how is video encoded? at DVinfo.net

Tim Goldman · September 17th, 2006, 09:00 AM

ok, just wondering, it video encoded in a wave form (not .wav the file but wave as in sine or square), or is it an actual frame of info sorta like film. I know digital is 1 and 0's (high quality digital includes 2's (haha j/k)).

So when my tape passes over the head of the machine, what happens exactly?

Emre Safak · September 17th, 2006, 09:26 AM

You have to be very specific about what kind of video you have in mind, because each format is encoded differently.

Tim Goldman · September 17th, 2006, 09:50 AM

ahh ok, see i was not knowing this till you told me. each brand of maker, or each format?

lets say beta cam
and then the good old vhs tape

just to get some ideas

J. Stephen McDonald · September 18th, 2006, 12:16 AM

Quote:

Originally Posted by Tim Goldman

ahh ok, see i was not knowing this till you told me. each brand of maker, or each format?

lets say beta cam
and then the good old vhs tape

just to get some ideas

Not to answer this almost impossibily broad question, but one aspect is, that even digital video signals are laid on tape with an analog carrier signal, although mostly using higher frequencies than analog recordings to do it.

David Kennett · September 22nd, 2006, 11:34 AM

Mistake! see next post.

David Kennett · September 22nd, 2006, 11:53 AM

Here's a little more detail - hopefully not too much!

First the signal (audio-video-whatever) is digitized. Imagine going to the beach, walking out into the water a little, and sticking a yardstick into the sand. You could then start reading the instantaneous depth of the water as the waves splashed against the yardstick. How precisely your yardstick is calibrated relates to the number of bits used per sample, and how often you read the yardstick relates to the sampling rate. An eight bit sample would have 256 discrete levels (two to the eighth power). CDs are sampled 44,100 times per second. Video is sampled millions of times per second.

Let's do a little arithmetic. A 720 by 480 frame of video has (multiplying 720 times 480) 345,600 pixels (picture elements). To have enough colors for photographic color, 24 bits (three bytes) are used for each pixel. This gives us 16,777,216 possible colors (two to the 24th power). Since three bytes are used for each pixel, we need 3 x 345,600, or 1,036,800 bytes for each frame. There are 30 frames per second, so we'll need 30 x 1,036,800, 0r 31,104,000 bytes each second. There are 60 seconds in a minute, so one minute of video would require 60 x 31,104,000, or 1,866,240,000 bytes per minute. multiply that by 60, and you get about 112 GIGABYTES per hour.
These numbers are just getting out of hand!

Enter compression! This is the most difficult to understand. Let's just say we are going to do some fancy arithmetic to reduce the data for each frame (JPEG), and take advantage of the similarity of adjacent frames (MPEG) to drastically reduce the amount of data required. There is always a compromise between the amount of data used, and the quality of the picture.
A remarkable amount of compression can be achieved with little effect on the picture. Try to do too much, and the ominous "blocking" occurs.

Finally, the data is re-configured to allow it to be recorded onto tape, a DVD, or be sent over a network, or even broadcast. This involves a mathematical process to allow some error correction. DVDs and digital cable or satellite are usually encrypted as well so they can't be easily copied.

Hope this helps a little!

Emre Safak · September 22nd, 2006, 12:29 PM

Nice, David, but he was asking about analog video.

Giroud Francois · September 22nd, 2006, 12:40 PM

basically, all signals are recorded the same way.
the signal is filtered (some frequencies cutted) and eventually sliced as component (could be RGB or YC) , then the signal is sliced again by time sample (usually per line), then sliced again (samples per line).
then you can record like any audio signal on tape.
Since the tape speed required for such high volume of data would be too great, usually the signal is recorded diagonally on the tape with a rotating head.
The great thing with analog recording is the data recorded at one place dos not depend from data coming before or after, so a little scratch on the tape will not cause a big loss, but this loss will be hardly correctable.
Additionally, an analog signal would "fade" while a digital signal would "disappear or switch (1 becomes 0)".
So years later, you still can read a bad analog tape just by amplifying the signal, while corrupted digital signal would just be unreadable.
Another good thing with analog signal is they are pretty close from the natural phenomenon they record too.
So recovering an unknown analog signal is pretty easier than deciphering meaningless digital 0 and 1.
you can almost "read" a video signal on a oscilloscope, while reading the inside of a digital file would just give you headache.

September 17th, 2006, 09:00 AM	#1
Tim Goldman Regular Crew Join Date: May 2006 Location: Marietta, GA Posts: 135	how is video encoded? ok, just wondering, it video encoded in a wave form (not .wav the file but wave as in sine or square), or is it an actual frame of info sorta like film. I know digital is 1 and 0's (high quality digital includes 2's (haha j/k)). So when my tape passes over the head of the machine, what happens exactly? __________________ Wish I had something witty to say

September 17th, 2006, 09:50 AM	#3
Tim Goldman Regular Crew Join Date: May 2006 Location: Marietta, GA Posts: 135	ahh ok, see i was not knowing this till you told me. each brand of maker, or each format? lets say beta cam and then the good old vhs tape just to get some ideas __________________ Wish I had something witty to say

September 22nd, 2006, 11:34 AM	#5
David Kennett Major Player Join Date: Jul 2003 Location: North Ridgeville, Ohio Posts: 407	Mistake! see next post. __________________ Dave

September 22nd, 2006, 11:53 AM	#6
David Kennett Major Player Join Date: Jul 2003 Location: North Ridgeville, Ohio Posts: 407	Here's a little more detail - hopefully not too much! First the signal (audio-video-whatever) is digitized. Imagine going to the beach, walking out into the water a little, and sticking a yardstick into the sand. You could then start reading the instantaneous depth of the water as the waves splashed against the yardstick. How precisely your yardstick is calibrated relates to the number of bits used per sample, and how often you read the yardstick relates to the sampling rate. An eight bit sample would have 256 discrete levels (two to the eighth power). CDs are sampled 44,100 times per second. Video is sampled millions of times per second. Let's do a little arithmetic. A 720 by 480 frame of video has (multiplying 720 times 480) 345,600 pixels (picture elements). To have enough colors for photographic color, 24 bits (three bytes) are used for each pixel. This gives us 16,777,216 possible colors (two to the 24th power). Since three bytes are used for each pixel, we need 3 x 345,600, or 1,036,800 bytes for each frame. There are 30 frames per second, so we'll need 30 x 1,036,800, 0r 31,104,000 bytes each second. There are 60 seconds in a minute, so one minute of video would require 60 x 31,104,000, or 1,866,240,000 bytes per minute. multiply that by 60, and you get about 112 GIGABYTES per hour. These numbers are just getting out of hand! Enter compression! This is the most difficult to understand. Let's just say we are going to do some fancy arithmetic to reduce the data for each frame (JPEG), and take advantage of the similarity of adjacent frames (MPEG) to drastically reduce the amount of data required. There is always a compromise between the amount of data used, and the quality of the picture. A remarkable amount of compression can be achieved with little effect on the picture. Try to do too much, and the ominous "blocking" occurs. Finally, the data is re-configured to allow it to be recorded onto tape, a DVD, or be sent over a network, or even broadcast. This involves a mathematical process to allow some error correction. DVDs and digital cable or satellite are usually encrypted as well so they can't be easily copied. Hope this helps a little! __________________ Dave

September 17th, 2006, 09:26 AM	#2
Emre Safak Trustee Join Date: Feb 2004 Location: Brookline, MA Posts: 1,447	You have to be very specific about what kind of video you have in mind, because each format is encoded differently.

September 22nd, 2006, 12:29 PM	#7
Emre Safak Trustee Join Date: Feb 2004 Location: Brookline, MA Posts: 1,447	Nice, David, but he was asking about analog video.

September 22nd, 2006, 12:40 PM	#8
Giroud Francois Inner Circle Join Date: Feb 2004 Location: switzerland Posts: 2,133	basically, all signals are recorded the same way. the signal is filtered (some frequencies cutted) and eventually sliced as component (could be RGB or YC) , then the signal is sliced again by time sample (usually per line), then sliced again (samples per line). then you can record like any audio signal on tape. Since the tape speed required for such high volume of data would be too great, usually the signal is recorded diagonally on the tape with a rotating head. The great thing with analog recording is the data recorded at one place dos not depend from data coming before or after, so a little scratch on the tape will not cause a big loss, but this loss will be hardly correctable. Additionally, an analog signal would "fade" while a digital signal would "disappear or switch (1 becomes 0)". So years later, you still can read a bad analog tape just by amplifying the signal, while corrupted digital signal would just be unreadable. Another good thing with analog signal is they are pretty close from the natural phenomenon they record too. So recovering an unknown analog signal is pretty easier than deciphering meaningless digital 0 and 1. you can almost "read" a video signal on a oscilloscope, while reading the inside of a digital file would just give you headache.