|
|||||||||
|
Thread Tools | Search this Thread |
January 24th, 2006, 07:33 AM | #1 |
Trustee
Join Date: Sep 2004
Location: Bristol, CT (Home of EPSN)
Posts: 1,192
|
Could someone please explain "Colorspace"?
I try to not ask these kinds of questions, but I keep seeing reference to 4:2:2 and 4:2:0 and can't seem to figure out what this means.
I switched from DV to HDV and from what I've read, there is a difference. But what? |
January 24th, 2006, 08:29 AM | #2 |
Inner Circle
Join Date: Jun 2003
Location: Toronto, Canada
Posts: 4,750
|
http://www.adamwilt.com/pix-sampling.html
kind of explains it. |
January 24th, 2006, 08:48 AM | #3 |
Major Player
Join Date: Feb 2004
Location: Philadelphia, PA, USA
Posts: 548
|
Color space refers to the values a computer uses to define color.
Physical space, for example, is defined in 3 dimentions, usually referred to as X, Y and Z. Color space can be defined a variety of ways. Red, Green, Blue (RGB); Hue, Saturation, Brightness (HSB) for example. Most color selection dialogs list values in both RGB and HSB since these are the most common color spaces used in computing. Video, however, typically records and stores data in a format that (roughly) corresponds to one channel for brightness, and two additional channels for color. The abbreviation for NTSC color space is YIQ and and for PAL it's YUV. The values you see for video resolution (4:1:1, 4:2:0, 4:2:2, 4:4:4) refer to the data compression used in video colorspace. Luma/brightness is rarely compressed. This is the first "4" and is the basis for defining the color compression. 4:1:1 and 4:2:0 store the same quantity of data, 1 pair of IQ/UV values for every 4 luma values. The data is simply arranged differently as shown in the link above. 4:1:1 is used for NTSC DV, where 4:2:0 us used for PAL DV. 4:2:2 is common in broadcast formats and HDV. 4:4:4 (no color compression) is generally relegated to HD "digital film" cameras like the CineAlta F950, Dalsa, and Thompson Viper. Hope this helps. Have fun. |
January 24th, 2006, 08:50 AM | #4 |
Capt. Quirk
Join Date: Apr 2002
Location: Middle of the woods in Georgia
Posts: 3,596
|
I'm sorry, but obviously I am drinking defective coffee... could you dumb it down a bit more for me? I guess I totally missed Adam's point, even with the pictures.
__________________
www.SmokeWagonLeather.us |
January 24th, 2006, 09:03 AM | #5 |
Major Player
Join Date: Feb 2004
Location: Philadelphia, PA, USA
Posts: 548
|
Here's another try. :)
Video stores images in pixels. Each pixel can have up to 3 values to define it's exact appearence. 1 value for brightness, 2 additional values that define color. (chromanence (sp?)) Since, when all this stuff was being invented, computers and even analog broadcasting tools, couldn't keep up with the data rate needed to handle all three values for every pixel of every frame, some really clever people figured out a cheat. Our eyes perceive brightness more accurately than color, so, these clever people figured out that if they simply "skipped" recording 50-75% of the color data, pretty much nobody watching the videos would notice.... and they were right. So, we have broadcasts and recording equipment that are operating 4:2:2 compression. For every 4 pixels, there are 4 brightness/luma values (for our eyes to perceive clarity) and 2 pixels that have color values (which they "share" with the other two pixels, giving us to general idea about the color of all four pixels). HDV is also recorded to this standard. DV cheats a little more. There is only one pair of color values saved for every four pixels, but since all four luma values are recorded, it's generally "good enough" for most common video needs. The arrangement of which color values are stored in NTSC vs PAL are shown in the page on Adam's site. Clear as mud? |
January 24th, 2006, 09:17 AM | #6 |
Capt. Quirk
Join Date: Apr 2002
Location: Middle of the woods in Georgia
Posts: 3,596
|
Much better, but... I understand the first value is brightness- 4, and the second value is color- 1,2, or 4, depending on how much info is stored in the pixels... but what is the 3rd value? Is it a redundancy, or for interlaced images?
Also, is there a way to boost mini DV to the higher color values?
__________________
www.SmokeWagonLeather.us |
January 24th, 2006, 09:41 AM | #7 |
RED Problem Solver
Join Date: Sep 2003
Location: Ottawa, Canada
Posts: 1,365
|
http://www.nattress.com/Chroma_Inves...masampling.htm
Explains pretty much all about chroma sampling. And yes, there's things that you can do about it! I sell tools for FCP, G Nicer and G Chroma Sharpen that fix DV chroma very well, and soon these tools will also be available, GPU accellerated in FinalTouch too. Graeme
__________________
www.nattress.com - filters for FCP |
January 24th, 2006, 10:17 AM | #8 |
Capt. Quirk
Join Date: Apr 2002
Location: Middle of the woods in Georgia
Posts: 3,596
|
Nice Graeme, that really did show a difference. Are your filters going to ever be made available to us non-FCP folks? I'm still using Premiere 6.
__________________
www.SmokeWagonLeather.us |
January 24th, 2006, 10:25 AM | #9 |
RED Code Chef
Join Date: Oct 2001
Location: Holland
Posts: 12,514
|
That seems to be a very nice text you've made on that subject Graeme. I'll
read it shortly. Thanks!
__________________
Rob Lohman, visuar@iname.com DV Info Wrangler & RED Code Chef Join the DV Challenge | Lady X Search DVinfo.net for quick answers | Buy from the best: DVinfo.net sponsors |
January 24th, 2006, 10:37 AM | #10 | |
Major Player
Join Date: Feb 2004
Location: Philadelphia, PA, USA
Posts: 548
|
Quote:
I'll be reading for days to catch up. :) Thanks for sharing. |
|
January 24th, 2006, 01:45 PM | #11 |
RED Problem Solver
Join Date: Sep 2003
Location: Ottawa, Canada
Posts: 1,365
|
Thanks all!
No plans for PC versions of anything. I have my work cut out keeping up with FCP stuff, and the new FinalTouch plugins. Graeme
__________________
www.nattress.com - filters for FCP |
| ||||||
|
|