|
|||||||||
|
Thread Tools | Search this Thread |
July 1st, 2005, 08:15 AM | #16 |
Trustee
Join Date: Mar 2004
Location: Milwaukee, WI
Posts: 1,719
|
yes but how and why are the 3 color channels split and seperated on a single chip? Is it because of the filter? Do digital still cameras suffer from this same color seperation?
|
July 1st, 2005, 08:33 AM | #17 |
Obstreperous Rex
|
See the "How Color is Created" section about halfway down the page at this link:
http://www.extremetech.com/article2/...1157576,00.asp This is a good overview of digital still camera image sensor technology that begins here: http://www.extremetech.com/article2/0,1558,15465,00.asp |
July 1st, 2005, 09:32 AM | #18 |
Trustee
Join Date: Mar 2004
Location: Milwaukee, WI
Posts: 1,719
|
ok thanks for the info Chris.
So from what I understand it isn't so much that the three colors are not lined up but that the pixels for each color alternates and is then interpolated. So in a way certain color details only really have information that is somewhat like 4:2:0. Although in this case there isn't a perfect 4 for luminace since that is also interpolated. I say 4:2:0 because first the colors alternate every other pixel. This gives us kind of a 4:2. Next blue and red pixels are only sampled every other line so that leaves us with the similar way 4:2:0 works. So in terms of HDV then for color after using 4:2:0 compression there wouldn't be that much of a difference between a single chip and a 3 chip camera except for the fact that the luminace channel would be interpolated in the single chip camera. In the 3 chip camera you would get more color detail but every other pixel would be thrown out during compression. On top of that the compression would alternate chroma channels on every other line. So even if these new single chip cameras have uncompressed output the image would be interpolated somewhat like 4:2:0 anyways? |
July 1st, 2005, 09:34 AM | #19 |
Trustee
Join Date: Mar 2004
Location: Milwaukee, WI
Posts: 1,719
|
Is this also why single chip camera with a Bayer filter are considered to have a loss of resolution since every single pixel is interpolated?
|
July 1st, 2005, 10:13 AM | #20 |
New Boot
Join Date: Jun 2005
Location: Paris-France
Posts: 20
|
hello,
On the other hand, zdnet states : ------------ .... at that price level, it competes with three-chip models from Panasonic, such as the AG-DVC30, which produce much better color accuracy and dynamic range .... ------------ http://reviews-zdnet.com.com/Sony_Ha...2.html?tag=tab |
July 1st, 2005, 10:16 AM | #21 | |
New Boot
Join Date: Jun 2005
Location: Paris-France
Posts: 20
|
Quote:
Of course it helps ! Thanks a lot Chris |
|
July 3rd, 2005, 04:41 AM | #22 |
Regular Crew
Join Date: Jan 2005
Location: Belgium
Posts: 139
|
There is a good reason why 3 CCD's are still preferable to 1 CCD cameras: light loss.
As you can see on the picture here ( http://en.wikipedia.org/wiki/Dichroic_prism ), the prism in a 3 CCD camera will split incoming light into blue, green and red. A single CDD uses a Bayer filter (or another compareable system) with a filter atached to each individual pixel ( http://en.wikipedia.org/wiki/Bayer_filter ). Filters block light and thus a single CCD camera won't be as sensitive as 3 CCD one. The reason why 3 CCD's aren't used in Digital Photography is fourfold: sensitivity isn't as important as it is in Digital Video, the single sensors are bigger and thus more sensitive by themselves, 3 CCD systems are impossible to package in a traditional SLR and the potential for mechanical misalignment issues between the sensors is bigger because of the higher amount of pixels required. The sensitivity arguments needs an explanation. Photographers have been content to work with 100 and 200 ASA film for ages. The FX1 has been regarded as OK but not excellent concerning low light sensitivity. Well, this FX1 has been rated at about 350 ASA by some on this forum. The difference between photographers and videographers is that the photographer can fall back to very slow shutter speeds. He can also uses flashes quite easily while a videographer in a run-and-gun situation will only use extra light (f.i. on camera lights) as a last resort. This run-and-gun situation is exactly where low light sensitivity matters the most. |
July 4th, 2005, 08:04 AM | #24 |
Major Player
Join Date: Nov 2004
Location: Houston, TX
Posts: 331
|
Poor Manual Control
I guess this is OK for home users, but lack of manual aperture control surely drops any interest most pros have in this camera. This could have been a great 2nd camera for some situations, but not without full manual control.
|
July 5th, 2005, 08:06 AM | #25 |
Major Player
Join Date: Sep 2002
Location: Belgium
Posts: 804
|
All single CCD camcorders and digital camera's (except Foveon types) have spatially offsetted pixels. Green pixels shift techniques, though optically totally different are on the image reconstruction level a subset of the mozaik reconstruction algorihms: smart combinations of neighbouring pixels allow higher (luma)resolutions and less aliasing effects than what would be expected with the same amount of coinciding pixels.
|
July 5th, 2005, 03:51 PM | #26 |
Major Player
Join Date: Jan 2005
Location: Vancouver, BC
Posts: 359
|
Pete, this camera is clearly aimed that the consumer market. If they gave you full manual controls then a lot of prosumers and pros would choose this one over the FX1.
They know the HD1 suffers from the same problem, so they're not too worried about it...
__________________
Do or do not, there is no try. |
| ||||||
|
|