How to use audio card to decode video? (How this video output is called?)

Question

My digital camera is shipped with this cable to connect it to TV:

When I connect it with that cable to laptop:

I hear audio output on left channel and "hear" video on right channel:

1. What is this video output called? Where can I find its description (number of pixels, timings, etc.)?
2. Is there already programs to decode it? Expected usage: arecord -f cd -r 192000 | some_filter > video.yuv
3. What picture quality should I expect when decoding video in that way?

Answer 1

I was asking if you were trying to hack the whole thing but I saw your tag. OK let's have fun.

You simply want to avoid buying a capture card. Sounds good, but...

You can't sample video from audio card

Simply that. Why? OK I'm going to tell you. Audio signal is sampled at human ear's Nyquist frequencies, which is double the bandwidth the ear can hear (...bad...): 44100Hz. With high-quality cards you can sample at 96KHz or even 192KHz for maniacs or professional applications. I see no reason about sampling at 96KHz, if ultrasound are at 22KHz nobody would be able to hear frequencies over about, say, 30KHz (which is far higher), but still some people claim 96KHz is cleaner in sound than 44.1KHz.

With video, the problem gets bigger and bigger. In order for your camera to transmit all the complex information about motion pictures (thanks Luke for having decoded NTSC with your brain only), you definitely need more bandwidth. Wikipedia says it's 6MHz.

So, definitely, you are unable to sample a 6MHz signal with a device that is capable of sampling up to 192KHz without losing most of the quality and maybe losing visibility of images.

The bandwidth required to analog encoding of a video signal is function of the frame rate (25FPS) and the frequential components in the video, meaning that a person with a fancy dress like this is a high-frequency component that may be drop, instead a plain white background is a zero-frequency component in the image.

[Edit] high-frequency samples in an image mean that the colour/lightning changes significantly across pixels (in a digital image) or across small parts of continous signal (analog signals). The truth is that an NTSC signal is the analog interpolation of an already-sampled signal.

I studied JPEG encoding to tell you about frequencies in video signals.

I hope I have been of help. Grab your pocket and buy your capture card :D:D:D:D:D

[Edit 2] To briefly answer your previous numbered questions:

1. It's analog video. You must know that prior to decoding. Luke said it's NTSC, so it's a standard. I believe it too, so if it's a standard you need to find out how it's encoded
2. No
3. Veeeeeeery poor, I told you why