Friday, February 17, 2006

Cornell fooled by vegetation again

I've been doing some more Luneau video analysis, and I just noticed what appears to be another massive error on Cornell's part.

At this link, I watched the second, magnified clip.

Note the black, nearly vertical smudge labeled "D" in the picture below (you can use the Luneau bird "A" to orient yourself; please ignore points B, C, and E for now).


Twenty-three frames after the above screen capture, the Luneau bird aligns with the smudge labeled "D", so that Cornell "saw" a black head and tail that resulted in the incorrect sketch below (from Figure 2 of their Science paper).


I think the sketch above is a complete misinterpretation of our view of the Luneau bird. The sketch shows a "high-angle" view of the back and both upperwings of the Luneau bird; in reality, we basically have a "rear-view". We are looking at both underwings of the bird (with wings below horizontal).

I'm completely unconvinced that we have a view of the bird's back from this angle, and I think we have no evidence that the Luneau bird had white dorsal stripes.

12 comments:

Anonymous said...

Well fair enough and I might add,
pretty good. No one will be able to say whether this bird isn't an aberrant Pileated but the wing wrapping around the tree prior to flight does seem to show a white upperwing. And even the little black dimple to the left of the white looks much like the black central body showing through.
Much like the specimen they show
with the wing draped in a similar manner. Yet some people believe
the bird is actually nearly facing the camera during that frame.
We know Pileateds like to get a look at "danger" before they go!
That's something we've all seen
because often they will just hide "'round back" once they've established that the danger isn't too near.
I guess I would believe, in the absence of sufficient pixels, that
it could be a pileated anchored
on the far side and peering over.
Then the bird wouldn't have to be
double-jointed to extend that wing and show us the underwing.
Then he presses away from the tree
and launches his flight completely
out of sight of the camera.
The red crest of a pileated disappears at this resolution especially in a frontal view.
I admit, I'm no ornithologist.
Though a lot of the science from Cornell starts out "If we assume this based on plausibility of the model shown...." Then the science part begins.
The science of saying: There's
a dark smudge and it hangs around
for 23 frames and then the bird
flies right into it, is perfectly
good science. Observation *is* good science

Paul Sutera, New Paltz, NY

Anonymous said...

First, is this really one of the frames that Cornell is using to support the white dorsal stripe hypothesis? I know that the drawing is showing that, but from my recollection, this isn't the frame they're using to support that. I apologize if I'm incorrect, but this point is not specifically relevant to the following

With that said, I don't think the black mark talked about in this post accounts for the black "head and tail". I use quotation marks there because I don't think that black smudge can be attributed to the bird's head, actually. But if you look in the first frame (the one with the letters added), there's still a black spot that corresponds to the tail or body. There is no spot corresponding to it in the latter frame (on the tree, that is; there is one at the "tail/body" position).

Also, looking at the "labelled" frame, notice the bend in the small tree located below the 'Lu' in 'Luneau' in line D. The topside of the black area to which D points is only slightly higher than this bend. Then looking at the latter frame, the top of the bird and the alledged "head" is substantially higher than this bend.

Based on these two points, I don't think "D" explains the black spots associated with the bird.

To me, the limited extent of it in the first frame, plus the shape of the wings, makes it appear that it's showing the birds topside and upperwings.

Anonymous said...

Good job. At least as valid as Cornell's evaluation of those frames.

This also illustrates how shaky, at best, their "proof" of the bird's size is. With that kind of pixilezation and blurriness, and with how close Pileateds and IBWOs are in size, and with how easily bird parts and vegetation and even shadows are confused, their measurements are nothing more than a guess.

Also, as noted, there are two more "6-Pixel Birds" that are indistiguishable from Cornell's "real" example.

The blurrier an image is, the easier it is to make it whatever you want it to be. That's why the only photographic evidence they have is blurry.

Anonymous said...

First, is this really one of the frames that Cornell is using to support the white dorsal stripe hypothesis?

Yes, that is frame Cornell is using to support the white dorsal stripes. I agree that the angle of the bird in the drawing is wrong in any case, and it looks to me that they mistook the black mark for the head of the bird. I think their "tail" is likely just a rear view of the tail/body.

It's just a guess, for me, and for Cornell.

Without better evidence, we're back to where we were before the Sparling sighting.

Anonymous said...

"Yet some people believe
the bird is actually nearly facing the camera during that frame.
We know Pileateds like to get a look at "danger" before they go!
That's something we've all seen
because often they will just hide "'round back" once they've established that the danger isn't too near.
I guess I would believe, in the absence of sufficient pixels, that
it could be a pileated anchored
on the far side and peering over.
Then the bird wouldn't have to be
double-jointed to extend that wing and show us the underwing.
Then he presses away from the tree
and launches his flight completely
out of sight of the camera.
The red crest of a pileated disappears at this resolution especially in a frontal view."

Yes, Paul you're right! That is what I see. I've observed Pileateds on several ocassions make that "peek a boo" motion. I've actually had a bird do this several times from both sides of a tree as I stood beneath it and shifted my position from left to right. It made me laugh and I was so grateful for the experience.

This motion is what I picked up on viewing the enhanced video ( something I couldn't do with the original). They really should focus on those initial first frames, because I feel that is the key to the whole identification.

Anonymous said...

it could be a pileated anchored
on the far side and peering over.
Then the bird wouldn't have to be
double-jointed to extend that wing and show us the underwing.
Then he presses away from the tree
and launches his flight completely
out of sight of the camera.


I don't think this hypothesis is all that reasonable. The biggest reason is that you can see the tail swing out from behind the tree while some white is still visible. Notice the angle that the edge of the white forms with the tail. The bird would have to be extending the base of the wing (the portion with the secondaries) but still holding its wrist in so that the primaries are actually folded in towards the body. I've never noticed a bird do this before. Birds always seem to flip the primaries out into a flying position while the rest of the wing is also being extended.

I also have to wonder if the white of the underwings would appear this bright if the bird was just beginning to open it's wing. Wouldn't it mostly be in shadow at this point, particularly since it's so close to the tree that there wouldn't be much light reaching the underwing? Of course, there are issues with exposure and white balance with a digital video camera, but it seems that the white shouldn't be that brilliant under the circumstances.

There's one thing I don't understand about this hypothesis. Are you suggesting that the head is visible? If so, does it extend further to the left than the white? (If so, shouldn't it be more obvious in the video?) I think if a bird was peering around the tree, the head would be leaning out further than any other body part or at least be even with the shoulder. Try it yourself. And if the bird was doing this, opening its wing would seem to be counter-productive. First, it would expose more of its body to the "predator", and second, it would move it's center of gravity further outward, away from its support against the tree. If you are trying to peek around a corner, you don't reach out with the arm on the side you're leaning towards. If anything, you reach out with the opposite arm to maintain your balance.

I wonder if measurements could disprove this hypothesis. I agree that measuring differences of only about an inch or so is not very reliable given the quality of the video. But I'm assuming that you are suggesting that the highest point of white is the birds wrist. If so, what's the distance between that point and where the white on the primaries ends? Should be expect to see where that white ends? I think the answer may be yes, considering that the white appears to extend down to where the base of the tail is as it flips out.

I have a few other points, but before I write an entire dissertation, I'll wait to see if I'm understanding the order of events in under this explanation.

Anonymous said...

This still does explain the apparent white blob on the bird's back, as the smudge is entirely black. The apparent white portion of the stationary smudge in the first image is part of the left wing of the flying bird, not a fixed part of the smudge feature. In the second image, some light-colored freature is in front of the black smudge, located between and above the light areas on the bird's wings.

Anonymous said...

"There's one thing I don't understand about this hypothesis. Are you suggesting that the head is visible? If so, does it extend further to the left than the white? (If so, shouldn't it be more obvious in the video?) I think if a bird was peering around the tree, the head would be leaning out further than any other body part or at least be even with the shoulder. Try it yourself. And if the bird was doing this, opening its wing would seem to be counter-productive. First, it would expose more of its body to the "predator", and second, it would move it's center of gravity further outward, away from its support against the tree. If you are trying to peek around a corner, you don't reach out with the arm on the side you're leaning towards. If anything, you reach out with the opposite arm to maintain your balance.

I wonder if measurements could disprove this hypothesis. I agree that measuring differences of only about an inch or so is not very reliable given the quality of the video. But I'm assuming that you are suggesting that the highest point of white is the birds wrist. If so, what's the distance between that point and where the white on the primaries ends? Should be expect to see where that white ends? I think the answer may be yes, considering that the white appears to extend down to where the base of the tail is as it flips out"

Yes that is what I'm suggesting.

What I see in the video and what I've come to notice of Pileated behavior in the field makes my analysis of the video more plausible. I can't say if an IBWO would react in the same manner for the simple fact that I haven't been blessed with a sighting of one.

But if you slowly manipulate the slider on the QuickTime player forwards and backwards at the point where the bird is just visible you can see it make that "peek a boo" motion. What Cornell says is the birds' wrist and where they take their measurement is actually the birds' head.The white is actually the birds' underwing which it flares.

The bird is now almost perpendicular to the tree trunk and slightly leaning back. It then peeks out again - this is where you get an almost full profile of the bird but the bill is not visible - and that's too bad.

It then pivots to the right with only its tail now visible as it launches from the tree. This I would say would be a more normal and fluid flight kinetic.

The aberrant white on this bird is what's causing all the problems here. That's why I asked if any were spotted with an abnormal nape stripe that extends onto the chest area of the bird. This is something I picked up on the video also.

Yes the head/crest is visible as the bird does its "peek a boo" prior to pivoting to the right.

Advance the video slowly forward then backwards,pause at that point keeping in mind the orientation of the bird that I've stated.

Anonymous said...

It seems like you're seeing a lot more than even Cornell in this video. :)

Anonymous said...

Though it is not evident in the first frame you present, the black "smudge" in the background is clearly revealed in the immediately preceeding frames to be solid black without light features. As the bird passes in front of it, a light area situated between and above the wings (above in reference to the video frame) is clearly evident in several consecutive frames, and is traveling with the bird. Your alternative explanation still fails to account for this feature. It gives every appearance of being a light feature on the bird, which becomes evident when the bird passes in front of the dark background.

Anonymous said...

As I read a lot of this, you are saying that when a bird is flying up and away from me, it's head is pointing DOWN and its tail is pointing UP?

Anonymous said...

"It seems like you're seeing a lot more than even Cornell in this video. :)"

LOL it does seem that way.

What we need is a Smudgeologist specializing in blobs to go over this video ;)