This is one normal PAL/NTSC scanline as seen on an oscilloscope screen.
Time in microseconds
0 1 2 3 4 5 6 0 1
0123456789012345678901234567890123456789012345678901234567890123012345678901
___________________________________________________ ______
| ^-- White level | |
| | Color |
| Arbitrary picture data | burst--v |
| | |
_MM_|___________________________________________________|_ _MM_|______
| WW ^-- Black level | | WW
| |____|
Front porch --^ ^ ^-- Back
Horizontal | porch
synchronization --+
Here are both the PAL and NTSC Macrovision 'magic' lines that do the trick. Both are shown at their maximum amplitudes.
0 1 2 3 4 5 6 0 1
0123456789012345678901234567890123456789012345678901234567890123012345678901
_ _ _ _ _ _ _ _
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | |
_MM__ | | | | | | | | | | | | | |__________________ _MM__ | |
| WW | | | | | | | | | | | | | | | | WW | | |
| |_| |_| |_| |_| |_| |_| |_| |____| |_| |
0 1 2 3 4 5 6 0 1
0123456789012345678901234567890123456789012345678901234567890123012345678901
__ __ __ __ __
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
| | | | | | | | | |
_MM__ | |__ | |__ | |__ | |___________|||||_______ _MM__ | |
| WW | | | | | | | | ||||| | | WW | |
| |_| |_| |_| |_| |____| |_|
The lines drawn above are quite similar. Both try to present false synchronization pulses to the VCR the first 40 microseconds or so. The rest of the line is black, because false syncs there would trigger the sync circuits in monitors/TVs and consequently the top of the picture would be very unstable. Some TVs really do suffer even now, I have seen it myself.
There are a few cycles of high frequency triangle wave in the latter part of the NTSC line (denoted by |||'s), but they shouldn't be contributing to the protection effect.
But just the bright pulses are not sufficient. If they had a constant amplitude, it would be quite easy just to increase the amplitude of the video signal and get a decent picture. Therefore the false back porch voltage level is varied according to some simple rules in order to get the brightness changes as annoying as possible.
The following pictures show how the false back porch amplitudes change with time. The lowest level is black, the highest is "super-white". The false syncs (below black level) do not change their amplitude. The perceived brightness of the TV picture is the inverse, e.g. the highest level in the diagram means the darkest picture on the screen.
_____________________ ___...
/ \ R1 R1 R1 /
/ \ ___ ___ ___ /
/ \ || | | | | | | | /
/ \ || | | | | | | | /
/ \ ||___|R2 |___|R2 |___|R2 |___| /
/ \|| |___| |___| |___| |/
: : : :: : : : : : : : :
: 2s : 7s : 2.3s ::32f:32f:32f:32f:32f:32f:32f: 2s :
: : : :: : : : : : : : :
10f 2f
<-- ~9s -->
R1 = lines in region 1
R2 = lines in region 2
f = frames
Here is where the two regions differ. When R1 rises to ~60% of max amplitude, R2 goes to black. Otherwise they change in parallel.
____________________..._________________ _____...
/ \ /
/ \ /
/ \ /
/ \ /
/ \ /
/ \_______/
: : : : : :
:3.5s : 22s : 5s : 4s :3.5s :
: : : : : :
As can be seen, the NTSC-Macrovision cycle is very simple.
All the slopes and the stable regions between them are timed in seconds, because a) the timing is not so critical and b) it is difficult to say in which frame a slope starts or ends.
Below is an example of what a recorded Macrvision line looks at its worst. The spikes are adjusted to have the amplitude of a normal sync pulse by the AGC, so the real sync pulses are in turn vanishingly small. Because of that the TV usually loses sync and the picture starts to roll.
0 1 2 3 4 5 6 0 1
0123456789012345678901234567890123456789012345678901234567890123012345678901
__ __ __ __ --
| | | | | | | | | |
.-MM--._| --._| --._| --._| -----------------------.____.-MM--._|
Here is a small disclaimer-type paragraph: I have built an eliminator and have used it for backing up my own precious videotapes. Try, for example, get a replacement for your damaged "The Little Mermaid" videotape. You're lucky, if you succeed. I almost never rent videotapes; the picture sucks and they are usually "pan-and-scan" transfers. And the last reason: I'm a hardware hacker, so I did it just for the heck of it.
Macrovision elimination is VERY simple, if you have some knowledge of electronics. My primary inspiration was:
"Macrovision decoder/blanker"
Elektor Electronics, October 1988, pp. 44-47.
(Note: it features an older version of Macrovision; not that different, though.)
I built roughly an equivalent circuit myself, but it was higly unsatisfactory. Reasons:
Below is a block diagram of my currently satisfactorily working device. It resembles only remotely the EE one. The basic idea is that the Macrovision pulses are replaced with a black level. I challenge anyone to make a simpler device.
video _______ _____ _____
in >------+----->| sync |>------------>| |>-->| |>-+
| | sepa | vsync |delay| | MMV | | Pulse
| | rator | | | | 1 | | that
| |_______|>-------+ |_____| |_____| | lasts
| LM1881 burst | region | the
| | start | whole
electr. 0| | | Macrov.
switch ___|_ | _ | region
| o | insert_black | | |<-------------+
+-----o/ |<---------------------------->|&|
| |___o_| | |_|<--------+
| | | |
| 1| black | |~45us pulse
| | level | | that covers
| | | | the false
| __^_____ | | syncs
| | | | |
| | sample |sample_now | line_start _____ |
+--->| and |<----------------+---------->| |>-+
| | hold | | MMV |
| |________| | 2 |
| sample black level |_____|
| from back porch
| ________
| | |
| | video | video
+--->| output |>--------->
| buffer | out
|________|
Some explanations:
I do have the whole circuit [...] which you can get here:
Note: The PostScript files are only available compressed due to the huge savings in space (40KB->8KB).
One method that I have seen on some old rental cassettes is the changing of horizontal sync amplitude in the middle of a frame. When dubbed, the AGC circuits change the gain to keep the sync amplitude constant, and the resulting picture has very noticeable bands of bright and dim picture.
--------------
Original frame
v--- constant grey level
...........................................................................
___________________________________________________________________________
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | | | | |
(The amplitude change is somewhat exaggerated)
-------------
Dubbed result
.............. ............
. . . .
.............. .................... ...........
___________________________________________________________________________
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
Disadvantages of this method:
.............. ...................... .........
................ ..............
______________ ______________________ __________
| | | | | | | ________________| | | | | | | | | | | ______________| | | | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
Remedy: clamp to front or back porch and re-create sync.
Another method I have seen is to place a color subcarrier burst at the bottom of the horizontal synchronization pulse:
..._ _MM___...
| | WW
|MMMM|
WWWW
The only effect I noticed was that the colors changed a little, when switching between 'clean' sync and 'bursted' sync. Does anyone guess what it is supposed to be doing?
OK, this is the end. Thank you for reading. Happy hacking.