Jump to content
Sign in to follow this  
Guest

Photogrammetry 101: Solving The Trackway Triangle

Recommended Posts

Guest

I think the casting footage was shot at Bluff Creek and was definitely filmed with the same K-100 camera as the PGF. We have an opportunity to use physical rulers in this footage to estimate distances from the camera using photogrammetry and geometry. This should settle what lens was on the K-100 to shoot the casting footage.

RogerK100_zps536e8bc6.png

The numbers point strongly to a 25mm lens for the casting footage. So what is the likelihood that Roger switched lenses between the PGF and the casting footage? Just more grist for the mill.

Edited by Gigantofootecus

Share this post


Link to post
Share on other sites
Bill

Just Curious:

 

There is no connection between camera speed (in frames per second) and lens on camera. They are entirely unrelated functions of a camera.

 

 

If anyone here has noticed my conspicuous absence from the lens issue discussion, I am still working things through and don't plan on commenting until I am confident that I actually have a final solution. Hopefully, I may be able to do that in the near future.

  • Upvote 1

Share this post


Link to post
Share on other sites
masterbarber

So you're still sticking with this:

 

"I remain absolutely confident the lens on Roger's camera was not a 25mm lens but the specific focal length may not be the 15mm I described in my report release."

Share this post


Link to post
Share on other sites
Bill

MB:

 

I can't recall when I wrote that, but no, I am not presently "absolutely confident" about any size for the lens. It's hard to explain exactly why this is so complicated (at least for me) but I do hope to finally publish something that shows the many considerations I've explored and the ways that these varied considerations have been so contradictory.

 

That quote sounds like something from around 2009, and so much has changed since then, I wish I could go erase all of the outdated ideas, but the miracle of the internet doesn't let me. So these old ideas endure.

 

I will re-iterate something I've said before and stand by still: In all the data I have, there is an error, and I haven't found it yet. Virtually every fact must be re-checked to see if it is correct or in error, and it's an exhausting process. It might be green's measurements on site, or dahinden's. It might be lens distortion, or lens field of view calculations. I'm challenging every base assumption, and every piece of data used in this process.

 

Sorry that I can't be more conclusive about the issue at this time. But I am determined to resolve this question, since I was the person who raised it.

  • Upvote 1

Share this post


Link to post
Share on other sites
OntarioSquatch

I think that's totally understandable. It's normal to change one's opinion after new information comes along. That comment is originally from The Munns Report Release Number 1B on page 10, which was released 6 years ago I think. 

Edited by OntarioSquatch

Share this post


Link to post
Share on other sites
JustCurious

Duh, I was having a real brain farth there!!!  I actually know that, so I must not have been thinking with both sides of my brain.

 

Thanks for not calling me an idiot! :crazy:

Share this post


Link to post
Share on other sites
SweatyYeti

^

 

It's o.k., JC.....the important thing is, you were thinking with the honest side of your brain. ;)

 

When it comes to this particular subject.....we don't always see that happening. 

Share this post


Link to post
Share on other sites
SweatyYeti

Giganto, as a check on the scaling of Roger, in the 'cast pour' footage, in your graphic...I tried another method of determining scale for the image. 

 

Using a 'cast pour' footage Frame in which Roger's face is visible....I rotated the image, and scaled it to match the size of his head in both pieces of footage...

 

 

GigantoGraphic-RogerOverlayAG1B_zpsykp2p

 

 

I then took that image...rotated it back to it's original orientation, and overlaid it onto your graphic....to compare how our scalings of the 'cast pour' footage compared...(they are a near-perfect match)...

 

GigantoGraphic-RogerOverlayAG2B_zpsspusr

 

 

 

In the top animation, it looks like I may have Roger's 'cast pour' image slightly over-scaled. But I decided to leave it at that scaling, because when I transferred the image onto your graphic...it was a near-perfect match to your 'cast pour' image. 

 

So, it looks as though your scaling of Roger is very accurate....probably within 1 or 2% of what it should be.  :)

Edited by SweatyYeti

Share this post


Link to post
Share on other sites
masterbarber

Thanks Bill. GF makes a pretty compelling case for the 25mm lens being used, based on what I'm seeing here.

Share this post


Link to post
Share on other sites
Wheellug

so.. What if calculations were presented using various lens sizes?   Using objects such as Roger, Patty, caste size and more?  What other images are present in the film that can be used... prior to the final footage?  

Using the diorama also, distances and known objects as well.

Just do it all for 25mm, then for other possibilities?    

This would allow for.. If a 25mm.. Patty is x tall, if other size lens, she is y tall.  

If I recall, something was done once,but not to an extreme for multiple common objects.  

Share this post


Link to post
Share on other sites
SweatyYeti

Wheellug wrote:

 

so.. What if calculations were presented using various lens sizes?   Using objects such as Roger, Patty, caste size and more?  What other images are present in the film that can be used... prior to the final footage?  

Using the diorama also, distances and known objects as well.

Just do it all for 25mm, then for other possibilities?    

This would allow for.. If a 25mm.. Patty is x tall, if other size lens, she is y tall.  

If I recall, something was done once,but not to an extreme for multiple common objects.  

 

 

 

There is only one other possible lens, Wheellug....20MM.  The 15MM lens has been definitively ruled-out, as a possibility.

Edited by SweatyYeti

Share this post


Link to post
Share on other sites
Bill

The following foundation work is still needed to be done:

 

1. Bench testing multiple lenses on the K100 (like at least 5 of the 25mm Cine Ektar and five of the 20mm Anastigmat lens) to document the degree of precision of each type of lens. Any variance found among the same type lenses is the margin of error when estimating Roger's lens, which we are not testing as we haven't located it yet (we do know it's serial number so we may find it yet).

 

2. Doing a distortion analysis on the lenses to measure that factor with precision. When we calculate angles from any specific film frame, we need to know if any factor of distortion widens or narrows any specific angle, vertically or horizontally.

 

3. Putting the film image scanning lens through a similar distortion analysis and comparing those scans to high resolution flatbed scanning of the same frame images.

 

4. For comparisons of the McClarin footage, we need to find what lens is on Green's Revere camera. The photo suggests it is a 25mm Wollensack Cine Raptor lens, but I have bench tested one and it specs out at a slighting shorter focal length that the 25mm Cine Ektar, but the film frame analysis I get shows John's lens needs to be longer, by at least 1mm, not shorter by about 0.5mm. That issue needs to be cleared up. Also the problem of bench testing the 25mm Raptor lens on a Revere Camera (I tested it on a K-100) is that the film magazines are no longer available in any film stock current labs can process, so either I or somebody working for me must learn how to load these magazines by touch only in total darkness, with a film stock we can get processed today. Without these magazines loaded with current film, we can't bench test the Revere camera.

 

5. The fact that true photographic film is a dying technology and films and film processers are vanishing little by little imposed challenges on this research as well. I can by magazines loaded with Kodachrome film by Kodak, but the last lab processing Kodachrome closed about 1-2 years ago, so the film can't be processed anymore (except somebody who does it as black and white, not color, and it's an amateur operation.) Green's filming of McClarin (and the second walk footage by another cameraman and using a different camera as yet unidentified) are valuable forms of corroboration necessary to crosscheck against the studies of the PGF. All angles and object measures should check out perfectly in both films, for a prefect solution.

 

There are other issues and challenges, but this will give you some idea of what is complicating the task.

  • Upvote 2

Share this post


Link to post
Share on other sites
Drew

I think the casting footage was shot at Bluff Creek and was definitely filmed with the same K-100 camera as the PGF. We have an opportunity to use physical rulers in this footage to estimate distances from the camera using photogrammetry and geometry. This should settle what lens was on the K-100 to shoot the casting footage.

The numbers point strongly to a 25mm lens for the casting footage. So what is the likelihood that Roger switched lenses between the PGF and the casting footage? Just more grist for the mill.

 

If both Patterson and Green used the same lens, why was JM so far behind the supposed trackway?  Just an eyeball estimate shows him 10-20 feet beyond where Patty walked.  If he was following the trackway, why was he so far back?

Share this post


Link to post
Share on other sites
SweatyYeti

^

 

There wasn't much left of Patty's footprints, Drew....so part of Jim's pathway was guesswork.

 

There is a very clear, and undeniable difference in the paths the two of them walked....since, after the 'look back'....(when they both turned their heads away from the camera operator)....Patty turned her body back to a 45-degree angle path.....while Jim simply continued walking straight ahead....(parallel to the plane of the camera).

 

From what I've deduced from comparing their paths...it looks to me like Jim was somewhere in the range of 15-25 feet further back than Patty was, at the F352 spot. 

Edited by SweatyYeti

Share this post


Link to post
Share on other sites
Bigfoothunter

The first time Jim went to the film-site - there were around two dozen tracks visible - the following year when Jim was there with Green, Jim saw plaster residue still on the ground where tracks had been poured. He walked within inches just beyond those tracks and guessed at the rest.

Share this post


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

×
×
  • Create New...