Week 9 and 10 Flashcards
what is a video
a sequence of still images (photographs) that create the illusion of movement when played in succession
what is each still image called
a frame
movies on film played from ___ to ___ fps
24-30
tv was originally ___fps
29.97
computer displays video at at least ___fps
12-15
in digital video, each frame is a ___ graphic stored as 0 and 1
bitmapped
describe the sampling and quantization processes of film
sampling = each frame sampled into a discrete sample and each sample becomes a pixel
quantization process = assigning a value to each colour
what can we “sample” with MOTION
- Timing of the motion
- Sampling = frames
- Higher FPS = more accurate motion but a larger file size
still images use rgb but videos use
YUV (YIQ) or YCbCr for MPEG compression
what does the Y and UV in YUV stand for
Y = luminance Brightness UV = CbCr in YCbCr = chrominance (colour or hue)
black and white tv only used which part of YUV
Y because it had no chrominance (colour)
when we went from black and white to colour TV, we had to lose ___ because colour increased the file size of each frame
fps
the human eye can detect changes in ____ more easily than changes in __
brightness than colour
it is common for the ____ to be sampled less often because ___
chroma sample (UV) changes in colour are less noticeable
what is the format of colour sampling method
Y:U:V
u and v are colour hue
y is brightness
4:4:4 is no compression
4:4:4 colour sampling is how much compression
none
4:2:2 colour sampling is how much compression
33% on hue, reduced from 12 samples to 8
4:2:2 subsample rate means that out of four colours, two of them will disregard their own colour and take the colour of their adjacent pixel
- lose information
- and save space in your media
4:2:0 colour sampling is how much compression
50% on hue, reduced from 12 samples to 6
4:2:0 subsample rate means that 3 out of four colours will disregard their own colour and take the colour of the last one to save a lot of space in the end
if you want to compress colour, you ….
keep the brightness but take the averages of the colours
what kind of display did original tvs have
Interlaced display
what is interlaced display
uses 48 horizonal scan lines, our eyes see phosphor dots on the screen and an electron beam excites the dots and gun scans through the dots horizontally
only scans every other line odd and starts back up at the top and draws the even lines
- takes two passes, one pass= feild
what is a complete scan in interlaced display
starts at top left and scans each row until bottom right
each pass in an interlaced display is called a
field
what is a progressive display
what we use instead of interlacing
means the whole image is shown at once instead of displaying in passes
480i means
interlaced is i and 480 is the number of feilds
720p and 1080p mean
progressive= p means the whole image is shown at onced 720p is 1280x720 (720 indicates number of rows) 1080p is 1920x1080 number indicates number of pixels
4K compared to 480i, 720p, and 1080p
whats the difference
what is the resolution of 4k and 8k
for 480i, 720p and 1080p, look at horizontal rows but for 4k and 8k you look at vertical columns
7680x4320 - 8k
3840x2160 - 4k
what do TV specs mean -
the tv specifications.
The 50” means 50 diagonal inches (not sure
why it is not metric )
The 4K or 1080 or 720 means number of
pixels
◦ Originally TVs before 1980s had 480 horizontal scan
lines
◦ Then HD TVs had either 1080 or 720 horizontal
scan lines
◦ NOW 4K or 8K TVs have ~4000 or 8000
VERTICAL scan lines
You may see i or p – interlaced or progressive
what does 4k refer to
width
high definition resolutions started at __ and then went up to ___, ___ and finally ___
720, 1080, 1440 then 4K
apple suggests that small devices should be held at ___ inches away from your face and that the screen resolution should be at least ___ ppi
10-12, 300 ppi
at 300 ppi, pixels are so close together and that at 10-12 inches from your eye, things start to look like continuous curves
this is refered to as high retina display
what is high retina display
Apple suggests that small devices (iPads etc.) should be held at 10-12 inches away from you face and that the screen resolution should be at least 300 ppi to look crisp.
Retina Displays have a pixel density high enough so that your eyes cant detect pixels at a normal viewing distance.
for this you must consider
- pixel density
viewing distance
display size
what are some things to things to think about before exporting a video
where will i be putting my video
- on the web, dvd video etc
who is my audience
will i still need to edit it later on or can i compress it all
why should you compress a video
they are fucking long
what is bit rate / data rate
amount of video processed per second
Bit rate = amount of data used to encode video or audio - megabits per second (Mbps) kilobits per second(kbps)
how do you calculate bit rate average
average bit rate = file size/length in seconds of video
bit rate is normally measured in
bits not bytes
don’t just consider the file size of the video, also consider how many bits you have to get across in a given amount of time
think about file size as water in pool
think about bandwidth as hose to fil pool
think of pressure as bitrate (filesize/length in seconds)
how to compress a video: 2 ways
take each image/frame and compress each frame individually
take all frames together and see if there are any spots where its not really changing … you can store that info in a single frame
compression strategies
lower frame size of vid
lower frame rate of video (removing frames or adding seconds)
pick a codec that does higher compression
lower picture quality of video
lower colour depth
play with the audio but this probably wont do much
what is a codec
is a piece of code that compresses a video or audio as it is created (exported from the editing software) and then when displaying it to the user decompresses the video or audio (also stands for the coder/decoder)
what codec does youtube use
H.264
examples of codex
DivX and H.264
if you get a vid on your machine and cant play it, what does this probably mean
that youre computer is missing the appropriate codec
compression concepts- spacial vs temporal
spatial compresses each frame individually and uses same compression technique as JPG, codecs that do spatial compression are animation, planarRGB
temporal compression does better compression but is not good for sports games because it saves info on specified key framesand the other frames just save differences from frame to frame
what is temporal compression - video
better compression
good for talkshow not for action games
- Save info on selected frames (called keyframes)
- all the other frames just save the differences from the previous keyframe
- good when the difference between current frame an keyframe is small
- codecs that use temporal compression: Sorenson video, H.264
spatial compression - video
Spatial
- Compresses each frame individually
- uses the same techniques as JPG compression
- Codecs that do spatial compression are Animation, planarRGB
lossy vs lossless compression in video
lossy lowers video quality but gets better file size and better data/bit rate (filesize/length)
lossless looks for large blocks of pixels that are the same to do Run length encoding
Lossy vs lossless depends on the codec