Wednesday, February 15, 2006

More analysis of the Luneau video

An anonymous contributor sent me this excellent analysis:
Attached is a comparison of some images from the recent Cornell website defending the identification of Luneau's video as IBWO. First is the image they themselves posted on the website to show how "unlike" the underwing patterns of the Luneau bird and confirmed Pileated are. Note the extensive black on the trailing edges of the PIWO wing. Also note that they are all flying in an angle perpendicular to the viewer, whereas the Luneau bird is flying directly away.

Below the black stripe, I have added comparisons of some screencaptures of frames from the videos posted by Cornell of confirmed PIWOs flying away from the viewer. Somehow, the video analyst wizzes at Cornell seem to have overlooked that these images are much more like those in the Luneau video in the direction of flight... and, amazingly, they are also more similar in overall pattern! Huh. Well, nevermind, if you just overlook that minor detail, you'll see why Luneau's bird is still clearly so different from a PIWO. Right?
(Click to enlarge)

20 comments:

Anonymous said...

Great bit of work. In particular the second from bottom on the right side appears to have a white trailing edge, when it doesn't. And this is a sharper video than the Luneau one.

Anonymous said...

And what about the upperwing white?

Anonymous said...

You could interchange the two videos quite easily, they could be the same (albeit one is sharper than the other).

Anonymous said...

I just wonder if in any of these,
the PIWO is flying UP and away, or
just across or even down in the
field of vision.
And is the Luneau bird actually
gaining enough in altitude that we
wouldn't be seeing underwing but
overwing? And at this distance
I don't know if I'm always seeing
black border or the camera artifact
fuzz that surrounds everything in the Luneau. For example when the
overwing shows (if it does), am I
seeing an all white trailing edge
with a black camera artifact. It seems most convincing that the underwing has a black border, just like a Pileated.
The whole approach taken by the analyst who submitted this is pretty good.
My biggest evidence that this is not a normal Pileated is the wing that
shows around the tree just before
flying. That sure does look like
a white wing-back to me. And Cornell assures me that the W'pecker would be double-jointed if that were in fact an underwing.
What fun this video has turned out
to be?!

Paul Sutera, New Paltz, NY

Anonymous said...

“And what about the upperwing white?”

What upperwing white? Cornell references a spot of white in a single frame as the bird first emerges from behind the tree… might be upperwing… I don’t know. Please note though, this (and the Franklin Gull discussion) are the special case of a bird seen at an oblique angle showing one upperwing and one underwing simultaneously. To symmetrically see the upperwing of both wings is a different matter altogether...

Cornell references upperwing white in some of the later wingbeats as the bird travels away from the observer, presumably symmetrical (upper on both wings). It seems unlikely though that this is upperwing. The bird is well above the horizon at this point. With the exception of highly acrobatic flight, a bird orients the plane of it’s wings to gravity (or the horizon) more so than it does to its line of travel. The plane of the wing is tipped with the leading edge down, particularly on the downstroke. This is what keeps a bird moving forward. An accelerating bird, above the horizon, gaining altitude, and flying away from an observer is not going to tip it’s upperwing back toward the viewer, particularly on the downstroke. This make no aerodynamic sense, and would propel the bird backward not forward.

You might see upperwing white on a bird flying away from you above the horizon as it lands in a perch, but this is precisely because it’s decelerating in order to perch... essentially, starting to fly backwards.

Anonymous said...

What he said.

Anonymous said...

a bird orients the plane of it’s wings to gravity (or the horizon) more so than it does to its line of travel.

Then how does a bird gain altitude?

Anonymous said...

"Then how does a bird gain altitude?"

Mostly by adding power to its stroke, while maintaining a horizontal body and wing postion. A bird is not an airplane. Fixed-winged airplanes align their bodies and wings with the direction of their travel, so when an observer watches an airplane takeoff away from their postion, they see the top side of the wing tipped back toward them. Typically, not so with birds.

Anonymous said...

I'm rather new to the blogosphere, but I must say it is an interesting experience where everyone's opinion is no better nor worse than anyone else's. Fortunately science doesn't operate based on majority (or minority) opinion, but on the evidence. Most important in the scientific method is first developing what appears to be the most parsimonious explanations of that evidence which exists while also developing alternative hypotheses to test and attempt rejecting. Science is not about proving something as much as setting up tests to disprove all alternative explanations.

Obviously, the existing evidence here can be interpreted in different ways, by different people with differing levels of experience, which is why of course we're all sharing our opinions here.

So, here's my opinion about this latest post. Am I alone here (at least so far) in thinking this "More analysis of the Luneau video" post is actually less than excellent and not much of an analysis? In fact I'd say it is remarkable for being unremarkable, or more simply unenlightening.

First of all, a number of folks have commented and seem excited on how similar the frames from the Luneau video are with many of the frames selected from the known Pileateds also from the new Cornell site.

Well, dah, they're all images involving big woodpeckers that show black and white patches. On that point at least, I would hope all of us agree. So yeah, all the frames pieced together on this post are similar, and intriguingly so I'm sure with the folks at Cornell happily agreeing with that assessment. To claim the "analyst wizzes at Cornell seemed to overlook this" appears to miss the point of comparing all these video clips with the original Luneau video in the first place on the new Cornell website.

In none of the selected original Luneau frames here where color can be discerned is there anything but white trailing edges on the bird's wings while it flys away (regardless of whether we are looking at upper or underwing, latter forward or trailing views, or undetermined). In contrast for all the known Pileated videos,in almost all (and in my view all) frames we see nothing but black trailing edges (exactly the opposite pattern on what is shown in the original Luneau clip). That would appear to be a huge and consistent difference (despite a frame here and there taken individually that some claim the opposite interpretation is possible). This to me is especially important since the known Pileated frames in fact are often superficially similar in many other respects with the original Luneau video, which I believe was Cornell's reason for showing these clips in the first place (at least they say this is so from what I have read).

Again, may be with some imagination we can mentally pencil in black trailing edges on the bird in the original Luneau video, but that would take much more imagination than some folks on this blog (and elsewhere) have been claiming the Cornell folks to have in their very detailed treatment of clearly a less than perfect piece of evidence.

I am still waiting to see how any big woodpecker in North America, other than an Ivory-bill (well, may be this is really a misplaced Imperial Woodpecker?), flying away from the camera, regardless of angle, would show nothing but white facing the camera on the vast majority of frames. All Pileated footage I've seen (confirmed also by fuzzy re-enactments of model Pileated vs. model Ivory-bill), regardless of position or angle with respect to the camera, indicates black trailing edges on the vast majority of frames.

That is, seeing anything other than black trailing edges rather consistently from footage of a normal Pileated may be impossible, at least based on all the existing evidence so far. In order to fit this alternative interpreation in with the original Luneau video, we would have to imagine something "whiting-out" the black trailing edge that should be there in most frames. Perhaps we just can't see the black trailing edges for some unknown reason , only in this one specific video, but "they got to be there(!)" for this explanation to work.

Cool theory, but a plausible explanatory mechanism for how it would be possible to miss seeing black trailing edges has yet to emerge from anyone here or to my knowledge in the scientific literature. By the way, Dr. Jackson's comments in the Auk and on NOVA on this point are completely unsatisfactory in my view, but these may have not been the best forums to explain what he meant to the scientific community. In other words, lots of bluster and hard feelings (which may or may not be understandable), but little light is shed on these critical points, at least not yet.

Jackson's opinion is simply like those we have here in the blogosphere and carries no more or less weight at this time in the scientific process than anything here, especially in the medium it was published (the "perspectives" section, not among the scientifically peer-reviewed articles). In other words, Jackson could have backed up his brief verbal "explanation" with an actual explanation and hard evidence (or at least the illustrations and photos of funny looking Pileateds he mentions exist, somewhere anyway). If he had been able to publish a detailed scientific peer-reviwed reanalysis then this would be a whole different discussion, but he chose not to for reasons he has yet to explain. So his opinion, which he deservedly has a right to, is no more or less important than any other opinion (including mine) given here.

That doesn't mean someone (including Jackson)isn't waiting in the wings with the silver bullet explanation about how a normal Pileated really is what Luneau filmed to show how this or another alternative explanation may be possible. And "may be" is the key phrase here in terms of Cornell's not having eliminated all other "possible" explanations, at least in the minds of many of the skeptics. However, the "possible" explanations I am aware of remaining are getting more-and-more convoluted and not yet scientifically vetted in any peer-reviewed journal. Minus a real rigorous scientifically based rebuttal and response from Cornell for all of us to form better science based judgements, what's published in Science and elaborated on Cornell's new website is the best available information now available.

Until there is published a detailed reassessment (with an obligatory response from Cornell) on how a normal (or even an abnormal) Pileated can show a pattern identical to the original Luneau video, then Cornell has to date provided the most parsimonius explanation regarding the identity of the bird in the original Luneau video, in scientific terms at least.

This "conclusion" may be unsatisfactory to those interested in conducting opinion polls here or elsewhere, but opinion polling on scientific data isn't science itself(otherwise we might be teaching in some states that the earth is less than 6000 years old, there's no such thing as continental drift, or creationism instead of evolutionary theory as the best basis for explaning biological processes to our future medical doctors and farmers). Although there is often science behind the statistical designs of the more rigorous opinion polling, the results of opinion polls on what people believe, while having socio-political ramifications, has little relevance to advancing scientific discourse.

So to advance scientific discourse, surely someone here or elsewhere has the right video camera equipment, alot of Pileateds to film in their backyard, some cloudy days to film Pileated, and can come up with a scientifically based demonstration of how it would be possible on a normal (or even an abnormal) Pileated Woodpecker to show only white on the trailing edges of both under and (depending on one's interpretation)upper wing frames to compare with the bird caught on the film by Luneau. As Paul in this string comments "What fun this video has tunred out to be!?" That is something I think we all can again agree on.

By the way, being new to blogging I am unfamiliar with how copyright infringement applies but perhaps the posting source and the blogmaster here received permission from Cornell for use of these copyright protected images. In case that is not true, ya'll may be interested to learn gaining permission is possible following the instructions on the new Cornell site:

Note: All images and text on this and the following pages are copyrighted and may not be used without written permission from the Cornell Lab of Ornithology or other named copyright holder. For information on permission to use text and images, contact Jennifer Smith at jls39@cornell.edu.

Thanks

Anonymous said...

"Then how does a bird gain altitude?"

Mostly by adding power to its stroke, while maintaining a horizontal body and wing postion. A bird is not an airplane.


I'm sorry, I'm just not grasping the physics of this. Would not simply adding power to the stroke just cause a greater forward velocity? If adding power to the stroke resulted in greater altitude, then a Gyrfalcon would have a very difficult time successfully pursuing a ptarmigan, for example. Wouldn't the falcon just end up a long ways up above the ptarmigan as it accelerated towards its prey?

It seems that there has to be an additional factor that translates more power either into altitude or forward speed.

Anonymous said...

“Mostly by adding power to it’s stroke”

Yes, mostly. There are fine adjustments to wing pitch, but this is where we come full circle… the bird makes the adjustments “relative to the direction of gravity (or the horizon) more so than to its line of travel” A bird flying “uphill” generally does not incline its wing and body “uphill”. Take a walk, look at the birds, look at how they hold their bodies, and try to see the back or the upperwing of a bird that is both above the horizon and flying directly away. Video it if you like -particularly if it’s a Gyrfalcon.

Anonymous said...

In response to I'm rather new to the blogosphere

Science is not about proving something as much as setting up tests to disprove all alternative explanations.

This is where Cornell has, in my opinion, failed miserably. The author of this blog showed, for example, that there are numerous artifacts in the video indistinguishable from the infamous "6 pixel Ivory-Bill." Simply because the object was "there" when the video was made, and absent later, does not mean it was an Ivory-Bill; if so, there must have been a small flock there.

In their original paper, Cornell says: We are unaware of any examples of extensively and symmetrically pibald pileated woodpeckers in museum
collections or the literature
(I didn't get copyright permission from Jennifer Smith to use that quote.)

Cornell apparently didn't do very thorough research. Noel Snyder saw a Pileated with symmetrical white secondaries years ago. Fred Collins wrote a paper in which he said In the 70s there was a partially albino Pileated Woodpecker that was very much patterned like an Ivorybill. It required a visit. Already since the Arkansas sighting another bird like this has been reported in Baytown.

As far as the “double raps,” experience has shown that people confuse any number of sounds for double raps (gun shots in at least one case.) Cornell has not shown how they can be sure that the kent calls are not recordings of recordings. It could very easily have happened by mistake or by a hoaxer. They didn’t come close to disproving alternate explanations, and in my opinion, the alternate explanations are far more likely.

I don't think any of us needed your lecture on opinion polls in science. At least in the beginning, skeptics were treated by many to be a "lunatic fringe," so I think it's fair to point out that that simply isn't the case, and that more and more professional scientists are publicly disputing Cornell's claims. Their opinion doesn't change the facts, but it's at least newsworthy.

In order to fit this alternative interpreation in with the original Luneau video, we would have to imagine something "whiting-out" the black trailing edge that should be there in most frames.

According to Cornell and others, in video like this a black and white object will show more white than is actually present. That effect is more extreme as the video becomes more pixilated.

Based on Cornell's flawed use of a stiff-winged model in the reenactment, I am willing to bet that the video of Pileateds they used was significantly different from that of the Luneau video. The Cornell video and frames that most of us see are a small part of the full frame. Since there are so few pixels in the images of the Luneau bird, it will show more white than the real bird showed. It will also show more white than Cornell’s Pileated footage, in which the bird is likely much larger in the original frames. In the Luneau video, the narrow band of black may not "possibly" disappear, it is LIKELY to disappear in many, if not most, frames.

The bottom line is that Cornell can repackage their weak evidence any way they want, but more words do not improve their current evidence. If they can claim that smudges of white that can be “seen” in some frames are dorsal stripes, it is certainly fair to claim that the black trailing edges that can be “seen” in others are Pileated secondaries. You need imagination to truly accept either theory as “proof” of what the video actually shows.

Less imagination, less interpretation, more evidence.

Anonymous said...

"Less imagination, less interpretation, more evidence. "

Interesting thoughts and opinions and I agree with this parting statement quoted above, but I'm sure we may disagree about who is applying the most vivid imagination to their interpretation. Now can someone (including you) back up your explanations about what the original Luneau video actually shows with (1) a scientifically based demonstration, (2) submitted via a scientifically written manuscript and (3) get it published in a reputable journal? Perhaps, just haven't seen it yet.

Until there is a scientifically based peer-reviewed rebuttal and obligatory Cornell response (which may be coming any day, or may never come, who knows?), the Cornell interpretation will continue to stand as based on the best science available, regardless of what any of us (including me, having some doubts about the identification myself) believe.

Given the assurdness of those who can explain everything Cornell has claimed away "so easily," should be just as easy to demonstrate all the points made here, so someone please do so for all our sakes and get the results published. Should be simple, no?

Thanks.

Anonymous said...

"Given the assurdness of those who can explain everything Cornell has claimed away "so easily," should be just as easy to demonstrate all the points made here, so someone please do so for all our sakes and get the results published. Should be simple, no?"

Evidently it isn't and that's the whole crux of the matter.

If the evidence you have to work with is of such marginal quality that both sides can argue their case to the same degree of assuredness then all you have left at the end of the day is ambiguity.

and...

Ambiguity has no place in science.

The whole body of Cornell's evidence did not warrant such an overblown expedition as this. If anything it should have been used as a starting block for a scaled back preliminary investigation.

I guess everything just started snowballing bigger and bigger.

What will happen now if no further proof is found? I do hope that someday it will be, if not in Arkansas, maybe Florida. I just don't know anymore.

Sigh,

Ambiguity sucks !

Anonymous said...

If the evidence you have to work with is of such marginal quality that both sides can argue their case to the same degree of assuredness then all you have left at the end of the day is ambiguity.

Very true. I believe that no one is going to come up with the magical experiment that determines with finality what the video shows. Debate will go on, and we'll never know for sure what the video shows.

In the big picture, "the truth" will be determined not by Cornell's paper, or a rebuttal, or a rebuttal to that rebuttal, but by:

1. Some high quality, irrefutable photos or video or the like, or,
2. Time passing, with no solid evidence produced by anyone.

In my opinion, the experience of the past 62 years makes the latter far more likely.

Anonymous said...

"Ambiguity sucks!"

That about sums it up for most folks here I suspect. But in combination with all other existing evidence, inaction would have been irresponsible too. If nothing new turns up, so be it, another example to add to future books regarding whether all the past "near-misses" could have been all wrong. "The truth is out there" (The X-Files and Jerome Jackson).

Anonymous said...

"If the evidence you have to work with is of such marginal quality that both sides can argue their case to the same degree of assuredness then all you have left at the end of the day is ambiguity."

There is one big difference between the two sides in terms of assurdness. One side's (i.e., Cornell's) assurdness is based on a peer-reviewed and published assessment of what they interpret to "definitively" not be a Pileated Woodpecker (normal or aberant) and have interpreted with all their evidence and through experimentation to be consistent with a normal Ivory-billed Woodpecker.

In contrast, the other side's assurdness is at this time based on opinion and no experimentation (or demostration). The opinion is largely derived from the belief (justified or not)that Cornell got it wrong. There have been some (in my opinion) very interesting critiques (including many here on this blog), but they are still collectively not very coherent in demonstrating how a normal (or aberrant) Pileated could be filmed and show all the characteristics that we have been discussing for a long time now.

There is nothing published (yet anyway) in a reputable peer-reviewed journal to support the notion that the evidence derived from the original Luneau video is actually consistent with either a normal or aberrant Pileated Woodpecker, or something else (Imperial Woodpecker?). Obviously, though, there are a number of folks that believe that the subject bird has to be either a normal or aberrant (which is it I'm not clear on) Pileated and of course have a right to that opinion.

Cornell folks have followed standard scientific protocols that led to their conclusion (which may not be correct). We are still awaiting a scientifically valid re-assessment from others that along with Cornell's response would lead to (1) reconfirming the Cornell intepretation, (2) negating it, or (3) some synthesis of interpretation. That's science in action, not necessarily pretty, but essentially it is (1) thesis, (2) antithesis, and if warranted (3) synthesis.

The antithesis may come soon or it may not come at all (probably the former), but until then the Cornell thesis represents the scientific "last (but not necessarily final) word" on the subject. Everything else is opinion, for better or worse.

For those that "believed" in the Cornell interpretation uncritically, that may be excuseable to a point, but for thoses critics/skeptics from the near-beginning that felt like they were being branded as heretics all I can say is that should never have happened from the perspective of scientifically based discourse. There's plenty to be skeptical of in this "crummy" video.

On the other hand, all the criticisms here and elsewhere have not been pulled together (yet) in a way that from a scientific point of view would negate Cornell's interpretation, no matter how strongly those opinions are expressed.

First it is important to show that point by point that one normal (or aberrant) Pileated can actually show all the characteristics folks here and elsewhere claim that Cornell has misinterpreted to not be consistent with Pileated and consistent with ivory-bill. That has not been done yet (sorry Tom, but you can try publishing your various critique to reactify that).

Those that believe there has been a colassal mistake by Cornell may be correct, but until there is a published rebuttal and response, they are subject to practicing as much faith-based orntihology as the believers who uncritically accepted Cornell's word for it without reading the details for themselves.

There is nothing wrong with expressing opinions, but science has a higher standard. Whether or not Cornell's interpretation stands the test of time in the scientific arena, well...only time will tell. Whether it stands the test of time in terms public opinion is another matter. Both are important, but separate processes.

Anonymous said...

Do the researchers have in their possesion a video and/or still of an aberrant Pileated with a nape stripe that extends across its chest, either partially or completely?

It may have been seen by the researchers themselves.

I feel that is the bird in the video.

Anonymous said...

Those that believe there has been a colassal mistake by Cornell may be correct, but until there is a published rebuttal and response, they are subject to practicing as much faith-based orntihology as the believers who uncritically accepted Cornell's word for it without reading the details for themselves.

Firstly, rumor has it that there is at least one official rebuttal paper "in-the-works."

Secondly, examining the available facts, and reaching a conclusion based on scientific principles, may not be "published science," but it is real science. I think it's unfair and untrue to call unpublished science of this nature "faith-based."

There is little doubt that future published rebuttal papers (if any) will use some of the ideas and arguments first introduced in this blog or elsewhere.

Anonymous said...

"There is little doubt that future published rebuttal papers (if any) will use some of the ideas and arguments first introduced in this blog or elsewhere."

Perhaps this is true. But in order to thoroughly address the point made by the Cornell folks, for which experimentation and demonstrations were used, the rebutters will need to do the same to show that alternative explanations for the whole sequence are in fact possible. Its one thing to draw alternative explanatiosn for every frame, and another to show that when strung together that it is possible that one normal or aberrant Pileated can be used to provide a sound alternative explanation.

It may be possible to do that, but it doesn't seem likely to me. Until that is done, all of this is interesting opinion, which has value, it's just not science (yet).

I once asked a colleague on a rare birds/checklist committee his opinion once seeing the video on a big screen. His reaction (and that after taking closer looks) was that there is no way that the original Luneau video shows a Pileated Woodpecker. I then asked if he thought it was an Ivory-bill. He just smiled and said, "I don't know."

Oh well, if this can't be a normal or aberrant Pileated, what alternatives are left? If the answer is Ivory-bill, this doesn't prove it is an ivory-bill, scientifically, but if you can eliminate Pileated and interpreted characteristics are consistent with ivory-bill, well in scientific terms that is good enough to publish and move forward until better scientific infomation emerges.

Which it might, sooner or later.