Jump to content
norseman

Thinker Thunker size comparison of Patty

Recommended Posts

Gigantofootecus

The Arm Ruler used to calc the distance from the camera to Patty at frame 352.

 

 

352_DFCa.png

  • Upvote 1

Share this post


Link to post
Share on other sites
Gigantofootecus

 

SY, if Patty was 118 feet from the camera at frame 352 then you can extrapolate along her trackway to frame 426 (shadow of TC-3 on Patty's back) and calc the Camera-352-426 triangle so you can fit TC-3 into the mix.

 

TreeShadows1.thumb.png.54dab90bdaebf27d4fdb6d255e14b4da.png

 

Frames 352 -> 426 = 74 frames  @11.2 frames/step, 41" per step = 22.6 ft

 

Here is the Camera-352-426 triangle:

 

352-426_triangle.thumb.png.979fead9f48c0003d6b37ca69ce6edec.png

 

 

Note the distance to frame 426 is 134 feet and the on-site measured distance to TC-3 was 95 feet. Quite a discrepancy (40 ft), which leads me to wonder if everyone got Roger's camera position wrong. After all, he had footprints all over the place but everyone measured from the wrong prints, which were at least 20 feet closer to Patty than he was actually filming her. How would they have known which footprints represented the camera position for any given frame? They wouldn't have, which is why it is more likely that the on-site measurements were inaccurate, than a non-standard lens was used to film the PGF. This would resolve the lens controversy and we can carry on with our calculations based on a 25 mm lens.

  • Thanks 1
  • Upvote 2

Share this post


Link to post
Share on other sites
hiflier

I know there is a lot to this and a lot of work on it has been done to date and I thank you all for it.

 

This is probably so far off base since I'm no where near the level you guys are as to be laughable but I'll go ahead anyway. We can determine the pixel width of Patty in any given frame but since the parameter goes from 352->426 I'll stay with that. The pixel width of TC's 1,2,and 3 are know from pixel width comparisons to the hard measurements on site. between all of that can anything be determined buy comparing Patty's pixel width in frame 352 with that of Patty's pixel width in frame 426. It would seem that Patty's angle to the camera is slight enough? Could the method be used for other determinations? If there are known distances to TC's 1,2, and three then pixel width should also be determinable.

 

The sun angle is good for figuring out time of day but as far as distances one would thing that pixel differences in moving object could be an important factor when looking at distance issues. My guess is that all of you have already thought of this and rejected it but since I don't know that I thought I would ask. And I will thank you for your patience ahead of time :) 

Share this post


Link to post
Share on other sites
Gigantofootecus
1 hour ago, hiflier said:

I know there is a lot to this and a lot of work on it has been done to date and I thank you all for it.

 

This is probably so far off base since I'm no where near the level you guys are as to be laughable but I'll go ahead anyway. We can determine the pixel width of Patty in any given frame but since the parameter goes from 352->426 I'll stay with that. The pixel width of TC's 1,2,and 3 are know from pixel width comparisons to the hard measurements on site. between all of that can anything be determined buy comparing Patty's pixel width in frame 352 with that of Patty's pixel width in frame 426. It would seem that Patty's angle to the camera is slight enough? Could the method be used for other determinations? If there are known distances to TC's 1,2, and three then pixel width should also be determinable.

 

The sun angle is good for figuring out time of day but as far as distances one would thing that pixel differences in moving object could be an important factor when looking at distance issues. My guess is that all of you have already thought of this and rejected it but since I don't know that I thought I would ask. And I will thank you for your patience ahead of time :) 

 

You have good instincts hiflier. You're describing the foundation of photogrammetry, which is how the distance from the camera affects an image. Measuring objects in photos is quite complicated because you are projecting 3D objects onto a 2D medium (film). This involves ortho-rectification, which requires a lot of algorithms, which you just have to take at face value. There are few peers that can critique a formal photo analysis, especially on non-scientific media platforms such as the BFF.

 

The following shows how photogrammetry and geometry can be used as you suggested:

288_480.png.a8dd45bb6053d295dbc01e71911ab937.png

 

With this info we know that Patty was 1.5x farther away from the camera at frame 480 than at frame 288, which is vital info for constructing the 17 step triangle formed between these frames. Resolving the Camera-288-480 triangle will tell all, IMO.

 

Edited by Gigantofootecus
  • Thanks 1
  • Upvote 1

Share this post


Link to post
Share on other sites
hiflier

Beautiful really, G! Thank you. And thank you for your kind compliment. Your clarification of the issue meant a lot.

Share this post


Link to post
Share on other sites
SweatyYeti
16 hours ago, Gigantofootecus said:

 

SY, if Patty was 118 feet from the camera at frame 352 then you can extrapolate along her trackway to frame 426 (shadow of TC-3 on Patty's back) and calc the Camera-352-426 triangle so you can fit TC-3 into the mix.

 

TreeShadows1.thumb.png.54dab90bdaebf27d4fdb6d255e14b4da.png

 

Frames 352 -> 426 = 74 frames  @11.2 frames/step, 41" per step = 22.6 ft

 

Here is the Camera-352-426 triangle:

 

Note the distance to frame 426 is 134 feet and the on-site measured distance to TC-3 was 95 feet. Quite a discrepancy (40 ft), which leads me to wonder if everyone got Roger's camera position wrong. After all, he had footprints all over the place but everyone measured from the wrong prints, which were at least 20 feet closer to Patty than he was actually filming her. How would they have known which footprints represented the camera position for any given frame? They wouldn't have, which is why it is more likely that the on-site measurements were inaccurate, than a non-standard lens was used to film the PGF. This would resolve the lens controversy and we can carry on with our calculations based on a 25 mm lens.

 

 

Thanks for doing more work on this geometry puzzle, Giganto.  Your input is of great value, in this type of analysis. 

 

There are a couple of figures in your post, though....that I think may need some 'tweaking'. 

 

First, regarding Patty's 'distance to camera'....I don't think she could have been quite that far away....due to the very small level of facial detail, detectable on Patty.

 

I have been thinking, recently, that her 'distance to camera' was greater than 102'. The reason I think that, is because the Photogrammetry solution produces a 'walking height' of about 5' even....(using the 25MM Lens @ 102'.)  Keeping the 25MM Lens' focal length in the equation...and increasing the 'distance to camera' can reasonably produce a 'walking height' result of 5 1/2'...to close to 6'...(though, I don't know by how much the 'distance to camera would need to be increased, to get to that height).

 

So, I think her 'distance to camera' was most likely greater than 102'....but, I find it hard to believe that it could have been as great as 118', given the very fine detail visible on Patty's face...(such as the lips, and the whit-ish spot right below her right eye.)

 

Secondly....with regards to this 40 ft. discrepancy...

 

Quote

Note the distance to frame 426 is 134 feet and the on-site measured distance to TC-3 was 95 feet. Quite a discrepancy (40 ft),

 

The distance to TC-2 (what you have labeled as Tree 3) was measured by John Green to be significantly closer than 134'....he measured the distance as 115'...

 

John-Green-Filmsite-Measurement-Diagram1

 

Patty's 'distance to camera', @ F426….would have been just about the same...possibly just a few feet further away.

 

Rene Dahinden had measured the distance from 'camera-to-TC-2' as being 95 feet...but he had Roger's camera position too close...by about 20'.

 

I can post more on this, in the next day or two. And I'll spend more time looking over your work, Giganto.  I'm sure that we can eventually get these distances worked-out....to where everything 'fits', and makes sense. :)

 

 

 

 

 

 

Edited by SweatyYeti
  • Thanks 1
  • Upvote 1

Share this post


Link to post
Share on other sites
Backdoc

Great stuff by you guys on this distance, height, angles and so on.

 

I think of a person going to an eye doctor.  They go in and when they leave they have a Rx for glasses if needed. The glasses end up being the right 'fit' because the person can usually see perfectly provided the eye doctor got it right.    If that patient went to another eye doctor, the same Rx would result because the other doctor would use the same scientific anaylsys to come up with the Rx. If there was a diff the diff would be very very very small.

 

Based on this thinking, it would seem to me the kind of analysis being offered here on the PGF would be no different.  It would be interesting to take TV Time to devote to having some 'skeptic' analysis along those same lines.  The math is always out in the open so it would be harder to hide behind the math facts on arm length, ratios, height, and so on.

 

I wish these shows on TV would devote at least 5 min to such a segment in a "the legend meets science" way.  Maybe this work will start that process.

 

I don't wish to give you guys more homework but I would make just one suggestion:

 

It would be my wish you could put together a video of the step by step process of this analysis.  What is written is great but a video would be even better.  What can I say, I want even more.

 

 

Share this post


Link to post
Share on other sites
Gigantofootecus
On 9/24/2019 at 10:51 PM, Gigantofootecus said:

The Arm Ruler used to calc the distance from the camera to Patty at frame 352.

 

 

352_DFCa.png

 

SY, I posted the numbers so we can test them. We have to assume the following:

 

1) The Arm Ruler is 28" as shown upthread.

2) The placement of the Arm Ruler vectors in frame 61 matches 352.

3) We know the lens eqn parameters needed to determine the distance to frame 352

 

Using the above criteria, note the following placement of the Arm Ruler vectors to calculate 102 feet from the camera. Doesn't look right to me.

 

Arm_Ruler_102ft_From_Camera.thumb.png.805212c654eeac9d641dde1cc89d1871.png

 

Post where you think the Arm Ruler end points should be and re-calc the distance from the camera. Keep in mind that we're talking about +/- 2 feet for every inch the Arm Ruler is out. 

 

But 1st, why do you believe the distance to TC-3 was 95 feet?

 

 

Share this post


Link to post
Share on other sites
SweatyYeti
1 hour ago, Gigantofootecus said:

 

SY, I posted the numbers so we can test them. We have to assume the following:

 

1) The Arm Ruler is 28" as shown upthread.

2) The placement of the Arm Ruler vectors in frame 61 matches 352.

3) We know the lens eqn parameters needed to determine the distance to frame 352

 

Using the above criteria, note the following placement of the Arm Ruler vectors to calculate 102 feet from the camera. Doesn't look right to me.

 

Post where you think the Arm Ruler end points should be and re-calc the distance from the camera. Keep in mind that we're talking about +/- 2 feet for every inch the Arm Ruler is out. 

 

 

Thanks for your post, Giganto.  I'll look over your 'arm ruler' line placements, and get back to you on it, later. 

 

I don't think that Patty's 'distance to camera' was 102'.....I think it was further than that....maybe closer to 110'.  

 

I consider the 102' figure to be very unreliable...since, as far as I know....it was a figure that Rene Dahinden came up with, in 1971. But, all of the footprints had disappeared by then....Patty's and Roger's. So, I don't know how he could have made such a determination....and hence, I don't see why that figure has been used as much as it has been...over the years.

 

Rene had made a couple of mistakes, in determining that distance...he had Patty's position too far back on the sandbar....(by about 20')…..and he had Roger's position on the sandbar wrong, by about 20', also. He had Roger's position too far forward....(closer to the 'main log' than he actually was).

 

 

Quote

But 1st, why do you believe the distance to TC-3 was 95 feet?

 

I'm a bit confused, here. Which tree are you referring to as "TC-3"??...

 

Tree-Shadows-on-Back-Two-Frames1-Sharpen

 

 

 

9 hours ago, Backdoc said:

Great stuff by you guys on this distance, height, angles and so on.

 

I think of a person going to an eye doctor.  They go in and when they leave they have a Rx for glasses if needed. The glasses end up being the right 'fit' because the person can usually see perfectly provided the eye doctor got it right.    If that patient went to another eye doctor, the same Rx would result because the other doctor would use the same scientific anaylsys to come up with the Rx. If there was a diff the diff would be very very very small.

 

Based on this thinking, it would seem to me the kind of analysis being offered here on the PGF would be no different.  It would be interesting to take TV Time to devote to having some 'skeptic' analysis along those same lines.  The math is always out in the open so it would be harder to hide behind the math facts on arm length, ratios, height, and so on.

 

I wish these shows on TV would devote at least 5 min to such a segment in a "the legend meets science" way.  Maybe this work will start that process.

 

I don't wish to give you guys more homework but I would make just one suggestion:

 

It would be my wish you could put together a video of the step by step process of this analysis.  What is written is great but a video would be even better.  What can I say, I want even more.

 

 

 

 

Thanks, Backdoc…..not to worry, there will be plenty more on this line of analysis. :) 

 

I do plan on putting together some videos on my analysis....hopefully, in the not-too-distant future.

Edited by SweatyYeti

Share this post


Link to post
Share on other sites
SweatyYeti

I tried working with your figures, Giganto….but I'm not familiar with the Lens Equation you used, in your Arm Ruler graphic.

 

Can you post the equation, itself? Thanks.  :) 

Share this post


Link to post
Share on other sites
Gigantofootecus

SY, here is a quick primer for applying the lens eqn:

 

The following graphic demonstrates the relationship between the camera/lens parameters and the vertical angle of view for 16mm film:

 

VAV.thumb.gif.3ba36e216da607b91c63c24e1cdd6804.gif

Photogrammetry attempts to restore 3D to 2D film. We must restore it to accurately measure projected objects on the film as if they were in 3-space. That process is called ortho-rectification. In this case, we are working with "close range" photogrammetry where X number of pixels = Y number of inches and the discrepancy between ortho-rectification becomes negligent.

 

My first stab at the lens eqn to est Patty's height was the following:

 

dist2cam25mm.png.e9fb22849a285095faa98a1030786b7b.png

 

This is a template for you to apply the lens eqn but we need to fine tune "hf". Otherwise, the results line up with my Arm Ruler estimate. Film only gives us angles based on the camera/lens parameters. Distances and heights must be derived.

 

The following shows the angles of the 17 step (16.8) triangle frames 288->480. The angles are the only thing we "know". Add the 41" step length into the mix and complete the triangles to "ortho-rectify" Patty. Apply the lens eqn to map out each of her steps and solve the 17 step triangle. This will also solve her height, distances from the camera, etc. This tees up the PGF for a true re-enactment.

 

17_steps_angles.thumb.jpg.540e1a8ce15a8363cbbce7d96f55b32f.jpg

 

ps. The only thing good thing M.K. Davis ever did (IMO) was to scan in Patricia Patterson's Cibachrome positive of frame 352, which is the Rosetta stone for all future analysis of the PGF.

Edited by Gigantofootecus
  • Thanks 1

Share this post


Link to post
Share on other sites
Bill

For consideration, here's the lens formula as published in the American Cinematographer's Manual.

12-04crop.jpg

  • Thanks 1

Share this post


Link to post
Share on other sites
SweatyYeti
8 hours ago, Gigantofootecus said:

SY, here is a quick primer for applying the lens eqn:

 

The following graphic demonstrates the relationship between the camera/lens parameters and the vertical angle of view for 16mm film:

 

VAV.thumb.gif.3ba36e216da607b91c63c24e1cdd6804.gif

Photogrammetry attempts to restore 3D to 2D film. We must restore it to accurately measure projected objects on the film as if they were in 3-space. That process is called ortho-rectification. In this case, we are working with "close range" photogrammetry where X number of pixels = Y number of inches and the discrepancy between ortho-rectification becomes negligent.

 

My first stab at the lens eqn to est Patty's height was the following:

 

dist2cam25mm.png.e9fb22849a285095faa98a1030786b7b.png

 

This is a template for you to apply the lens eqn but we need to fine tune "hf". Otherwise, the results line up with my Arm Ruler estimate. Film only gives us angles based on the camera/lens parameters. Distances and heights must be derived.

 

Thanks, Giganto...for the additional explanation, and graphics. 

 

I'll work with the numbers tomorrow....and see what I get. 

 

 

Quote

The following shows the angles of the 17 step (16.8) triangle frames 288->480. The angles are the only thing we "know". Add the 41" step length into the mix and complete the triangles to "ortho-rectify" Patty. Apply the lens eqn to map out each of her steps and solve the 17 step triangle. This will also solve her height, distances from the camera, etc. This tees up the PGF for a true re-enactment.

 

ps. The only thing good thing M.K. Davis ever did (IMO) was to scan in Patricia Patterson's Cibachrome positive of frame 352, which is the Rosetta stone for all future analysis of the PGF.

 

 

There is a comparable method to the '17 Step triangle', Giganto…..which is simpler, and would involve fewer variables. It is using the measured distance across the base of the far hillside...

 

F352_Lens_Cross-Field_measurements1.jpg

 

 

Steven Streufert, and his crew, measured the distances between the large trees on the far hillside, back in 2012.  Those measurements aren't ideal, though, for use as a cross-field measurement....since the locations of the trees form an arc across the hillside.

 

A better measurement, which can still be made...would be a straight-line measurement, as shown by the green line.

 

Edited by SweatyYeti

Share this post


Link to post
Share on other sites
Bill

Sweaty:

 

I received this information from a friend who was just at Bluff Creek a week ago, and was kind enough to do some measurements for me that are precisely what you asked for (and what i wanted for a very long time).

 

His email to me:

 

Bill,

We completed the task at Bluff Creek today.  I'll get the exact numbers to you tomorrow. However, I recall that from the large tree (we used center mass to center mass of the trees) to Laurel was 77' and to Hardy it was 82'

I'll give you more details, etc tomorrow when my brother and I sit down and go back over thing.

V/R

Mark Riendeau.

  • Upvote 1

Share this post


Link to post
Share on other sites
SweatyYeti
13 hours ago, Bill said:

Sweaty:

 

I received this information from a friend who was just at Bluff Creek a week ago, and was kind enough to do some measurements for me that are precisely what you asked for (and what i wanted for a very long time).

 

His email to me:

 

Bill,

We completed the task at Bluff Creek today.  I'll get the exact numbers to you tomorrow. However, I recall that from the large tree (we used center mass to center mass of the trees) to Laurel was 77' and to Hardy it was 82'

I'll give you more details, etc tomorrow when my brother and I sit down and go back over thing.

V/R

Mark Riendeau.

 

 

That's great, Bill!  Thanks for mentioning it. :) 

 

That measurement should go a long way, in determining which size lens was on the camera...since that distance is about 3/4" of the full width of the film frames....and, because of it's long distance from the camera. 

Share this post


Link to post
Share on other sites

×
×
  • Create New...