Film-Tech Forum ARCHIVE: Very tricky changeovers

This is topic Very tricky changeovers in forum Film Handlers' Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=1;t=011781

Posted by Leo Enticknap (Member # 534) on 02-10-2015, 02:04 AM:

Just finished playing a 35mm archival print of To Be or Not to Be (1942), and am waiting for the last bulb to cool. The R1-2 changeover happens as Carole Lombard is walking away from the camera in medium shot, and in the middle of a line of dialog when the "over" cue appears! I clipped one syllable and saw the tiniest jump cut when I ran the first two reels this afternoon to set focus, framing and volume. So on the actual public screening I threaded reel 2 four frames further from the start of the action than I normally do (assuming that the heads and tails have not been cut), and this time (and without wanting to sound overly smug) it was bang on - even her footstep looked fluent across the change.

All the other changeovers in this show were almost as critical - they were all in the middle of a scene, some were in the middle of a continuous shot and most had movement within the shot taking place during the over cue. I find this weird - in the '30s and '40s, the standard Hollywood studio practice seems to have been to put a fade to black at the end of a reel wherever possible, to minimize the on-screen impact of the changeover happening very slightly early or late. Not in this show! This is the first film I've ever come across in which every changeover has to be done almost frame-perfect in order for it not to be noticeable to the audience (assuming that they don't see the cues and know what they mean).

Anybody know of any others?

Posted by Marcel Birgelen (Member # 6801) on 02-10-2015, 06:28 AM:

It seems like you've got some nice projectionist training material there [Wink]

.

But I think the editor is to blame here too. I guess you'll always want to have a changeover at a non-critical point in time, if remotely possible. Putting it in the middle of a line of dialog or in a moving shot just feels moronic or incompetent or both...

Posted by Rick Raskin (Member # 1561) on 02-10-2015, 06:44 AM:

I once ran a print of "Thoroughly Modern Millie" where a C/O occurred right in the middle of a scene where Julie Andrews is in the middle of a line of dialogue. What I saw most evident was a horizontal misalignment of the machines. I didn't run the print enough times to sync any dialogue issues.

Posted by Scott Norwood (Member # 30) on 02-10-2015, 07:28 AM:

Vertigo and GWTW both contain changeovers where music carries across the reel change.

Presumably, the assumption at the time was that no frames would be cut out of the print at the reel ends, so it would be possible to make frame accurate (or nearly so) changeovers.

The fade-out/fade-in thing also requires nearly frame-accurate changeovers in order to not destroy the effect.

Posted by Louis Bornwasser (Member # 3063) on 02-10-2015, 07:36 AM:

A number of Elvis shows were shot VistaVision. When released to the drive ins, they used regular 35mm. Changeovers were in the middle of the songs!

Also, remember back in the 30s that sound changeovers were made with a big knob that had zero in the middle and volume on both sides. It was impossible not to fade down and then back up. This was physically large and coupled to at least one other knob with a shaft and gearing. There was a certain amount of inertia in this. Words would have been lost.

Posted by Richard P. May (Member # 3598) on 02-10-2015, 10:52 AM:

AN AMERICAN IN PARIS has the changeover from reel 6 to reel 7 in the middle of the long ballet sequence. It always works pretty well, as the image is very different, but the music is continuous.
THE RED SHOES, in contrast, has the entire ballet on one reel. The curtain goes up, and at the end of the sequence, it goes down.

Posted by Tony Bandiera Jr (Member # 2365) on 02-10-2015, 12:40 PM:

CHARIOTS OF FIRE had dialog on most of the changeovers....I ran that on Davis-Monthan AFB during my first few months of being a projectionist. It ran three times that weekend so after clipping a few words or so on my first show I had time to adjust and nail the changeovers for the rest of the shows.

There were some others I have run but the titles escape me right now.

Posted by Brad Miller (Member # 2) on 02-10-2015, 02:34 PM:

Manual changeovers. How silly. [Razz]

My automated ones are always frame accurate with no effort.

Posted by Martin McCaffery (Member # 37) on 02-10-2015, 04:44 PM:

The China Syndrome had a change over while the emergency warnings were going off and on in a steady pattern. If you didn't hit it just right it was like an aural jump cut.

Posted by Leo Enticknap (Member # 534) on 02-10-2015, 06:21 PM:

quote: Marcel Birgelen
But I think the editor is to blame here too. I guess you'll always want to have a changeover at a non-critical point in time, if remotely possible. Putting it in the middle of a line of dialog or in a moving shot just feels moronic or incompetent or both...

I'm wondering if the reel ends in this print were determined by the archive that made it rather than the original editor and/or negative cutter. Only one of the reels had negative lab cues - the rest were all Clint Phare-type scribed cues. It was also a 100-minute movie in five reels (it's very unusual for every reel of a feature print to be the full 20 minutes, especially from that era - 16-18 minutes is more common), and there was nitrate decomposition artifacting printed from the negative in one section. So I'm wondering if the source element from which this was printed was assembled from multiple sources itself, and that the person(s) assembling it simply didn't know what doing a manual changeover involves and therefore didn't give any thought to the positioning of the reel ends.

quote: Scott Norwood
Presumably, the assumption at the time was that no frames would be cut out of the print at the reel ends, so it would be possible to make frame accurate (or nearly so) changeovers.

Good point. After 2K reels became the norm in the mid-30s but before platters and towers, no-one would ever have needed to cut a leader or tail or leave ID frames, and so release prints would simply not ever have lost any frames from the start and end in normal use. For the picture changeover at least, the ramp-up speed of the motor on a given projector is an unchanging, known quantity, and so the operators at each theater would eventually work out what point in the leader to thread to such that the first frame of the action is in the gate when the over cue appears. Once you've done that, the accuracy of the change is limited only by the speed of the human reflex.

quote: Louis Bornwasser
Also, remember back in the 30s that sound changeovers were made with a big knob that had zero in the middle and volume on both sides. It was impossible not to fade down and then back up. This was physically large and coupled to at least one other knob with a shaft and gearing. There was a certain amount of inertia in this. Words would have been lost.

Wow! And I thought we were old fashioned, having separate picture and sound change buttons! In that case, no wonder the major studios liked to end their reels with a fade-out. It must have taken a lot of skill to do that and the pix change on the nail, unless the booths in those days were so lavishly staffed that there was a separate projectionist to change over the audio.

Posted by Scott Norwood (Member # 30) on 02-10-2015, 06:31 PM:

quote:
And I thought we were old fashioned, having separate picture and sound change buttons!

Is that an old-timey thing? I have worked in booths that were installed anywhere from 1938 to 2012 and they are pretty much evenly split on having coupled vs. separate controls for picture and sound c/o. I never bothered to correlate it with the date of the installation.

On the most recent install, I was able to specify the configuration, and I requested footpedals for picture c/o and pushbuttons for sound c/o. The picture c/o is also connected to the douser in the video projector in that booth so that seamless changeovers from film to video and back can be made easily. They have the Dolby remote boxes, too, so that it is easy to switch between film and video sound formats.

Posted by Marcel Birgelen (Member # 6801) on 02-10-2015, 07:01 PM:

quote: Leo Enticknap
I'm wondering if the reel ends in this print were determined by the archive that made it rather than the original editor and/or negative cutter. Only one of the reels had negative lab cues - the rest were all Clint Phare-type scribed cues. It was also a 100-minute movie in five reels (it's very unusual for every reel of a feature print to be the full 20 minutes, especially from that era - 16-18 minutes is more common), and there was nitrate decomposition artifacting printed from the negative in one section. So I'm wondering if the source element from which this was printed was assembled from multiple sources itself, and that the person(s) assembling it simply didn't know what doing a manual changeover involves and therefore didn't give any thought to the positioning of the reel ends.

The absence of lab cues on all but one reel is probably the most glaring evidence this was reconstructed from multiple sources. The lab monkey tasked to create it probably only cared about the most efficient way of handling things [Wink]

.

I know some directors often make it hard to get it always at the end of a scene, but even during the mass-production 35mm years, I never liked sloppy editing work in regards to reel endings.

quote: Scott Norwood
Is that an old-timey thing? I have worked in booths that were installed anywhere from 1938 to 2012 and they are pretty much evenly split on having coupled vs. separate controls for picture and sound c/o. I never bothered to correlate it with the date of the installation.

Well, I'm actually quite curious. For what reason would you actually want to have those controls separated?

Posted by Sam D. Chavez (Member # 1841) on 02-10-2015, 08:07 PM:

There is audio trailing the picture at the ends of reels and you can delay the sound c/o ever so slightly in order to hear it all. Also, there is almost never audio in advance of the first frame of picture on the incoming reel so it all works out nicely as the trailing audio at the tail of the outgoing reel was actually the sound for the incoming reel.

Louis was mentioning the swing arm changeover from the '30's but this was only Western Electric, others brands were made more directly just like today but sound c/o was always separate from picture. It was only when cinemas begin to automate that the c/o functions were tied together for expedience.

Posted by Monte L Fullmer (Member # 2797) on 02-10-2015, 09:28 PM:

What I used to do: is douser/sound delay on occasion.

I'd hit the douser button first on second cue, then, in a split second later, do the sound change knowing of the added sound on the tail.

Made for a good smooth changeover.

Posted by Mike Blakesley (Member # 26) on 02-10-2015, 09:33 PM:

Our old system did the douser and sound with one button.

Posted by Paul H. Rayton (Member # 1817) on 02-10-2015, 10:38 PM:

The film prints of Ken Russell's movie "Tommy" (1975) had at least two changeovers in the course of the film that required precision in timing at the c/o, to get the outgoing chords of the ending music and the precise initial instants of the incoming reel. Careful threading was imperative, and God forbid your finger slipping on the button when the motor start cue appeared! I ran it during the early commercial runs of that show at the Bruin in west L.A., and had to pay super-close attention to several changes each show.

All that is solved in the DCP version that now circulates!

Posted by Stephan Shelley (Member # 8587) on 02-11-2015, 08:55 PM:

Ran an Emerson, Lake and Palmer music film many years ago. It was a video transfer and no thought about where the changovers were. Most were right in the middle of a music number. If you did it right you did not loose a beat,

Posted by John Eickhof (Member # 323) on 02-12-2015, 07:44 PM:

Through the years I have seen and operated in a couple booths that had no electric changeovers! One had a slim plywood sheet mounted on and across the projection ports! It had rectangular cutouts and was mounted in slides with a handle in the middle and right end for it to be slid back and forth opening one port or the other! You had to be quick and it made an interesting dissolve in the picture! The other was actually all metal and was branded by weaver!

Posted by Jerry Axelsson (Member # 3175) on 02-13-2015, 12:17 AM:

Once upon a time when we we ran "THE LAST BOY SCOUT"-1991, the projectionist which unpacked the print was in a hurry due to a very late arrival of the print.
At the end of reel 5, the change-over cue is just before a bad guy falls into the rotor-blades of a helicopter.
As it turned out, the Swedish censors had cut out this splatter scene between the start cue and the change-over cue which ruined the function of the original cue marks...

Posted by Manny Knowles (Member # 1171) on 02-17-2015, 01:10 PM:

It's been a while, but I think the first c/o in JURASSIC PARK happens in the middle of a musical cue.

Also, IIRC the first c/o in JAWS has a line of dialogue right up to the last frame of the outgoing reel.

The c/o in CRY OF JAZZ also occurs in the middle of a music cue.

There's a reel in SPHERE that begins or ends with exactly one frame of computer monitor static.