Frank Agbat

Here is some information that Gary Mack was kind enough to share with me when I discussed this particular Muchmore print ........... http://www.aparchive.com/aparchive/pages/p...news/ln_jfk.htm The work was done on an "Archangel" machine (not a Spirit as I told you) in England, then converted to US broadcast standards. The conversion introduced visual artifacts and it was also transferred at the wrong speed. The "restored" version runs much faster than normal. Slowing down the tape introduces even more confusing visual artifacts. Gary Mack The process of converting this version of the film from PAL (European) standards to NTSC (US) DID introduce artifacts AND it ended up at the wrong speed. This was due to error(s) in the conversion process, not the conversion process itself. Edit -- they have also done some zoom/crop in places. I don't know if this is a result of stuff in the analog domain (copy from 8mm version to 16mm) or digital (Discovery channel doing some "production effects" or simple ineptitude).

The Discovery channel DVD version of the Muchmore film has some issues. It does not seem to have the same issues with any of the other included films. It appears that the Muchmore clips may have been telecined on a 768 x 578 (or 720 x 576) 50Hz field-rate PAL system w/interlacing. There is nothing wrong with this, as this is a supported and standard configuration. However, in NTSC-world, this needs to be converted to 720x480 59.94 system prior to distribution... This is where I think things might have gone wrong... From the look of things, the original (PAL telecine) was NOT inverse telecined. Instead, it appears to have been de-interlaced using a bob/weave filter. This produced some of the ghosted images that people have seen in their clips. When the clip was authored to the NTSC DVD, interlacing was added back in. However, it was added back in to the previously "bobbed" version. Thus, some of the ghosted "fake" frames are interlaced with real frames, and some real frames ONLY appear as interlaced frames. The side-effect is that it MAY not be possible to recover all of the frames in full-frame detail (or it is going to take some time and effort to do so). Another side-effect is clips that will appear "jumpy" if one attempts to reduce the clips to purely progressive images using methods that would work with a normal telecined DVD. I'm going to start digging in to this literally field by field (not frame by frame). In the interim, I have to caution against drawing any conclusions using clips and/or stills from this source. As we've seen in other threads, NTSC/PAL craziness can be the source of anomalies...

Chris, Telecining adds a duplicate frame (or frames) to compensate for different frame rates. A purely duplicated frame (or frames) would be quite easy to detect -- even when watching a film, and most certainly so when looking at it frame-by-frame. Missing frames are harder to detect, but as we have shown on the "Missing Nix Frames" thread, not impossible. This is why I was theorizing about the clever use of blur to allow the insertion and removal of frames. Just about every DVD will have the issue that you ran in to. In general, extracting purely progressive frames isn't too difficult, especially once you know what to look for and have the right software handy. Most, if not all, software DVD players de-interlace frames using some combination of filters known as "Bob and Weave". This de-interlaces the video before displaying it on your monitor (computers are progressive scan devices). If you have one that allows you to single-step through frames, you'll notice some have the ghost images. These are the result of the bob & weave filters and aren't on the original film. For example, here are two frames from the Hughes film: However, on the DVD (in this case, the discovery channel show) there is a "frame" in between these two "real" frames. This frame is interlaced -- portions from A and B are on alternating lines (see the example below -- left picture). When you play a DVD on your computer, it applies the filter(s) mentioned above so that you don't see the annoying horizontal lines on your monitor. On the right side of the lower picture is the result of such a filter: The "Field Bob" frame is less displeasing to the eye than the horizontal lines when played on a computer. However, neither frame is actually on the Hughes film.

Hi Chris, You posted two animations previously -- one of which had some ghosting effects, the other one was clear. The ghosting effects were almost certainly not on the original film. They are the result of the process necessary to convert them to a format that is playable on Televisions. Film is originally a progressive source -- nothing but a series of complete and sequential pictures one after another. Traditional televisions don't work quite that way, so film sources must be modified (processes to make frame rate corrections and a concept called interlacing) to be viewable. Essentially, duplicate frames (to correct frame rate) and "special" frames (interlaced) frames are added. This process is called "Telecining." The result is a film that is playable on a television, yet runs at the correct frame-rate. When a DVD is ripped (or frames are advanced one-by-one in DVD player software and captured), if the added frames are not removed (a process known as inverse Telecine, or IVTC), they will show up when you frame-advance on a computer (which is a progressive device). That is why one must be careful to work with only the purely progressive frames when making computer animations from DVD frames of a film source (gasp!). To answer your question -- one would not want to, nor would it be necessary to, create ghosting or interlace effects on a film source (unless you were doing something intentionally for special effects, etc).

Eugene, The upper left photo is Stoughton, the lower right is Rickerby, if memory serves me correctly.

Jack, You're right about the framing of the projector; I meant to comment on this previously, but got distracted with other things. The rounded corners of the images *could* be a result of the telecine process used. Unlike taking photographs of individual frames (e.g. MPI's method), the telecine process actually "plays" the film. However, while we're on the subject of frames, etc. Jack -- how difficult would it be, in your opinion/experience, to intentionally utilize motion blur to add or substitute frames in a short movie sequence? For example -- what if one did something like this: ORIGINAL: Frame 1 (Clear) Frame 2 (Clear) Frame 3 (Some motion blur) Frame 4 (Clear) REPLACED WITH: Frame 1 (Original frame - Clear) Frame 2 (Replace with motion-blurred frame 1) Frame 3 (Original frame - Some motion blur) Frame 4 (Replace with motion-blurred frame 4) Frame 5 (Original frame 4 - clear) What has been accomplished here is that the original frame 2 is gone completely, replaced with a blurred version of frame 1 (it could even be slightly panned depending on available frame space). The original frame 4 is moved one frame later, with a blurred version of itself replacing it. I've removed the data from frame 2, yet managed to add time to the film. The question is -- how possible is it using optical technology to intentionally create motion blur?

Chris, It looks to me like you do not have truly progressive frames. The 2/3 frames show signs of interlacing and/or improper de-interlacing. This may be why you are seeing ghost images.

Robin, Thank you for sharing the frames. The obvious side-effect is to note that there are obviously better quality versions of the film(s) out there than appear on the NFV/Groden DVD. Hopefully, my Discovery DVD will arrive quickly -- I'm eager to see the footage it contains.

The only major problems I see with establishing sync with this clip are: 1) The shortness of the clip 2) The low frame rate of the camera (12fps) 3) The distance from the camera to the limo. Ostensibly, Toni Foster might, again, make a reasonable starting point for establishing relative sync/timing.

Bronson sequence - original version: http://i49.photobucket.com/albums/f280/fra...-orig-small.gif Larger version of the stabilized sequence: http://i49.photobucket.com/albums/f280/fra...l-enh-anim2.gif

Bronson sequence - stabilized / color corrected / gaussian deconvolution. Source: NFV/Groden

One has to wonder if MPI took into account the PIXEL aspect ratio of a television (arguably their target for the DVD production) vs. that of a monitor. Computers use (almost) square pixels, whereas video (TV, et al) use pixels that are rectangular. It would be nice to think that MPI would have corrected for this (After Effects, Premier, et al, have plenty of facilities to handle Pixel Aspect Ratio and to produce correct videos for both Digital and Analog (TV) targets). Considering they couldn't even get the frame numbering correct, it is rather a stretch to imagine them working out the subtleties of aspect ratios. Of course, the entire process is fraught with danger. The analog/optical portion of the process can introduce any number of issues even before the digital domain enters the equation. The transparency or film scanning process is *probably* going to be a square-pixel device. Again, one would like to think (!) that MPI managed to preserve the original aspect ratio through these two steps. The result should be a digital image with the correct aspect ratio in a square-pixel format... Hopefully. If properly authored, the DVD product will have the correct aspect ratio for rectangular pixels... Hopefully. The NFV sources are another matter entirely, as I'm now convinced that the DVD originates from a Video source that was captured and digitized. This is why (among other reasons) I strongly caution against using these types of images for microscopic examination. They are useful for many, many things, but not for all things.

You are correct! (I'll edit my post accordingly). In the missing final frame, one could easily argue some combination of "lazy" and/or "thinking for us" i.e., MPI determined that there was nothing of value in this frame and merely ignored it. The more I closely examine their work, the more I wish they'd make the individual frames in all their various zooms, etc, available for purchase in full-resolution.

Speaking of missing frames... MPI's version of the Z-film is missing the following frames: 155, 156, 208, 209, 210, 211, 341, 350. (edit - also frame 486 is missing -- thanks for point this out, David) If I understand the provenance correctly, it is claimed that the MPI version was taken from the "camera original" in the archives, and that the damage was caused while the film was in the custody of Time/Life. (Inexcusably, MPI does NOT correct for the missing 341 and 350 frames, so all the frames after 340 are mis-numbered). Apart from the utter incompetence of Time/Life to allow such an important film to be damaged at all, I got to thinking about the particular frames in question. Specifically, why would certain frames end up getting damaged? What causes such damage? What is special about those frames? The timing problems with the preposterous WC fantasy forced the search for evidence of a first shot prior to the "Stemmons Road Sign-vicinity" shot. This could conceivably explain some of the 155/156 damage (probably from wear and tear, bulb-heat damage, etc). 208-211 seems fairly obvious as well -- the search for the magic bullet shot (or shots)... 341 and 350, though, are curious. Unlike the aforementioned frames, there is no evidence of a pending problem in the preceding frame nor in the following frames. They're simply missing from MPI's collection of frames in any of the formats provided. So what is special, if anything, about those frames, or did some knucklehead at Time-Life simply break the film in two places and spliced around it?

An extremely sane position. Perhaps the only one that will preserve said sanity... While we're on the subject of the autopsy photographs... From what I've read and/or interpolated, I think it is possible that some of the "autopsy" photographs may actually be post-autopsy in timing. Perhaps they were taken during reconstruction work by the morticians after the doctors had completed their work. There was, I believe, some discussion about having the body lie in state (open casket), but this idea was eventually vetoed. This makes "Fox #5" neither a fake, nor a genuine autopsy photo. (I *think* this opinion is based on "In the Eye of History", but it has been a while since I read it -- so weigh accordingly).

There is actually very little missing in the frame. The top and bottom of the spliced area do not register well with the previous frame, as I discovered in my attempt to "repair" the frame. Additionally, there appears to be some warping/non-linear distortion in the area near where the break occurred, but only on the lower half of the frame. I attributed this to heat, but that was merely a theory. I don't really know why the splice is as it is other than "poor workmanship" on the part of the splicer.

With all due respect, Frank ... the very first post that fathered this thread mentions a film that is not supposed to have any missing frames during the shooting as having missing frames. The removal of film frames would be an alteration by definition. "The red lines between the black lines are my estimates of missing frames. When the film runs the missing frames give an appearance of smooth continuity. When properly placed in space on the composite the jerky leaps of speed at the missing frames is obvious." I think you showing the films in sync was an important task to complete and I commend you for it, but it was that needed task to answer that alteration question once and for all concerning the possibility of the assassination films being tampered with .... least ways that is the way I read it. Now I will step back and observe your continued posting on the Muchmore film in the "Missing Nix frame" thread because I think it too, is part of authenticating these assassination films running in sync with each other. Also, thanks to Robin - for his film grabs are really good ... some of the best I have seen pertaining to the Muchmore film. Bill Miller Sorry to get huffy, Bill. You're right to say that alteration is the crux of, and is germane to, this thread. I'm a little tired and short-tempered tonight, and it came through in my message. When that happens, it is usually time for me to call it a night.

Robin, Excellent -- thank you. I'm very eager to get my Discovery Channel DVD -- this version appears to be vastly superior.

Okay... We have 2.3 x 1023 threads where you guys can call each other names and argue about alteration, one's lineage, etc. Back to the task at hand... 1) I'm going to fix all the damn numbering on MPI's brain-damaged frames to correct for their blithering errors. Unless anyone has any strong objections, I'll put a Z-time field on the frames that have surrounding black space (the "under sprocket" edition, e.g.). 2) I'm going to re-do any sync animations that involve any z-frames after Z340. 3) I'm going to rip the Z frames from two other sources: the NFV/Groden DVD and JFK the movie. 4) I'm going to keep beating my head against the wall that is Muchmore sync...

The answer ios "Yes" ... I believe Costella got them correct. Also, the best image source that I can offer is to load the frames from the movie JFK because that is one of the better and clearest images of the assassination you will get to use at this time. Those images are far better than MPI's IMO. Bill Thanks, Bill. I'm going to spend a little time re-numbering MPI's frames to match Costella's version. This isn't a big deal for the frames that do not have a number juxtaposed on them. I also plan to generate individual frames from the NFV DVD as well as "JFK". I just wish MPI would have employed some rather simple quality control to their efforts. The total and correct number of frames is 486, right? MPI is missing 155 & 156 (but skips numbering accordingly) MPI is missing 208, 209, 210, & 211 (but skips numbering accordingly) MPI is missing the correct Z341. What they show and have numbered as 341 is actually 342 MPI is missing the correct Z350. What they show and have numbered as 349 is actually 351 MPI stops at what is numbered 483 but is in reality 485. MPI does not include 486. Anything I missed? (Edit -- Brain damage -- corrected to show the correct numbering... sorry)

Another thing to remember is that a very small angular error extrapolated over a long distance results in a very large error. This is why I try very hard to avoid relying on angles, perspective assumptions, lens-corrections, etc, to determine which frame corresponds to which. I'm reasonably certain that there weren't any UFOs or strange warps in the space-time continuum in DP that day. Physics is the ultimate workaholic -- it just never takes a break. However, what you're seeing cannot and should not be dismissed. You're looking at things in a different way, and it is good to do so.

MPI made mistakes in their numbering. Also, what happen with Altgens legs can be seen in other places on the Zfilm, as well. The couple at the top of MPI Zframe 344 is just one example. Take notice of the peoples legs in Z301 and Z302 that were walking behind Toni Foster. Bill We should produce a comprehensive list of erroneous MPI frames. I knew about the missing frames in the early portion of the film -- the numbering even skips correctly. It is fairly easy to make a "corrected" version that includes these missing frames. However, to fix the higher numbered frames, I'm going to have to fix the numbering on each individual frame (sigh) in all 4 versions (sigh, again). Needless to say, I only want to do this one time... Is Dr. Costella's version correct? If so, I'm going to use it as a sanity check for re-numbering. This is also going to cause us to re-consider our film sync sanity checks that involve higher numbered frames. Mathematically, I'm not too worried, as a frame is just an event that happens in time. However, when we go to visually compare frames, we'll need to be certain that what we are seeing really *is* what happened at the particular moment... MPI (or whomever is holding them) needs to release the full-resolution stills of each frame. Between poor frame registration, mis-numbered frames, and MPEG2 compression, we have less than the ideal images with which to work.

I previously posted some frames... However, I realized that I had several more versions of the Z-film to consider. Dr. Costella's combined edit has a Z341 that does not appear to be in the MPI version. Also, the frame you posted above (Jennings version) is Z350 in the Costella combined edit version. The "MPI Z349" that you show, is Z351 in Costella.

Muchmore frame 42 from the NFV/Groden DVD, optically zoomed version. (This frame shows some of the damage that lead to the splice in frame 43). Nix frame 23 from the NFV/Groden DVD optically zoomed version

John, I must admit an inordinate level of frustration created by attempting to sync the Muchmore film to either Nix or Zapruder... The following cryptic image is another attempt to understand what is going on with the two films: The lines represent the height of the top of Agent Hill's head in the two films. The yellow lines are from Nix frames 41-50 (41 is on the right, largely obscured by a red line). The Red lines are MM frames 59-66. I noted that there is a "low spot" in both graphs and tried to more-or-less line up that point. The low-point in Nix is NS-46. The low-point in Muchmore is MS-64 If this point is lined up and run through the time computations, it leads to the same problem I keep running into with Muchmore... Aligning MS64 with NS46 ends up mathematically matching MS41 with NS23 exactly (both cameras supposedly run at the same 18.5fps rate). This makes no sense to me, as there is no skull damage in M41, but NS23 CLEARLY shows splatter... However, aligning M42 with NS23 causes disagreement with both this chart AND the headturn sequence. I think it is highly likely that MM and Nix may be some fractional frame apart from one another, and the only way we're going to be able to determine this is by finding a Zapruder alignment (or Bronson, yet unexamined) that gives us a better time base. Either that, or the two cameras are not really operating at the same fps and are drifting...