I have seen this when film was loaded into a projector or camera backwards. Most projectors will also run in reverse.

John

Never seen this myself either however agree with others. The old system of using 16mm and running it twice through the camera could explain what happened. Imagine the film was not at the end and the cameraperson swapped the reels and then swapped them back again. People did weird things back then.

Bob.

[edit] read my next post -- mystery solved.

I knew about the cameras that used 16mm film, but only exposed half the film in one pass. However, once the film was slit, that system should produce normal 8mm film.

Hmmm .... You've all got me thinking ...

Could this be 3D??

This film is from someone who owned a prominent camera/photo store in the center of Hollywood, and had access to lots of stuff. More to the point, something I didn't mention in my first post is that I ended up throwing out exactly half the footage because it appeared to be a duplicate. I never thought to look for a subtle offset between the "duplicates."

If it is 3D, will I even be able to detect the subtle offset between right and left lenses? I guess I will have to find some frames with objects close to the camera because those should have the most offset between right and left.

Yeah, the more I think about this, the more it makes sense. You could construct a stereo camera that took two images across the top half of a 16mm frame. Since they would each be half normal width, it would make sense to make them half normal height (to maintain, what was back then, the "normal" 4:3 aspect ratio). The camera drive mechanism would still be stock 16mm, so the frame would advance the entire distance of a normal 16mm frame, thus leaving the bottom half of the frame unexposed. You would then, if I am thinking correctly, put the film back in the camera and take the other half of the film which, now that the film is running backwards, would have the unexposed film in the upper half of the frame, just where it would need to be in order to be exposed.

The problem was, I'll bet, that once the film was exposed, someone forgot to have it processed as normal 16mm film, and it was instead processed with the technique all of you have described.

This is going to be a fun challenge. Of course if it turns out to be 3D, how will I ever display it?

Time to call James Cameron.

OK, thanks to everyone who helped. Sorry, no 3D (although I did find out that there was a Bolex 16mm camera that DID take 3D back then).

The problem was simple: It was normal 16mm film that someone mistakenly had processed into 8mm. Since I captured it with a Workprinter, which in essence gives me a scan of each frame, I think I can reassemble the original 16mm frame in Vegas by lining up four instances of the capture, with two pairs offset by one frame, and two other pairs reversed and flipped, and using track motion get the frames to align. At worst I'll have a little "Cinerama" effect at the boundaries.

My one remaining question is this: is there a way in Vegas to stop Vegas from expanding the video to fit the project? I want to create a project that is 1440x960 (double NTSC in each dimension) and then have each of my 720x480 captures fill exactly 1/4 of the frame. So far, I've been unable to stop Vegas from trying to scale the video up to fit the space.

It's still an neat challenge.

Thanks everyone!

I have shot and transferred a lot of reg 8mm which was 25 feet of 16mm perforated on both edges at the 8mm frame pitch. Half the film width was exposed on the first pass through camera and then the takeup and supply reels were swapped and the file run through the camera a second time exposing the other half of the film width. The fim processor split the film and splced the ends making the 50 foot roll. I have seen double exposed, half exposed , where the user forgot the run the file through a second time. I have no explanation odd frames on one roll and even frames on the another roll. The frames running upside down and backwards we did fro fun by flipping the processed clip end for end and front to back and splice it in. The framing would be off. :)

Old Ed

John,

When you get a change take a look along the perforations on the film and see who processed the film. I will be surprised if it is Kodak. Regular 8 is 16mm film but it has twice the perfrorations as 16mm film.

ed

You might face quite a challenge then.
You'd need to scan right to the edge of the film with an enlarged gate. You'd still have a small section of the frame damaged by the cut to wrangle as well.

Bob.

The perforations are definitely regular 8mm: big holes (compared to Super8) and tightly spaced (compare to 16mm).

You'd need to scan right to the edge of the film with an enlarged gate. You'd still have a small section of the frame damaged by the cut to wrangle as well.Fortunately, the Workprinter has an enlarged gate, no shutter, etc., so I get a scan of everything (I can even get the image that shows up between the sprocket holes on film from some cameras).

I'm still trying to figure out how to stop Vegas from enlarging the video to fit the project. I guess I'll use the 1440x960 project setting and then set the pan/crop for each of the four events to that. So far, the video looks like it is going to line up pretty well.

Of course we'll have to wait to see how it looks on the big silver screen.

You'd still have a small section of the frame damaged by the cut to wrangle as well.Here's what it all looks like, assembled in Vegas (nice use of pan/crop, track motion, and parent track motion):



As you can see, the top & bottom halves align absolutely perfectly. In fact, there is extra overlap, which I don't quite understand. However, the left/right (where the film is slit) doesn't get close.

I have to go back and re-capture, making sure I don't crop either horizontal edge. However, it looks like I may have a bigger gap than I can finesse.

Always a new challenge five minutes ahead.

Very interesting thread!

"The problem was simple: It was normal 16mm film that someone mistakenly had processed into 8mm."

John, this seems incompatible with what you've written in a subsequent post with reference to the sprocket hole spacing, which you've indicated was consistent with 8mm film. Splitting a roll of 16mm film wouldn't have any bearing on the pitch of the sprocket holes.

Is the spacing on 16mm film twice that of 8mm? If so, it seems more likely that a roll of double-8mm was filmed in a 16mm camera and then split as if it were 8mm, to produce the results you've described.

Edit: At any rate, with the image you have, you could tell viewers that it was filmed from an 'obstructed view' seat...

Read this.

Aha -- that's what I thought...

Thanks Jay. That completely explains it, and is 100% consistent with what I was seeing. This was definitely "Double 8."

I contacted Roger Evans (the inventor of the Workprinter) and confirmed that while my Workprinter XP has an enlarged gate, and while I can see clear to the sprocket holes and also to the top and bottom of the frame, the film still must have some support on the right side to keep the film from buckling, so there is some small part of the film that is obscured by the gate on the right side. This is the side where the film was split, and the part of the film that is obscured unfortunately ends up being in the center of the reconstructed frame.

So, the "obstructed view" effect (nice term, BTW) is how it is going to be.

I just found out that while this film was given to me by a contemporary, her father -- who actually took these films and is now over 90 -- is going to come up here personally from L.A. (350 miles) to pick up the film. He has obviously never seen this the contents of this film, and since it does contain images of the pope, it will be really fun to see his reaction.

This business can certainly be rewarding at times.

I think cbrillow's got it! Double 8 shot in a 16mm camera, then split by the lab. If you could take a file to your gate, it looks like you could restore it to 16mm. LOL
Ralph

There is no problem running double-8 film in a 16mm camera, but one would do that only in an emergency or for a special purpose, and then with very particular instructions to the processors. The reconstructed frame seems to be different in magnification left and right. I'm still wondering about the odd/even frames (in the apparently 8mm) being of different scenes. Doesn't seem to match the outcome of running double-8 in a 16mm camera. Maybe someone was trying to do a 2 projector 8mm cinerama.

The "different scenes on even and odd frames" was not true in the end.

Since John's scan was done as an 8mm film, one frame got just the top half of the 16mm frame (actually the top left corner of the full frame), the next scanned frame came from the bottom half of the same 16mm frame. When viewed as 8mm frames, looked like two different scenes on alternate frames, but really was upper and lower parts of the same scene.

John - regarding your image above:
Was the Pantheon actually used as a bar in the 1950's??? (lol)

Of course, the red "BAR" letters in the image are on posts sticking out from the corner of the sandstone building at the left edge of the frame, but it looks like they're fastened to the old Pantheon itself (or at least I'm assuming that's what that is...)

Tim L

Of course; I wasn't paying attention to the significant difference between top and bottom of scenes. The reconstructed frame doesn't match across the split, but I presume that image was generated with a mismatch in frame time which is easily corrected..

John,
did you get an answer to your question about filling the screen?
I think it is in the pan/crop function - there is a button to maintain aspect ratio or not...
if it is on, it will take the timeline video and fill the screen... creating a distortion for very narrow or tall images/video...
if it is off, it will leave the original aspect ratio alone and show black edges on the shorter edges (top and bottom or on both sides) ...

I'm sure there is a setting in the preferences that should change the default for that function...

I hope I understand the question correctly...

Until that time... Earl J.

Both Tim and Serena got it right: it was the top and bottom half of the original 16mm frame I was seeing. When split and then projected on 8mm equipment, it takes two 8mm frames to display the entire original 16mm frame. Tim figured it out pretty quickly, but when I originally looked at it, I thought the frames were from completely different scenes: one frame showed a bunch of people walking on the street, and the next frame was a shot of a building, seemingly a much darker exposure. Of course the street was fully lit, and the building above the street was in shadow.

There is no difference in magnification between the top and bottom half of the left frame. Same thing for the top and bottom half of the right frame (unless I nudged the pan/crop numbers, although I typed them in just to make certain the magnification stayed the same.

I downloaded the image and tried to trim each side to match... no dice...
There is a piece hidden in the black portion of the left side that connects to the right that is barely visible... primarily, the center portion of the car and the connecting pieces of the building and roof.
* * *
So. in my mind it becomes a matter of masking each side to the precise point where it will match the other side ... then, maybe as simple as picture-in-picture, matching each side...

Once matched, the completed composite might require a bit of color and contrast adjustment to iron out the slight discrepancies between the sides...

Please keep us posted on the progress and the solutions you devise...

Quite the challenge... I'm very interested in how you meet and overcome this wonderful challenge - wonderful because it is a great puzzle for Vegas as well as a wonderful project for the client. . .

Until that time... Earl J.