Missing Nix frames

[Robin Unger]

Similar positions re: Newman's in Nix and Muchmore frames.

Edited December 22, 2006 by Robin Unger

[John Dolva]

Using Franks Table N11=Z301 and using a suggested M17 as Z287 then N11 would be M31. The M frame posted is M16 so that would be Z286 or N-5 which doesn't exist.

[Frank Agbat]

The Muchmore film has a splice/break right around the location of the headshot (naturally...).

Mr. Groden's DVD has a Black and White version that was made before the original(?) film was broken. This version, although very poor quality, allows us to establish whether any Muchmore frames were lost as a result of the break/splice.

On the DVD, the B&W version starts one frame later than the Color version. However, in both cases, the films run in sequence with one another. In other words, the splice appears to be repair due to a break, rather than deletion of frames.

For the purposes of frame numbering, I am calling the splice frame "MS-42" (where MS-1 is the first frame in the "shooting sequence" portion of the Muchmore film). The color version of the NFV/Groden DVD runs from MS-1 to MS-65. Lacking any more precise information in the interim I'm going to work under the assumption that the camera operates at 18.0 fps, and will adjust accordingly if better information becomes available.

[Ed O'Hagan]

2. something moves behind the pergola and light glints

John,

From what I can see, light was being reflected from camera lenses in that location.

[Frank Agbat]

I've been observing various aspects of the Nix film.

For this example, I have stabilized a series of frames. The source of these frames is the NSV/Groden DVD, 1st-generation copy of the Nix film. *NOTE* - I *have* adjusted the saturation and luminance values in these frames to make the background more visible. The frames have also been run through a deconvolution filter using a Gaussian blur model. (This filter is similar, but vastly superior, to an "unsharp mask" or sharpen algorithm. It is useful to correct small focus problems without the nasty artifacts and edges that normal sharpen filters create. The downside is that this type of filter is painfully slow)

Please consider the following sequence, and note especially the motorcycle cops:

Source frames NS23 - NS70

...and a close-up of the sequence:

Source frames NS23 - NS66

...two of them brake HARD shortly after the headshot and look directly to their right. They react almost simultaneously, presumably in reaction to a shot. For those keeping track of time, their reaction starts about 25 frames after NS22 (~1.3 seconds).

[Frank Agbat]

I found myself with the rarity that is known as "free time" today...

The Muchmore film (at least the version on the Groden/NFV DVD) has a break in frame #42 (where 1 = camera restart). The film has been repaired, via a splice, and comparison to an intact (but poor quality) B&W version of the film, no frames have been lost. The result is that the color film has a large gap in the center of one frame (MS-42). Other frames have damage (bulb heat, most likely). The good news is that the limo area is not included in the damage or in the splice.

Here is the damaged frame:

I got the crazy notion that most of this image is actually intact, and that the splice introduces an inordinate amount of white gap. I split the damaged frame into two images, and aligned the top and bottom halves against fixed landmarks. The result of that is:

The following mini animation is very much a work in progress and is far from final. However, it shows a stabilized sequence of Muchmore frames. In this sequence, I started with the "de-spliced" frame 42 and had a little fun filling in the gap area. It isn't perfect (I'm pretty good with digital image processing, etc, but am not much of an artist!), but it does allow viewing the limo without the splice causing the frames to jump. The frames have also been through my favorite deconvolution filter...

[Frank Agbat]

The following table is proposed, but I lack the certainty of it that I had for the Nix frame alignment, as I cannot find a definitive frame-rate for the camera. Lacking this, I have it currently operating at 18.0 frames per second, as that is the published spec for the model of camera used (Keystone K7 s/n 20648). This table assumes that MS17 is most closely aligned with Z289, which is necessary to make the head shot work. Lacking any solid alignment point, I set MS-41 = Z313. The numbers are then back-reconciled to be expressed in terms of the Z325 zero-time.

John Dolva postulated MS17 = Z287 based on the child's hand position. The problem with this alignment and an 18fps camera is one of simple math. If these frames are in perfect alignment, this creates the situation where MS39 falls closest to Z313. Over 26 frames (1.421 seconds), there could be nearly 1/2 frame worth of drift due to differences in frame rate (assuming 18.0 for Muchmore) -- less if the cameras have more similar frame rates. So the most that could possible happen is Z313 aligning to MS40... However, there is no evidence of head injury or ejecta/splatter in MS39 or 40.

The result of this theory is the following sequence of events:

MS-40, Z312, N22, -- MS-41, Z313, N23, -- MS-42, Z314, N24

MS-Frame   Time (Z-325 reference)

01		-2.8888888889 (closer, timewise, to Z272 than to Z273)
02		-2.8333333333
03		-2.7777777778
04		-2.7222222222
05		-2.6666666667
06		-2.6111111111
07		-2.5555555556
08		-2.5000000000
09		-2.4444444444
10		-2.3888888889
11		-2.3333333333
12		-2.2777777778
13		-2.2222222222
14		-2.1666666667
15		-2.1111111111
16		-2.0555555556
17		-2.0000000000 (close to Z-289, which is at -1.9672131148)
18		-1.9444444444
19		-1.8888888889
20		-1.8333333333
21		-1.7777777778
22		-1.7222222222
23		-1.6666666667
24		-1.6111111111
25		-1.5555555556
26		-1.5000000000
27		-1.4444444444
28		-1.3888888889
29		-1.3333333333
30		-1.2777777778
31		-1.2222222222
32		-1.1666666667
33		-1.1111111111
34		-1.0555555556
35		-1.0000000000
36		-0.9444444444
37		-0.8888888889
38		-0.8333333333
39		-0.7777777778
40		-0.7222222222
41		-0.6666666667	  (Head shot)
42		-0.6111111111	  (Splice due to break)
43		-0.5555555556
44		-0.5000000000
45		-0.4444444444
46		-0.3888888889
47		-0.3333333333
48		-0.2777777778
49		-0.2222222222
50		-0.1666666667
51		-0.1111111111
52		-0.0555555556
53		 0.0000000000	   (NS-35 / Z325 - our reference)
54		 0.0555555556
55		 0.1111111111
56		 0.1666666667
57		 0.2222222222
58		 0.2777777778
59		 0.3333333333
60		 0.3888888889
61		 0.4444444444
62		 0.5000000000
63		 0.5555555556
64		 0.6111111111
65		 0.6666666667

[John Dolva]

Frank, this will be interesting to resolve as 17 and 287 are ones where the hand is extended and the frames on either side it is not.

I wouldn't automatically assume the headshot is 41-42. (there are as yet puzzling reasons to consider even 36-37) however, am looking for a second matchpoint. There's also reason to think that she took a step forward at about halfway through the filming. More later.

[Frank Agbat]

Frank, this will be interesting to resolve as 17 and 287 are ones where the hand is extended and the frames on either side it is not.

I wouldn't automatically assume the headshot is 41-42. (there are as yet puzzling reasons to consider even 36-37) however, am looking for a second matchpoint. There's also reason to think that she took a step forward at about halfway through the filming. More later.

John,

I think there may be a parallax issue to consider, based on the location of the two cameras. To my eye, in MS-16, the young man's right hand is extended outward. In MS-17, it moves closer to his body (perhaps he is clapping -- uncertain). In MS-18 it has vanished from Muchmore's viewpoint. It remains hidden in MS-19.

In Z-287, his hand is well outside his body, and in Z-288 it is *still* outside his body area, although it has moved closer. Unfortunately Z289 and Z290 are both blurry AND under the sprocket. However, by 290, even with the blur, it is obvious that his hand it no longer extended.

Because of angles, I think his hand will disappear from Muchmore's view *before* it will be in front of his body from Zapruder's view. This is why I felt that 289 was a better candidate than 290. However, I think 287 is too early for the match point. I could very easily be wrong about this, though...

Another potential match point that I looked at, but was unable to resolve to my satisfaction is the wind-gust that puff's the right-side of Jean Hill's bright red coat. However, this area seems to be obscured by one of the motorcycle cops...

[John Dolva]

I should stop saying good work, it seems invariably to be so. So, plain 'business' language.

I ride bikes, though never a harley, and never on the right side of the road. Curious about what the cop does with his hand? Is there some reason to reach like that? Debris to dislodge, comm switch?

I'll just mention there seems there may be reason to consider more frames missing in M. Just mention incase similar indications are found.

Interesting also it's the three far cops that react, by slowing down, changing the formation from the arrow to a diagonal running north.

[Robin Unger]

Hi Frank/John

Is this the same SPLICE frame.

[Ashton Gray]

Frank, your work is technical expertise realized as artistry without the slightest compromise to the technical expertise—which combination approaches real magic.

Please consider the following sequence, and note especially the motorcycle cops:

Source frames NS23 - NS70

...and a close-up of the sequence:

Source frames NS23 - NS66

...two of them brake HARD shortly after the headshot and look directly to their right. They react almost simultaneously, presumably in reaction to a shot. For those keeping track of time, their reaction starts about 25 frames after NS22 (~1.3 seconds).

Without wanting to take anything at all away from the "reaction to a shot" hypothesis, I still feel it responsible to amend as follows:

• 1. I don't know of any testimony of any of the policemen at issue saying that they saw or heard or reacted to a shot from there. If it exists, I would like to be directed to it. And I think it would make a valuable augmentation to what you've shown above.
  2. Reaction to a report (shot-like sound) might be a more comprehensive statement, leaving other possibilities in consideration, such as possible intentional misdirection.
  3. At the end of the first sequence above, at least two of the three men who were on the knoll steps have started running up those steps, one of them pretty far up them by the time the camera tracks there. That running away motion, in the confusion and the apparent shooting of the President as Clint jumps onto the back of the limo, could have commanded police attention to some degree.
  4. Several of the motorcycle policemen had been hit with considerable debris from the head shot.
  5. This is the most troublesome: at some point very soon after the shooting, Sheriff Decker broadcast on DPD channel two for all available men to go immediately to the railroad tracks east of Elm—which is the direction that the motorcycle cops look. The only time stamp for Decker's broadcast that I know of is 12:30. That embraces 60 of the longest seconds conceivable, and there is no record that I know of that would allow verification of how soon after the head shot he made this broadcast. That makes it troublesome. That, and the fact that I don't even know if the motorcycle cops were monitoring that channel. Maybe it's irrelevant to what is shown in the sequence, but I thought it at least should be mentioned. (In fact, I'd love to know just how quickly Decker made that broadcast. But that's something else entirely.)
  

Really outstanding work that defies adequate thanks and acknowledgment.

Ashton

Edited December 26, 2006 by Ashton Gray

[Robin Unger]

[John Dolva]

Robin, Frank may have to answer. I haven't reached that yet. Off hand it looks a composite of two frames where the limo is centered and the top frame is semitransparent.

Very nice second frame, is it from a complete set?

[Robin Unger]

Motorcycle cop already looking RIGHT at z-312