Talk:Composite video

Illustration

Would it be possible if someone made an illustration or graph of how it is done getting 3 signals (YUV) in one "plug" with only 2 connectors? I'm very curious about that, atleast, but dont nearly have enough background knowledge to do it. anon 10:56, Jun 17, 2010 (UTC)

Noise

The process of modulating RF with the original video signal, and then demodulating the original signal again in the TV, introduces several losses. This conversion also typically adds noise or interference to the signal as well. Other than noise, what are the losses? In theory, one can modulate an ideal carrier, and demodulate it again. The low signal levels of RF (it has to match what one could receive on an antenna) means noise is significant. (You can't escape thermal noise.) Generating a proper vestigial sideband signal is also tricky, but again, it can ideally be done without loss. Spectral errors aren't really losses, either. Gah4 (talk) 16:59, 28 May 2016 (UTC)

In theory, no problem but in the real world it is an additional two steps and the advice to avoid unnecessary modulation/demodulation when possible is sound. ~Kvng (talk) 17:03, 25 April 2021 (UTC)

modern day

The article indicates that Modern day devices with analog outputs have typically omitted the modulator option. I suppose, but a very popular modern day device is the HDTV converter, which most often does have RF output. But yes, I suspect that they are less common on DVD players and such than they used to be. Gah4 (talk) 22:40, 26 August 2019 (UTC)

I have removed the modern day claims. They were dated and uncited. ~Kvng (talk) 17:03, 25 April 2021 (UTC)

Reference to analog TV tuners

this part: "TV sets sold these days no longer have analog television tuners and cannot accept channel 3/4 from a modulator. But because composite video has a well-established market for both devices that convert it to channel 3/4 outputs, as well as devices that convert things like VGA to composite, it has offered opportunities to repurpose older composite monitors for newer devices."

Is nonsense. Maybe analogue TV tuners don't exist in the author's country anymore. But my brand new sony tv has one.

I assume by "channel 3/4" the writer is referring to north america. — Preceding unsigned comment added by 2A02:A442:581E:1:3595:B7C4:F757:236D (talk) 22:19, 16 March 2022 (UTC)

I have made some improvements ~Kvng (talk) 20:09, 5 April 2022 (UTC)

AHD

Should AHD be mentioned as an upgrade for Compsite? Having lot of problems finding a spec. https://www.delko.si/en/kamere-in-video-nadzor/ahd-high-definition-standard.html --Kitchen Knife (talk) 14:12, 20 June 2022 (UTC)

I think AHD is required to establish notability first, as indicated by its own article. I don't think that's warranted currently. --Zac67 (talk) 14:31, 20 June 2022 (UTC)

Of the several competing systems(CVI * TVI) it does seem to be coming out on top, though with so many acronyms and so little data on the protocols themselves I may have missed something.--Kitchen Knife (talk) 19:00, 20 June 2022 (UTC)

Modern updated info and real world context

Harry munday and I have been discussing recent changes to this article on my talk page at User_talk:Kvng#Composite_video. It would be better to have that discussion on this page but maybe too late for that. ~Kvng (talk) 18:47, 2 June 2024 (UTC)

Following on from User talk:Kvng#Composite video discussion regarding the info box and formats.

``So the reasonable compromise taking a day to think about this would be "analogue successors S-Video & Component" and then another row "digital successors SDI & HDMI" if we are covering all our bases.``

I think the info box fields may need to be edited for this change to look clean?

My edit also included the following, which I don't think there is much debate from a practical information perspective.

1. Adding D1 4fsc sampling info in MHz. (makes sence on the D1 side with the info below as to what 4fsc is)
2. Adding a luma/chroma decoded full 4fsc example (via ld-analyse, from public domain cvbs-decode demo samples)
3. Added the sampling non-square pixel numbers for "standard" SD NTSC/PAL, IMX stanards, and the 4fsc frame information that they are extracted from, intestingly this info also applys in context to how the above AHD stuff works in the real world, as it took advantage of the extra dead space that could be used visually at cost of more bandwith and VBI area.
4. Clarity of digital sampled waveform (it looks diffrent on a analog scope so DSO or digital store ossiliscope should be defined)

I should notate that ld-analyse is the only & a open-source viewing tool, broadcast and consumer users have today for properly inspecting digitally sampled composite or s-video style signals in a true visual manner, better then you would/could on a broadcast CRT video monitor with H/V shift mode, and it gives real world example of viewing 4fsc sampled media by the field/frame or line by line.

I think that image should be added and maintained on this page for high educational factor, its essential to add for people who will never touch a CRT let alone a higher end one in years to comming decades, there is also notably a growing world of archives in raw baseband and raw FM format from video tapes to be decoded/viewed by anyone in a fully preserved manner like this.

Have been working on a 4fsc page draft as this info should have been added years ago, so getting that all in order math formatting wise and citations. Harry munday (talk) 21:23, 8 June 2024 (UTC)

Here's the disputed revert. I would prefer to continue to work the issues one at a time.

WRT #1 in your list above, your changes related to 4fsc removes the definition of the term four times the color subcarrier frequency (4fsc). You also add the acronym a few other places where I assume it is technically correct but may not be necessary or helpful to readers. ~Kvng (talk) 14:44, 9 June 2024 (UTC)

4fsc is the common term, it has 3 correct definitions in context of CVBS signal sampling:

1. English - four times the color sub-carrier frequency
2. In actual frequency 14.2MHz NTSC & 17.3MHz PAL (28msps / 35msps respectively in Analog to Digital sampling hardware terms)
3. In non-square pixel values 910x525 NTSC & 1135x625 PAL

I suppose introducing it in English makes the most sense, but it the outer edits cover the visual the math can be worded in better context in relation to the sub-carriers of NTSC 3.58MHz / PAL 4.43MHz.

It should also be noted the Composite RF signal can be sampled from 28msps+ (million samples per second) but the image frame is always sampled at 4fsc this is why math and or visual context needs to be presented together or it's just truly obscure to the lay person. Harry munday (talk) 15:04, 9 June 2024 (UTC)