Jump to content
Sign in to follow this  
Guest

Patty's Height

Recommended Posts

Guest

^^Nothing coincidental about it. I scaled Patty's legs to match Bob H's, since we can't see Patty's feet in frame 352. I see that I got it right.

Share this post


Link to post
Share on other sites
Guest

  Somebody needs to do a little photo experiment, that has the equipment to do this and end all this discussion about how tall patty really was.

   Somebody please make a cardboard cut out of a man shape about 6ft tall and a cut out of a foot that is 14.5".  Place the foot off the ground at about 2 feet high and 2 feet closer to the camera.  Then shoot it at close range say 12 feet with a wide angle lens, then with a long range zoom lens at say 120 feet  or 10 times as far, scale them to the same height side my side and see how their foot sizes match up.

Share this post


Link to post
Share on other sites
MikeZimmer

You can also use someone's known height to measure their arm length in a photo. In the following example Bob H has an arm length of 128 / 430 = 30% of his 6'2" standing height = 22", which was correct.

parnsbob.png

 

 

 

At first glance, this a bit tangential to the main topic, but I have my reasons. I am looking at Giganto's picture of Bob H. Here is how Bob H. stacks up on one anthropomorphic ratio, the brachial index:

 

Radius = 55% +/- U1%

Humerus = 73 +/- U2%

 

However, the values U1 and U2 stand for the unknown percent error in the measurements. If the error is big enough, say +/- 10% based on some work of mine a few weeks back, it might change our conclusions.

 

Calculations:

 

Brachial index for Bob H. = (55/73) *100 = 75.3

 

Brachial index for Caucasian Male: Sample N = 10, Mean = 73.4, Standard Deviation from the Mean = 1.96 (from Aiello and Dean,  An Introduction to Human Evolutionary Anatomy)

 

So, Bob H. is 75.3 - 73.4 =1.9 units away from the mean. This is about 1 standard deviation above the mean. This makes him a common sort of man in his arm proportions, provide that:

 

1 - the reference data is sound

2 - my calculations are sound

3 - I transcribed the numbers correctly

4 - the measurements of limb length are sound

5 - we can quantify, probabilistically and with a range, the degree of uncertainty in the measures

 

I took the reference data from a standard source, so if I read the correct column, and the underlying science was well done, we are still left without analysis of error in the dimensions of the bones of Bob's arm. He looks pretty human to me, but if some solid upper and lower bounds were set, say 95% confident about the range, we could re-do the analysis, and find out how Bob H. stacks up. Maybe he really does fit in the suit. ;-)

 

This seems to be the crux of many of the measurement issues, ratios or absolute; the size of the error. I keep in mind that I may have missed some things of great importance. Others may disagree with my conclusion.

Share this post


Link to post
Share on other sites
SweatyYeti

Sweaty, I have rescaled Roger as if he was filmed >100 feet from the camera. This is the only way you can use the casts as rulers to measure his height. This is why you got a low estimate. Add 5% to your height estimate to compensate for the closeness of the object to the camera. 

 

 

 

Thanks for your explanation, Giganto. :)

 

The reason why I posted the 'Cast Ruler' graphic, was to provide an example of that type of distortion in which both the length of the foot, and the height of the subject were known quantities...in order to demonstrate the distortion very clearly.

 

 

 

 

The "bloom" of the foot in frame 72 can be estimated by comparing it to frame 61.

 

 

I just made that comparison myself, a week or so ago. There is a measurable difference. 

 

 

 

 

You mean angled in frame 352? Patty could be upside down in frame 352 and her legs would not be foreshortened. She was walking parallel to the camera plane at the time. In frames 61 & 72 yes, but not 352.

 

 

No, I'm talking about Patty's upper-leg being angled in Frame 72....thusly...

 

F72-AngledLeg1_zps76326c53.jpg

 

 

As a general rule...when Patty's lower-leg is straight vertical...her upper-leg is angled, to some degree.

 

 

 

You must mean the "tilt" of the camera which affects the scale of the images by the cosine of the angle of tilt, which is negligible in this case.

 

 

Yeah, that's right.  I don't know about it being negligible though, Giganto...in Frames 61 and 72, Patty's head looks a bit squashed

 

Here is an overlay of a similar image of Patty, from the middle sequence...(F315)....with Roger on the same level of ground. Note how, at this particular scaling of F315, Patty's body length (from the calf to the shoulder) appears slightly shorter.....yet her head appears higher...

 

F72-F315-VerticalForeshorteningAG2_zpspc

 

 

I don't think there is any way to re-scale F315 to get all of Patty's body details to match-up. In those early Frames, Patty's head appears out of (proper) proportion to the rest of her body. :)

Edited by SweatyYeti

Share this post


Link to post
Share on other sites
MikeZimmer

The full Glickman report is available on Photek Imaging's website HERE: http://www.photekimaging.com/Support/rptcol2.pdf

I do think it would be interesting if Bill Munn's were able to obtain that scanned copy of the film to compare with what he currently has in his inventory.  It sounds as though the scan used was from film that had been in storage for 25 years and that was back in 1993/1994.  What I noted was that the 953 frame count was present in this first generation copy made one month after the film was made.  In that case, it would have been made before much of the subsequent showing of the original film done by Patterson and Dahinden.

 

I wonder if it the data is even readable after 25 years? If they used archival quality media, and stored them in appropriate surrounding, maybe? Does anyone even know how to get ahold of this stuff?

Share this post


Link to post
Share on other sites
Guest

Thanks for your explanation, Giganto. :)

 

The reason why I posted the 'Cast Ruler' graphic, was to provide an example of that type of distortion in which both the length of the foot, and the height of the subject were known quantities...in order to demonstrate the distortion very clearly.

Yes, we need to account for all the variables for these photos especially when they are relatively close to the camera.

 

I just made that comparison myself, a week or so ago. There is a measurable difference.

 

I agree. This is why I use frame 61 instead. If you can see the toes, then you don't have appreciable overexposure/bloom. 

 

No, I'm talking about Patty's upper-leg being angled in Frame 72....thusly...

As a general rule...when Patty's lower-leg is straight vertical...her upper-leg is angled, to some degree.

Don't even bother to try to measure foreshortened body parts. Nobody buys the unforeshortened body parts, which is why this whole exercise is probably another waste a time, unfortunately.

 

Yeah, that's right.  I don't know about it being negligible though, Giganto...in Frames 61 and 72, Patty's head looks a bit squashed

 

Here is an overlay of a similar image of Patty, from the middle sequence...(F315)....with Roger on the same level of ground. Note how, at this particular scaling of F315, Patty's body length (from the calf to the shoulder) appears slightly shorter.....yet her head appears higher...

Scale the body parts to a common length and compare the rest of the body. Otherwise, the litmus test is whether anyone gives a ****. I feel like we are spinning our wheels here, like all the BF researchers before us.

 

I wonder if it the data is even readable after 25 years? If they used archival quality media, and stored them in appropriate surrounding, maybe? Does anyone even know how to get ahold of this stuff?

Glickman apparently got access to the ORIGINAL FILM and scanned frame 352 for his height estimates. This scanned frame was the gold standard for all photogrammetric analyses to follow. In the end he never knew Patty's distances from the camera so he couldn't estimate her height. However, the rest of his analysis was damned impressive!

Edited by Gigantofootecus

Share this post


Link to post
Share on other sites
Guest

^

 

I thought the original was lost when it was borrowed out to ANE in 1971 and then after Patterson died it wasn't returned to Patricia Patterson and ANE went into liquidation and the original hasn't been seen since.

Share this post


Link to post
Share on other sites
JustCurious

I double checked and I stand corrected, Glickman's scan was done from a copy of the original made one year after the fact.  That's the Eastman 78 Safety film Bill Munn's was working with I believe.  The scans were digital and amounted to 30 GB of data.  They didn't have such large hard drives back in 1993/1994, so it was put on CDs.  All the details are included in the report.

 

Something that makes this analysis difficult is that Patterson wasn't equidistant, nor at the same angle from the subject throughout the filming.  So not only don't we know the distance from the camera, we don't know the distance from the camera for individual frames.

Share this post


Link to post
Share on other sites
SweatyYeti

 

Don't even bother to try to measure foreshortened body parts. Nobody buys the unforeshortened body parts, which is why this whole exercise is probably another waste a time, unfortunately.

 

 Otherwise, the litmus test is whether anyone gives a ****. 

 

 

 

Whether or not anyone "gives a ****", Giganto... :) ...the one aspect of the PGF that can be settled, is what Patty's 'body height' was. I think that is something worth accomplishing.
 
Most importantly, though...we need to determine what the upper-limit is, on Patty's height...because if that upper-limit is determined to be well under 7'...then her exact height becomes a moot point. Patty's height only has significance, as to what she was, if it was 7' or greater.  

 

 

 

 

I feel like we are spinning our wheels here, like all the BF researchers before us.

 

 

I don't think so, Giganto....we are making progress in narrowing-down what Patty's height was. And a definitive determination of which size Lens was on the camera will tell us what Patty's height was, within a very small range. :)

Edited by SweatyYeti

Share this post


Link to post
Share on other sites
Bigfoothunter
Something that makes this analysis difficult is that Patterson wasn't equidistant, nor at the same angle from the subject throughout the filming.  So not only don't we know the distance from the camera, we don't know the distance from the camera for individual frames.

 

Patterson changed locations twice after is initial filming of the creature walking away from the creek. As I recall, there was a measurement recorded from Roger's prints to the filmed subject's tracks at a certain point in the film. Without looking for the reference - it may have been 88 feet.

 

In speaking with McClarin yesterday, he said that the plaster residue was still present on the sandbar when he and Green went there a year after the film was taken. Jim told me that he walked within an inch or two of the residue at the look back location . So knowing the camera's and their lens options - they should allow someone to get a pretty accurate idea of her height in relation to Jim's by way of Photogammetry. There should be an additional 1 - 2 inches of height added to Patty as her feet went deeper into the ground than McClarin's. So while Jim was indeed further off the correct path when he passed the stick embedded in the sandbar that Patty incidently stepped on - Jim was certain that he was no more than an inch or so to the wooded background in the film than was Patty.

 

do you remember if you walked next to the particles on John's side of the trackway or the other side as that question has come up
197724_10151521157032550_1298130117_n.jp
Other side, inches farther away than Patty's tracks.

Share this post


Link to post
Share on other sites
SweatyYeti

 

In speaking with McClarin yesterday, he said that the plaster residue was still present on the sandbar when he and Green went there a year after the film was taken. Jim told me that he walked within an inch or two of the residue at the look back location .

 

 

That isn't possible, Bigfoothunter. Jim couldn't have walked that close to Patty's trackway, at the F352 point...because, if he did...the path he continued on would have passed by in front of tree T-C2, and then right through the debris pile.

 

Here is a crop from Bill Munns' '20MM Lens' diagram....illustrating Jim's path being close to Patty's...

 

MunnsReport-JimPathwayDiagram-20MMLensC_

 

 

As it shows...if the line he is walking is continued straight ahead, Jim would pass by in front of T-C2, and thru the debris pile. 

 

Here is a smaller version of the graphic I posted recently...rotated, and with some added highlighting. It shows the same path contours as Bill's diagram, for Jim and Patty...

 

KrantzTrackway31_zpsea760cd4.jpg

 

 

As Patty turned-away from looking at Roger....only one step after F352....she turned her body in mid-step, and resumed the 40-45 degree angle path that she had been walking on. She had to do that to avoid walking right into the debris pile. 

 

After Jim's 'look back'...he continued walking straight ahead for several steps, and did not walk through the debris pile. Obviously, he couldn't have been within inches of Patty's trackway at F352.

Edited by SweatyYeti

Share this post


Link to post
Share on other sites
Bigfoothunter

^^

 

I do not know if the angle at which Green filmed the walk to that Roger had makes a difference or not. What I do know is that I copied and pasted Jim's response to me from our Facebook chat.

 

I asked Jim if he approximated where the tracks had been when doing the walk with Green and he replied:

 

Jim McClarin - Hi Bill, It was a fairly close approximation since particles of plaster of paris from the Titmus castings were visible around the track perimeters.

 

Bill Miller - Are you thinking of your being there in 1967 soon after the film was shot .... I am talking about it being a year later when you went with Green. Are you saying plaster particles was still there a year later?

 

Jim McClarin - Yes, it was there the following summer. No reason for it to get up and vacate the footprint sites.

 

Bill Miller - Do you remember if you walked next to the particles on John's side of the trackway or the other side as that question has come up

 

Jim McClarin - "Other side, inches farther away than Patty's tracks."

 

 

When asked if he ever filmed or took photos of the trackway ........

 

Jim McClarin - Bill, I recall making a super-8 movie of the trackway the first time I visited it with Henry and a radio reporter whose name escapes me (possibly Bob Barnett?). I also tried to do my own version of the comparison film of me that Green filmed next summer. Whatever happened to the movie I don't know but I sorta think I sent it to Green, he analyzed it, said it was taken from the wrong location, and that motivated him to come down the following summer with frames from the P/G film so that, by aligning objects in the foreground and background the way they appeared in the frames he was able to have his own 16 mm camera in virtually the same location as Patterson's. Green does not remember my sending the film to him or having seen and rejected it. If he sent it back to me it was lost by either me or by Loren Coleman to whom I gave my casts, files, etc. when I "got out of the business." Coleman does not have it and he has been asked by me and others about it. He says in his many moves he has lost all sorts of stuff. All I've gotten on it is dead ends. At the time I'm sure I lost interest in my own "recreation film" since it was not useful for making a height determination.

Edited by Bigfoothunter

Share this post


Link to post
Share on other sites
MikeZimmer

 

... Most importantly, though...we need to determine what the upper-limit is, on Patty's height...because if that upper-limit is determined to be well under 7'...then her exact height becomes a moot point. Patty's height only has significance, as to what she was, if it was 7' or greater.  ....

 

 

If I am computing correctly, a height of 7' for Patty require Giganto's 6' 3" estimate to be low by 12%.  His estimate method was composed of multiple steps, and each one would have an accompanying error range. It must be the case that errors propagate and are cumulative, but I wonder just how we could test to see if it is reasonable for his number to be out by 12%.  You might think that overall, errors would go in both directions, some low, some high. I am thinking of really looking at his approach again, to make sure that I properly understand it, and see how each stage in a multi-step procedure might contribute to error and uncertainty.

 

I have been thinking about the use of averaging, recommended by Backdoc in a different context, and also mentioned in a couple of places at least by Giganto.

 

 

My intial thought on averaging when it was brought up by Backdoc was that it might give us a good grouping, with a centre, but could not remove the effect of consistent bias. For an average to approximate a true value, I think that the errors must be randomly distributed about that central value. So, as an example from my youth, when using a rifle on a bench, shooting a target, you might get a nice grouping of shots, but the centre of the group might be considerably off target, maybe due to the sights being off, maybe due to windage or other factors.

One of the things that I find makes it hard to follow these discussions is the reference to frame numbers, as if they were old friends. Unfortunately, they are not on my friend list, so lack a solid referent.

Edited by MikeZimmer

Share this post


Link to post
Share on other sites
MikeZimmer

Great topic and post, Gf!  The torso, limb sizes and proportions of the animal in this image are striking: 

 

attachicon.gifPattyBlowUp_reduced.jpg attachicon.gifPattyRVCompilation.jpg

 

And then the PGf 'moving pictures' show that the inhumanly proportioned individual above moves in a very smooth and very natural fashion which is indicative of absolutely no costume ever.  I agree with your 6'3" height estimate and that her width (and/or, ASH ratio) is the kicker!  8 ^).  thx!

 

 

 

I was looking at an old thread recently and found a few posts by norseman on the effect of stuffing a kid in a snowsuit, the effect being to make it pretty impossible for the kid to lower his arms and then move them properly. This is a pretty common place observation for people with kids who live where there is winter.

 

The point here is that given the obviously massive thorax of Patty relative to height, whatever it is, a mime such as Bob H., or the vast majority of folks, would not be able to fit in such a padded suit, and then move the arms from the shoulder joint in the natural and easy swing that we see in the film. It just could not happen. This is a common sense point, that needs little beyond the admission of the obvious, Patty is built like a masonary reinforced outbuilding.*

 

 

* Yeah, I know that is not the most common way of putting it, but in deference to the power of the moderators, I try to maintain a little decorum. ;-)

Edited by MikeZimmer

Share this post


Link to post
Share on other sites
MikeZimmer

...

You can also use someone's known height to measure their arm length in a photo. In the following example Bob H has an arm length of 128 / 430 = 30% of his 6'2" standing height = 22", which was correct.

parnsbob.png

...

 

Obtaining Arm Length from a Known Height

 

with some additional considerations for error estimation and foreshortening

 

 

 

Based on work by Gigantofootecus

 

1: You can use someone's known height to measure their arm length in a photo.

2: Bob H. has a humerus of 73 pixels as measured from his image

3: Bob H. has a radius of 55 pixels as measured from his image

4: Note that there would be foreshortening for the arms, particularly the lower arms, and I don't see that this is being taken into account. His lower arm especially will appear to be shorter than it really is. Try standing the same way as you see in the picture and see the angle of the lower arm in particular. You would have to estimate this angle and use the trigonometric cosine ratio to make a correction, on both upper arm and lower arm, for a better estimate.

My rough and ready estimate is 10 degrees from the vertical for the upper arm, maybe +/- 5 degrees. The cosine of 10 degrees appears to be 0.98. So, adjusting for foreshortening in the upper arm gives 73 x 0.98 = 71.9 pixels.

My rough and ready estimate is 20 degrees from the vertical for the lower arm, maybe +/- 5 degrees. The cosine of 20 degrees appears to be 0.94. So, adjusting for foreshortening in the lower arm gives 51.7 x 0.9 = 49.5 pixels.

5: Bob H. has an unadjusted arm length of 73 pixels + 55 pixels = 128 pixels

6: Bob H. Has an adjusted arm length of about 72 + 50 = 122 pixels, a bit off the other calculation, and also this does not include the uncertainty around the degree of foreshortening. This estiamte could be tightened up with better measurement, probably being best done with two people, and some consistent method.

7: Bob H. Has a standing height of 430 pixels, as measured from his image

8: His arm length is (128/ 430) X 100 ~= 30 % of his standing height using the unadjusted figure

8: His standing height is 6' 2†= 74â€

9: Arm length = 30% x 74 = 22â€

10. Using the adjusted figures, His arm length is ((71.9+49.5)/ 430) X 100 ~= 28 % of his standing height, translating into 21â€.

If we look at the uncertainty we may have to add some upper and lower limits for the height, upper arm, and lower arm in pixels. Presumably, an intuitive understanding of human anatomy and external landmarks allowed us to come to some reasonable approximation of where the measurement end-points should be. However, in measurement there is always uncertainty. We can use logic to set limits in some cases, and also use informed guesswork to set limits. There will be anatomical landmarks to make reference to. A trained anatomist might make better estimates of boundaries for people.

In principle, you want to set some measure of confidence for the upper and lower limits. This is in the form of a probability. We can normally find limits with a probability of 100%, just specify impossibly large starting and ending positions, or at least wholly unlikely numbers. However, we really want to make our interval as tight as possible, while staying at a high level of confidence, say 95%. Going to 99% would be better. So, sharpen your pencils, and establish logical justification for coming up with tight limits that are still at a high level of confidence.

Research Question

Once we have established limits to accuracy at some measure of confidence, what are we to do with them? With one measurement, we have three values: lower limit, middle value, and upper limit. With two measurements, we have three more. Altogether we now have 3 x 3 possible combinations. With the 3 measurements, plus the uncertainty around the angles of foreshortening, I believe that we have 3 to the power 5 combinations, which appears to be 243 combinations.

The question arises as to how to use these error estimates. Here I propose that you really want a computerized solution of some sort, at least a darn good spreadsheet, to combine the various numbers and come up with a range for your final estimate. I confess to not knowing if this is how it would be done in practice by professionals, it just make some sense to me. I also do not know how you would assign a probabilistic level of confidence to the result, since you are combining five ranges, each with its own level of confidence. It may be that the overall level of confidence goes down through multiplication of probabilities, but this is a novel situation for me, and I don't really know how it might best be handled.

 

Share this post


Link to post
Share on other sites
Sign in to follow this  

×
×
  • Create New...