WEBVTT Kind: captions Language: en 00:00:00.080 --> 00:00:03.989 When you hear the term “copy protection”, these days you’re probably more likely to 00:00:03.989 --> 00:00:08.960 think along the lines of DRM or some clever way to prevent copying a digital something-or-other, 00:00:08.960 --> 00:00:12.450 such as a music file, DVD, Blu-ray, or whatnot. 00:00:12.450 --> 00:00:15.940 If this is your frame of reference, you might be a little surprised to learn that a form 00:00:15.940 --> 00:00:21.520 of copy protection had been in use for pre-recorded video cassettes as far back as 1985. 00:00:21.520 --> 00:00:29.140 ♫ sweet music with tuba ♫ 00:00:33.620 --> 00:00:39.400 The Macrovision company, wh… which later became the Rovi corporation, which later bought 00:00:39.410 --> 00:00:44.680 TiVo and decided to absorb that identity so is now TiVo, came up with a way to exploit 00:00:44.680 --> 00:00:49.050 the automatic gain control circuitry in a videocassette recorder, and create pre-recorded 00:00:49.050 --> 00:00:53.660 tapes which would mess around with that circuitry to prevent copying them. 00:00:53.660 --> 00:00:58.190 Now to be clear, this wasn’t so much a copy protection scheme as it was a copy deterrent. 00:00:58.190 --> 00:01:00.520 This wouldn’t exactly prevent you from making a copy, 00:01:00.520 --> 00:01:03.800 but that copy would be almost unbearable to watch. 00:01:03.800 --> 00:01:05.260 So first let’s see what it does. 00:01:05.260 --> 00:01:10.080 I have this VCR plugged into this VCR, and the TV is currently displaying the output 00:01:10.080 --> 00:01:15.420 of this VCR through This VCR, and I’m gonna stick a prerecorded tape into this one and 00:01:15.420 --> 00:01:16.690 a blank tape into this one. 00:01:16.690 --> 00:01:19.030 I want to make a copy of Rain Main. 00:01:19.030 --> 00:01:23.341 So now, I’m going to hit record on this VCR, which will copy Rain Main onto the blank 00:01:23.341 --> 00:01:24.750 tape inside it. 00:01:24.750 --> 00:01:28.260 Pay attention to what happens as soon as I hit record. 00:01:32.660 --> 00:01:34.740 That’s not exactly a good sign. 00:01:34.750 --> 00:01:39.310 Well, aside from that, it seems to be working...OK, no apparent issues so far. 00:01:39.310 --> 00:01:42.870 But when you play the copy back, in addition to the generational loss brought about by 00:01:42.870 --> 00:01:48.170 being a second generation analog copy, the brightness of the image changes wildly, and 00:01:48.170 --> 00:01:49.750 it’s not at all stable. 00:01:49.750 --> 00:01:50.979 Why is that happening? 00:01:50.979 --> 00:01:52.890 And why only when I’m recording? 00:01:52.890 --> 00:01:57.000 Well, what Macrovision does is screw around with the vertical blanking interval of the 00:01:57.000 --> 00:02:01.890 video signal in such a way that a TV won’t really be bothered, but that a VCR attempting 00:02:01.890 --> 00:02:05.550 to record would incorrectly interpret as a change in signal strength. 00:02:05.550 --> 00:02:09.280 I’m not going to go too much into detail here, as I’ve already made a video explaining 00:02:09.280 --> 00:02:13.590 the basics of analog television, but at the start of every field there are a number of 00:02:13.590 --> 00:02:15.590 lines in the signal that are blank. 00:02:15.590 --> 00:02:19.910 They are transmitted at the highest intensity which creates a blacker than black segment 00:02:19.910 --> 00:02:24.580 that the TV can use to synchronize itself to a broadcast image and display it properly. 00:02:24.580 --> 00:02:28.020 To see the vertical blanking interval, you need to get yourself an old TV that allows 00:02:28.020 --> 00:02:29.900 you to adjust the vertical hold. 00:02:30.760 --> 00:02:33.420 This particular set to the rescue once more! 00:02:33.420 --> 00:02:37.420 I have the vertical hold ever so slightly off so the image will roll. 00:02:37.420 --> 00:02:41.910 This dark bar is the vertical blanking interval, and the really dark portion is the sync pulse 00:02:41.910 --> 00:02:43.810 that the TV is looking for. 00:02:43.810 --> 00:02:47.800 You can tell that it almost catches it each time it makes its way to the bottom, but with 00:02:47.800 --> 00:02:51.560 the adjustment just a tad off, it won’t stay in one place. 00:02:51.560 --> 00:02:55.680 Many videocassette recorders would use this pulse for determining the signal strength. 00:02:55.680 --> 00:02:59.370 Not all channels come in with the same strength, which in the land of video means they could 00:02:59.370 --> 00:03:02.840 appear brighter or darker depending on how clear the signal is coming in. 00:03:02.840 --> 00:03:08.420 A TV doesn’t care too much about the relative signal strength, but a VCR really does. 00:03:08.420 --> 00:03:13.170 Just like audio tape, a recording onto video tape needs to be made at the correct level. 00:03:13.170 --> 00:03:17.599 Too much saturation and detail is lost through clipping, and too little saturation and the 00:03:17.599 --> 00:03:19.040 picture becomes noisy. 00:03:19.040 --> 00:03:23.260 By calibrating its signal gain to the level of vertical blanking interval, the VCR would 00:03:23.260 --> 00:03:25.629 get a known reference point for brightness. 00:03:25.629 --> 00:03:29.360 This meant that no matter how strong the signal was or how bright or dark the scene it was 00:03:29.360 --> 00:03:33.640 recording was, a VCR would be able to make a recording correctly. 00:03:33.640 --> 00:03:37.400 So then, with this knowledge in their back pocket, Macrovision Solutions Corporation 00:03:37.400 --> 00:03:42.459 devised a way to mess with the blanking interval in order to confuse the gain control circuitry 00:03:42.459 --> 00:03:43.480 of a VCR. 00:03:43.480 --> 00:03:47.819 Let’s look at how Macrovision changed the blanking interval.This copy of the Magic School 00:03:47.819 --> 00:03:50.379 Bus wasn’t encoded with Macrovision. 00:03:50.379 --> 00:03:52.400 Notice how there’s really nothing in there. 00:03:52.400 --> 00:03:54.650 So what does a tape with macrovision look like? 00:03:54.650 --> 00:04:00.150 Well, with the Wrong Trousers, the BBC took care to encode their tape with Macrovision. 00:04:00.150 --> 00:04:02.900 Notice these new blocks inside the blanking interval. 00:04:02.900 --> 00:04:04.900 That IS Macrovision. 00:04:04.910 --> 00:04:08.830 You’ll see that next to the black blocks, which are actually high intensity pulses because 00:04:08.830 --> 00:04:13.190 a TV signal is inverted, are white blocks that keep fading in and out. 00:04:13.190 --> 00:04:16.810 This is likely what causes the brightness changes in the copied tape. 00:04:16.810 --> 00:04:20.749 A VCR that is recording this video signal keeps seeing the brightness of the vertical 00:04:20.749 --> 00:04:24.689 blanking interval go up and down, and it will treat it as though the entire signal were 00:04:24.689 --> 00:04:26.669 getting stronger or weaker. 00:04:26.669 --> 00:04:30.610 This means that it keeps erroneously adjusting the strength of the recording, which causes 00:04:30.610 --> 00:04:32.490 the recorded image to suffer. 00:04:32.490 --> 00:04:35.820 Warning: the following section of the video contains some speculation. 00:04:35.820 --> 00:04:38.110 OK, I ran into a wall regarding some specifics. 00:04:38.110 --> 00:04:43.529 A lot of the information about Macrovision you can find online seems incomplete and doesn’t 00:04:43.529 --> 00:04:45.779 really jive with what I’ve observed. 00:04:45.779 --> 00:04:49.800 Most descriptions of Macrovision will talk about the high intensity pulses and the automatic 00:04:49.800 --> 00:04:54.590 gain control, but this only explains the image getting brighter and darker. 00:04:54.590 --> 00:04:58.720 This might be particular to these two VCRs, but I tried dubbing in both directions and 00:04:58.720 --> 00:05:03.750 the resulting recordings were wildly unstable--not just too light or too dark. 00:05:03.750 --> 00:05:07.439 Because you can see the white blocks change in intensity when you look at the blanking 00:05:07.439 --> 00:05:11.669 interval, it would seem that this is what causes the image to go light and dark. 00:05:11.669 --> 00:05:12.669 And that’s fine. 00:05:12.669 --> 00:05:16.759 But in addition to these pulses, which are described as “excessive voltage pulses”, 00:05:16.759 --> 00:05:18.840 there’s a little something else. 00:05:18.840 --> 00:05:23.259 These handy oscilloscope captures on the Wikipedia page show that the intensity of the high voltage 00:05:23.259 --> 00:05:27.469 spike goes up and down, and the lower portion here stays constant. 00:05:27.469 --> 00:05:31.610 So this means that this part is the white boxes we can see appear and disappear. 00:05:31.610 --> 00:05:35.869 Assuming these two graphs show these two states of the vertical blanking interval, then the 00:05:35.869 --> 00:05:39.830 only component that changes is the level of these upper pulses. 00:05:39.830 --> 00:05:43.129 This make sense because the light portion follows a dark portion. 00:05:43.129 --> 00:05:45.330 Within the lines, you can see dark, light, grey. 00:05:45.330 --> 00:05:46.460 Dark, light, grey. 00:05:46.460 --> 00:05:50.560 And the graph does the same--dark light grey, dark light grey. 00:05:50.560 --> 00:05:54.850 I think that not enough attention is given to these lower pulses, as these appear to 00:05:54.850 --> 00:05:59.409 be at the same level as the synchronization pulse in the vertical blanking interval. 00:05:59.409 --> 00:06:03.309 If we go back to the tape without Macrovision, you can see this very dark portion here in 00:06:03.309 --> 00:06:04.310 the blanking interval. 00:06:04.310 --> 00:06:08.930 That is the high intensity pulse that the TV locks onto for vertical synchronization. 00:06:08.930 --> 00:06:11.520 But the rest of the blanking interval is just a grey color. 00:06:11.520 --> 00:06:13.060 That’s just the black level. 00:06:13.060 --> 00:06:16.650 Many black and white televisions including this one can’t produce a true black unless 00:06:16.650 --> 00:06:20.340 there’s a good deal of contrast in the scene, which is why it appears grey. 00:06:20.340 --> 00:06:23.960 It will however always produce true black during the sync pulse, because that pulse 00:06:23.960 --> 00:06:27.899 is blacker than black--meaning the electron gun is entirely shut off. 00:06:27.899 --> 00:06:32.289 When you go back to the tape with Macrovision, you can see that the added black bars are 00:06:32.289 --> 00:06:34.970 also at that blacker than black intensity. 00:06:34.970 --> 00:06:39.449 It seems to me that these probably screw around with a VCR’s ability to maintain vertical 00:06:39.449 --> 00:06:40.830 synchronization. 00:06:40.830 --> 00:06:45.259 These dark boxes extend the sync pulse farther into the blanking interval than it should 00:06:45.259 --> 00:06:48.879 be, and the pulses appear multiple times within a single line. 00:06:48.879 --> 00:06:52.080 The white bars that appear and disappear are likely what interferes with the automatic 00:06:52.080 --> 00:06:55.939 gain control and will cause the brightness issues, but that shouldn’t mess with tracking 00:06:55.939 --> 00:06:57.550 or image stability. 00:06:57.550 --> 00:07:00.530 There seems to be more than just that going on. 00:07:00.530 --> 00:07:05.189 Some descriptions of Macrovision call dubbed tapes simply annoying to watch, but this? 00:07:05.189 --> 00:07:06.610 This is unwatchable. 00:07:06.610 --> 00:07:10.229 Just to check that it wasn’t simply this TV that was struggling to maintain synchronization 00:07:10.229 --> 00:07:14.830 from the copied tape, I played the tape back through the LCD TV in my kitchen, which doesn’t 00:07:14.830 --> 00:07:19.039 use the vertical blanking interval in quite the same way a CRT television does. 00:07:19.039 --> 00:07:23.639 It had a more stable image, but not only did the image repeatedly become darker, it lost 00:07:23.639 --> 00:07:28.599 all color completely and the VCR was clearly struggling to maintain tracking. 00:07:28.599 --> 00:07:31.660 You can hear the pitch of the audio waver continuously. 00:07:31.660 --> 00:07:36.162 ♫ severely warbled music ♫ 00:07:36.162 --> 00:07:40.220 The signal got so bad that the TV kept blanking the screen and giving up. 00:07:40.220 --> 00:07:44.509 It seems possible to me that not only does Macrovision mess with the automatic gain control, 00:07:44.509 --> 00:07:48.449 but it might also mess with automatic tracking circuits of a VCR. 00:07:48.449 --> 00:07:51.449 Which may help to partially explain the next part: 00:07:51.449 --> 00:07:56.259 Regardless of the exact specifics of Macrovision, there’s one interesting quirk to the system. 00:07:56.259 --> 00:08:00.499 Really old VCRs, which don’t have the same automatic gain control circuitry of their 00:08:00.499 --> 00:08:03.909 modern counterparts, are impervious to Macrovision. 00:08:03.909 --> 00:08:08.089 (every time I need to move this thing, I regret it.) 00:08:08.089 --> 00:08:12.270 This beast just really doesn’t care what’s going on in the vertical blanking interval. 00:08:12.270 --> 00:08:14.389 It will dub that tape just fine for ya. 00:08:14.389 --> 00:08:17.659 See, it’s recording right now, and the TV doesn’t show anything weird. 00:08:17.660 --> 00:08:19.940 Let’s play the copy back to be sure. 00:08:19.940 --> 00:08:21.960 Yep, just fine. 00:08:22.680 --> 00:08:24.280 Hey, it’s me from the future. 00:08:24.289 --> 00:08:28.169 While shooting the B-roll I needed for this video, I noticed that the VCR which was copying 00:08:28.169 --> 00:08:32.029 Rain Man was making really odd sounds as it recorded. 00:08:32.029 --> 00:08:34.340 Take a listen. 00:08:35.180 --> 00:08:44.780 [sound of VCR recording but with wild speed fluctuations] 00:08:46.220 --> 00:08:50.880 Normally you never hear a VCR make noises like this, except for maybe when it’s adjusting 00:08:50.880 --> 00:08:52.350 its tracking. 00:08:52.350 --> 00:08:56.210 This really seems like the Macrovision is confusing the VCR on where the start of the 00:08:56.210 --> 00:09:00.550 video field actually is, and it keeps trying to compensate for that. 00:09:00.550 --> 00:09:04.900 The video head drum needs to spin in perfect alignment with the video signal, and it seems 00:09:04.900 --> 00:09:09.780 like this VCR is being fooled into thinking the start of each frame moves as the macrovision 00:09:09.780 --> 00:09:11.510 blocks fade in and out. 00:09:11.510 --> 00:09:16.460 The old VCR probably isn’t affected because it doesn’t try to chase a wavering signal-- 00:09:16.460 --> 00:09:20.420 I would imagine it has much less of an ability to adjust the speed of its video head and 00:09:20.420 --> 00:09:25.460 tape transport to signal fluctuations, and as a result Macrovision just doesn’t affect it. 00:09:25.460 --> 00:09:26.440 Anyway, 00:09:26.440 --> 00:09:30.460 Even into the days of Macrovision, there were some VCRs that weren’t affected by it. 00:09:30.460 --> 00:09:35.170 It was estimated that 85% were, but that leaves a significant number of VCRs floating around 00:09:35.170 --> 00:09:36.870 which just don’t care about it. 00:09:36.870 --> 00:09:40.790 Also, for those that were so inclined, getting a Macrovision defeat device from the grey 00:09:40.790 --> 00:09:42.710 market wasn’t all that hard. 00:09:42.710 --> 00:09:45.280 In fact, you could build one yourself. 00:09:45.280 --> 00:09:49.670 Because Macrovision can’t entirely eliminate the pulse for vertical synchronization, otherwise 00:09:49.670 --> 00:09:54.400 the TV couldn’t display an image at all, a defeat device just needs to remove the altered 00:09:54.400 --> 00:09:58.360 parts that were messed up with the boxes and replace it with standard black lines. 00:09:58.360 --> 00:10:02.620 Now, when DVD came along, Macrovision was incorporated here, too. 00:10:02.620 --> 00:10:06.750 Because DVDs are a digital format, they don’t contain the video fields like a videocassette 00:10:06.750 --> 00:10:10.461 does, so the player actually generates the Macrovision in the vertical blanking interval 00:10:10.461 --> 00:10:12.300 of its composite output. 00:10:12.300 --> 00:10:16.470 It will only do it so long as the disc asks for it, which not all of them did. 00:10:16.470 --> 00:10:20.690 For example, this copy of the Princess Bride doesn’t use Macrovision, but this copy of 00:10:20.690 --> 00:10:22.260 Spirited Away does. 00:10:22.260 --> 00:10:26.330 The fact that the analog Macrovision system was shoved into the DVD format caused some 00:10:26.330 --> 00:10:28.840 problems for early DVD adopters. 00:10:28.840 --> 00:10:32.220 Remember how the image got messed up only after I had hit record? 00:10:32.220 --> 00:10:36.090 That happened because unless it’s recording, the automatic gain control circuitry isn’t 00:10:36.090 --> 00:10:39.430 active and it passes the signal straight through to the TV. 00:10:39.430 --> 00:10:44.640 But many VCRs weren’t so kind and instead applied automatic gain control to whatever 00:10:44.640 --> 00:10:48.500 input they received, regardless of if it was recording or not. 00:10:48.500 --> 00:10:53.361 Many TVs only had one composite input, so for anyone who hooked their new DVD player 00:10:53.361 --> 00:10:58.500 into their TV through their VCR, and if their VCR didn’t simply pass-through the video 00:10:58.500 --> 00:11:01.820 from the DVD player, they might get a horrible mess. 00:11:01.820 --> 00:11:05.860 Imagine how annoying it would be to play your first DVD and all of the sudden have the image 00:11:05.860 --> 00:11:07.590 come out like that. 00:11:07.590 --> 00:11:11.680 Many people thought their new DVD players were somehow broken, but thankfully newer 00:11:11.680 --> 00:11:17.150 VCRs like this one won’t cause a fuss unless you tried to record from the DVD player. 00:11:17.150 --> 00:11:21.180 And while we’re at it, this attempted copying of Spirited Away shows a second thing added 00:11:21.180 --> 00:11:24.920 to Macrovision, which was called Level II ACP. 00:11:24.920 --> 00:11:28.640 This would invert the phase of the colorburst occasionally, which produced stripes of the 00:11:28.640 --> 00:11:30.800 wrong color on a copy. 00:11:30.800 --> 00:11:34.280 Quite honestly that’s the least annoying of the issues with this recording. 00:11:34.280 --> 00:11:38.330 Now these days, far more complicated copy protection schemes are in use, much to the 00:11:38.330 --> 00:11:43.080 chagrin of anyone who wants to make legal copies of their media which is, by the way, 00:11:43.080 --> 00:11:44.040 allowed. 00:11:44.040 --> 00:11:46.920 But Macrovision was I think quite clever. 00:11:46.920 --> 00:11:50.880 By messing with the fundamental structure of the video signal just enough to keep a 00:11:50.880 --> 00:11:57.250 recording VCR from working correctly, but not enough for a TV or a playing to become 00:11:57.250 --> 00:12:00.120 confused, they actually had a pretty elegant solution. 00:12:00.120 --> 00:12:02.480 Thanks for watching, I hoped you enjoyed the video! 00:12:02.480 --> 00:12:05.620 If this is your first time coming across the channel and you liked what you saw, please 00:12:05.620 --> 00:12:07.660 subscribe to Technology Connections! 00:12:07.660 --> 00:12:09.580 You can also support the channel on Patreon. 00:12:09.580 --> 00:12:12.880 Patrons of the channel are what keeps these videos coming, and if you’d like to join 00:12:12.880 --> 00:12:16.940 the awesome people who make this possible, please check out my Patreon page. 00:12:16.940 --> 00:12:19.140 Thanks for your consideration, and I’ll see you next time!