WEBVTT Kind: captions Language: en 00:00:00.500 --> 00:00:03.237 Class, teacher’s gotta bit of a hangover. 00:00:04.134 --> 00:00:05.948 I... I can’t do it today, so 00:00:05.948 --> 00:00:08.233 we’re gonna watch a mooooovie! 00:00:08.843 --> 00:00:13.182 Let me just whip out the ol’ Filmosound, here… 00:00:15.658 --> 00:00:20.996 [projector whirrs to life, and muffled speech is heard coming from it] 00:00:21.355 --> 00:00:24.921 This is a Bell & Howell 16mm film projector. 00:00:24.921 --> 00:00:28.077 As you might have guessed from its name Filmosound, 00:00:28.077 --> 00:00:30.845 plus the fact that it has a volume and tone control, 00:00:30.845 --> 00:00:32.653 plus the fact that sound is coming from it, 00:00:32.653 --> 00:00:39.538 this thing is not only capable of recreating the illusion of movement through rapidly projecting a series of still images onto a screen, 00:00:39.538 --> 00:00:46.682 but it can also play a soundtrack synchronized with that moving imagery to recreate a true audio-visual experience. 00:00:46.682 --> 00:00:49.025 Just like the movies in a theater! 00:00:49.025 --> 00:00:51.500 But where is that sound coming from? 00:00:51.500 --> 00:00:53.528 Film is an optical medium, 00:00:53.528 --> 00:00:57.057 and doesn’t sound come only on discs or magnetic tape? 00:00:57.057 --> 00:01:02.354 Well Jimmy, it’s time we learned a littlebit about sound on film. 00:01:02.354 --> 00:01:08.730 The 16mm film that this projects was a very commonly used sort of mid-tier format. 00:01:08.730 --> 00:01:12.427 As a matter of fact it’s still used in some production environments. 00:01:12.427 --> 00:01:16.858 Sitting between the cheap-but-limited 8mm film used for home movies 00:01:16.858 --> 00:01:21.644 and the expensive-but-highly-detailed 35mm film used in Hollywood, 00:01:21.644 --> 00:01:26.135 16mm found widespread use where that middle-ground was appropriate 00:01:26.135 --> 00:01:30.997 such as for television productions, industrial and educational films, and the like. 00:01:30.997 --> 00:01:37.829 But, as ever, exceptions abound and some theatrical releases were shot on 16mm film 00:01:37.829 --> 00:01:42.337 and plenty of television productions were shot on 35mm film 00:01:42.337 --> 00:01:46.097 so there are no hard and fast rules here. 00:01:46.922 --> 00:01:48.067 *slurp* 00:01:48.067 --> 00:01:54.610 Looking at the film itself we can see a series of images in the center, plus sprocket holes on one side of the film... 00:01:54.610 --> 00:01:58.079 and an Audacity waveform on the other. 00:01:58.079 --> 00:02:01.273 That’s… interesting. 00:02:01.273 --> 00:02:03.934 I wonder if that’s the soundtrack. 00:02:03.934 --> 00:02:04.921 It is. 00:02:04.921 --> 00:02:09.503 But before we get to how the projector reads that, which is trust me - very exciting 00:02:09.503 --> 00:02:14.274 we should probably talk a little bit about how the projector handles the film it projects. 00:02:14.274 --> 00:02:17.167 Because that itself is pretty fascinating. 00:02:17.167 --> 00:02:19.179 The basics are pretty straightforward. 00:02:19.179 --> 00:02:26.822 Obviously the projector shines a bright light through the film so that a lens can throw a big blown-up image of that film at a screen, 00:02:26.822 --> 00:02:32.789 and it’ll just show one image after the other real fast and boom - The Illusion of Movement. 00:02:32.789 --> 00:02:34.153 But be warned. 00:02:34.153 --> 00:02:37.697 It might be tempting to simply move the film through the projector smoothly. 00:02:37.697 --> 00:02:40.061 But try that and you’ll be disappointed. 00:02:40.061 --> 00:02:46.142 Instead of seeing anything coherent on the screen, you’ll see a blurry mess of shapes moving quickly upward. 00:02:46.142 --> 00:02:51.806 To actually see anything, every image on the film needs to be still while it’s projected, 00:02:51.806 --> 00:02:57.996 and that’s why projectors like this are elaborate mechanical contraptions which make that trademark projector noise. 00:02:58.857 --> 00:03:02.670 [that trademark projector noise] 00:03:03.639 --> 00:03:06.629 Here, let’s look behind the lens. 00:03:06.629 --> 00:03:11.064 This part is known as the film gate and it’s absolutely critical. 00:03:11.064 --> 00:03:17.362 The lamp and its condensing lens blasts light through this hole and thus through a single frame of the film, 00:03:17.362 --> 00:03:23.817 and when the projection lens is in place it essentially magnifies that tiny frame and throws it at a screen. 00:03:23.817 --> 00:03:27.197 To the left of the light hole are three metal fingers. 00:03:27.197 --> 00:03:29.988 This is the film advance mechanism, known as the shuttle, 00:03:29.988 --> 00:03:33.529 and those fingers engage with the sprocket holes on the film. 00:03:33.529 --> 00:03:42.780 If I slowly turn the projector mechanism by hand, you’ll see that these fingers repeatedly stick out, move down, then retract and move back up. 00:03:42.780 --> 00:03:46.123 An awful lot like the feed dog on a sewing machine. 00:03:46.123 --> 00:03:49.936 With the help of this pressure plate to provide a bit of friction, 00:03:49.936 --> 00:03:54.166 the result is a stepped, rather than continuous, film motion. 00:03:54.166 --> 00:03:58.828 Every frame of film is held stationary so it can be projected clearly, 00:03:58.828 --> 00:04:03.933 and after a small fraction of a second the film is quickly advanced to the next frame. 00:04:03.933 --> 00:04:12.543 To keep you from seeing that movement, a rotating shutter synchronized with the mechanism blocks out the lamp’s light while the film is in motion. 00:04:12.543 --> 00:04:18.769 Once the next frame is in position and the film has stopped, light is allowed to pass once more. 00:04:18.769 --> 00:04:23.722 In a motion picture camera, the same mechanism is used - however in that case, 00:04:23.722 --> 00:04:28.702 the film is collecting light from a lens to capture photographic images, 00:04:28.702 --> 00:04:33.188 and the shutter not only prevents it from getting exposed while advancing to the next frame 00:04:33.188 --> 00:04:40.070 but also can change the length of time it is exposed through varying the size of the opening on the shutter - 00:04:40.070 --> 00:04:45.938 and that’s why they use "shutter angle" in the movie business and not fractions of a second. 00:04:45.938 --> 00:04:53.808 Anyway, in a projector the constant opening and closing of that shutter means that the projected image is in fact flashing. 00:04:53.808 --> 00:04:55.674 Early on this was very noticeable 00:04:55.674 --> 00:05:00.368 and the frantic flicker on the silver screen is where the slang term “flick” comes from, 00:05:00.368 --> 00:05:01.569 fun fact. 00:05:01.569 --> 00:05:10.120 Later projectors, in an effort to make this less noticeable, have shutters which block the light multiple times with each projected image. 00:05:10.120 --> 00:05:16.750 Although the frame rate is still 24 frames per second, this projector displays each frame three times 00:05:16.750 --> 00:05:23.509 meaning that the image actually flashes 72 times per second, which is too fast for most people to notice. 00:05:23.509 --> 00:05:26.713 Since the film can only move when the shutter is blocking the light, 00:05:26.713 --> 00:05:29.597 and that happens three times per frame, 00:05:29.597 --> 00:05:36.111 this means that the film advance movement happens in one half of one third of 1/24 of one second - 00:05:36.111 --> 00:05:38.881 or 1/144th of a second. 00:05:38.881 --> 00:05:41.388 Or one grossth of a second. 00:05:41.388 --> 00:05:45.372 That means the film advance mechanism is yanking the film down… 00:05:45.372 --> 00:05:47.349 real fast. 00:05:47.349 --> 00:05:52.610 Now, when you actually do the math it works out to about 2 and a half miles an hour so it’s not… 00:05:52.610 --> 00:05:56.758 that fast but still, that’s pretty stressful on the film. 00:05:56.758 --> 00:06:06.176 And to help with that, the film follows an elaborate path through a pair of rotating sprockets that are, again, mechanically linked to the film advance and shutter. 00:06:06.176 --> 00:06:08.663 Threading this thing is… a pain 00:06:08.663 --> 00:06:11.088 and involves slipping the film over this sprocket, 00:06:11.088 --> 00:06:12.414 down through the film gate, 00:06:12.414 --> 00:06:14.346 then up over this second sprocket, 00:06:14.346 --> 00:06:17.149 around this mysterious thing, 00:06:17.149 --> 00:06:23.346 then back under that second sprocket, before finally taking a spin around these roller guides on its way to the take-up reel. 00:06:23.346 --> 00:06:27.782 After it’s all threaded you close in these little film guides and pray. 00:06:27.782 --> 00:06:30.426 I’ve only messed this up several times! 00:06:30.426 --> 00:06:36.487 Anyway, the sprockets advance the film smoothly and you’ll notice that there’s an area above and below the gate 00:06:36.487 --> 00:06:39.267 where the film just sorta hangs loose. 00:06:39.267 --> 00:06:46.956 These are there to provide some slack before and after the gate in order to reduce the stresses caused by the jerky film advance. 00:06:46.956 --> 00:06:49.876 If you forget to leave this slack there when threading, 00:06:49.876 --> 00:06:53.017 congratulations you’ve just chewed up some film! 00:06:53.017 --> 00:07:00.663 But when threaded correctly and operating, we’re left with a relatively straightforward (if mechanically elaborate) device. 00:07:00.663 --> 00:07:09.060 The first sprocket steadily pulls film from the supply reel, maintaining a few frames of slack as the film turns downward to pass through the gate. 00:07:09.060 --> 00:07:16.647 There it is repeatedly advanced one frame at a time, and between movements the shutter opens to project the image on the film. 00:07:16.647 --> 00:07:18.870 After it exits the gate at the bottom, 00:07:18.870 --> 00:07:23.597 another short length of slack is maintained as it turns again into a second sprocket which, 00:07:23.597 --> 00:07:29.126 moving at the same exact speed as the first, pulls the film steadily away from the gate. 00:07:29.126 --> 00:07:36.280 Looking at the transition space between the herky-jerky gate and the smoothly-turning sprockets is fascinating. 00:07:36.280 --> 00:07:40.398 It comes across better in-person, in fact it’s almost freaky in-person, 00:07:40.398 --> 00:07:45.322 but you’ll notice that near the sprocket (and in fact anywhere else along the film path) 00:07:45.322 --> 00:07:47.236 the film is just a blur. 00:07:47.236 --> 00:07:51.872 It’s moving smoothly so you can’t focus on any of the images it contains. 00:07:51.872 --> 00:07:57.592 As it gets closer to the gate, though, its periodic stillness allows you to pick out film frames, 00:07:57.592 --> 00:08:02.613 and right near the gate you can actually see motion in the imagery on the film. 00:08:02.613 --> 00:08:06.826 It appears upside-down because the lens flips it during projection. 00:08:06.826 --> 00:08:13.742 You can see the same thing as it exits the gate, but it becomes a blur again as it approaches the second sprocket. 00:08:13.742 --> 00:08:17.961 Now, if this were a silent projector this would be the end of its job - 00:08:17.961 --> 00:08:21.751 the film could simply go straight to a take-up reel after this point. 00:08:21.751 --> 00:08:24.991 But this ain’t no silent projector, now is it? 00:08:24.991 --> 00:08:25.788 No it’s not. 00:08:25.788 --> 00:08:29.044 But before I explain the sound this thing reproduces, 00:08:29.044 --> 00:08:33.459 I think it’s important that we go over how flicks became talkies. 00:08:33.459 --> 00:08:40.599 The technologies of sound reproduction and the motion picture were both fairly well understood by the turn of the 20th century. 00:08:40.599 --> 00:08:44.305 We knew how to capture and reproduce sound with phonograph records, 00:08:44.305 --> 00:08:48.410 and we knew how to make movies to be projected in a theater. 00:08:48.410 --> 00:08:52.101 It seemed pretty obvious that these two technologies should be connected, 00:08:52.101 --> 00:08:55.110 but there were two big problems. 00:08:55.110 --> 00:08:58.694 The first was that early acoustic sound reproduction technology 00:08:58.694 --> 00:09:03.070 wasn’t suitable for the large audiences you’d find in a theater. 00:09:03.070 --> 00:09:07.661 The devices of the time simply weren’t loud enough to reach the people in the back. 00:09:07.661 --> 00:09:12.999 But the second and arguably bigger problem was sound synchronization. 00:09:12.999 --> 00:09:14.657 That would prove tricky. 00:09:14.657 --> 00:09:19.959 Recording sound and filming action at the same time wasn’t a difficult concept. 00:09:19.959 --> 00:09:22.909 In fact, people were trying to make talking movies happen 00:09:22.909 --> 00:09:30.557 for about as long as movies and talking machines coexisted, but satisfactory results were stubbornly elusive. 00:09:30.557 --> 00:09:35.125 Every time you were finished shooting you were left with two separate recordings: 00:09:35.125 --> 00:09:37.789 one visual, one sound. 00:09:37.789 --> 00:09:41.085 Playing them back together wasn’t exactly rocket surgery, 00:09:41.085 --> 00:09:46.573 but for a convincing experience, they needed to be reproduced at the exact same time 00:09:46.573 --> 00:09:54.036 and crucially neither one could drift from the other so the playback speeds had to be exact. 00:09:54.036 --> 00:09:57.649 Otherwise, y’know, Singin’ in the Rain, and all that. 00:09:57.649 --> 00:10:02.324 By the 1920’s electronic sound amplification was starting to be a thing, 00:10:02.324 --> 00:10:04.319 and that solved the volume problem. 00:10:04.319 --> 00:10:07.790 But we still had the synchronization problem. 00:10:07.790 --> 00:10:12.214 Some early sound films did use phonograph records as their sound source, 00:10:12.214 --> 00:10:17.668 including the technically first feature-length talkie The Jazz Singer. 00:10:17.668 --> 00:10:23.493 These systems involved mechanical interlocks between projector and phonograph to keep them in sync, 00:10:23.493 --> 00:10:30.173 but it required projectionists to carefully align each device with cue marks at the start of every reel 00:10:30.173 --> 00:10:36.130 and was susceptible to a bumped record player completely ruining the moviegoing experience. 00:10:36.130 --> 00:10:40.016 But what if the sound could be stored on the film itself? 00:10:40.016 --> 00:10:46.012 Putting both visuals and audio onto the same physical object would guarantee they’d stay in sync, 00:10:46.012 --> 00:10:47.955 but how could we do that? 00:10:47.955 --> 00:10:51.957 Magnetic tape won’t be invented until 1928 so that’s not gonna work, 00:10:51.957 --> 00:10:54.880 and besides that would be complicated to manufacture. 00:10:54.880 --> 00:11:00.471 You’d have to, like, bond this theoretical tape stuff to the film somehow. 00:11:00.471 --> 00:11:05.825 Now film works by blocking light in specific parts to make an image - 00:11:05.825 --> 00:11:09.149 what if we used that concept to store sound? 00:11:09.149 --> 00:11:12.421 Great idea, several people in 1919. 00:11:12.421 --> 00:11:16.413 In a classic case of contrived collaboration, Lee de Forest - 00:11:16.413 --> 00:11:21.824 who you may recognize as an electronics pioneer who invented the first amplifying vacuum tube - 00:11:21.824 --> 00:11:28.695 would work with Theodore Case and Earl I Sponable to invent an optical sound-on-film process, 00:11:28.695 --> 00:11:32.229 building on earlier work in optical sound transmission. 00:11:32.229 --> 00:11:36.605 See, we had already learned that we could modulate light to carry sound - 00:11:36.605 --> 00:11:42.751 in fact Alexander Graham Bell, the telephone guy, had demonstrated this in 1880. 00:11:42.751 --> 00:11:47.267 That concept was then used by the navy.... and fast forward to 1919 00:11:47.267 --> 00:11:54.539 and de Forest is granted a patent which uses the photographic properties of motion picture film to carry a sound signal. 00:11:54.539 --> 00:11:59.497 You might have noticed that 1919 was before 1927. 00:11:59.497 --> 00:12:07.024 Yes, in an odd twist, sound-on-film predates synchronized sound-on-disc by… 00:12:07.024 --> 00:12:08.969 well it kinda does and kinda doesn’t. 00:12:08.969 --> 00:12:12.642 Hollywood for whatever reason was much more enamored with discs, 00:12:12.642 --> 00:12:16.989 that is once they were finally convinced talkies might just be more than a fad, 00:12:16.989 --> 00:12:25.998 and that’s why the Jazz Singer used the Vitaphone system, but the Phonofilm system was in use in 1923… 00:12:25.998 --> 00:12:27.927 just not in Hollywood. 00:12:27.927 --> 00:12:32.736 Anyway, glossing over a ton of history, let’s get on with it. 00:12:32.736 --> 00:12:38.576 This projector uses essentially the same sound-on-film technology that de Forest patented. 00:12:38.576 --> 00:12:48.709 In fact, from the 1930’s and into the 1990’s sound technology in movies shown in theaters remained largely unchanged. 00:12:48.709 --> 00:12:53.088 A long, long time ago I showed you what the film looks like. 00:12:53.088 --> 00:12:56.375 Along the edge opposite the sprocket holes is the soundtrack 00:12:56.375 --> 00:13:02.148 and its resemblance to the waveforms you might see in modern sound recording software is no coincidence. 00:13:02.148 --> 00:13:05.520 This is a physical representation of an audio signal, 00:13:05.520 --> 00:13:11.490 and thanks to the fact that the film moves, we can reproduce it with elegant simplicity. 00:13:11.490 --> 00:13:18.261 You might have thought that this projector has but one lamp: this nasty little 1,000 watt light bulb 00:13:18.261 --> 00:13:23.448 that will light up the screen and also burn a nice little hole in the film should it get stuck. 00:13:23.448 --> 00:13:26.363 But when you switch on the audio amplifier, 00:13:26.363 --> 00:13:30.000 you’ll discover that there is in fact a second one. 00:13:30.511 --> 00:13:34.504 That one’s not important though, it’s just to help you see a little bit in a dark projection booth. 00:13:34.504 --> 00:13:40.540 Wait a minute though and you will find a third lamp lighting up this little red dot. 00:13:40.540 --> 00:13:42.130 This is the exciting one. 00:13:42.130 --> 00:13:44.228 This lamp is called the exciter 00:13:44.228 --> 00:13:48.624 and its job is to shine a light through the film’s soundtrack. 00:13:48.624 --> 00:13:53.806 That mysterious thing I mentioned during the threading process is the sound section, 00:13:53.806 --> 00:14:00.935 and when threaded the soundtrack on the film lies between the exciter lamp and a light-sensitive doohickey of some kind 00:14:00.935 --> 00:14:05.199 that will generate a voltage or maybe current proportional to how much light hits it. 00:14:05.199 --> 00:14:08.400 I’d be more specific but the technologies changed a bunch over the years 00:14:08.400 --> 00:14:14.000 and I couldn’t tell you if this is a photodiode or a selenium light cell or a photoresistor or what. 00:14:14.000 --> 00:14:17.377 It doesn’t really matter, though, the specific specifics aren’t important, 00:14:17.377 --> 00:14:22.259 what is important is that the light sensor, when it picks up changes in brightness, 00:14:22.259 --> 00:14:25.830 will generate a signal which we can amplify. 00:14:25.830 --> 00:14:30.759 The light sensor lives here, tucked behind this large silver drum thing. 00:14:30.759 --> 00:14:33.818 The large silver drum thing, called the sound drum, 00:14:33.818 --> 00:14:40.029 serves a similar function to the capstan in a tape recorder, and the film is threaded around it. 00:14:40.029 --> 00:14:43.244 It’s a little different as it’s not powered, 00:14:43.244 --> 00:14:48.860 but inside the projector we find a large flywheel that the drum attaches to. 00:14:48.860 --> 00:14:55.860 Like a tape recorder’s capstan, the rotating mass of the flywheel smooths out any small variances in film speed that would be audible as flutter - 00:14:55.860 --> 00:15:01.486 and since everything in here is driven by mechanically-locked gear trains, 00:15:01.486 --> 00:15:09.063 while the motion of the sprockets might look smooth, they’re actually much too jittery for pleasant sound reproduction. 00:15:09.063 --> 00:15:16.570 When threaded properly, these spring-loaded rollers put tension on the film to keep it snugly in contact with the drum, 00:15:16.570 --> 00:15:21.231 and once the drum is up to speed they’ll just sort of float up and down as needed. 00:15:21.231 --> 00:15:28.924 In essence, the floating rollers constantly absorb the tiny bits of slack introduced by the slightly jittery turning of the sprocket, 00:15:28.924 --> 00:15:35.709 and the inertia of the drum keeps the film moving as smoothly as possible through the sound section. 00:15:35.709 --> 00:15:40.147 This is by far the most fiddly part of the threading process on this projector 00:15:40.147 --> 00:15:42.750 and I hate it! 00:15:42.750 --> 00:15:47.719 You need there to be enough tension on the film for those floating rollers to actually float, 00:15:47.719 --> 00:15:54.937 which means you kinda have to slip the film in the top section of the sprocket, lock in the guard, bring it around the drum, 00:15:54.937 --> 00:16:02.200 and then keep constant tension on it as you slip it under the bottom half of the sprocket and the film engages with the teeth. 00:16:02.200 --> 00:16:05.561 You only have about one film frame of tolerance, here, 00:16:05.561 --> 00:16:12.411 and if there’s enough slack that those rollers don’t ride on the film the sound will either sound muffled and warbly 00:16:12.411 --> 00:16:14.846 or just won’t come through at all. 00:16:14.846 --> 00:16:18.071 Now, to be clear, not every projector works like this one does - 00:16:18.071 --> 00:16:25.339 this is an older and relatively simple projector, in fact it’s old enough to have a vacuum tube amplifier. 00:16:25.339 --> 00:16:31.519 Newer and fancier units featured automatic threading or just a generally less annoying design. 00:16:31.519 --> 00:16:35.887 Anyway, now that we have a way to make the film move smooth as butter, 00:16:35.887 --> 00:16:37.786 we’re done, right? 00:16:37.786 --> 00:16:39.520 Well, not quite. 00:16:39.520 --> 00:16:45.650 Simply shining light through the soundtrack section and onto a light sensor won’t work well. 00:16:45.650 --> 00:16:51.520 The sound signal on the film is, in this case, modulated using the variable-area technique. 00:16:51.520 --> 00:16:55.603 The width of the clear sections corresponds to the signal amplitude - 00:16:55.603 --> 00:17:01.698 more clear area means more light will pass through the film and thus we get a stronger signal from the sensor. 00:17:01.698 --> 00:17:05.153 And of course that varies along the length of the film - 00:17:05.153 --> 00:17:11.266 combine the changing width and linear movement of the film and you’ve got an amplitude that changes with time - 00:17:11.266 --> 00:17:14.328 just like sound pressures do in the real world. 00:17:14.328 --> 00:17:17.975 But in order to reproduce that signal with any degree of fidelity, 00:17:17.975 --> 00:17:23.592 we need the sensor to be able to discern rapid changes in amplitude like you see here. 00:17:23.592 --> 00:17:32.164 To do this, the exciter lamp shines through a slot and lens which projects only a tiny slit of light onto the film. 00:17:32.164 --> 00:17:41.660 That literally narrows the area being scanned which makes the signal we receive more precise and allows us to recreate the nuances of the waveform. 00:17:41.660 --> 00:17:45.031 And the fidelity that’s possible here is frankly not bad, 00:17:45.031 --> 00:17:51.798 especially considering that this is 16mm film which only moves at a hair over seven inches per second. 00:17:51.798 --> 00:17:55.130 Here are some samples taken from this projector. 00:17:55.812 --> 00:18:00.904 [Narrator] Printing is a process of transferring an image from one surface to another. 00:18:02.914 --> 00:18:04.182 [sad bagpipe toot] 00:18:04.792 --> 00:18:10.042 [normal bagpipes] 00:18:11.172 --> 00:18:19.434 Note that 35mm film, which is what you’d be seeing in a commercial theater, moves at a much faster 18 inches per second. 00:18:19.434 --> 00:18:24.156 With more than twice the speed, high-frequency sounds are reproduced much more accurately, 00:18:24.156 --> 00:18:26.808 and the fidelity is greatly improved. 00:18:26.808 --> 00:18:33.820 Still, though, the sound on 16mm is perfectly acceptable - and with a fancier projector, I imagine this would sound better. 00:18:34.896 --> 00:18:41.100 [Narrator with acoustic guitar in background] For right handers: feet apart, left foot forward. 00:18:41.100 --> 00:18:45.710 Bring the ball in front of your chest, then above your shoulder. 00:18:46.643 --> 00:18:48.658 Steady the ball with your left hand. 00:18:49.519 --> 00:18:51.549 Hold the ball with your right. 00:18:52.662 --> 00:18:56.238 Shift your weight to your left foot as you throw. 00:18:57.565 --> 00:19:00.150 [70’s flutes in background] There’s nothing there I really don’t understand. 00:19:00.150 --> 00:19:01.210 [Narrator] Excellent. 00:19:01.210 --> 00:19:04.530 Then perhaps you can tell us what this code does. 00:19:04.530 --> 00:19:05.530 How should I know? 00:19:05.530 --> 00:19:11.270 Sure, I can read the language… but how can I understand unless you tell me what the names stand for? 00:19:12.024 --> 00:19:16.919 [more very groovy mid-seventies instructional flute-music] 00:19:18.390 --> 00:19:20.403 Something doesn’t add up. 00:19:20.403 --> 00:19:24.230 And that’s exactly what Crookes thought when he invented this device. 00:19:25.091 --> 00:19:27.240 Now, nature isn’t wrong. 00:19:27.240 --> 00:19:29.295 The theory’s wrong. 00:19:29.295 --> 00:19:32.477 Instead of demonstrating the pressure of light, 00:19:32.477 --> 00:19:35.301 he must have been demonstrating something else. 00:19:35.301 --> 00:19:39.650 [Narrator] The correct position for the head is slightly tilted back. 00:19:39.650 --> 00:19:41.725 When the arms are in their correct position, 00:19:41.725 --> 00:19:48.330 a pole can be placed on the bridge of the nose and the ends of it will rest between the thumb and forefinger of each hand. 00:19:48.330 --> 00:19:54.370 All the windows in our house have wooden doors outside. 00:19:57.635 --> 00:20:02.610 They must be opened every morning, and that’s my job. 00:20:02.610 --> 00:20:05.228 [Narrator] Communicore, with its central location, 00:20:05.228 --> 00:20:07.190 is the hub of Future World. 00:20:07.190 --> 00:20:10.678 In a sense, it’s also the heart of EPCOT Center, 00:20:10.678 --> 00:20:15.033 for here are the computers that control many of the attractions. 00:20:15.822 --> 00:20:19.572 [Narrator] Sometimes he went stalking with his Highland uncles. 00:20:20.288 --> 00:20:22.619 [birds chirping] 00:20:23.625 --> 00:20:29.647 Once, after staying still for a long time, Wee Gillis couldn’t hold his breath 00:20:29.647 --> 00:20:31.123 [exhales] 00:20:33.137 --> 00:20:35.151 [clopping hooves] 00:20:35.900 --> 00:20:38.418 and he frightened away the deer. 00:20:40.690 --> 00:20:46.144 [more incredibly ‘70s flute music, but now with harpsichord!] 00:20:52.854 --> 00:20:55.159 [Narrator] It’s much like the Golden Rule. 00:20:55.751 --> 00:20:59.009 If you read programs critically when you get a chance, 00:20:59.009 --> 00:21:01.184 apply some of the ideas we’ve covered, 00:21:01.184 --> 00:21:03.699 and use tact with personal criticisms, 00:21:03.699 --> 00:21:06.784 you and your colleagues will be a lot more successful 00:21:06.784 --> 00:21:09.336 and so will your programs. 00:21:10.125 --> 00:21:13.326 There are some interesting things to point out about the sound, here. 00:21:13.326 --> 00:21:19.674 Of course, that buzzing is mainly the result of an old amplifier that probably could do with new capacitors. 00:21:19.674 --> 00:21:24.165 But did you hear a sort of metallic sound throughout the recordings? 00:21:24.165 --> 00:21:29.315 One of the downsides of analog optical audio is that it’s susceptible to vibrations - 00:21:29.315 --> 00:21:35.341 if the film physically moves even just a tiny bit in a way that affects how much light hits the sensor, 00:21:35.341 --> 00:21:37.708 you’ll hear that movement. 00:21:37.708 --> 00:21:41.960 And the sound drum on this machine doesn’t spin silently. 00:21:41.960 --> 00:21:45.351 If I give it a quick spin we hear a similar sound. 00:21:45.351 --> 00:21:51.220 I believe that is getting transferred into the signal, though it could also be one of the rollers. 00:21:51.220 --> 00:21:53.418 In any case, it’s just interesting. 00:21:53.418 --> 00:21:57.090 Something else to point out is what signal noise sounds like. 00:21:57.090 --> 00:22:03.245 For instance, when the translucent leaders at the head and tail of a film travel through the sound section, 00:22:03.245 --> 00:22:10.536 what we hear is a scratchy sound not unlike a phonograph needle sliding on the surface of a record before it locks in the groove. 00:22:10.901 --> 00:22:14.702 [buzz as amp switches on, then scratchy surface noise] 00:22:14.702 --> 00:22:17.610 [clicking and popping] 00:22:17.610 --> 00:22:22.264 Little bits of dust or scratches on the soundtrack will have a similar effect. 00:22:22.264 --> 00:22:26.854 And since even a silent section will have tiny little variations in it, 00:22:26.854 --> 00:22:30.574 there is a noise floor just like you find on vinyl records. 00:22:30.574 --> 00:22:37.784 Anyway, it’s just interesting that the noise is much more like a phonograph and much less like magnetic tape. 00:22:37.784 --> 00:22:42.028 You may have already guessed that since the sound section is here 00:22:42.028 --> 00:22:46.197 but the projected frame is all the way back there, 00:22:46.197 --> 00:22:49.808 that the sound is recorded onto the film ahead of the images, 00:22:49.808 --> 00:22:52.465 in fact by 26 frames. 00:22:52.465 --> 00:23:00.310 At the beginning of many films there will be a brief beep on the soundtrack 26 frames ahead of a marker which will flash on the screen, 00:23:00.310 --> 00:23:02.942 and that helps assure that the sound is in sync. 00:23:02.942 --> 00:23:07.386 Threading mishaps can cause film to bunch up in unexpected ways 00:23:07.386 --> 00:23:10.191 so it’s nice to have that confirmation. 00:23:10.191 --> 00:23:16.280 However, the sync can easily be off by a frame or two thanks to the slack section after the gate. 00:23:16.280 --> 00:23:21.960 It’s easy to vary how much film is left floating here, and that will affect sound sync. 00:23:21.960 --> 00:23:26.341 Luckily, if it’s off by just a couple of frames most people won’t notice. 00:23:26.341 --> 00:23:29.306 Like how it was slightly off in this entire shot. 00:23:29.751 --> 00:23:32.832 Another thing to point out about these soundtracks is that, 00:23:32.832 --> 00:23:38.253 while you might assume these are in stereo thanks to the two waveforms right next to each other, 00:23:38.253 --> 00:23:40.920 this is in fact just a mono recording. 00:23:40.920 --> 00:23:44.059 The two parallel waveforms are identical. 00:23:44.059 --> 00:23:47.300 So, why bother having two of them? 00:23:47.300 --> 00:23:49.511 Is there some magic to be had? 00:23:49.511 --> 00:23:50.820 Yes, actually! 00:23:50.820 --> 00:23:56.679 This, like most of the films I have, is a dual-bilateral variable-area soundtrack. 00:23:56.679 --> 00:23:58.760 Say that five times fast. 00:23:58.760 --> 00:24:03.204 Two parallel soundtracks help reduce what’s known as azimuth distortion - 00:24:03.204 --> 00:24:09.450 which is basically how well or poorly centered the scanning beam is on the soundtrack. 00:24:09.450 --> 00:24:14.477 With a single waveform running in the center, a slightly misaligned scanning section 00:24:14.477 --> 00:24:19.449 (which could happen for a number of reasons including a slightly out-of-tolerance exciter lamp) 00:24:19.449 --> 00:24:25.735 might miss the edges of the waveform and thus that sound information wouldn’t get through. 00:24:25.735 --> 00:24:32.042 Having two identical soundtracks right next to each other means that if the scanning beam is slightly off, 00:24:32.042 --> 00:24:38.687 the signal is less likely to be damaged as at least one of the two waveforms will be completely captured. 00:24:38.687 --> 00:24:42.501 Some manufacturers would go on to take this concept to the extreme 00:24:42.501 --> 00:24:47.213 and produce soundtracks with four or even six identical waveforms. 00:24:47.213 --> 00:24:48.415 Did it help? 00:24:49.097 --> 00:24:50.097 Maybe. 00:24:50.671 --> 00:24:56.929 OK, so we understand how the projector makes sounds from those wiggly waveforms on the film, 00:24:57.324 --> 00:25:00.391 but how did they get there in the first place? 00:25:00.391 --> 00:25:04.123 Well, have you ever seen one of those laser light show thingies? 00:25:04.123 --> 00:25:06.670 Think that, but itty bitty. 00:25:06.670 --> 00:25:14.274 A device called a mirror galvanometer was used to bounce a tiny spot of light onto a strip of photographic film. 00:25:14.274 --> 00:25:21.059 That mirror, when fed an audio signal, would vibrate and wiggle that spot of light left and right. 00:25:21.059 --> 00:25:27.770 The stronger the signal, the more it vibrated, and thus the wider a line it would appear to make. 00:25:27.770 --> 00:25:32.560 Without a signal it wouldn’t vibrate at all and it would remain a spot. 00:25:32.560 --> 00:25:39.230 Running photographic film past this wiggly light would produce waveforms like this once developed with a reversal process. 00:25:39.918 --> 00:25:44.877 I’d love to know more about the specifics of how films like this were mass-produced. 00:25:44.877 --> 00:25:51.670 For instance, most of the color films I have are quite severely faded, but the soundtracks are not. 00:25:51.670 --> 00:25:57.099 Indeed, the soundtracks were made using conventional black-and-white, silver-halide film chemistry 00:25:57.099 --> 00:26:02.485 and then bonded somehow to the dye-based color film print. 00:26:02.485 --> 00:26:06.137 However that was done, the soundtrack must be very very thin 00:26:06.137 --> 00:26:09.470 as you can’t feel a ridge of any kind. 00:26:09.470 --> 00:26:16.150 Or perhaps the film was made with two emulsions on it, but given what I now know about how film is manufactured 00:26:16.150 --> 00:26:17.727 (thanks Destin) 00:26:17.727 --> 00:26:19.605 that seems unlikely. 00:26:19.605 --> 00:26:22.723 Anyway, I don’t want to dwell on this but I am very curious. 00:26:22.723 --> 00:26:29.019 Oh, and while all of my films have variable-area soundtracks, that wasn’t the only way to do it. 00:26:29.019 --> 00:26:34.546 In fact, de Forest’s original Phonofilms used a variable-density scheme. 00:26:34.546 --> 00:26:38.733 Here, rather than wiggle a constant light to make different shapes, 00:26:38.733 --> 00:26:42.544 we change the brightness of a light source projected through a slit. 00:26:42.544 --> 00:26:52.270 In essence it’s the same exact concept as we see in the projector, but with a fast-acting light source exposing the film to record sound onto it. 00:26:52.270 --> 00:26:57.687 That makes a soundtrack in which the recorded sound looks more like banding than a waveform 00:26:57.687 --> 00:27:01.405 but it will block light in much the same way when being reproduced 00:27:01.405 --> 00:27:06.373 so it is entirely compatible with this or any other optical sound projector. 00:27:06.373 --> 00:27:09.731 And, although much rarer in the grand scheme, 00:27:09.731 --> 00:27:15.718 there are films in which the audio section is indeed a strip of magnetic tape bonded to it. 00:27:15.718 --> 00:27:19.742 They probably offered higher fidelity especially on 16mm film, 00:27:19.742 --> 00:27:22.919 but they required a compatible projector. 00:27:22.919 --> 00:27:26.994 Speaking of projectors, before we get to the part about Digital Sound On Film - 00:27:26.994 --> 00:27:29.106 that’s right, that’s a thing - 00:27:29.106 --> 00:27:33.648 I want to talk a little bit more about the lost art of projection. 00:27:33.648 --> 00:27:39.819 You might be wondering how long of a film the projector can show before it runs out of film. 00:27:39.819 --> 00:27:48.006 Well, a big 15 inch reel like the one I’m using as a takeup-reel can hold 2200 feet of film which, side-note, 00:27:48.006 --> 00:27:51.240 did you know that’s where the term “footage” comes from? 00:27:51.240 --> 00:27:54.024 As in, “where’s that sasquatch footage?” 00:27:54.024 --> 00:27:56.907 Motion picture film is measured in feet! 00:27:56.907 --> 00:27:59.519 And some quantity of it is footage! 00:27:59.519 --> 00:28:02.354 There are even little gauge marks on the reels. 00:28:02.354 --> 00:28:06.787 Anywho, since 16mm film has relatively small frames, 00:28:06.787 --> 00:28:12.451 at 24 frames per second a full 2200 foot reel will run for an hour. 00:28:12.451 --> 00:28:20.804 That’s a pretty uncommon size, though, and the 1600 foot reel, which ran for 44 minutes, was commonly used for feature films 00:28:20.804 --> 00:28:26.575 with 800 foot reels like these common for short-subjects or educational films. 00:28:26.575 --> 00:28:30.540 This means that feature films came on multiple reels. 00:28:30.540 --> 00:28:39.188 And on 35mm film with its much larger frames, a 90 minute movie might be split up onto five 2000 foot reels 00:28:39.188 --> 00:28:42.188 which held a maximum of 22 minutes runtime. 00:28:42.188 --> 00:28:49.409 And for many years 1000 foot reels were standard, meaning the same film would need to be on 9 reels. 00:28:49.409 --> 00:28:55.110 So… how did movie-going audiences get to watch a seamless performance? 00:28:55.469 --> 00:28:59.364 Well, have you ever watched a home-video release of an old movie 00:28:59.364 --> 00:29:03.610 and caught one of these little dots showing up in the corner? 00:29:03.610 --> 00:29:08.170 If you’re too young you probably haven’t as these are generally edited out these days, 00:29:08.170 --> 00:29:10.901 but those are cue marks. 00:29:10.901 --> 00:29:16.299 See, although a projector in a cinema could only show 11 or 22 minutes of film at a time, 00:29:16.299 --> 00:29:18.591 through the Magic of Having Two Of Them, 00:29:18.591 --> 00:29:25.689 you could have another projector standing by ready to go and switch to it the moment the first one runs out of film. 00:29:25.689 --> 00:29:31.030 Those cue marks appear at 8 seconds and 1 second before the end of a reel. 00:29:31.030 --> 00:29:34.410 The first signals to the projectionist to get ready to switch, 00:29:34.410 --> 00:29:37.029 and the second says “now’s the time”. 00:29:37.029 --> 00:29:39.245 This used to be entirely manual, 00:29:39.245 --> 00:29:46.837 and projectionists would have to quickly cover the first projector’s lens with a shutter and open up the second which was rarely seamless, 00:29:46.837 --> 00:29:52.296 but eventually this process was semi-automated and merely required the push of a button. 00:29:52.296 --> 00:29:57.886 As you can imagine, back in the day a projectionist’s job was pretty hands-on. 00:29:57.886 --> 00:30:01.492 They’d have to constantly be getting the next projector ready - 00:30:01.492 --> 00:30:09.063 as soon as they hit the changeover, they had to immediately unload the first projector, then start threading the next reel and cue it up. 00:30:09.063 --> 00:30:13.442 Then they could rewind that first reel (assuming they had a separate rewinder) 00:30:13.442 --> 00:30:18.355 and after that they’d be on the lookout for the cue marks to do it all over again. 00:30:18.355 --> 00:30:22.970 These folks were really watching the same movies again and again and again 00:30:22.970 --> 00:30:29.611 which was undoubtedly tedious but also meant they could get really good at timing the changeovers. 00:30:29.611 --> 00:30:32.771 If they were good enough, the audience wouldn’t even notice. 00:30:32.771 --> 00:30:37.091 Unless they were looking back at the projection booth like some kinda nerd. 00:30:37.091 --> 00:30:38.606 Like me. 00:30:38.606 --> 00:30:47.897 Eventually, though, most theaters would upgrade to large platter systems which could store an entire spliced-together feature plus trailers. 00:30:47.897 --> 00:30:50.694 And those could be loaded from the center, 00:30:50.694 --> 00:30:56.750 meaning the film just shuttled between two platters and didn’t ever need rewinding. 00:30:56.750 --> 00:31:00.508 This meant a projectionist only had to start a film 00:31:00.508 --> 00:31:05.626 and could wander around to other screens in a multiplex and just keep tabs on them. 00:31:05.626 --> 00:31:12.299 Plan some sort of schedule for screen times and you might only need a single person to manage a dozen screens. 00:31:13.053 --> 00:31:15.709 That’s what they call progress, I guess. 00:31:15.709 --> 00:31:23.230 Speaking or progress, starting in the ‘80s people got a taste of sound made not from shapes, grooves, or magnetic flux 00:31:23.230 --> 00:31:25.707 but from ones and zeroes. 00:31:25.707 --> 00:31:31.244 They also liked the idea of sound coming from many different places in a theater. 00:31:31.244 --> 00:31:36.420 The advent of digital surround sound meant that sound-on-film was gonna need an update. 00:31:36.420 --> 00:31:39.228 And it looked like this! 00:31:39.228 --> 00:31:41.867 This is a 35mm film trailer, 00:31:41.867 --> 00:31:46.497 and you probably recognize our old friend dual-bilateral variable-area! 00:31:46.497 --> 00:31:49.230 Though in this case, it actually is stereo - 00:31:49.230 --> 00:31:55.905 professional projection systems used much more precise sound heads to correctly scan both waveforms separately. 00:31:55.905 --> 00:31:59.129 By the 1970’s, stereo sound was pretty standard 00:31:59.129 --> 00:32:08.898 and Dolby Stereo - a way to derive Left, Right, Center, and Surround channels from a cleverly-modulated stereo source - offered basic surround sound using this technology 00:32:08.898 --> 00:32:11.461 starting in 1976. 00:32:11.461 --> 00:32:14.822 But anyway, that’s not the only soundtrack on here. 00:32:14.822 --> 00:32:16.472 In fact it has not one, 00:32:16.472 --> 00:32:17.549 not two, 00:32:17.549 --> 00:32:18.696 not three, 00:32:18.696 --> 00:32:21.428 but four different soundtracks! 00:32:21.428 --> 00:32:28.159 To the left of the analog soundtrack we find Dolby Digital sound hanging out in the middle of the sprocket holes. 00:32:28.159 --> 00:32:35.966 First used in 1992, these QR-lookin’ matrix barcodes, complete with an adorable little Dolby logo in the middle, 00:32:35.966 --> 00:32:41.132 encode a Dolby AC-3 bitstream at 320 kilobits per second. 00:32:41.132 --> 00:32:48.636 There’s some error correction built in, of course - by my rough count I think the raw data rate is closer to 530 kilobits a second 00:32:48.636 --> 00:32:58.524 but anyway to upgrade your cinema a new digital sound head that scans this area using what you can describe as basically a tiny high-speed camera… 00:32:58.524 --> 00:33:01.287 (sigh) that’s gonna ruffle some feathers. 00:33:01.287 --> 00:33:08.513 There’s a scanner thingy that uses the same tech that’s in a camera sensor to read the datastream with the help of backlight, let’s just leave it at that - 00:33:08.513 --> 00:33:15.468 that thing gets installed somewhere on your projector’s film path and now you’ve got digital surround sound! 00:33:15.468 --> 00:33:22.449 And a fun thing about being digital is that you no longer needed to put the sound head in exactly the correct place. 00:33:22.449 --> 00:33:29.309 Just like 16mm film, oddly enough, the sound data is 26 frames ahead of the projected image. 00:33:29.309 --> 00:33:34.440 But you could easily put the sound head a little bit further ahead and simply buffer the data 00:33:34.440 --> 00:33:39.049 to delay the sound by however many fractions of a second you need to attain synchronization. 00:33:39.659 --> 00:33:40.840 That’s pretty neat. 00:33:40.840 --> 00:33:46.808 Oh, and the fact that the physical location of the sound is well ahead of the projected image 00:33:46.808 --> 00:33:51.320 gives the MPAA preview approval cards a second purpose: 00:33:51.320 --> 00:33:58.394 at every film splice, there’s gonna be a little over a second of silence in the audio before the picture catches up, 00:33:58.394 --> 00:34:03.217 and since multiple trailers would get spliced together onto the start of a main feature, 00:34:03.217 --> 00:34:06.191 there will always be a series of gaps. 00:34:06.191 --> 00:34:08.432 Might as well use those for this thing. 00:34:08.432 --> 00:34:11.412 To be clear that also happened with analog audio 00:34:11.412 --> 00:34:14.760 but I only thought of it at this point in the scriptwriting process. 00:34:14.760 --> 00:34:19.983 Now, to the left of the Dolby Digital soundtrack you’ll find another very digital one. 00:34:19.983 --> 00:34:22.379 In fact this one is even more digital! 00:34:22.379 --> 00:34:24.530 As in, it has a higher bitrate. 00:34:24.530 --> 00:34:27.060 This is Sony Dynamic Digital Sound, 00:34:27.060 --> 00:34:30.171 also known as SDDS, also known as 00:34:30.171 --> 00:34:34.495 of course Sony made another sound format they just can’t help themselves! 00:34:34.495 --> 00:34:43.758 Hitting theaters in 1993, it’s scanned in much the same way as Dolby Digital and encodes 8-channel audio at 2.2 megabits per second, 00:34:43.758 --> 00:34:45.777 much greater than Dolby. 00:34:45.777 --> 00:34:51.279 However, very few movie studios embraced mixing in that particular format 00:34:51.279 --> 00:34:55.974 and since the data is placed on the edge of the film where it’s vulnerable to damage, 00:34:55.974 --> 00:35:04.876 reportedly this often suffered from dropouts and yeah it was the least popular format for cinemas when film projection was still the norm. 00:35:04.876 --> 00:35:06.646 Are you really surprised? 00:35:06.646 --> 00:35:09.091 But I said there’s a fourth format. 00:35:09.091 --> 00:35:10.688 Where’s that one? 00:35:10.688 --> 00:35:12.999 Why, it’s right here! 00:35:12.999 --> 00:35:21.280 This is Digital Theater Systems surround sound, also known as DTS, which also first hit the screen in 1993. 00:35:21.280 --> 00:35:26.960 You may notice that this is hardly any data at all compared to Dolby Digital and SDDS. 00:35:26.960 --> 00:35:30.670 Did they discover some sort of really magic compression format? 00:35:30.670 --> 00:35:36.086 No, as a matter of fact this is merely a timecode as DTS did a throwback 00:35:36.086 --> 00:35:38.209 to sound-on-disc. 00:35:38.209 --> 00:35:45.853 Special CD-ROMs containing a 5.1 channel mix encoded at 882 kbit/sec stored the sound, 00:35:45.853 --> 00:35:53.390 and this data stream simply identified what film or trailer was playing and the exact time of the projected frame. 00:35:53.390 --> 00:35:59.143 Offboard equipment with the loaded discs would simply monitor this code and cue up the correct audio. 00:35:59.143 --> 00:36:02.345 This made the scanning equipment much simpler 00:36:02.345 --> 00:36:09.787 but meant that the sound was technically not on the film anymore and of course required more off-projector hardware 00:36:09.787 --> 00:36:12.170 plus the wrangling of the discs. 00:36:12.170 --> 00:36:17.018 But anyway, one nice benefit of the DTS system from a producer’s point of view was that, 00:36:17.018 --> 00:36:21.930 like Dolby Digital, it was also a 5.1 channel system. 00:36:21.930 --> 00:36:25.221 Since it used ~essentially~ the same sound mix, 00:36:25.221 --> 00:36:33.950 a production studio could master in either one and get essentially the same experience between a DTS or Dolby-equipped cinema. 00:36:33.950 --> 00:36:37.140 Which was another nail in the coffin for SDDS. 00:36:37.140 --> 00:36:38.905 But don’t be blue, Sony. 00:36:38.905 --> 00:36:41.671 Eventually you’ll win one of those format wars. 00:36:41.671 --> 00:36:47.639 Oh, and while off-film discs might have made the projectionist’s job a bit more complicated, 00:36:47.639 --> 00:36:50.480 that technique was theoretically more future-proof 00:36:50.480 --> 00:36:56.011 as it could be easily upgraded to DVD-ROM which could store more channels with more bits, 00:36:56.011 --> 00:37:01.073 though I’m not sure this was ever done in theaters before film projection faded away. 00:37:01.073 --> 00:37:03.651 But at this point, I think we’re done. 00:37:03.651 --> 00:37:06.348 And that means it’s time to rewind the film! 00:37:06.348 --> 00:37:11.403 On this projector, that means depressing this latch to lift the take-up reel into this position, 00:37:11.403 --> 00:37:15.954 then grabbing the tail of the film and wrapping it around the supply reel a few times. 00:37:15.954 --> 00:37:18.802 After that’s done, set the projector to reverse. 00:37:18.802 --> 00:37:22.068 Oh, did I mention you can watch anything backwards? 00:37:22.068 --> 00:37:28.103 That’s fun and/or necessary sometimes because a piece of masking tape got caught in the gate somehow. 00:37:28.126 --> 00:37:30.083 .terr sall wus yeep ziT 00:37:30.083 --> 00:37:31.687 .ssip moes sey vaaawW 00:37:31.687 --> 00:37:32.905 .sez yeep ziP 00:37:32.905 --> 00:37:35.476 .swup g'nosh gelwozwod wilno wunnondazI 00:37:35.476 --> 00:37:36.928 .herri hunai pumeH 00:37:37.251 --> 00:37:44.609 Anyway, after it’s set to reverse you can hit this button which engages a clutch to make the supply reel go backwards real fast! 00:37:44.609 --> 00:37:46.508 Reel. Fast. 00:37:46.508 --> 00:37:50.898 Then at the end, you can watch it go fwip fwip fwip fwip for as long as you like. 00:37:50.898 --> 00:37:54.081 Though you should probably shut it off before too long. 00:37:54.081 --> 00:37:56.695 This stuff is fragile, y’know. 00:37:57.628 --> 00:38:00.098 ♫ cinematically smooth jazz ♫ 00:38:02.459 --> 00:38:03.501 [inhales] 00:38:06.041 --> 00:38:07.490 wheeeeee 00:38:11.259 --> 00:38:12.356 Eugh 00:38:12.864 --> 00:38:13.957 The 16... 00:38:13.957 --> 00:38:14.728 The - hmph. 00:38:14.728 --> 00:38:15.537 Heh heh. 00:38:15.537 --> 00:38:16.443 Heuh!! 00:38:16.443 --> 00:38:20.008 As a matter of fact it’s still used in some production environments. 00:38:20.008 --> 00:38:23.382 Sitting between the [film moves] that’s gonna be really loud, isnt it? 00:38:23.382 --> 00:38:25.990 Every frame of film is held stationary. 00:38:26.668 --> 00:38:28.234 [confusion] 00:38:28.869 --> 00:38:32.185 Stationary still... that’s what’s written there? 00:38:32.185 --> 00:38:33.794 “stationary, so.” 00:38:33.794 --> 00:38:35.625 There's just a redundant word. 00:38:35.625 --> 00:38:37.721 Thanks, spell check for not catching that. 00:38:37.721 --> 00:38:40.731 You kinda have to slip the film in the top section of the shproc… 00:38:40.731 --> 00:38:43.259 shection of the shprocket. 00:38:43.259 --> 00:38:45.650 A shection of the shprocket. 00:38:47.810 --> 00:38:51.063 In a world where Technology gets Connected 00:38:51.063 --> 00:38:54.544 the fate of all humanity lies in the hands of one man. 00:38:54.544 --> 00:38:58.725 He must convince people to buy powdered dishwasher detergent. 00:38:58.725 --> 00:39:01.039 Will he succeed? 00:39:01.039 --> 00:39:03.550 In theaters this Christmas.