Monday, March 13, 2006

Luneau video, frame 33.3

An anonymous reader emailed me the following:
I think frame 33.3 (that Fitz et al interpret with a sketch of the upright woodpecker behind the tree) is particularly key in the discussion of the Luneau video evidence. You do a really good job of going through the video on this page, but your interpretation really jumps out when you look at a blown up and slowed down section of the video.

With the framecaptures, all you get is blobs, but when you look at the blob move you get the gestalt (the sequence in motion is greater than the sum of its parts). It really becomes evident that this is an underwing (complete with black trailing edge and no
black median stripe) on a down stroke and not a bird sitting up with a folded wing. There is no way that his sketch of the bird half hidden behind the tree is accurate.
...
The wings are also "bending" because this bird is flapping hard to gain altitude, it isn't gliding down - it has to leap up and beat its wings for all it is worth to get lift and steering etc ...

So, what is needed is a clip like the one below - in slo mo -AND a sketch (also below) - that shows the way the bird is positioned. Your case shows best when it is in motion ... the freakin taxidermy wing isn't moving and the feathers aren't being forced to bend the way that a bird's wing does ...
Please take a look at this video clip from the Luneau video, blown up and slowed down:

Luneau video clip, MPG2 format
Same Luneau video clip, MOV format

In the picture below (left side), the emailer provided a very crude sketch to show the bird's general position.


(The sketch at left was overlayed onto Figure S1 from Cornell's Science paper. Please note that you can purchase a copy of the complete Luneau DVD here.)

Please note that I've written more about the interpretation of frame 33.3 under item 7 here.

47 comments:

Anonymous said...

Is that sketched wing anywhere near big enough to match the bird that appears to the right of the tree a few frames later?

Anonymous said...

The sketch isn't done by a trained bird artist ... it was done with a computer mouse after drinking a few beers ... all it shows is the very general position of the bird in contrast to the nice sketch that was published in SCIENCE.

If you play the slow motion, loop clip ... just sit there and watch it a dozen times .... you can clearly see that this is not a bird perched on the side of a tree it is a bird that is leaping away and opening its wings as it takes off ... it is all in the gestalt of the clip ... you can draw a caribou on the still capture and make it plausible ... but you can't come to cornell's sketch if you see the thing in motion ... it is a Dan Rather moment for Fitz et al.

Anonymous said...

This is hardly new. And in fact that sketch shows that it does NOT work. The interpreted tail and wing tip (the only parts of the interpreted bird actually visible) are far too close together for the size of the bird when it emerges from behind the ree trunk. And it is implausible that the wing could have gotten fully extended as sketched without having revealed more of the underwing in the previous frame. Wings unfold laterally, they do not pop up vertically.

Anonymous said...

Yes, this does help. It's clear when watched "in motion" that the bird is not perched as Cornell says. That white patch is a moving underside of the wing.

But in any case, I cannot believe that anyone at cornell would stake their reputation on any of this rather flimsy "evidence". There may be Ivory-bills left in the wild but you wouldn't know it by anything that Cornell has done.

Anonymous said...

Yes! Watch this slowed down video over and over again. Let it sink into your brain. Now Cornell would have you believe that the bird is not visible, then suddenly he leaps, steps, or magically appears just sitting there. Then what? Walks back behind the tree before taking flight?

No! Watch as the video instead shows the tail and wing flap of a bird taking fight. Not a sitting bird.

Anonymous said...

Wings unfold laterally, they do not pop up vertically.

What if the bird had it's head more or less away from the camera. and tail more or less towards the camera?

To me it seems that some people strongly seem an IBWO, some people strongly see not an IBWO, and neither side can convince the other. That must mean both have plausible cases, the video doesn't rule out Pileated. If the video doesn't rule out Pileated (or any other explanation), it's inconclusive... And if it's inconclusive, isn't saying it's an Ivory-bill without further evidence and pretty big risk?

Anonymous said...

I was promised a black trailing edge, yet I do not see it. I don't even see it in the still in this post.

Anonymous said...

Wow, that's pretty clear. It's a darned Pileated--dark trailing edge and all! Is this the smoking gun that will blow the lid off Cornell's Ivory-billed Cold Fusion Confusion!

Anonymous said...

Personally, I don't neet the "six pixel bird", or the "perched bird just prior to take off," or the "size estimate." The obvious white pattern on the upperwing in the later flaps, the evident white on the dorsum in a few frames, and the flight style and wingbeat rate are sufficient for me. I have read the alternative analyses for all these points and are utterly unswayed by them.

So dispute the "six pixel bird" and "frame 33.3" all you like; they are not necessary. To my mind there is sufficient evidence of the bird's ivorybill-ness in the rest of the video to make those points unnecessary. If the Sibley et al. article does not adequately address the rest of the video, then it will not have actually provided a comprehensive demonstration of why Cornell's fundamental conclusion (the video shows an ivorybill) is correct. It will have merely picked a few nits.

Anonymous said...

Thanks for this video clip. I have been following this blog and other discussion lists, and have remained agnostic about the bird's existence. I have nothing to gain if the bird doesn't exist, have no axe to grind, and am just enjoying the whole (or)deal.

However, I have to say that this is the first piece of evidence that I have found compelling and conclusive. That's a Pileated Woodpecker. There's no need to discuss or go through some sort of mental gynastics to interpret it. If you cannot see that, then have a look on a bigger monitor or one with higher resolution. Again, I have no interest in convincing anyone or proving anyone wrong.

The id of the bird in the video is (finally!!!) clear as day - a plain old Pileated Woodpecker.

I still don't know if the Ivory-billed is out there, but I do know now that it was not captured on video.

Anonymous said...

It's not the black trailing edge that pileated believers must show. It's the ivory bill of a true Ivory-billed woodpecker that is missing so far.

Cornell, show us the ivory bill!!

Anonymous said...

The "black trailing edge" is a video artifact. All bright objects in the video are given a spurious dark fringe by the camera. Look at the oar and the passenger's hand in the foreground in the original video (they have been cropped out of the posted image)

Anonymous said...

Watch the video being played, it's very hard to see from the still. A thin black area appears AND disappears with the appearance of the white.

Besides what a previous commenter said about the black edge being an artifact which is apparent on other objects as well, I must ask why is it a thin black edge? On a Pileated, the black edge is fairly thick all around the white (except on the leading edge, of course). Why is it visible on the alledged wing tip but not around the secondaries (or at least much thinner according to your theory)? Are you going to say that it's because those feathers are flexing due to the wing being in a downstroke? How could the bird be flapping it's wings downward given that it's right next to a tree? There's no room for a downstroke. Also note that there are areas where the green of the background meets the white with no black between.

Anonymous said...

It's not the black trailing edge that pileated believers must show.

If you're going to say that there's a black trailing edge, you do have to show it.

Anonymous said...

Can you prove it is an artifact of the camera? You can't, anymore than anyone can prove it's the trailing edge. Besides, only the TOP of the canoe paddle shows a black margin, and it's much wider than the band shown by whatever part of the bird is in the video. INCONCLUSIVE!!

Bill Pulliam said...

I have watched the video through several times, and I have to say my opinion about that first glimpse of the bird has changed. I was never quite comfortable with the Cornell sketch, because the white looked too extensive to be a folded Ivorybill wing. However, the vertically-extended pileated wing suggestion does not seem sensible to me either.

What I have noticed now is that the timing and location of that white flash does seem to be perfect for it to be the white of a wing that is just starting to be opened. It fits the beat and location to be the start of the first upstroke of the sequence of wingbeats visible after the bird clears from behind the tree. However it does not extend far enough vertically to represent a fully-extended wing. So I now would interpret it thusly:

What we are seeing is a flash of underwing white on the bird's right (our left) wing as it just begins to open its wings and launch itself to its left (our right) away and behind the tree. The wing is just opening, not fully extended, nor fully closed. And the view is of underwing, not folded upperwing. The black edge does indeed appear to be very narrow or nonexistent, within the limits of what can be resolved under the "black fringe" artifacts.

Unfortunately, if the wing is just begining to open, and we are viewing its undersurface, we are probably looking at the wrist area, not the trailing edge. Both species would show a major flash of white in this case, and I don't think the video resolution is adequate to make out enough detail to distinguish the two. Thus, we are left with only the rest of the video, and the underwing-upperwing arguments that have been thoroughly covered (and irreconcilably disgreed on, it seems) here. Evidently the smoking gun has fizzled and failed to fire in either direction.

Anonymous said...

All bright objects in the video are given a spurious dark fringe by the camera.
Not all of them. Look in the background. There are several large vertical light patches. If you look at the one directly above the paddle handle, for example, you'll see (if you freeze the video) that most of the left side of the shape is not bordered by black. In addition, the width of the black around the paddle and hand are uniform (the spurious dark fringe), whereas the black around the bird's wing is thicker near the tip of the wing, which is what you would expect from a pileated's wing. The thickness around the wing also changes with each frame, which makes it less likely that it's an artifact of the camera instead of the movement of the black feathers on the wing of the bird. At best, the spurious fringes appear inconsistently around the bright objects in the video.

Anonymous said...

"At best, the spurious fringes appear inconsistently around the bright objects in the video"

Which is a real pain when the two species in question are largely distinguished in flight by the black-white pattern

Anonymous said...

I find the Luneau video, overall, immensly frustrating. It refuses to give up definite marks. Is that a black trailing edge or an artifact? It looks like there is white on the back in flight, but where is the black band under the wing? If it were just "bad" and "useless" we'd not still be doing this nearly a year after it was made public. The same goes if it would reveal ANYTHING that nails the ID, either way, without argument. It appears to many (including me) to show some strong ivorybill features, but they just refuse to go over the line to pass the "grandparents" test (you can show it to your non-scientist grandparents and they will understand it too). So no matter how convinced we may be, we will never convince all reasonable people either way.

And so we reach a scientific stalemate. It will be resolved in the future when either (a) no one has ever been able to get better evidence, and even the most stalwart of the rational believers will concede that even if it was here in 2005 that was the last gasp, and it is gone now; or (b) we get an indisputable image, a dead chick in an abandoned nest, a fresh eggshell, something, anything that will not require fighting.

Anonymous said...

Which is a real pain when the two species in question are largely distinguished in flight by the black-white pattern

True! But let me try to clarify my point and leave it at this.

1) Not all light regions in the video have dark borders around them.
2) Those that do, which we know to be artifacts (e.g. paddle and hand), have a border of fixed width. (And I think that the fact that the width is uniform is partly what makes it look artificial. Well, that and the fact that we know it shouldn't be there ;) )

From this I concluded (rightly or wrongly, you can decide...) that if a light region has a black border that is not uniform in width, it's not an artifact of the camera.

3) The black border around the bird's wing is not uniform in width - it's thinner around the secondaries than on the wing tip.
4) The width changes as the bird's wing moves.

So, I don't think the black trailing edge or wingtip is an artifact. It looks like a normal pileated's wing. And the fact that the black is thicker at the wing tip, just like a pileated's wing, I believe, is very strong evidence that it is, in fact a pileated.

Otherwise, the camera has somehow both added more black where it "belongs" (the wing tip), and then changed the thickness of the black artifact when the wing moves. And it has added it there but not in other light patches in the video. These just seem to be too much of a coincidence to be considered plausible.

In any case, let's hope the upcoming article sheds some more light on the video.

Anonymous said...

It's a Pileated! The odds that some dude in a canoe would get the last "bigfoot" video of the last Ivory-billed Woodpecker at this time and in this universe are so huge as to be laughable.

Bill Pulliam said...

"So, I don't think the black trailing edge or wingtip is an artifact. It looks like a normal pileated's wing. And the fact that the black is thicker at the wing tip, just like a pileated's wing, I believe, is very strong evidence that it is, in fact a pileated. "

If it is a fully-open wing, yes. But if it is a wing just beginning to open, and the highest part we see is the wrist, not the wingtip (the geometry makes more sense to me that way, the first move a bird usually makes when it is taking off is to push the wrist away from the body, then open the outer wing and the primaries), I'm not sure how different the wings of the two species would look in that case. Especially blurred.

Anonymous said...

And the fact that the black is thicker at the wing tip, just like a pileated's wing, I believe, is very strong evidence that it is, in fact a pileated.

...except the Sibley guide shows the black edge of the secondaries is only marginally thinner than at the end of the wing (and if you consider that this includes the "fingers" of the primaries which likely might not show up in a blurred video, the black of the end of the wing might appear thinner). The National Geographic guide shows a black area that's of equal thickness around the entire wing. I've seen photographs supporting this, so is it really safe to say that the black trailing edge is thinner around the secondaries, thin enough to practically disappear in a video where the black is visible at the end of the wing?

Anonymous said...

If it is a fully-open wing, yes. But if it is a wing just beginning to open...

I agree with your point, here. However, it seems clear to my eyes that we are seeing the black tip of the wing. The wing has begun it's downstroke before disappearing behind the tree. The trailing edge of the wing changes direction and begins to move to our right. The downstroke would not occur until after the wing has fully opened, right?

That the wing is beginning its downstroke and is fully opened seems so obvious to my eyes to the point of not needing to be argued for, but here we are. It's a bit frustrating trying to argue this point because we seem to disagree on the fundamental raw data, even before trying to interpret it. I don't know if anyone else is seeing it as I describe. If there is some disagreement here, then the earlier poster's point about not being able to resolve this among reasonable people is well taken.

Anonymous said...

That the wing is beginning its downstroke and is fully opened seems so obvious to my eyes to the point of not needing to be argued for, but here we are. It's a bit frustrating trying to argue this point because we seem to disagree on the fundamental raw data, even before trying to interpret it. I don't know if anyone else is seeing it as I describe. If there is some disagreement here, then the earlier poster's point about not being able to resolve this among reasonable people is well taken.

Again, how does a bird that has its body right next to, or attached to (with its feet), a tree begin a downstroke, particularly with a wing that is about a foot long? Is there any room for that downstroke? Seems like it would just end up injuring its wing against the tree.

But I agree that it seems that we've reached a point where the outstretched wing crew won't be able to change the mind of the folded wing posse and vice versa.

Until Thursday, my friends, I bid you good night.

Anonymous said...

...except the Sibley guide shows the black edge of the secondaries is only marginally thinner than at the end of the wing...

Well, I looked at the big Sibley guide and measured the longest black segment of in the secondaries area and compared it with the longest black section in the "fingers" area. By my measurement, the "fingers" are about twice as long as the secondaries. I would characterize this as "noticeably" rather than "marginally". However, I think you still make a fair point, because the Nat'l Geo guide does portray them as being of similar thickness, and a quick Google search of pileated wing images shows a good variety of relative thicknesses, which I wasn't aware of.

What I see in the video looks like the upper wing of the bird on the right in the painting on this page by Louis Agassiz Fuertes (third from the top):

http://www.50birds.com/givorypileatedcompare.htm

is it really safe to say that the black trailing edge is thinner around the secondaries, thin enough to practically disappear in a video where the black is visible at the end of the wing?

I take it your point here is that if it's a pileated, then the black would be equally thick around the wing. And if it is equally thick, then how can I explain why one part of the blackness nearly disappears (secondaries) while the other is still visible (primaries)? (If I've misunderstood your point, my apologies.)

As far as the relative thickness goes, it looks like there is some variability, though the "fingers" appear longer in Sibley and elsewhere, so I would venture to say that that's not out of the ordinary.

I think the trailing edge almost disappears because as the bird begins its downstroke, the trailing edge of the wing moves from being "open-faced" towards the camera, to the tips of the secondaries pointing to the camera. The primaries stay pretty much "open-faced" in the video.

I am not basing any of this on how I believe a wing flaps or any theory, or trying to come up with an explanation to support my argument. This is just what is apparent to me in the video, and the wing movement seems perfectly natural to me. To put it another way, I didn't watch the video with the belief that it was a pileated. I was unsure and hoping for ivory-billed. I believe it is a pileated as a result of having watched the video.

Anonymous said...

Hmm... I can see the point, but the difficulty I have with that being a fully extended wing is that the bird's tail appears to still be pivoting. So this would require the wing to actually be stretched forward, but wings don't generally do this. Granted there is so little of the bird visible this is hard to judge. If it is the underwing of the right wing fully extended, we have unobstructed views of this part of the bird about 7 and 8 frames farther on. And there the controversy starts again, because the underwing appears almost entirely white with only black tip, and no black lateral bar.

Anonymous said...

I have Luneau's DVD and I've pulled the 4x tiffs into Photoshop in multiple layers so I can quickly flip back and forth between frames. I also use the video on Cornell's site for scrubbing the frames:

http://www.birds.cornell.edu/ivory/rediscovery/videowide/IBW4xFull

The video from the original post here was of poorer quality. I would avoid using it.

Let me also say that I have shot and viewed around 300 hours of miniDV video, not that that makes me an expert.

I use the millisecond frame naming from the 7th page of the Cornell/Science supplement pdf in interpreting what I think can be seen of the bird:

http://www.sciencemag.org/cgi/data/1114103/DC1/1

0.0ms - nothing definitive

16.7ms - a very narrow band of white first appears to left of tree. Seems unlikely to be an artifact since it doesn't show up in 0.0, it is restricted to the same general area that white will appear in 33.3, and it's brighter than the surrounding pixels. There isn't any detail to it, but it would show smooth motion of the white from 16.7 to 33.3. There is no surrounding darkness to its left. But it is extremely narrow so nothing's definitive.

33.3ms - bold white of the wing appears (definitely not an artifact :) ). Given the high contrast of the white and background one needs to be particularly careful interpreting what appears along the edges since this is where artifacts are most likely to be introduced (for an obvious example of this, look at the upper edge of the aluminum of the paddle). It looks like there might be some dark above the white, it is darker and more sharply black but with the high contrast can't be sure. The notch in the white above looks similar to the 4x notches in the man's fingers. To the left of the white it is tough to make a call. It is darker there but I think it's subject to interpretation as to whether that is bird or artifact. A bit below the white I think you can make an argument that you can see a tail pointing down when viewed in sequence with 50.0 and 66.7, nothing definitive though.

50.0ms - The bold white of 33.3 has been diminished to a paler white which I would interpret as motion blur where the white was present for only part of the frame as the wing moves to our right and behind the tree. The white is thicker at the top than it is at the bottom which looks like the white shape morphing or twisting from 33.3. There is a dark patch above the white which looks like it might not be an artifact, some darkness to the left and below the white that could be the bird in motion but it's not definitive. Further below the white is the first definitive sign of what I interpret to be the tail. This is not an artifact, it is too large, dark and sharp to be an artifact. It remains in view in 66.7. Anyone have a theory other than the bird's tail?

66.7ms - The white of the wing and any dark immediately surrounding it is now behind the tree. The tail looks as it it has moved with motion consistent with pivoting up from its position in 50.0. There is a fanning of the back edge consistent with a pivoting motion. The darkest pixels in this frame are lighter than those in 50.0 (I measured them) as you'd expect from increased motion blur.

83.3-133.0ms - All 16-19" length and 29-33" wingspan completely hidden behind the 12" tree.

150.0ms - hint of white appears to right of tree where wing appears in 166.7.

166.7ms - white of wing in flight definitely appears to right of tree.


the veeb

Anonymous said...

and where is the black median stripe on the wing ... it is on the cornell model of a ivory bill with masonite wings
, but it doesn't appear in the wing in 33.3 clip?

I'm unclear if anyone who has watched the slo motion 33.3 clip sees a bird perched on the side of the tree that was drawn in the figure of the Science article?

It seems like the comments are discussing the pattern on the underside of the wing open wing ... cornell says that this is a perched bird with a closed wing ...

Anonymous said...

Here are my criteria for judging sketches or explanations for 0.0ms-166.7ms:

Sketches should be drawn to scale. It really serves no useful purpose to draw a sketch that fits the pixels but is totally out of scale with the reported diameter of the tree trunk (~12.5 inches). The sketch of the original post is ridiculously small in proportion with the trunk if you accept Cornell's measurements.

If there is a tail pivoting into view in 50.0 and 66.7, the tail in 33.3 should be consistent with this tail motion. If a tail is seen in 50.0 and 66.7 where is it in 33.3? If you don't think it is a tail in 50.0 and 66.7 what is an alternate theory for what is appearing in those frames.

Sketches must be consistent with getting a 16-19"l/29-33"w woodpecker from a vertical orientation clinging to the back of a tree with a 12.5" diameter to flying away from the camera in frame 166.7ms. Sketches and or explanations should be provided that account for the visible bird pixels for as many frames as possible taking into account physiology and physics. An isolated sketch just doesn't cut it.

I still don't think that'll ID the bird but it does allow a better assessment of a sketch or explanation.

the veeb

Bill Pulliam said...

OK I've been looking through the video clips of Pileateds launching in to flight that are collected on the Cornell video analysis web page. I finally feel that I have a handle on the gymnastics here. All these birds launch in a similar fashion:

Starting from the vertical, they swing the upper body away from the tree and sideways to about a 45 degree angle from horizontal.

Then both wings are rapidly and fully extended, essentially in to full upstroke position. But, since the bird is still swinging away from the tree trunk, its back is pointing sideways away from the tree, and both wings are extended away from the tree, not upwards. The upper wing points away at about a 45 degree angle; the lower wing points sharply down.

As the wings are extended, the bird launches away from the trunk and begins the first downbeat, which propels it away from the tree. It takes a wingbeat or two for the bird's to roll back to an orientation squared with the ground. This is visible 10 frames after frame 33.3 when the bird is at full downstroke, and the left wing is actually wrapped almost underneath the body (this would be the second downstroke, the first having happened behind the tree). You can watch the bird finish rolling to level flight over the next wingbeat

If you followed all that description, you will understand that what we are seeing in the frame in question is in fact the bird's right wing, fully extended but pointed away from us at roughly a 45 degree angle. The left wing is also fully extended, pointing down, and invisible behind the tree. The fact that the right wing is pointed away explains its seeming short length and awkward angle.

So what we are seeing is a forshortened view of the fully extended underwing. the part of the wing that we can see is that behind (rearward) of a diagonal from the wing tip to the posterior wing base.

Just my opinion, others will doubtless disagree, but that section of the wing is awfully white. The black trailing edge is awfully faint if it is present at all, especially considering the forshortening. I also note that in the subsequent frame, the trace of the wing that still peeks out from behind the tree appears to be white, not black. This would be the underwing trailing edge.

Anonymous said...

"Even if you think Cornell's scientific work is extremely flawed, do you seriously think that they'd have a photo of a Pileated with symmetrically white secondaries and still go forward with the Ivory-billed announcement?"

You're looking at it in the video.

Oh, and Mr. Gallagher you and your colleagues assesment of the birds' position on the tree trunk is spot on. But your gamble of the kent calls, foraging signs, and "qualified" sightings did not lead you to the real Ivorybill you all hoped would be found in Arkansas.

Here's what Tim Gallagher says about the Luneau video in "The Grail Bird", pages 224 and 225(the bold font is mine):

"In the blown-up film, I could see what appeared to be a large bird with a black-crested head and a white bill peering out from behind a tupelo...I was completely floored. Virtually all of the ivory-bill's major field marks were there, albeit fuzzy."

Oh! really, I see a Pileated playing peek a boo with Mr. Luneau.

Why did Cornell draw up that sketch of the bird when it is already pivoting behind the tree when a full profile of the bird would have been more conclusive? That is/was your interpretation of the video?

Wrap it up in Arkansas and go give Mr. Collins a hand in Louisianna. I think he's on to something over there.

Anonymous said...

For god's sake people...stop. It's over, or as a very wise person said on another thread of this blog...

"This is about the deadest horse I have ever seen beaten so badly."

Anonymous said...

I agree with Bill Pulliam's description of the gymnastics that the birds generally follows having watched the Cornell Pileated videos also. The bird starts to pivot the body from vertical to horizontal and then quickly extend the wings away from the tree. The subsequent downstroke then takes the bird away from the tree and the bird eventually rights itself.

That is why I have been so interested in what I interpret to be a pivoting tail in frames 50.0 and 66.7. Because that would be consistent with pivoting the body from vertical to horizontal. And if that is the case, you can interpret how the body is oriented at those moments. And from there you can figure out how the wings might be oriented.

Bill's description of the right wing pointing away from the viewer seems a plausible explanation of the pixels given the physics and physiology (i.e. gymnastics), and the scale makes sense. Seems a viable explanation. The first detailed one I've seen for the extended underside of the wing (but I may have missed something earlier).

I still think that the full extension could happen a frame or so later and that you can't rule out seeing a dorsal side of the wing in 33.3 but I want to think about that and Bill's explanation more.

I think the power stroke happens at around 66.7 or 83.3. If it happened earlier I don't think you would see any tail in 66.7.

Need to check the timing coming out the left side to see how that agrees with.

the veeb

Anonymous said...

It appears that the visible portion of the wing is roughly 5 cm in width at its fattest give or take a bit depending on fuzziness and fringing artifacts. How much white is there on the trailing 5 cm of a pileated underwing?

Anonymous said...

I retract part of my earlier comment about scale.

I had been assuming that the bird was perched on the tree it appears behind, that is how Cornell depicts it. But it seems more likely that it is perched on something further behind the tree. That makes it a lot easier for it to take off while staying mostly out of sight behind the tree in front. It is also more consistent with its size and flight path when it emerges on the other side.

Mea culpa.

I still like Bill Pulliam's takeoff scenario better than what was depicted in the sketch of the original post. A sketch to compare would help.

the veeb

Anonymous said...

I don' think you need another tree behind that one. The bird is not oriented perpendicular to the line of sight; it is flying away at an angle hence it is foreshortened. Trying to measure approximately aong the length of the wing when it emerges from behind the tree , assuming the same scale as the tree, gice a wing length of very very roughly 30 cm, which is in the ballpark for either species. It is also only hidden for the first stroke, and in the next two strokes it never seems to open its wing fully. I think there is plenty of room to hide a fleeing pileated or ivorybill behind that tree for the time required to make that first escape stroke.

Bill Pulliam said...

The total length of a single wing when it emerges from behind the tree, compared to the 30 cm tree trunk, looks like about 25-30 centimeters, which as was mentioned above is in the ballpark for either species. I've scribbled pencil sketches to fill in the "missing frames" (five of them from when the wing disappears to when the wing reappears, one of which does show the tail tip). It appears that the bird's center-of-body draws a nice arc upwards and then downwards, onto the line of travel apparent when the bird emerges (away and slightly downwards, but higher than the initial takeoff point). It also pivots sharply away from the camera; try to picture it simultaneously rotating clockwise and away from the observer. It is a tight squeeze, but it does appear that naturalistic wing motion (based on the visible wingbeats later in the video) does just barely keep a pair of 25-30 cm long wings that are just beyond the 30 cm tree truck hidden for 5 frames (less than 0.1 sec). As the above post mentions, the wings are not fully extended on the downstroke, which keeps them close to the body. They are never extended to anything close to the full 70-cm (or so) wingspread in those first beats. The wings in fact wrap around (as though around a barrel) as they come down; a motion reminiscent of a person reaching forward with both arms to scoop in a big armful of something.

Bill Pulliam said...

Now that everyone has gone away...

What do I infer about species ID from this, given that at least the two of us still here seem to be in agreement as to what part of the bird we are seeing?

Accepting that we see only the underwing in these first escape wingbeats:

The thing that strikes me about frame 33.3 and the immediately following one (66.7?) is that the wing always shows white. Even in the latter frame, when the wing is vanishing behind the trunk, the sliver that remains shows white. This is in notable contrast to the first glimpse of tail, which is clearly black. Six frames later when the first hint of wing appears from behind the trunk, it also shows white, though I suspect this would be expected for either species. It is difficult for me to imagine how you could realistically position a pileated underwing to show white even in its vanishing sliver as it is eclipsed by the tree. Especially considering that when the bird comes into view from behind the trunk, the underwing is still presented rather well towards the camera, not edge on.

So, I'm afraid, I do get "ivorybill" out of this sequence, not pileated.

I'm off until the midde of next week, by which time we will have some new published papers to read and discuss, if the rumors are correct.

Anonymous said...

So let's get this straight. Out of all this analysis, out of all these pixelating wingbeats of white and black, you actual claim to see Elvis's rediculous bejeweled cloak peaking out from behind that tree?

So this is what it has come down to? The evidence for UFO's, Elvis, Bigfoot et al is sufficient even at the slimmest of margins.

Here's the scientific way to do this. It is a Pileated until they find and photograph the whole bird with a white beak. Until then, the correct question is how can this pileated possibly show so much white on the underside of the wing.

You say "It is difficult for me to imagine....." Well, there it is. The wing in takeoff from behind a tree with the bird's body vertical to the ground is complicated! Believe in probabilities. Be scientific. What are the odds that that bird is a Pileated and not an Ivory-billed? Virtually certain.

Easy for us though. Just think of Fitz. Would you want to be where he is on this? Ask yourself, you are a famous ornithologist. Well respected. Would you stake your reputation on what you just wrote above.

I think not!

Bill Pulliam said...

Fine, whatever. I spelled out my logic, I gave my reasoning, I said what it looked like to me just based on what was on my screen. You wanna start ridiculing and hurling words like "rediculous", UFOs, and Bigfoot, fine. I'm out of here again. Guess I should have just stayed gone. For a moment there it looked like we might actually be having a discussion without insults. But evidently this is impossible, and not just because of the "true believer" cult as you call it.

Anonymous said...

So there does seem to be two types of people in the world. Those who find that evidence for UFO's must be very very strong. They want their videos to show the little green fellows with clearly visible antennae on their heads.

And then there are those who say they are satisfied that Elvis can be identified by a few pixels.

Well, I think that that is very fine and well. But the latter group has got to realize that they are going to be wrong 99.99999999% of the time.

It's really not complicated. It's just that simple.

Anonymous said...

Yeah, I really think attacking someone's views and reasoning with insults is cheap (but not unexpected). The one thing I like about this site (besides monitoring the rumblings of the coming storm) is the alternate explanations of what's in the video. That's attempting to do something that approaches science.

Personally I don't think there are enough pixels for me to say what's going on in the video as far as wing markings beyond white undersides and black tips. I don't see much in the video that is convincingly a trailing edge of black, mostly what I do see more likely could be artifact. But with the motion and number of pixels I'm not sure you could see the trailing edge anyway. I definitely think that any wedge of black on the underwing would be lost in the blur. Somebody name 1 or 2 frames that you think most convincingly show a trailing black edge?

As Bill said hopefully we'll have a published paper soon so we have some new analysis to discuss. More science, less insults (yeah right :) ).

the veeb

Anonymous said...

Let me restate another alternate theory for 33.3 that I had mentioned earlier on another thread. The theory is that the white that shows up in 33.3 is actually the underside of an opening wing, not a fully extended wing. I had mentioned Cornell's frame 3 of comparison clip 6 before but I think frame 4 of clip 5 may be a better example:

http://www.birds.cornell.edu/ivory/rediscovery/support/compareflight#HD4

That frame doesn't show the correct orientation of the wing but it shows the general shape and timing I'm trying to make work. I think it's 10-20ms earlier in the wing opening sequence than what Bill Pulliam was proposing.

Not sure I can make the timing work to get the bird to the other side of the tree in synch with the wing movement but I think the tail movements on takeoff make more sense with this timing.

I think this theory would point more toward Pileated than Ivory-billed since Pileated would have more white in the right location for the pixels that show...

the veeb

Anonymous said...

That does look plausible. There is a problem in that field 50.0 (next frame) appears to show the wing in a similar orentation, just farther to the right and more behind the tree. In the Pileated comparison clips, the wing passes through that configuration very rapidly on the way to full extenaion.

Anonymous said...

It amuses me that arguing probabilities is "attacking someone's views" and discussing a pixel or 2 in some frame of the video and concluding from that that "Elvis Lives" is "science".

Put your faith in probabilities people. What are the odds that this is an Ivory-bill? No chance in H_ll.

Anonymous said...

That does look plausible. There is a problem in that field 50.0 (next frame) appears to show the wing in a similar orentation, just farther to the right and more behind the tree.

Earlier in this thread I described 50.0 as being paler from a motion blur and thicker at the top than in 33.3. I think that would still be consistent with the wing rapidly being extended up and to the right. Need to sketch it... and look at the PIWO video more...