Compatible Color: The Ultimate Three-For-One Special

[@Clay3613]

Then RCA decided that vinyl was a good format for movies too and wasted about 17 years on that, ultimately bankrupting the company.

[@TechnologyConnections]

Ah yes, the Capacitance Electronic Disc. I have a CED player; it's a very interesting concept but so so _sooooo_ obviously flawed that it's amazing they just kept pursuing it. Oh well. That will probably become a video some day.

[@VisionThing]

Technology Connections Yes please do a video on that, had no idea of this.

[@scottlarson1548]

There are already several great KZbin videos on CED players. I think TechMoan did a very funny one.

[@Clay3613]

Doesn't mean he can't expand upon it.

[@justmytake4152]

I think the big reason for Capacitance Electronic Disc's failure was that it come to the market too late I think it would have worked commercially if it came out earlier so it would have a chance in the video rental firms but I think VHS & betamax which although the recorders/players were more expensive it biggest selling point was you could record TV. Then any advantage it may have had with picture quality over them was lost when Laserdisc and CD (as it was predictable this would reduced the cost to mass produce the discs which were virtually the same just smaller compared to vinyl and Capacitance Disc) eventually was released first.

[@Korpsaws]

No music, no intro, a true champ. Doesn't need any hype to be excellent :)

[@mrflamewars]

Korpsaw yes! There are far too many long, stupid " hey fucker look at me! " Intros on KZbin

[@cesarrodriguz7549]

You sir, get a like. :)

[@RecycleBin0]

YEAH!!!!!

[@daaknait]

Exactly!

[@betaneptune]

Totally agree. I find music on videos like these to be distracting and totally unnecessary. Thanks!

[@DreamOf944]

5:05 ... "That's Y" ... possibly the most complicated build-up to a pun ever.

[@andrewandrei3062]

ikr xd

[@eval_is_evil]

Or "Hertz to think about it"

[@amitraam1270]

...and that's why we love him :)))

[@Speedster04_BRA]

"That's the sort of amazing life-changing things you can only learn here, at Technology Connections!" - Alec

[@The-CNS-Channel]

​@@eval_is_evilthought I was first to hear that :(

[@lawrencedoliveiro9104]

6:36 Probably good to mention by this point that “I” stood for “in-phase” and “Q” for “quadrature”. That is, these were the 0° and 90° components extracted from the phase-modulated chrominance signal. If you think of this signal as a point moving in two dimensions, the direction of the point from the centre gives you the hue, and the distance from the centre is the saturation of the colour.

[@ScottGrammer]

Excellent explanation of a complicated system!

[@johnbell3621]

@@ScottGrammer And the local oscillator to extract the I and Q needed to be in sync with the transmitter - hence the colour burst signal.

[@jimharris9394]

And there was actually a specialized kind of oscilloscope, (forgot the name of it), that displayed the quadrature relationship between these signals in exactly that way, (It was a radial display), and if you were rich enough to have one for your TV repair business, you could tune the color to a gnat's hair. Since it was expensive as oil-rights, TV shops used the much less expensive, (and correspondingly less accurate), bar/line/dot generator to align color. Most people didn't notice. Those of us who did and could see the difference, it drove batty!

[@Vincent_Sullivan]

@@jimharris9394 The name of the specialized oscilloscope you are trying to remember is "vectorscope". This makes sense as it is displaying the vectors that Lawrence was talking about is his excellent description of the I and Q signals. On the vectorscope screen no signal produced a dot in the centre of the screen. When given a colour signal the dot would move away from the centre. The direction it moved indicated the hue of the colour and the distance from the centre indicated the saturation of the colour. The graticule of the vectorscope had various small roughly square areas marked on it and when given a standardized colour test signal various of the illuminated blobs on the screen should fall within the marked areas (preferably in the centre of the marked area) if all of the devices in the chain of colour processing equipment were adjusted correctly.

[@TimRrstrm]

YIQ obviously stands for 'you is questioning'

[@Roxor128]

Don't forget the system used by your Cold War rival: SECAM. Interesting fact about SECAM and PAL: The brightness parts are mutually-decodable, but the colour is not. In divided Berlin, the West used PAL and the East used SECAM and both sides would tune into each other's transmissions in black and white and their own in colour.

[@sebastiangorka200]

SECAM is pure nostalgia

[@Roxor128]

@@sebastiangorka200 Or perhaps _Ostalgie,_ for the Germans.

[@berndhutschenreuther8342]

Usually we had two-system-decoders decoding SECAM and PAL for a long time. Not in the very beginning. But soon after ...

[@MattMcIrvin]

The Atari 2600 game console had a reasonable PAL variant, but the SECAM version of it had a half-assed implementation of color encoding that gave most games silly and garish color combinations (the color was just determined directly by the luminance value!) I've always figured that was why competing consoles like the Odyssey^2 (or Philips Videopac, I guess) that presumably went to the effort to do this properly were much more successful in France.

[@melkiorwiseman5234]

NTSC = Never Twice the Same Color SECAM = Something Essentially Contrary to the American Method PAL = Peace At Last 😄

[@KlausWulfenbach]

"I'm sorry, this was only supposed to be two parts." No need to apologize for giving us yet more fantastic content. :)

[@ProfessorYana]

Hear, hear!

[@WPL410]

Make some Gmod content Tinkerer

[@Speedster04_BRA]

@The Mad Thinkerer *says that* - *watches the CED saga* - *takes that back*

[@vk3hau]

NTSC - Color PAL - Colour

[@JulianAlpsNews]

SECAM - Couleur ;) (Although it should really be "PAL - Farbe," since the system was invented in Germany.)

[@anonytube1679]

NTSC = Never The Same Color

[@justanotheryoutubechannel]

vk3hau PAL - Good Colour

[@firefish111]

Yeah right

[@robertoricardoruben]

@@JulianAlpsNews SECAM = System Essentially Contrary to the American

[@lawrencedoliveiro9104]

6:53 Actually you’ve got it upside down; higher signal levels correspond to lower luminance levels (darker), while lower levels are higher (lighter). It was done this way so that transients caused by interference would random dark dots instead of random light dots, and the former were considered to be less noticeable.

[@eac-ox2ly]

Damn, that's hella clever

[@TheXGamer969]

I believe that the inversion is a property of the picture tube itself, what you mentioned was just a side effect of this.

[@nakayle]

@@TheXGamer969 No, in analog TV, higher modulation creates a darker picture until it gets to about 72% at which point the picture is totally dark- this is called "blanking" level. Modulation beyond this point is for the sync pulses. But none of this applies to our current digital TV system which is entirely different.

[@davidg4288]

@@nakayle A totally black picture would overheat the final amplifier tube on the transmitter, especially UHF transmitters. The transmitter would convert a solid black picture to gray after a certain time period to prevent damage.

[@brentfisher902]

Actually, if you use negative modulation, it is easier get the voltages going to the video input at the right level, rectify the RF signal TV transmission after filtering out the audio carrier and pass it through a low pass filter followed by a slow automatic gain control, this way the most negative edges of the synchronization pulses are at a known level and it can have a slightly more positive threshold on the resulting signal which is the correct way up, once the new data point is more positive than that threshold while the previous point is lower than that threshold, an electronic timing circuit looks for the back porch signal where the color burst is after filtering out the color burst, and the gain is further adjusted until that time slot of the signal is at the regular NTSC CVBS voltage level. At the same time, the color burst that was filtered out gets its peak-to-peak amplitude measured which is used to set the color gain. If the TV signal was transmitted the right way up, it would be much harder to get the levels correct as it would be dependent on the brightness content of the transmitted TV image.

[@antonnym214]

I was four in 1964 when we had an old Admiral black and white TV and our neighbor got a color TV. I remember watching Popeye cartoons on their TV and really appreciating the color, and then watching on our black and white at home and still IMAGINING what all the colors were from what I had seen next door. Good videos! Definitely, yours is one of the two or three best channels on the -tube!

[@crist67mustang]

Wonderful experiencie. I was 11 in 1978 when color tv came to my country, American NTSC. Greetings from 🇨🇱

[@antonnym214]

Interesting to think of a color TV set as an analog computer. Those boys were brilliant in those days. Did so much with so little! The car automatic transmission was also an analog computer. Great stuff!

[@jeromeglick]

This is the best concise explanation of color TV that I've come across to date. Attempting to explain QAM is a tall order and you nailed it! By the end of high school I understood how black & white TV worked. Then in college I worked at a TV master control production center with racks of tape machines, TBCs, waveform monitors and vectorscopes, and it took me until the end of college to finally understand how color TV worked! It all comes down to math, algebra, physics, the manipulation of electrons. The more you study it, the more you realize how scientifically beautiful it is ....if you don't hurt your head trying! It's like a dream that it works!

[@notthatyouasked6656]

My family got our first color set around 1968 or 69. My Mom bought it and had it put in without telling anyone. I came home from school one day, walked by the usual TV viewing area, and was completely speechless at the sight of a color TV. I got my first personal TV as a Christmas present at age 13-ish, a 13-inch diagonal black and white set. I didn't own a color set myself until I was 21 or 22 - it was my college graduation present.

[@TechnologyConnections]

11:22 THOSE BEZELS! Hey, I'm sorry about the audio on this one--there are some annoying mic issues here and there. I experimented with a new method of securing the mic cord which really backfired. Oops.

[@noahmccann4438]

Technology Connections Those bezels indeed! I’m watching this on an iPhone X and can no longer complain about its comparably minuscule ones. It’s also pretty remarkable that we can learn about an old TV, while viewing it on an OLED screen capable of HDR. We’ve come a long way since shadow masks and QAM! Excellent video by the way! The diagram showing how the color burst worked was very helpful, I was originally unsure why the TV couldn’t *just* use its own oscillator, but with the help of the diagram I realized that could be out of phase with the one used to encode the data.

[@gregorykhvatsky7668]

T H I C C

[@MostlyPennyCat]

Technology Connections And I'll write it here as well, just in case. If you're going to get into PAL, please please do Teletext. It was an absolute phenomenon, digital interactive television in the 1970s! We had chatrooms, magazines, games, automatic VCR programming! We could even download software off of it for our home computers!!

[@shimes424]

I'm on a 4 yr old OLED with higher pixel density and better colors. Too bad Samsung decided it was too expensive with not enough return value. Great that the iPhone went OLED, mad that Samsung can but won't make a better screen for it.

[@Yes-ng6rf]

Another fantastic presentation, it is mind bending how they achieved all that back then with such tight restraints on bandwidth ETC, I laughed and threw up the first time I was subjected to (No Two the Same Colour), TVs with purple skies and orange people, but seeing that it was designed to operated on valve technology by people using slide rules is simply amazing. I was also unaware that NTSC was an abbreviation for the governing body, live and learn, again, thank you for the cliffhanger series.

[@Jaymac720]

He addressed this lightly in a later video, but for those curious about QAM, there’s one crucial aspect that was left out. The signal has to kinda squiggle around in order to get a phase offset from the colorburst. The squiggling around of the very top of the signal will be enough for a color television to determine the phase offset but not enough that it’ll bother a black and white television. Furthermore, it’ll apply a sin function and a cosine function to that phase offset in order to determine I and Q respectively. The sin function is called “in phase” and the cos function is called “in quadrature” as a cosine wave is 90° out of phase with a sine wave. That’s how modern QAM works to this day. If I got anything wrong, feel free to correct me. This is just my best interpretation from reading websites and watching videos on QAM

[@sjcb]

Some people were just born to teach. I've learned a LOT from this guy

[@johnrehwinkel7241]

In many color sets, the final dematrixing was done in the picture tube itself. The luminance signal would be sent to the cathodes, and the color difference signals to the grids (or, the other way around) and the CRT would do the final multiplication as well as displaying the result. The cleverness displayed in those color sets was amazing.

[@videolabguy]

@11:33 The reason the complexity of the consumer televisions escalated in the late 40s was due to all of the developments during WWII. This dumped thousands of well trained electronics technicians into the 1945+ job market as well accelerated electronics R&D thanks to those new fangled computing machines and other fundamental breakthroughs.

[@scottlarson1548]

The complexity was simply because it is complicated to decode an NTSC color signal reliably and they hadn't discovered the shortcuts yet. They weren't making them complicated for fun. Also, what were these "new fangled computing machines" in the early 1950s? We were still using mechanical adding machines well into the 1960 and electronic calculators weren't common until the early 1970's.

[@videolabguy]

As an engineer, my first order designs are based on the overall block diagram of the system. It usually contains a lot of redundancy. But, it allows each section to be studied in detail before another engineering pass goes after simplification and cost reduction. The NTSC color coding is traceable to the studies of human vision which was very precise but not very efficient from a manufacturers point of view. RCA had multiple analog computers doing mundane tasks that were monstrous tasks before. Like tuning antenna arrays. SEE: Antenalyzer in George H. Brown's famous book, “and part of which I was - Recollections of a Research Engineer”. These were still electric slide rules. But, they were greatly increasing productivity. I believe that engineers at RCA and Hazeltine both used analog computers to model various circuits that ended up in NTSC. Yes, correct. Most of the complexity was placed on the broadcaster end of the service to make TVs "simpler". The same was done with the arrival of HDTV.

[@scottlarson1548]

Take a look at the schematic of an RCA CT-100. It's 36 tubes but I see no redundancy or needless complexity. Decoding NTSC was difficult back then, especially when you have to make separate filters for the I and Q signals since they have different bandwidths.

[@karlscheel3500]

videolabguy, I studied electronics back in college. There, I remember being introduced to the "integrator" and "differentiator" circuits, which were simply a series RC-circuit powered by a pulse-waveform. If the output with respect to ground was taken across the capacitor, you had an integrator. If the components were reversed, and it was taken across the resistor instead, you had a differentiator. They were both a type of analogue computer. Imagine that: such simple circuits being able to perform calculus! :)

[@xsc1000]

Look at 60's TVs. They decoded the same NTSC, but the complexity is reduced. Sometimes it's reduced too much (to cutting cost as always), so the color fidelity was in fact worst then on CT100 (no correct DC restoring, subtracting Y and chroma directly on CRT, no IQ delay line...).

[@highvoltagefeathers]

These videos have the highest density of digestible, interesting information anywhere on youtube.

[@DavidMcGuireRaceVids]

The stuff that was done with analog signals is nothing short of amazing.

[@deusexaethera]

Regarding the Y, I, and Q ratios: I instantly recognize the "Y" ratios as the relative sensitivity of the rod cells in the human retina to different colors of light, which determines how we expect greyscale images to look. From what I can tell, the "I" value determines the skewing of the color towards orange or cyan, which allows for Y+I to form a 2-color TV signal. The "Q" value further skews the "I" signal to allow for magenta and lime-green colors, and by combining magenta (red + blue) , orange (2 parts red + 1 part green), lime-green (1 part red + 2 parts green), and cyan (green + blue), you can get the full spectrum that humans can see. Yes, green is heavily referenced, but that's because human eyes are primarily sensitive to green light, with red and blue only providing color, not providing the majority of image detail.

[@arnellpatscott6948]

And that's what it points to on a Vectorscope.

[@arnellpatscott6948]

Good book to read about color (colour) is "Full Spectrum" by Adam Rogers

[@leonerduk]

Oh wow, thank you :) For the past 20 years I've always thought that the colour burst contained the colour information, that awkwardly the TV set would have to remember for the rest of the scan line to apply later on. You managed to explain that it isn't, and how it actually works, in a way I've never really seen anywhere else before. Big thumbs up.

[@shmehfleh3115]

Excellent video! As you mentioned, the practice of broadcasting shows in color predates most people actually owning a color TV by at least a decade. Well, the same is true for production studios: The earliest color TV broadcasts, if they were recorded at all, were recorded using B&W kinescopes, so no color copies survived. However, the extra info a color TV signal contains introduces a visible pattern of interference on old B&W TV equipment, including those kinescopes. Today, this regular interference pattern can be analyzed by software and used to reconstruct the original color recording. I know a few early Dr Who episodes have had their original color restored using this technique.

[@catlover10192]

That's what the "chroma dots" he talks about at the end are.

[@Meow_YT]

It helped me to understand all the colourspace stuff to visualise it, seeing the RGB and YIQ cubes and how they related to each other really helped me more than the maths.

[@lracrellim2711]

Ha! It hertz to talk about it.

[@tjja7321]

i read that exactly when he said it!

[@CC-ke5np]

Now that Mega-Hurts!

[@lracrellim2711]

C64 C64 LOL

[@creature_of_fur]

Thought _exactly_ that

[@jpaugh64]

He does his own subtitling, instead of relying on KZbin auto-subtitles! I really appreciate that.

[@MostlyPennyCat]

If you're going to be doing PAL, please eventually get round to doing Teletext. It was a phenomenon here, digital interactive television in the 1970s!

[@kaitlyn__L]

not really interactive, purely receptive: the tv had to wait until the specific page you requested was broadcast, and then it loads into the display buffer of the decoder chip. of course, plenty of interactive-seeming things could be established by setting different choices to different pages. you couldn't send any data back directly on the service - though bt ran a service with very similar graphical specs over the phone lines, which did allow interactivity, with their hardware.

[@Ben-fr8gi]

+Kit Vitae I would say it's interactive in the sense you could display the page you wanted when you wanted, so although it was repeatedly broadcast, the user experience was one of choosing your own content as desired. Not greatly interactive by modern standards, but a big step towards the idea of on-demand digital information at the time. It took us a decent step beyond just changing channel as our only way of interacting with the TV.

[@Architector_4]

Ooooh, Teletext. That was an awesome thing to fiddle around with as a kid. Out of all channels we had, it was only the boring news one that actually broadcasted teletext stuff. I remember going around those digital colorful pages. There was one that would update continuously, that simply displayed white text on black background below, being an effective way of showing subtitles. Another one was on a teletext channel 888 that was just all kinds of tests and stuff, with rainbows, and flashing text, and every letter... It was hypnotizing just looking at that, while news woman in the background was reporting whatever happened... ._.

[@CZghost]

@@Architector_4 In Czech Republic, the only station that supported Teletext CC in the beginning of PAL standard was national television company CT (Czech Television), with commercial stations latching later on. Teletext is still used today in Czech Republic, though its interpretation is now better handled with digital signal replacing an old analogue TV we are kind of still used to. If you want to display CC, there are now two ways of doing that: eighter using Teletext, or it is transmitted as subtitles provided by the station for the program. I've never actually tried the Teletext CC, not even the subtitles. However, with new standard of digital broadast, there's now possibility to broadcast more than one sound carrier, allowing us to change languages of the spoken word (mainly Czech and English when it comes to our country). With subtitles being able to be turned on, it's perfect for watching movies in original language with Czech subtitles on. Only if all stations actually used that feature. Now only Czech Television actually uses this feature in foreign programs (except for Slovak programs 'cause there's no need to provide translations), I wish that commercial stations actually grabbed this opportunity to provide audience the ability to choose the language, without the need to enforce it by legislative means.

[@MaxOakland]

It’s so cool that you guys had that. I wonder why we never did in the US

[@antoniomaglione4101]

The simplest way I were able to describe color TV in one of my lessons, was that every color element was described with a complex number - basically, a number with an angle. The number represented the intensity (luminance) and the angle - between the zero reference provided by the subcarrier and the transmitted value - represented the color. As matter of facts, the color was encoded as a phase difference - an angle. It is obvious that the compatible color system was defined by a roundtable of geniuses. Telefunken improved the NTSC by switching the phase of 180 degrees at every line, requiring a color delay line. Thanks for the toughful video.

[@lidarman2]

Holy cow. I do signal demodulation all the time in my line of work but never knew this. It is amazing all the stuff that has come from signal modulation and mixing. All those retired engineers and laughing at how easy things are now usign digital signal processing.

[@argvminusone]

Digital video compression (MPEG, etc) is pretty wild too, though, isn't it?

[@Tmanaz480]

You can see now why NBC worked so hard to be the first full color network, even when so few people owned color sets. NBC was owned by RCA, the electronics giant who made and sold color sets.

[@Phredreeke]

fun fact: the field-sequential color system did find use in the later moon landings, although it ran at the standard 525/29.97 rate and was converted from field-sequential to standard NTSC once received on Earth.

[@jerryg50]

Excellent explanation about NTSC Television system. I worked in colour TV systems during the 70s through to the early 2000s. The circuit design for this system is complex. I like your simple explanation. The average person would be lost to understand this in any form of detail. This system was genius.

[@ErikPelyukhno]

I’m a huge fan of this channel, you find the right balance of piquing my interest in the technology and helping break down how the technology works on a deeper level than most dare to do. Thanks for making us viewers feel not patronized and really trying to explain difficult concepts in entertaining yet informative videos! I can’t wait for part 3.

[@ninoporcino5790]

Finally a clear explanation of how it works. I have been looking for a such understanding for a while, thanks! So in short, the luminance signal is no longer AM modulated but instead it's amplitude+phase modulated; which appears as normal AM when decoded by old B/W TV sets. And the short color burst during the back porch is used to create a clock reference for the incoming line. Cool trick.

[@JoshBrinson]

Wow! We really took color tv for granted. I had no idea how complicated it was, and this really speaks to engineering back then. And they did this with pencil and paper, not computers. Absolutely amazing.

[@ProfessorYana]

This is quickly becoming one of my favorite channels on KZbin. So much interesting information on the history of various bits of technology!

[@thenerdmanual]

Your show reminds me of The Star Hustler with Jack Horkheimer...I mean that as a compliment. Really loved the way he explained things, and enjoyed his personality. Same goes for you.

[@trondmm]

5:50 So, the colorburst frequency is 3.579545 MHz... That raises a question. The Motorola 68000 processor used in the NTSC version of the Amiga 500 runs on a clock frequency of 7.15909 MHz, which is exactly twice of that. Do you know if there's a connection, and if so, what that is?

[@MattMcIrvin]

Yep. Colorburst crystals were common and cheap and a lot of these early personal computers used colorburst crystals as their processor clocks, with a circuit to divide or multiply the frequency by some simple factor.

[@jordanhazen7761]

Besides colorburst crystals being common and cheap, any computer generating color composite video needed that specific frequency anyway, so using one crystal for both purposes (but dividing it differently) saved money. I know the A500 composite output was monochrome-only, but original Amiga 1000's had color composite out. PAL and NTSC Amigas ran with slightly different CPU clcoks, and just about everything in Amigas (audio sample fetches, etc.) was synchronized to the video timing. This is also why the original IBM PC ran at 4.77MHz, which is 14.31828MHz/3, while 14.31828/4=3.57MHz color burst (needed for CGA card's composite-out - unlike the higher-res monochrome one, this board had no crystal of its own! The pre-division 14.31828MHz clock was present on a bus pin going to all slots.

[@brentfisher902]

If you take a reference frequency and pass it through an amplifier with the volume cranked stupidly high, you would get clipping of the signal which would result in multiples of the frequency as harmonic distortion. Add a bandpass filter and a buffer amplifier to that and you get your higher frequency.

[@CaveyMoth]

I'm so addicted! I'm binging all of your videos! Color me educated!

[@c182SkylaneRG]

"It Hertz to think about it..." :) I see what you did there.

[@chrisko2337]

I just recently stumbled upon your channel. I must say, it's very informative and yet not boring. Keep up the good work, you definitely deserve some more viewers. Greetings from Germany.

[@oxillerate7992]

Could you do a series on Plasma televisions? I remember being in awe of how amazing the Pioneer Elite picture was, I must have sat for a few hours watching the promo movies at the Home Theater store over and over when they came out. Must have been...2001...2002...?

[@WarriorsPhoto]

You have my respect for the research done on this older tech. It’s amazing to me, how much tech we are still using this day.

[@kamrongrant]

Don't be sorry mate! Your in-depth content excluding no seemingly mundane detail is what I come here for. When I have a tech-itch, I need it scratched with someone passionate explaining every minute detail to a tee. It is kind of therapeutic learning, soothing my brain back into smart after coping with all the stupid around in life lol. Keep up the awesome work! :)

[@wilbur9416]

Thank you for the technology classes on KZbin. Always great lessons even when sometimes it is just a refreshing course

[@bitrot42]

Nice job tackling a very complex subject! It's pretty amazing they got this stuff to work using 1950s analog technology. "Compatible color" was a huge boon... right up to the end of the analog era, you could still buy little black and white CRT TVs.

[@jordanhazen7761]

yeah, It's a shame ATSC has rendered those impractical (the entire MPEG TS must be demodulated to get any video at all.. no opportunity any more to leave out significant circuitry), along with mobile reception in general due to its multipath sensitivity.

[@brentfisher902]

@@jordanhazen7761 I said this a zillion times...but I will say what Karl Marx says...capitalism destroys everything it touches.

[@crazyivan030983]

PLEASE, I AM BEGGING YOU!!! Do more info graphics :) You have very good calm way to explaining things :) greetings from Poland :)

[@dansanger5340]

I'm so old I remember the bad old days of early NTSC. When I was a kid, we had an older NTSC TV that was always causing problems. The "vertical hold" would constantly unlock and cause the picture to scroll continuously. If you couldn't fix it by adjusting the vertical hold knob, then sometimes banging on the side of the TV would work. The color was usually bad, so you had to constantly adjust the tint knob. When things got really bad, you'd open up the back of the TV, unplug all the tubes, and then take them down to the drugstore where they had a tube testing machine with a bunch of sockets on top. You'd take your tubes one at a time, plug it into the right socket, and a meter would tell you if the tube was still good or not. You'd buy replacements for the bad tubes, then take everything back home, plug them back in, and hope it worked better.

[@MattMcIrvin]

I think my parents got a solid-state TV pretty early on, but I remember drug stores all having those tube testers, with the big cartoon picture of a TV with a miserable face and a thermometer stuck in its mouth captioned "TV SICK?" And the old sitcom joke: "Did the repairman look at the TV?" "Yes, he said it was just a tube." "Oh, what a relief." "The picture tube!!" WHOMP WHOMP WAH WAAHHHH

[@melkiorwiseman5234]

6GV8 was the usual culprit. 😆 But honestly, I can remember that being the cause of frame collapse and vertical hold problems in multiple B&W TVs and I doubt that changed a lot with color TVs.

[@speedyink]

Great video! You did a good job explaining what is otherwise a ridiculously complicated to explain broadcast system. I don't mind the extension of the parts, all these videos are very interesting.

[@fluffy_tail4365]

I am actually happy about having an extra part, especially for history bits. I know the technological part so hearing the crazy stories surrounding their invention is the most fascinating part to me.

[@Waccoon]

Really great info, as usual. I'm surprised you didn't show a screenshot of an oscilloscope, though. Watching an NTSC line (or any other analog display signal) on a scope is a real hoot. Some day you should do a segment on computer modems. I'd love to see your take on the history of the devices, and all the complicated handshaking that goes on between the transceivers. Let's see how many whippersnappers have heard all those wonderful sounds!

[@Natalie-ez1zc]

Support

[@kaitlyn__L]

Second the desire to see moving footage on a scope. I've seen test patterns on them before though. It's interesting that you mention modems, since QAM was involved in those too - bringing the throughput to above 9.6kb/s while remaining at 9600Bd.

[@davidm.4670]

@@kaitlyn__L I remember 100baud ...- dialup !

[@seanrodgers1839]

I started using a modem in the early 80s. 300 baud, slower than you could type, or read.

[@Felixicity]

Thank you for these videos! They're really fascinating, and I feel like I learn a lot about older technology I didn't. I especially liked your christmas lights video. And it's a little silly, but my cat loves your videos too--he'll usually hop up so he can watch when he hears your voice.

[@TheAnubis57]

We brought a used color television from hotel chain sale in 1977. When plugged in and turned it on I was completely blown away. For 17 years I only saw black & white.

[@johnrehwinkel7241]

I remember reading about color matrixing and working through the circuitry, finding the Y signal, and R-Y and B-Y and wondering where the circuitry was that combined them. It turned out the picture tube itself was doing the multiplication! The luminance signal drove the cathodes, and the color difference signals drove the grids, producing the individually modulated R, G, and B electron streams directly in the CRT.

[@stevethepocket]

The color burst explanation completely went over my head, but it reminds me of the (definitely easier to understand) wizardry that enabled backward compatible stereo records.

[@techmage89]

Pocket Fluff Productions As I understand it, basically the color burst is fed into a resonator which continues to oscillate at the same frequency and phase for the duration of the line. This recovered carrier signal can then be used to demodulate the color without the color carrier actually having to be present during the active part of the line.

[@Ben-fr8gi]

Yes, that's what I understood from it. Despite its name, the colour burst does not contain the colour information. It is a reference frequency (aka carrier signal) needed to decode (aka demodulate) the colour hidden inside the video signal. Like all carrier signals it is is a constant repeating frequency. In rough terms, a mathematical function combines the modulated colour signal from the video and this carrier signal to extract the intensity of the three colours. Of course, to combine the two - both must be available at the same time. Rather than constantly transmit the carrier signal however - which would require more bandwidth and wasn't an option, a small section of is inserted in to the signal as the colour burst. Since this carrier signal is needed all the time, the circuitry in the set will continue repeating it until the next colour burst. Since it's a constant frequency, someone might wonder why broadcast it, the set could just produce its own? Well, it can, it does exactly that between the colour bursts. The need for the set to have it transmitted is that the carrier frequency must line up with the video signal. Because the carrier signal is always changing in amplitude (basically, a different number at any moment in time) the output of the function applied to the modulated video signal depends on the amplitude of the carrier at the moment in time it's calculated - so both carrier and modulated signal must come from the TV broadcast to ensure they line up. If they go out of synchronisation the function creates the wrong output and the colour is messed up. The challenge was to insert colour in the existing black and white signal while still retaining backwards compatibility, so some part of the existing signal has to be used. One challenge was to modulate the video signal to add colour information, the other was to transmit the carrier signal in a way that didn't adversely affect backwards compatibility. The idea was pretty ingenious, just briefly insert a section of the carrier signal in to the broadcast, and since it's a repeating signal that does not change frequency the circuitry in the set can keep repeating it after the colour burst has stopped being transmitted. But when to transmit it? They inserted the colour bust in to a section between lines of video. It's a quiet moment when there's no video information to corrupt, and happens often enough that almost as soon as you turn a colour TV on it will have a colour burst available to start performing the mathematical function and displaying the image in colour. Meanwhile, existing black and white TVs don't notice it since it's in a region of the signal they don't use for information displayed on screen, so it doesn't adversely affect their function. The modulation to the video signal does in fact show up on black and white TVs, but it's good enough imperceptible. Hopefully this makes sense?

[@allanrichardson1468]

And backward compatible stereo FM radio. The sum of the two stereo channels is transmitted as the main audio signal. The difference is double sideband suppressed carrier amplitude modulated onto a 38 KHz subcarrier, and a very low amplitude 19 KHz pilot subcarrier is added to provide a phase reference. All three components of the stereo signal, along with special SCA audio (either paid background music or reading for the blind), if present, are added together to modulate the FM carrier. The legacy (or extremely cheap) mono FM receivers ignored everything above 18 KHz and played the sum of the left and right channels. The stereo receiver splits the combined signal into audio below 18 KHz (the sum), the pilot subcarrier at exactly 19 KHz, which synchronizes the 38 KHz oscillator, which, added to the DSBSC modulated signal from 20 to 56 KHZ, demodulates into the difference between the channels. Sum PLUS difference gives one channel, sum MINUS difference gives the other, fed to two amplifiers and speakers.

[@brentfisher902]

On a mono disc record, the needle moves from side to side to record the sound. For stereo, the needle can also simultaneously move up and down...the up and down movement carries the difference that stereo has from the mono signal.

[@LittleMissDeath]

Every time my dad and I have a conversation about old TV sets, he always has to add (excitedly, as if it's still new information to him) the fact that he didn't know Star Trek was filmed in color until he watched it at his grandmother's house because she had a big ol' color TV.

[@glennso47]

Superman was photographed in color long before people had color TV sets. (The George Reeves version of the show. )

[@joevignolor4u949]

As an experiment I once pulled the final IF amplifier tube out of a color TV to kill the video signal. Then I placed a functioning B&W TV set next to it. Then I connected a jumper wire from the video output circuit of the B&W set to the video input circuit of the color set. Guess what happened? I got a normal color picture on the color set using the video signal coming from the B&W set.

[@brentfisher902]

This proves that a color broadcast is simply a black and white broadcast with color shoehorned in. Aside from encoding the color into a composite NTSC signal, no modification is needed at the TV transmitter or antenna...just playback the color CVBS signal at the test point where you would ordinarily playback a black and white one.

[@Hopeless_and_Forlorn]

Great explanation, as usual. One correction to make: At 7:00, the luminance signal shown is actually inverted in function. The higher the signal level, the lower the resulting beam current, and the darker the trace. When I learned television repair in the 1970s, I was gobsmacked to learn how the RCA system worked, and particularly that it was completely compatible with the black and white system. It was like learning that some alien super technology had taken over between 1941 and 1951.

[@jordanhazen7761]

Yes, this inverted negative modulation is beneficial for sending horizontal and vertical sync pulses ("blacker than black") at maximum power, so that even with very poor reception the picture won't usually tear or roll. Also, any impulse noise spikes from arcing motor brushes, etc. will tend to show up as black dots rather than white, less noticeable in most scenes. This was part of the original 1930s B&W RS-170A standard, carried forward into NTSC, and also used in PAL, but I believe France's SECAM (and its B&W predecessor) used positive modulation, as did the old British 405-line VHF system.

[@Nickferal]

This is why I swear I could guess colors in an old black and white TV, two different colors are actually different shades of Gray in b&w

[@MattMcIrvin]

The luminance values of those colors will display as grays. On very early black and white TVs showing a color broadcast, in regions of the most intense color, the chrominance signal would sometimes manifest as stray dots on the screen. B&W TVs made after the introduction of color TV had some circuitry to just filter that out, so you'd just see the luminance information.

[@StreetComp]

6:01 - Someone needs to learn After Effects 😂 one of my favorite things about your videos is these little messages you briefly flash onscreen. You have one of the best channels on YT, thank you and keep up the good work. Btw, if anyone actually reads this message AE takes some real time and effort but is very much worth it, can make a career with this skill if want to

[@Jaymac720]

I’m an engineering major and QAM still flies over my head. Granted I’m only a sophomore, but I’m going too need to understand this

[@harpoonlobotomy]

I've been listening to this in the background while I work and I have absolutely no idea what's being explained, but goddamn if I can't keep listening anyway.

[@AtrumNoxProductions]

Better a few long videos then one short not as informational video. This is really cool! I look forward to another one!

[@1990chrism]

Do you plan on doing one of these for PAL and SECAM ?

[@TechnologyConnections]

PAL will be addressed in my odd-and-ends video. I hadn't planned on talking about SECAM but perhaps I shall. SECAM is interesting but also weird.

[@1990chrism]

Fantastic, I look forward to watching :)

[@TheChipmunk2008]

Ah yes, NTSC: Never Twice the Same Color, SECAM: System Essentially Contrary to the American Method, and PAL: Perfection At Last

[@justmytake4152]

Are you also planning to look at system N (PAL-N) i.e 625-line in your odd-and-ends video as I have always wondered the choice for 6 Mhz channels while excluding system A (405 lines - 5 MHz) most other analogue TV system used 7 MHz (VHF) or 8 MHz (UHF)? Then the use of Vertical blanking interval particularly teletext might be interesting since its apparently in the 1970s it resulted in the BBC asking TV manufacture to develop a better TV remote control.

[@Redspl]

TheChipmunk2008 but SECAM is more compatible with PAL than NTSC, you know? (no audio(it's in another part of the signal) and no color (other method of generating it), but b/w video will appear)

[@RapiBurrito]

Man your enunciation is so relaxing... I can bake and watch these visa all day long!

[@jonathanpullen7439]

Best explanation of NTSC / compatible color I have ever seen.

[@ryanyoung2421]

"CBS's sequential color wheel system" How many takes did it take for that to come out so smooth?

[@robfriedrich2822]

11:50 Olympic games - in West Germany many people bought color TVs. In East Germany they tried to force, but the program most people want, was black and white on East German color TVs, because it was PAL and East Germany used SECAM

[@strakhovandrri]

LIFE COULD BE DREAM LIFE COULD BE DREAM I have YIQ as my homework assignment, and I am so glad that watching my favourite channel is not only a pastime but also is a process of University learning.

[@leandro842]

Here in Brazil they used a PAL-M color system before the digital SBTVD/ISDB-tb standard, the PAL-M analog color standard was created in 70's and consist in a video analog type M signal (525 vertical lines like in NTSC) but rather than NTSC color standard, PAL-M used RGB to encode colors (the same system used in european PAL) they reached this using a 6 MHz bandwidth (the same used by the grayscale television). Because it, brazilian standard was the only one to use 60Hz vertical frequency with full color range. Something curious about it is the fact that in PAL-M (where PAL is the color system and M de analog video standard) the colors was encoded using Phase Alternating Line, it prevents dissonance of the chrominance signals, therefore in PAL-M image was free os "ghosts" around the outlines a common phenomenom present in NTSC standadard, at the same time the fact of PAL-M worked at 60Hz avoided the flicker present in european PAL and SECAM standards.

[@leandro842]

In fact it uses the PAL-M system as well, the Star One C2 (its name) will be operational until 2023 in C band and currently it transmits 30 open channels with national scope. In the same satellite there are KU band transponders for private operators and C-band transponders transmitting digital open channels in the ISDB-Tb / SBTVD (FTA) system together with the Star One B3 and B4 satellites.

[@k.chriscaldwell4141]

I'm old enough to have lived in a home that had only *A* black and white TV. That is, ONE TV, and it was black and white. No remote either. After we got a color TV -- _"Movnin' on up..."_ -- I always wondered how the color signal displayed just fine on our old black and white TV. Now I know. Incidentally, my family was one of the first to have a microwave oven. Several time we hosted people in our home to show them how it cooked potatoes, boiled water, cooked TV dinners (on a plate), defrosted meat, softened ice cream, heated leftovers, cooked popcorn in a brown paper bag, etc. Strangers would sometimes knock on our door and ask if they could see it in action; Usually with a TV dinner or potato in hand--I kid not.

[@brentfisher902]

* Dire Straits intensifies *

[@Gribbo9999]

I think you showed the Sony Trinitron red blue and green bars. The early sets had triangular groupings of RGB phosphor dots. I am old enough to remember the improvement that Trinitron brought. I was also in a PAL country so had less of the Never The Same Colour (NTSC) issues.

[@nickwallette6201]

Thanks! I'm deep into studying retro game console hardware, and the NTSC video generation bits were making my head foggy. This actually really helps understand what's going on, and now the squiggles on my scope almost make sense.

[@brentfisher902]

You really want to make your brain fried? Look up how the Apple IIe computer displays color...the pixels are exactly 1/4 (or 1/2) of a color subcarrier cycle apart, resulting in text with bright color fringes when the color burst is being transmitted...

[@JimTheZombieHunter]

Although I'm old as heck, full tube only analog colour sets were pretty well gone by my teens, things to be parted out for tube projects - hybrid sets had been established, and by the 70's IC solutions had shrunk the chassis to small(ish) PCB's instead of multiple steel behemoths. I do recall opening a 50's ('58?) round tube colour Admiral with electronic ultrasonic remote though - that thing must have had over 60 tubes, and a dead 60's RCA in the early '00s which really didn't have that many (the colourburst crystal was in a 9 pin glass envelope!). I never really did understand how they worked beyond the abstract, but it's amazing that they made it happen with such a relatively low active parts count. So easy to slap a microcontroller behind a flashing light now. Why not? In production probably cheaper than a discreet multivibrator, and I fear that the future of electronics may come down to some future great chip war where only a handful of field programmable parts will fulfill pretty well all functions. But I digress. My only point: Although the application now long obsolete - In my opinion early colour broadcast should be a mandatory EE course.

[@pixoariz]

Nice job, highly informative. Plus a tip of the hat to the great work of the late Ed Reitan, who documented the history of color TV, archived over at the Early Television site.

[@betaneptune]

At 10:12, notice the spelling of "lose" in comic-strip panel. I thought misspelling it as "loose" was a relatively recent thing.

[@davidm.4670]

"lose" is escaped - missing - 'loose' is not tightly held = 2 different words ;-)

[@Luxcium]

Nice you are getting better and better man !!! From that regional boy with a jacket to this international man with nice backing (I like your other setup also but you are improving every day it’s nice)

[@utah133]

NTSC analog color sets used to have a service adjustment called "color killer." Adjusted properly, it kept the circuitry from trying to extract color from a monochrome signal. You'd see color "confetti" in the B and W program if it wasn't set right.

[@Cruznick06]

I'm excited you're going to talk about PAL. So thanks for making another video on this topic!

[@lizichell2]

Ffrightfully fascinating how colour is transmitted in the same bandwidth as the black and white signal

[@programagor]

9:03 That was some unnecessary mathphoboa there... QAM is beautiful in its mathematical form, realy just 2 simple lines with a lot of impact.

[@soundguydon]

Thank you for these videos! You're doing a great job, and I for one don't mind you continuing this into another video ;-)

[@static-san]

I can't find if I commented this or not but there's a genuinely interesting reason why the colour information actually fits in in the first place. If you look at the spectral response of a monochrome video signal, you'll see it looks like a comb. It turns out that the chrominace information *also* looks like a comb. The reason for that frequency of the colour burst was so that the colour information fits in between all the monochrome picture information. I daresay some very clever engineers figured this out just with mathematics! But it also means a TV channel transmitting in colour takes up *no more bandwidth* than if they were only sending black-and-white. That was a very big deal! This also explains why you might have heard about "comb filters" in high end analog video equipment a few decades ago. High quality comb filters were for separating the luminance and chrominance information from a composite signal with the least loss of quality.

[@sa3270]

I got so obsessed with this I wrote a computer program that generates and decodes a simulated line of NTSC video. (No, it doesn't do it in real time, this was for fun) The decoder can work with either a single composite (Luma (Y) + chroma) signal, or for better quality, separate Luma and composite signals. It watches for the horizontal sync, and at the appropriate time shortly after that, "charges up" two circular buffers representing I and Q reference loops, with the colorbust data. Y is computed as a running average over the color cycle if not provided separately. I and Q at each sample are computed as running averages by multiplying the chroma value by the corresponding entries in the I and Q reference buffers for the current phase, over half of a color cycle. Once the Y, I, and Q values are in hand, it is a simple matter to apply the conversion formula to convert them to RGB.

[@oskenso]

Thank you so much for these information videos. This in one particular helped me in my quest to write a "CRT" shader which took into account simulating both NTSC and crosstalk on the signals

[@chriswf88]

In the UK, we were in black and white for a long time. But it was very noticeable in the opening credits of 'A man called ironside', they used solarised single colour images, and there was a clear diagonal pattern on the black and white screen, so I knew (not sure how) that these were blocks of a single colour (I was too young to have worked out the electronics behind it). The pattern was the colour subcarrier, as we now know.

[@kevinsullivan3448]

My parents received an 19in RCA 'portable" Color TV as a wedding gift. It was definitely smaller than any other TV I had seen as a kid, and we would take it to other people's houses to watch important televised broadcasts, like the moon landing that occurred on my older brother's birthday in 1969. We had the only color TV on our block up to 1971.

[@Shamino0]

The luma/chroma components are (conceptually) similar to the "joint stereo" encoding used for stereo radio broadcasts, in order to remain mono-compatible. Instead of broadcasting left (L) and right (R), you broadcast L+R and L-R. L+R is the mono audio. L-R encodes the stereo separation. If you have a mono receiver, you just pass L+R to both speakers and you're done. If you have a stereo receiver, you add/subtract them to get left and right: (L+R) + (L-R) = L+R+L-R = 2L (L+R) - (L-R) = L+R-L+R = 2R The YIQ concept is similar, but in this case it is using matrix multiplication instead of simple addition and subtraction. So to extract RGB from YIQ, you just run it through another matrix multiply: R = Y + (0.9469 * I) + (0.6236 * Q) G = Y - (0.2748 * I) - (0.6357 * Q) B = Y - (1.1 * I) + (1.7 * Q) I also don't know how the numbers were derived. The decoding matrix is, I believe, the inverse of the encoding matrix, but I don't know how the encoding matrix numbers were derived. See also en.wikipedia.org/wiki/YIQ#Formulas

[@LakeNipissing]

11:22 . . . For those interested in this vintage RCA CT-100 TV, KZbin member Spatsbear is in the process of completing a full restoration of one of these televisions. In the video uploaded last week, all electrolytic capacitors have been replaced.

[@scottfranco1962]

This is a good overview. The basic underlying trick of color TV, or why they could stuff 3 colors into a black and white signal, was that still images (or slowly moving images) in TV don't really take the full bandwidth of a TV channel, but rather occupy a series of harmonics, or peaks in the signal spectrum at regular intervals. The color signal is fitted into the spaces between those peaks. What happens if the picture is moving? The color actually degrades quite a bit, which you can see if you hit stop during those times. The color becomes blotchy. They are taking advantage of the fact that you can't perceive color very well in any case in a fast moving image. The TV engineers effectively used signal compression, in analog, before digital TV made that standard. The system was pure genius at that time, but of course we are stuck with the re-recorded artifacts of old video to this day.

[@brentfisher902]

The NTSC color subcarrier's frequency is chosen so that if you take the whole composite video signal, make a copy of it and delay it one line period, then add it to the original signal, the result is only the black and white signal, providing that the image doesn't change much from line to line. The difference of those signals would be the color only signal, respectively.

[@Rock4896]

Just pledged on patreon to push you past your first goal! I love your content, keep it going!

[@TechnologyConnections]

Thank you kindly!

[@DanielRichards644]

man the more I learn about tech and having actually built some simple circuits 20 years back in Highschool the I appreciate just how insanely complex bordering on the level of Magic that modern technology is.

[@fredkilner2299]

The colorburst will show making the picture speckly on a very old B&W TV. Most B&W TV's recognized that it is a color transmission and somehow get just the brightness without showing the embedded colorburst frequency oscillation. You can see this for self easily if you happen to have a TV with RGB inputs. There is more than one type of RGB. If you have a composite video signal from a DVD or Blueray or VHS Tape player or HDTV tuner you can connect the composite signal to one of the RGB signals. It may only work for one. Anyway that composite signal is the brightness with the encoded wave at the colorburst frequency. The picture you will see will have speckleness from that embedded wave. It is very noticeable. I first noticed this accidently using the APPLE ][ computer with a B&W monitor which does not recognize color. So especially on low res graphics instead of interpreting the 16 different (sorry 15 since 2 colors are basically the same) colors as different shades of grey it shows the speckliness of the changing amplitude which a color monitor would interpret as color and a smarter B&W monitor (or every B&W TV I ever saw) would see the colorburst and thus interpret as smooth black/white/grey shades. When you implement an APPLE ][ in an FPGA like I and many APPLE ][ fans have done it forces you to become very familiar with how the video works. About half of the logic on the APPLE ][ motherboard is related to generating the video signals. The clock rate of every early 8-bit computer I know is so it can generate the pixels so has the correct pixel clock. APPLE 's 6502 runs at slightly above 1MHz and for every cycle outputs 7 pixels while most other computer run at slightly under 1 Mhz and output 8 pixels per cycle. APPLE Frequency/7=pixel clock. Most others/8=pixel clock.

[@ozrithclay6921]

Just want you to know. You make awesome content.

[@johnmonkus4600]

The Tournament Parade was displayed over the air in theaters with color sets all over the country.

[@dubsy1026]

I am very happy now the series is wrapped up for colour analogue TV

[@randystegemann9990]

I love the shot of the old TV chassis. 11:24

[@randystegemann9990]

@G Guest And that's only after correctly aligning all of the transformers and coils.

[@newq]

I remember wondering any this exact thing when I was little and just getting into electronics. Your videos kick ass! :)

[@kjamison5951]

I love your videos! Takes me back to 1990 when I was learning all about TV signals in University.

[@osmiumsoul9535]

Self aware humor, complex ideas in layman terms. Excellent youtube channel

[@mrflamewars]

Was looking forward to this one. Totally love your work.

[@Austin23191]

Why was this just now reccomended to me? Love it