WEBVTT
Kind: captions
Language: en

00:00:00.089 --> 00:00:01.089
Here we go.

00:00:01.089 --> 00:00:02.389
More stuff about the beginnings of color TV.

00:00:02.389 --> 00:00:05.680
We’re continuing from the last video so
click up above if you got here somehow without

00:00:05.680 --> 00:00:08.140
seeing those or find the link down in the
description.

00:00:08.140 --> 00:00:12.440
I’m back in a long sleeve shirt cause I’m
about to go to work, so let’s get cracking.

00:00:12.440 --> 00:00:16.410
NTSC often received the nickname “Never
The Same Color” or similar due to its reliance

00:00:16.410 --> 00:00:19.390
on the phase offset alone to determine hue.

00:00:19.390 --> 00:00:23.240
Unless the television set was dead on the
money with the way it handled the color burst,

00:00:23.240 --> 00:00:27.320
it would be very likely that it decoded the
chroma data slightly out of phase and the

00:00:27.320 --> 00:00:28.990
color would be messed up.

00:00:28.990 --> 00:00:32.860
The tint control on a television would allow
for manually shifting the reference point

00:00:32.860 --> 00:00:36.899
from the color burst to correct for errors,
with incorrect skin tone usually being how

00:00:36.899 --> 00:00:38.880
one identified a problem.

00:00:38.880 --> 00:00:42.719
After vacuum tube circuitry was replaced with
transistorized equivalents, most of these

00:00:42.719 --> 00:00:48.129
problems went away, and automatic tint control
eventually became standard on television sets.

00:00:48.129 --> 00:00:51.879
By the mid 1970’s, broadcast equipment had
progressed to the point that the color burst

00:00:51.879 --> 00:00:56.410
was always transmitted with the exactly correct
phase, pretty much removing the need for tint

00:00:56.410 --> 00:00:57.670
control altogether.

00:00:57.670 --> 00:01:02.670
PAL, the standard in most of Europe, reversed
the phase of the color encoding on each alternate

00:01:02.670 --> 00:01:05.999
scan line, hence its name Phase Alternating
Line.

00:01:05.999 --> 00:01:09.930
This made it such that a tint control was
not needed as the television would automatically

00:01:09.930 --> 00:01:14.630
find the correct phase offset through averaging
the alternate lines, and thus PAL was generally

00:01:14.630 --> 00:01:16.299
deemed superior.

00:01:16.299 --> 00:01:20.289
Europe received color television more than
a decade after the US, in part because of

00:01:20.289 --> 00:01:22.779
the more challenging geography of Europe.

00:01:22.779 --> 00:01:27.260
The NTSC system would drift color badly if
transmission conditions weren’t more or

00:01:27.260 --> 00:01:28.369
less ideal.

00:01:28.369 --> 00:01:33.009
So Europe waited patiently for a better system,
which PAL most certainly is.

00:01:33.009 --> 00:01:37.680
Telefunken in Germany patented PAL, the work
of Walter Bruch, in 1962.

00:01:37.680 --> 00:01:42.899
The first broadcasts of this new system occurred
in 1967 on BBC 2 in the UK.

00:01:42.899 --> 00:01:46.630
Apparently this was of the Wimbledon Championships
and let me just say that as an American, my

00:01:46.630 --> 00:01:49.569
reaction to learning this fact was “Of course
it was Wimbledon”.

00:01:49.569 --> 00:01:53.149
ON THAT NOTE, I was rather annoyed by this
sort of comment.

00:01:53.149 --> 00:01:54.749
And I’ll tell you why.

00:01:54.749 --> 00:01:58.809
Just pause for a minute as ask yourself, was
NTSC a choice?

00:01:58.809 --> 00:02:01.059
How could it have been a mistake?

00:02:01.059 --> 00:02:06.249
It was the first broadcast black-and-white
compatible color television transmission scheme

00:02:06.249 --> 00:02:07.929
on the planet.

00:02:07.929 --> 00:02:12.909
PAL is virtually identical to NTSC with the
exception of the alternating phase.

00:02:12.909 --> 00:02:17.370
Of course it would be better, it was developed
later, specifically to address NTSC’s weaknesses.

00:02:17.370 --> 00:02:21.819
We were stuck with it in the States because
it was already there, and we weren’t about

00:02:21.819 --> 00:02:23.829
to pull the plug on an existing standard.

00:02:23.829 --> 00:02:27.660
That’s the whole reason we spent so much
time making color TV compatible with black

00:02:27.660 --> 00:02:28.959
and white sets.

00:02:28.959 --> 00:02:33.739
I think many people appreciate PAL’s much
higher image resolution of 625 lines verses

00:02:33.739 --> 00:02:39.629
525 in the US, but that extra resolution came
at the expense of a reduced frame rate.

00:02:39.629 --> 00:02:40.989
And speaking of framerates,

00:02:40.989 --> 00:02:44.829
One commenter on the last video was singing
the praises of PAL because, with a framerate

00:02:44.829 --> 00:02:50.700
of 25 FPS, movies running at 24 frames per
second just need to be sped up by 4% to be

00:02:50.700 --> 00:02:51.980
converted to a PAL broadcast.

00:02:51.980 --> 00:02:56.099
I’m sorry, but I completely disagree with
that being “better”.

00:02:56.099 --> 00:02:58.580
I don’t want to just speed up the film!

00:02:58.580 --> 00:03:03.099
We used Three-Two pull down, which sort of
blended frames together by advancing the film

00:03:03.099 --> 00:03:05.249
frame between television fields.

00:03:05.249 --> 00:03:09.090
In effect some film frames were shown over
two fields, and some three.

00:03:09.090 --> 00:03:12.769
Most people, in fact I’d say virtually all
people, can’t see this, since the fields

00:03:12.769 --> 00:03:14.219
aren’t drawn together.

00:03:14.219 --> 00:03:19.120
And since it leaves the runtime and audio
as they should be within .1%, I’d much rather

00:03:19.120 --> 00:03:20.120
do that!

00:03:20.120 --> 00:03:24.500
And speaking of speaking of framerates, lets
discuss the 29.97 frames per second nonsense

00:03:24.500 --> 00:03:26.840
brought to you by NTSC.

00:03:26.840 --> 00:03:30.829
Matt Parker of Standup Maths made a video
about this very thing, which I shall link

00:03:30.829 --> 00:03:31.920
to now.

00:03:31.920 --> 00:03:35.930
Simply put, the carrier frequency of the audio
signal could interfere with the chrominance

00:03:35.930 --> 00:03:39.359
carrier and produce a visible dot pattern
on the screen.

00:03:39.359 --> 00:03:43.819
To keep this from happening, the audio carrier
frequency needed to be an integer multiple

00:03:43.819 --> 00:03:45.269
of the line rate.

00:03:45.269 --> 00:03:46.269
Which it wasn’t.

00:03:46.269 --> 00:03:48.459
So, something had to give.

00:03:48.459 --> 00:03:51.630
You couldn’t just change the audio carrier,
though, cause that would mess up existing

00:03:51.630 --> 00:03:52.680
TVs.

00:03:52.680 --> 00:03:55.269
But you could change the line rate, or frequency.

00:03:55.269 --> 00:04:02.290
By reducing the line frequency from 15,750
Hz to 15,734 Hz the audio carrier was now

00:04:02.290 --> 00:04:07.730
exactly 286 times the frequency of the line
rate, rather than some nasty number.

00:04:07.730 --> 00:04:11.379
This would minimize interference and prevent
the dot pattern from occurring, and existing

00:04:11.379 --> 00:04:14.980
television sets wouldn’t be bothered by
the change as their vertical hold circuitry

00:04:14.980 --> 00:04:19.000
was designed to compensate for slop in the
vertical retrace frequency.

00:04:19.000 --> 00:04:23.800
Of course, lowering the linerate lowered the
frame rate, as it’s still 525 lines per

00:04:23.800 --> 00:04:30.210
frame, so the frame rate was lowered to 29.97
frames per second entirely as a side-effect

00:04:30.210 --> 00:04:31.699
of compatible color.

00:04:31.699 --> 00:04:34.030
In PAL countries, this wasn’t ever an issue.

00:04:34.030 --> 00:04:37.440
It turns out backward compatibility wasn’t
ever on anyone’s mind in europe, or at least

00:04:37.440 --> 00:04:39.199
those disinterested in it won out.

00:04:39.199 --> 00:04:45.520
For example, in the UK, BBC2 started black
and white 625 line broadcasts in 1964, and

00:04:45.520 --> 00:04:49.800
owners of older 405 line televisions just
wouldn’t get to see it.

00:04:49.800 --> 00:04:53.680
This was about 10 years after NTSC had started
broadcasting.

00:04:53.680 --> 00:04:59.389
And PAL was based on the 625 line standard,
and likely due to the known framerate debacle,

00:04:59.389 --> 00:05:03.980
the audio carrier in this standard was specified
to be an integer multiple of the linerate,

00:05:03.980 --> 00:05:05.750
15,625 hertz.

00:05:05.750 --> 00:05:10.770
Thus the frame rate could stay an even 25
frames per second once color broadcasts began

00:05:10.770 --> 00:05:13.460
without the audio carrier affecting the image.

00:05:13.460 --> 00:05:17.250
In the previous video, I said that because
the chrominance carrier is suppressed during

00:05:17.250 --> 00:05:20.990
transmission, most of the interference between
chrominance and luminance is eliminated.

00:05:20.990 --> 00:05:22.680
But not all of it is.

00:05:22.680 --> 00:05:26.280
Black and white televisions made after the
changeover to color were designed to filter

00:05:26.280 --> 00:05:30.889
out this interference, but earlier televisions
or cheaper ones without the filter circuit

00:05:30.889 --> 00:05:35.030
would display a pattern of dots in highly
colored areas of the screen.

00:05:35.030 --> 00:05:38.470
The dots were a visible manifestation of the
chrominance carrier’s interference.

00:05:38.470 --> 00:05:42.520
Now, I’m going to make a speculative statement
because I couldn’t find anything to confirm

00:05:42.520 --> 00:05:46.850
or deny this after hours of poking about online,
so here goes.

00:05:46.850 --> 00:05:51.690
Because the luminance signal is used in calculating
I and Q, it seems to me that the luminance

00:05:51.690 --> 00:05:57.069
signal itself must be manipulated at precise
times to actually create changes in the calculated

00:05:57.069 --> 00:06:01.810
I and Q values, otherwise it would seem their
calculated output would always be the same

00:06:01.810 --> 00:06:03.190
for a given brightness.

00:06:03.190 --> 00:06:07.379
I believe that the chroma dots are the visual
manifestations of this happening.

00:06:07.379 --> 00:06:11.960
I have read that the dots are most visible
in highly colored areas, which makes sense

00:06:11.960 --> 00:06:16.879
as the amplitude of the chroma carrier dictates
the saturation of a color, and the phase difference

00:06:16.879 --> 00:06:18.069
the hue.

00:06:18.069 --> 00:06:22.699
It therefore sounds to me like the Y carrier
is constantly being altered along with color

00:06:22.699 --> 00:06:26.169
information, and that alteration appears as
dots.

00:06:26.169 --> 00:06:29.190
If that’s not at all why the dots appear
please comment below.

00:06:29.190 --> 00:06:33.090
ANYWAY, the dots would turn out to be useful
in the 21st century.

00:06:33.090 --> 00:06:37.520
Many older television programs were recorded
onto videotape masters that were re-used to

00:06:37.520 --> 00:06:39.270
lower production costs.

00:06:39.270 --> 00:06:44.120
With these master tapes lost, the only copy
of the TV shows that remained were on film,

00:06:44.120 --> 00:06:45.780
transferred via telecine.

00:06:45.780 --> 00:06:49.250
In many if not most cases, this film was black
and white.

00:06:49.250 --> 00:06:53.199
This had led to many of these early color
television shows only being preserved in black

00:06:53.199 --> 00:06:54.199
and white.

00:06:54.199 --> 00:06:58.610
But some TV studios used a telecine to capture
the output from a black and white monitor

00:06:58.610 --> 00:07:01.020
that didn’t filter out the chroma dots.

00:07:01.020 --> 00:07:04.840
Using modern software, these dots have been
used to reconstruct the color.

00:07:04.840 --> 00:07:09.379
By carefully examining the magnitude and position
of these dots, the color information can be

00:07:09.379 --> 00:07:10.439
extracted.

00:07:10.439 --> 00:07:15.140
This wasn’t an easy task, mind you, particularly
with the extra geometry component added by

00:07:15.140 --> 00:07:19.260
the curved glass screens of the monitors,
but it’s still an amazing thing.

00:07:19.260 --> 00:07:22.230
I’ll link to some more info down below if
you’d like to learn more.

00:07:22.230 --> 00:07:25.599
Now let’s talk a bit more about Guillermo
Gonzalez Camarena.

00:07:25.599 --> 00:07:30.090
Two videos ago I briefly talked about him
and his clever but ultimately inadequate color

00:07:30.090 --> 00:07:32.240
wheel adapter for black and white televisions.

00:07:32.240 --> 00:07:37.620
He’s also known for inventing a CRT with
two electron guns rather than three.

00:07:37.620 --> 00:07:40.720
This would supposedly make color televisions
cheaper to produce.

00:07:40.720 --> 00:07:45.810
However, there are two glaring problems with
attributing him to “inventing” that.

00:07:45.810 --> 00:07:50.259
First, if there were an individual who were
to have experimented in two color TV long

00:07:50.259 --> 00:07:53.919
before Gonzalez Camarena, who would you guess
that to be?

00:07:53.919 --> 00:07:54.919
Anyone?

00:07:54.919 --> 00:07:55.919
Anyone?

00:07:55.919 --> 00:07:57.800
I’ll give you a hint, he’s Scottish!

00:07:57.800 --> 00:08:00.349
That’s right, it’s JOHN LOGIE BAIRD.

00:08:00.349 --> 00:08:02.280
He’s a persistent little bugger, isn’t
he?

00:08:02.280 --> 00:08:07.270
Now I’m starting to see why people feel
so strongly about his contributions to television.

00:08:07.270 --> 00:08:11.610
So first, two-color “color” can be achieved
using complementary colors.

00:08:11.610 --> 00:08:15.789
In fact, the first Technicolor films were
done with a red and green film strip all the

00:08:15.789 --> 00:08:18.419
way back in 1916.

00:08:18.419 --> 00:08:22.520
Baird used a really wonky looking picture
tube called a Telechrome which instead used

00:08:22.520 --> 00:08:25.330
cyan and red-orange phosphors.

00:08:25.330 --> 00:08:29.860
These two colors can produce a realistic skin
tone and a limited range of other colors.

00:08:29.860 --> 00:08:35.470
He came upon the concept for the telechrome
in 1942 and demonstrated it in 1944.

00:08:35.470 --> 00:08:41.330
Gonzalez Camarena’s idea is dated to 1963,
with one source saying 1962.

00:08:41.330 --> 00:08:42.360
Links are below.

00:08:42.360 --> 00:08:49.100
But a far more damning issue against Gonzales
Camarena is the fact the the 1953 NTSC standard

00:08:49.100 --> 00:08:51.660
includes support for two color receivers!

00:08:51.660 --> 00:08:54.380
That’s right, that whole I and Q business?

00:08:54.380 --> 00:08:57.440
It turns out, I on its own is a two color
gamut.

00:08:57.440 --> 00:09:01.780
A television could be built with cyan and
red-orange phosphors and demodulate only the

00:09:01.780 --> 00:09:05.710
I signal, and you would get a functional two
color TV.

00:09:05.710 --> 00:09:10.060
This may also help to explain why I is given
three times the bandwidth of Q.

00:09:10.060 --> 00:09:14.300
If it were meant to be able to produce a reasonable
color image on its own, it would make sense

00:09:14.300 --> 00:09:17.070
to prioritize bandwidth to I over Q.

00:09:17.070 --> 00:09:21.790
But the fact of the matter is, two color “color”
is never as good as three color.

00:09:21.790 --> 00:09:24.290
Skin tone might look right, but not much else
will.

00:09:24.290 --> 00:09:28.430
Thus, since you’d need almost the same amount
of manufacturing complexity to produce a two

00:09:28.430 --> 00:09:33.190
color CRT as a three color--still need a shadow
mask and phosphor grid, you’re really just

00:09:33.190 --> 00:09:37.630
losing one electron gun--it was never deemed
to be worth the significantly worse color

00:09:37.630 --> 00:09:40.600
fidelity in exchange for a few pennies saved.

00:09:40.600 --> 00:09:44.560
Many people correctly commented in previous
videos that the color wheel from the CBS system

00:09:44.560 --> 00:09:48.160
survived in single chip DLP projectors and
televisions.

00:09:48.160 --> 00:09:52.650
But you might not have known that for a brief
time in the 1950s, you could purchase a kit,

00:09:52.650 --> 00:09:57.450
called the Col-R-Tel, to convert your existing
black and white TV into color.

00:09:57.450 --> 00:10:01.060
By performing a few wiring modifications,
you could add a device which would extract

00:10:01.060 --> 00:10:06.940
the color data from an NTSC broadcast, alter
the CRT’s output to a sequential field system,

00:10:06.940 --> 00:10:10.080
and use a color wheel to reproduce the color
sequentially.

00:10:10.080 --> 00:10:13.450
Gonzalez Camerena’s work would actually
be a thing after all.

00:10:13.450 --> 00:10:16.880
There’s a great link down below, along with
many many others, if you’d like to learn

00:10:16.880 --> 00:10:17.880
more.

00:10:17.880 --> 00:10:22.030
And for my last tid-bit, a brief word on shadow
masks, aperture grilles, and Trinitron.

00:10:22.030 --> 00:10:25.710
I’m still planning on making a separate
video showcasing Trinitron, but some people

00:10:25.710 --> 00:10:29.800
have correctly pointed out that the CRT set
I’ve been using for demonstrations doesn’t

00:10:29.800 --> 00:10:33.170
have the same type of shadow mask that CRTs
of the time did.

00:10:33.170 --> 00:10:37.880
I do have a device which has the original
type CRT, but it doesn’t really fit well

00:10:37.880 --> 00:10:39.220
into this discussion.

00:10:39.220 --> 00:10:40.480
Well, here, I’ll show you.

00:10:41.680 --> 00:10:43.360
♫ Macintosh Startup Chime ♫

00:10:43.680 --> 00:10:44.880
Yep.  That’s it.

00:10:44.880 --> 00:10:47.540
Doesn’t really fit into the video, now does
it?

00:10:47.540 --> 00:10:52.010
Most CRT computer monitors--if not all that
aren’t Trinitron--use a design similar to

00:10:52.010 --> 00:10:54.410
what you’d find on the earlier color sets.

00:10:54.410 --> 00:10:59.630
These CRTs don’t have the individual triads
like you see here, instead it’s just a repeating

00:10:59.630 --> 00:11:02.130
pattern of red, green, and blue circles.

00:11:02.130 --> 00:11:07.040
Because of the high image resolution needed
in a computer monitor, a very fine dot pitch,

00:11:07.040 --> 00:11:11.360
the term used to describe how many dots appear
over a given length, was required.

00:11:11.360 --> 00:11:16.410
The CRT from the GE set uses a more modern
phosphor and shadow mask design which increases

00:11:16.410 --> 00:11:21.140
the phosphor-to-mask ratio, but at the expense
of a coarser dot pitch.

00:11:21.140 --> 00:11:24.600
This design was used in televisions to better
compete with Trinitron displays.

00:11:24.600 --> 00:11:27.440
But again, we’ll talk more about that later.

00:11:27.440 --> 00:11:29.010
As always, thank you so much for watching!

00:11:29.010 --> 00:11:30.830
I hope you enjoyed the video.

00:11:30.830 --> 00:11:33.710
If you’re new to this channel and liked
what you saw, please subscribe so you won’t

00:11:33.710 --> 00:11:34.720
miss the next ones.

00:11:34.720 --> 00:11:38.680
I’d also like to thank all of my current
supporters on Patreon, especially these patrons

00:11:38.680 --> 00:11:40.930
who get their names in lights.

00:11:40.930 --> 00:11:44.980
Patreon supporters have allowed me to spend
much more time focusing on this channel, which

00:11:44.980 --> 00:11:47.180
is bringing videos to you more frequently.

00:11:47.180 --> 00:11:50.380
If you’re interested in helping out, please
check out my Patreon page through the link

00:11:50.380 --> 00:11:53.110
on your screen, or down below in the description.

00:11:53.110 --> 00:11:54.350
Thank you for your consideration.

00:11:54.350 --> 00:11:56.200
And, I’ll see you next time.

00:12:06.880 --> 00:12:07.740
(Exasperated) Wooh!

00:12:09.260 --> 00:12:10.380
Hoh, OK