Andrew Ray

Film making has never been limited to the rich or select few...but it has been limited to the resourceful and creative. Should be this statement " but it has NOT been limited to the resourceful and creative" ??

"the information is as good as it's source" Who said that? That is why I like the posts where people provide the links to the primal source of information. Other wise I always treat the post as a "IMHO" If someone spends the time to show us the exact numbers or facts, what's the problem to cut and paste the links to the post windows from the web site you are reading it all about? or put the reference where you got it from. I always do it if I can. We want to contribute here for good of everybody. Taking good information from this forum and delivering bad one is not a good contribution, is it? http://www.lib.berkeley.edu/TeachingLib/Gu...t/Evaluate.html Andrew

Richard, are you sure that the process of some guy taken out of the screening room was actually done because they wanted competitor out? Maybe there was just one person taken out to talk to or someone wanted to give him some material or who knows what was meaning of this event. How far you were from this person when he was asked to follow the bouncer? Did you hear the conversation between them when they were leaving the screening room? Andrew

I forgot one thing, these images are shown without smooth edges algorithm applied to it. In real post you will smooth the edges so you will not see the staircase effect, the shapes will be just bit softer. But then when do you have 100% contrast and 100% saturation in the image? OK it is worse case scenario.

Shooting blue on red will gain you resolution like from below attached images. Left picture is 4:4:4 where all pixels have RGB component for all horizontal 4096 pixels. The right one is 4:2:0 equivalent of Bayer sensor without green, 2048 pixels per horizontal line. So 4K format becomes 2K format. This will be the case in Dalsa sensor that is not oversampled sensor, it mean it has 4096 pixels per horizontal line. RED has 4900 pixels per horizontal line so the loss of resolution will be less visible. Drop of resolution from 4K to 2K I will not label ?image becomes very soft? First, it will happen only in the absence of green and near 100% contrast, second 2K in such picture is still very good image quality even for viewers sitting not farther then the width of the screen. Human eye will not recognize the single pixel on the screen at such distance if we have 3600 pixels per horizontal line. You can try it in front of your HDTV. When you are watching 1980 horizontal pixels from the distance equal to the width of the screen you will see the raster of the pixels. Yes, because there is less then 3600 of them. Go bit further and you will not see raster anymore. You need 20/20 vision for this experiment. So all guys out there with home theaters organized in such a way that screen is on the wall along the room not across the room have to switch to 4K projectors. 60 degree FOV is very nice especially when watching all this wide angle shots done with 12mm or less lenses but then you need 4K pixels for horizontal line to have it silky smooth. Note: At this moment RED uses 4900 pixels per horizontal line to show the safe area in the view finder and sends only 4086 pixels for the projected area, so no oversampling in such a case. However I hope that you can override this setup when you capture the data via high speed RAW port on the camera. Not good for normal film making since we need safe area but for underwater shots we do not have mikes hanging there and 20% oversampling will probably keep this 4K always 4K providing that you can get the 12mm FL wide angle lenses that can even resolve 4K. Andrew Below are actual 1:1 crops of 4K image, watch it from the distance equal width of screen to see the cinema experience of it for 4K display and if you have 2K or 1980 pixels in the horizontal line of your computer screen you have to watch it from ~1.7 width of the screen.

One more thing. To be honest, this sensor underperforms if there is no green in the picture. So reds on blue edges and purple on red or blue edges etc. will give you lower resolution. However any edge or detail on any color when green is present will give you almost full resolution of the pixel count.

Note how scanning the sensor for unique RGB value works. The unique design of Bayer pattern gives you always 3RGB pixels in each circle for only one pixel offset when you move in horizontal direction and also one pixel offset when you move in vertical direction. So each pixel move in 4096 horizontal pixel row will give you 4096 RGB groups of readings. (we have two green pixels in each RGB trio so the higher sensitivity of green is matching the human eye sensitivity)

Carl, it is amazing how brains of different people do follow the same pattern to discovery. Maybe because we read the same books and information material. We took some of the cameras here and tested maximum horizontal resolution. Both for 3CCD cameras and Bayer camera. Well you can follow the same experiment and you will get the same result. First I was surprised that prism does introduce so much distortions, both in terms of XY alignment and in terms of back focus plain. Higher camera resolution camera you pick more of these problems you see. So maybe you have 3 X 1920X1080 or 3X2K pixels but the prism and lenses working with the prism do screw up the picture big way. Because of this I don?t think there will be any 3CCD 4K system out there build. Carl, you can not discuss 3CCD systems without taking all elements under the account. If you take 3CCD 4K system based on prism and flat one surface of 4.9K Bayer sensor, You will see it gives you the same horizontal lines test resolution. Yes it looks like we are comparing 3X4K versus 1X4.9 K pixels. But remember these 3X4K CCDs are on top of each other in the same place and looking through the prism in to the lenses. Such arrangement doesn?t give you any more horizontal sensors does it? Yes Panasonic is shifting green sensor in its 3CCD camera but that is different story. On 2K cameras with 3CCD shifted sensor, YES it is equivalent to 4K Bayer. But again on 4K 3CCD shifted camera the prism will screw up the resolution so badly that it will compare to 4K Bayer anyway. It is like discussing 4K digital cinema versus 35mm cinema. Everybody is discussing the master copies. But reality is that we should discuss the final projection output in the cinema. This is what people do watch. 35mm film goes 3 times through the lenses before it hits the viewer eyes. Digital 4K will go only once during the acquisition, via camera lenses. Then if material is displayed on non projection display system the digital 4K stays 4K and digital copy is 100% mathematically lossless. Film material though, I don?t want even to mention how much resolution you loose when you do copy the film.

Look on the pictures above. D-20 and 4K sensor based camera will compare like Wavelett 2k and Wavelett 4k D-20 is 6MP sensor and RED is 12MP If anything to compare to, it is Phantom or Dalsa.

You can calculate the speed by presuming that there will be minimum 11:1 compression. Then take 3 (colors) X 4096 X 12 bits for each line X (times) number of lines per aspect ratio (for 2.35:1 there are 1743 lines) This way you get number of bits per frame X 24 or 30 for frame rate. So for 2:35 screen aspect at 30fps each second will contain following number of mega bytes. 3 X 12 X 4096 X 1743 X 30 /1024/1024/8 = 919.160Mbytes/sec = 7.35Gbits/sec colors X bit per color X horizontal pixels X vertical pixels X frame per second / convert to kilos / convert to megas So if each second of 4K/30fps delivers 919MBytes/s then after compression of 11:1 will be 83.55Mbytes (919/11) Remember 1K=1024 and 1MB=1024X1024 The RED RAW compression ratio for now is 11 to 13 times. BTW 4K you can shoot only at 30fps

Carl, I was as skeptical as you, when I had my first contact with RED back in December 2006. Since then I did read some material, upgraded my knowledge about the digital cinematography and guess what, we just placed the order for 2 RED cameras. However I like your method of learning, by throwing the wrench in the gear box. It is the fastest way to learn how gear box works. This way we do teach each other. More controversial questions you ask more people will answer to your controversy. I did post this in my previous post here but again. 4K format is defined as a 4096 pixel in each horizontal line of frame. The vertical pixel count depends from aspect ratio. Now I can ensure you that from 4,900 horizontal photosites of Bayer sensor you can get 4096 unique values of RGB for each horizontal line. http://en.wikipedia.org/wiki/Bayer_sensor It is bit tricky to see it right away, with old way of thinking, but try it. (Question: Canon DSLR camera 20D or 30D is it 8MP camera or 2MP camera?) So which part of this definition you misunderstood?

Matthew, this compression doesn?t really matter since in most project you will use RAW port anyway. And what is coming from the raw port can print the same film copies that now DI process is printing.So compression is just academic discussion here.

Alessandro, yes if RED CODE compression codec would have the compression ratio option say 6-1 then the final data output will be two times bigger as compared to 12-1, so it will reduce recording time by about 50% However the on camera circuitry, may not be fast enough to record/process the data. I think you can push the recording on RED to 6-1 with good safety margin but then when you use memory card recording media instead of hard drive it will cost you exponentially more. Andrew You will see the same dynamic range on the RAW output from the camera that clicks at 9GB/sec so I don't think so.

Alexander, at the current state of camera sensor technology the film is bit better in highlights (2-4 stops) The only way to beat it is to use two digital sensors, one servicing shades the other highlights. Some will argue with this statement because high end camera sensors are better in the deep shade so you could push it down and get improved dynamic range. There is a lot of money invested in the development of camera sensor technology right now, so I expect constant improvements there on almost monthly basis. Again, we are comparing film masters with the camera sensors. Since most movies right now are going through DI process anyway that uses similar digital sensors as the 4K cameras do, the final distribution copy will have the same limitations in highlight. ARRI scanners though use two passes process though, one for neutral and one for highlights. Andrew

Carl, why camera needs real time live output in its full resolution? Does film camera have live full resolution output? Actually if you connect 4K monitor with deBayer input that supports 10GB port that RED has on its body, you could see live RAW at full 4.9K (full sensor resolution). But why anybody will need raw deBayer output? Yet they did it just in case. I call it OCD output. (Obsessive Compulsive Disorder) http://www.miramax.com/aviator/ Andrew