Well, let's see...32bbp x 22,000,000 pixels = 704,000,000 bits / 8bits/byte = 88,000,000 bytes /1024 = 85,937kbytes/1024 ~= 84 megabytes.Originally posted by: Adul
droolage! how big a file would that produce?
If you can afford a camera with a 22MP CCD, you can afford a couple 1gb microdrives or one of those photowallets that holds like 3 GB of pictures 🙂In other words, my "big" 128MB memory stick suddenly holds 1.5 pictures. 🙁
Wow. Totally awesome! 😀Originally posted by: Mark R
It's little baby brother (the 16 MPx CCD) has a read out time of 2 seconds - I would guess that this one probably is 3 seconds. Sensitivity is quoted at 25 ISO for the back.
The camera back that this CCD goes in is very interesting - The 'light' model features just the CCD, but the standard and Hi-res models use some ingenious techniques to improve the quality of the image. All models, of course, feature standard methods of improving quality such as Peltier cooling of the CCD to reduce thermal noise.
Like most image sensors these CCDs use a Bayer filter on the front of the sensor - each pixel is made up of 4 seperate neighbouring subpixels (1 red, 1 blue and 2 green subpixels per pixel). These are arranged in a grid fashion. The problem with this is that you do not get full colour information for each point in the image - you have to interpolate the colour data from the subpixels to form full coloured pixels. You also may have problems with coloured fringing due to moire.
To get around this problem, the higher-end model backs, have tiny motors which move the CCD (by up to 1 sub-pixel width in each direction), so that full RGB data is available for each sub pixel (4 shot), and also for the space between each subpixel (16 shot).
What's more is that the CCD and back are specified to have a full 14 bit colour depth per channel (42 bit colour). With all the enhancements activated and the automatic tiling system (to enlarge the effective sensor area) files can be produced with 160million pixels - in 48 bit colour format these files are approx 1GB each (and that's for the old 16 MPx back!)
Originally posted by: owensdj
vetteguy, I think your calculations are correct, but that image size is only for a bitmap. The images will be stored using JPEG compression which makes them far smaller than that.
Considering that the CCD in question is designed for high end medium format cameras (presumably to be used in studios considering its low ISO speed), I doubt that the users thereof would want to reduce the quality of their image any by applying a lossy JPEG compression to it. Rather, they'd use TIFF with lossless compression, or some other format capable of storing the extra bits per channel of color information that the CCD is capable of acquiring (TIFF and Photoshop's proprietary format are both capable of up to 16 bits per channel, and maybe more). Of course, I'm really not "in the know" when it comes to high end digital stuff, but this is all based on simple logic, considering that the data from that CCD would be used to make 16x20 and possibly larger prints. 🙂Originally posted by: LethalWolfe
While you could probably choose to use JPEG compression I'm sure you could capture the images in the uncompressed TIFF format.Originally posted by: owensdj
vetteguy, I think your calculations are correct, but that image size is only for a bitmap. The images will be stored using JPEG compression which makes them far smaller than that.
Lethal