Kevin Sutton: "5K" video is 5120x2700 pixels or 13.8Mp per frame. The Red's 12 RAW frames per second stills is pretty impressive. It also does 4K video at 30 fps in RAW.
The Canon SLR's (current generation) and the Canon video camera only capture MJPEG making editing more difficult.
As a video system, the RED is miles ahead and is in a completely different market space than the Canon's.
The epic does 120fps at 5120x2160 but is limited to 96fps at 5120x2700.
RPJG: "Each lens supports 4K (4096 x 2160 pixels) resolution"
What does that mean - do Canon sell many lenses that *don't* support such resolutions, especially given the number of pixels on heir DSLRs?
For the price the c300 should have a 8192x4608 full frame diagonal sensor capable of 60p raw recording. The c300 is ridiculously over priced and under spec'd.
Strange they use the 4096x2160 resolution instead of 4096x2304 since the c300 is 16:9 why would they change this in the future? I wish the 4096x2160 resolution would be abandoned, it would make no sense to change the aspect ratio of future tv's again.
Roland Karlsson: I think the secret to the prices is that it shall be possible to focus without the FOV changing and it shall be possible to zoom without the focus changing. Thats why they are so expensive - and thats why they are so long. It takes lots of glass and mechanics to achieve those requirements.
If both the focus and zoom is motorized it is very easy to accomplish both with standard lens designs.These overpriced lenses will be obsolete in a few years for sure.
MonkRX: A lot of discussion on resolution here. Keep in mind that "HD", whatever the standard may be, has traditionally meant 720p, 1080p, 1080i.
Laptops with 1280x720 resolution are often advertised with "HD" displays.
So I think the article jumped the gun and assumed it would be "at least 1080 vertical" because it could easily be only 720 vertical (still perfectly fine for web use).
And 1920 x 1080 = 2073600 pixels. Thats slightly over 2 MegaPixels (Not 4 MP). Obviously if you have a 1920 x 1080 display, you could view the image 1:1, or 100%.
Many people seem to forget that camera megapixel numbers are subpixel numbers which is not comparable to the resolution of a display. A full hd display has over 6 million pixels and to get that resolution with a bayer sensor you would need an 8mp sensor since the greens are doubled.
BMWX5: Wow, very clear watching watching it in full screen 1080!
Lets hope they release better players then that select the right resolution setting automatically.I know vimeo allows 1080 too but i dont know what it does on lower resolution displays.Anyways thanks for the link, i will watch it one more time :)
Its 720 actually
Christian Roux: Thank you for this article, but I find it too turned to the Sony brand. It would have been necessary to mention the only real camera (a compact) that has existed for over a year now, namely the Fujifilm Finepix Real 3D W3. It would be useful to encourage camera manufacturers to produce good cameras (models for pictures, not only video cameras models) model with good sensors (Panasonic has announced a model for winter but it seems it should be not terrific in this aspect).It should be noted that among a long list of websites showing very few good photos in 3D (and anaglyph often) very few sites show to future users of 3D cameras what can be done with ... It would have be nice to present the MPO format too.May I suggest you to let some 3D experimented users to explain how make 3d photo concretelyHere are two galleries showing more than 1000 photos:Anaglyphe: www.pbase.com/christianswiss/3dMPO: www.christian-roux.ch/3d
Would you suggest displaying those images captured with very close convergence on a tv that is 3 meters away? If anything else is captured at a longer distance in that same shot it would require your eyes to diverge while your eyes are converging at 3 meters to see the very close object, it simply doesnt work like that. If you capture close objects with the lenses pointing straight ahead and then display those images seperated by the distance of your eyes the close object will automatically be displayed for close convergence of your eyes. The convergence distance of your eyes will then depend on the difference between the captured view-angle and the displayed view-angle. There is simply no logical reason for 3D camera's to have convergence. Displaying fixed images for 3D that are a 1 to 1 match to the display is a flawed concept. The two images need to be independent just like for HMD's which have no overlap at all while a cinema screen has almost complete overlap.
tulo: i never managed to understand the use of "toe-in"/converging optical axis for close focusing like in the Sony TD10. it is the OA of the viewer's eyes that are supposed to converge, giving a sense of depth - not the OA of imaging device, is it?
I agree with this.What matters is that recorded infinity is displayed on the screen separated by the distance of the viewers eyes which means that the image separation in terms of percentage of the whole image is dependent on the size of the display, in the cinema a 6cm separation is negligible compared to the size of the screen while on a computer display it is usually over 10% of the image width. It makes no sense at all that this separation is fixed into the recording, it should obviously be a customizable setting at home which for a tv where the size of the screen is known by the electronics could obviously have a very simple preset.Until this becomes the standard and they drop the convergence nonsense 3D will be fundamentally flawed.
And whats with the silly 2-3 cm lens separation in many of these camera's?