative to the background. The size of the bird's image is kept the same, I did not draw it shrinking with distance. Once again, this is a rough first cut...

344kB Quicktime animation

Notes --

The bird's positions and postures while it is behind the tree are of course conjecture. Yes, I realize the bird looks like a cross between a pigeon and a grouse in my drawings of these frames. The main point is to approximate the missing 1.25 wingbeats while the bird launches and rolls into flight attitude while remaining entirely concealed behind the tree.

I have rejected Cornell's closed-wing interpretation for fields 33 and 50 (the second and third frames) in favor of the extended-wing model shown. I agree with many commentators that this fits better with the rhythm of the visible portion of the flight sequence, the relative sizes and positions of the exposed parts of the bird, and the documented actions of Pileated woodpeckers launching in to flight from near-vertical perches. There are, of course, no movies of Ivorybill launch sequences available for reference. Note that, in my interpretation, the visible wing in field 33 is extended both upwards and away from the camera, hence its apparent length is forshortened. This behavior is demonstrated well in several of the Pileated comparison videos posted by Cornell on their Luneau video analysis web pages; plus it just makes sense that the bird will extend its wings away from the tree to prepare for the first power stroke that will carry it into clear space.

This sequence covers up to the point that the bird becomes more or less stationary relative to the background, just before it curves left and upwards. In the final frames of this animation, the bird is travelling almost exactly away from the camera. There is no point in this sequence at which any dorsal surfaces of either wing are definitely visible during a downstroke. This issue is not entirely clear for the left wing in the early downstroke frames of the first wingbeats (fields 233-250, 350-383) and has been the subject of controversy; I'll discuss that more in later postings. It is clear that the inner portion of the left wing in these frames is viewed close to edge-on; any slight shift could reveal dorsal or ventral surface. Again, more on this later.


Post a Comment

<< Home

Site Meter