In message <BAPId.1169$>, Oh, great; that all we need - outr cards being filled in the background. 8-bit TIFFs lose lots of highlight precision, and usually clip some of the top of the dynamic range if they are made to be immediately viewable. They are about 50% bigger than RAW, even if they use the LZW compression (and have less and less-precise information). For Bayer-sensor cameras, RAW is it; really. If they find better ways of losslessly compressing RAW, great, but keep your TIFFs. --
No. The truth is that a so-called "six megapixel" digicam has 1.5 million red sensors, 1.5 million blue sensors, and 3 million green sensors. It's not really "six megapixels" at all, but it's now become standard practice to rate and specify digicams this way. [Foveon sensors are another story...] The RAW file needs only to encode six million digital values -- ie., the readings from these six million sensors. Reconstruction into RGB "pixels" is done within the RAW converter. If the file were in TIF format, you'd need 18 million encoded values -- assuming you're calling this a "six megapixel" image. JPG is actually somewhat appropriate for encoding digicam images, but is not nearly as flexible as RAW. That is, with RAW, you can reinterpret the six million sensor readings at will. With JPG you must live with the effects of the original interpretation. rafe b. http://www.terrapinphoto.com
Speaking of TIFF, why don't more cameras offer a TIFF option for Tiff is about 300% larger than low ISO RAW for 8 bit files with my Canon dSLR cameras. 16 bit tiffs would be about 600% larger.
No. Lossless compression just doesn't work that well for most photographic images. To the extent that there's a lot of sharp detail in the image, the compression gets worse (ie., less effective.) Try it for yourself with WinZip. The compression will vary quite a bit depending on the nature of the image -- but typically isn't much more than 15-20%. JPG, on the other hand, is specifically optimized for images and to take advantage of the nature of human vision. So while it is regarded as "lossy" (and it is) the compression is terrific, and the actual losses (in practice) can be made almost negligible. rafe b. http://www.terrapinphoto.com
In message <>, It *is* six megapixels, because light levels are recorded at six million different 2D locations in the focal plane. A pixel does not equal R + G + B. --
You will NEVER convince him, EVER! You and I will sit here KNOWING that the Bayer system works really well (better than anything else tried for digital, so far) and that it aproximates the way the human eye works, but pixel peepers around wont be satisfied, EVER If you gave them a million dollars, they would complain that it is GREEN & WRINKLED.
In message <>, Gather them together, and we'll see if any of them can tell the difference between two foveon images in Lab mode with the 'a' and 'b' channels pixelized to two pixels, at normal viewing magnifications. --
Actually on my camera (Olympus E-1), TIFF takes more space than RAW, and there are other cameras where RAW saves even more due to being compressed (it isn't compressed on the E1). However, TIFF is still subject to a lot of problems that JPG has (in camera processing possibly losing details, clipping of dynamic range from 12-ish bits to 8). All that is avoided by using TIFF instead of JPG, is any artifacts caused by the JPG lossy compression.
Not at all, I've made the same arguments myself several times, right down to the explanation of the a and b channels in Lab space. I fully appreciate the cleverness of the Bayer sensor. Even so, by the time I'm screwing with the image in Photoshop, every pixel has to have a Red, Green and Blue coordinate (or some set of coordinates in whatever type of color space I'm using at the time.) A "pixel" in terms of my 10D's Bayer sensor isn't the same thing as a "pixel" in Photoshop. The point of my explanation was to explain the relative sizes of TIFF vs. JPG vs RAW files. So save your uppercase shouting for someone else, OK? rafe b. http://www.terrapinphoto.com
Perhaps the firmware for COMPRESSED tif files is prohibitive. Compressed 8 bit-channel tiff's using LZW or ZIP are generally still larger than RAW files. Plus you have the problem of opening them, many (maybe most) software apps won't open LZW compressed tiff files. I once tried compressing 16 bit-channel tiffs and they just don't compress well with LZW or ZIP, some of the "compressed" files were actually larger than the uncompressed ones and none had savings of even 10%. So for 16 bit files you're still looking at about 6x the memory space compared to RAW. RAW is the way to go, you can generate your tiff later, several different ways if need be, and it's much smaller on the cameras I've used.
Tiff is about 300% larger than low ISO RAW for 8 bit files with my Canon dSLR MB per 8 bit/channel tiff (for those easily confused by %'s).
I am chagrined, however, the common claim of 6, or 8 megapixels for a Bayer camera is generally accepted to mean 6 or 8 megapixels of luminance information, nowhere on any of the packaging, manuals, or advertising do any of the (Bayer) manufacurers claim that these cameras give 6 (or 8) million RGB pixels..Except for Fuji, where the picture size and pixel count is accomplished by magic dwarves living inside the camera. (but they still dont claim anything about RGB pixels. People who spend the bucks to buy a digital camera should have done enough research to understand this BEFORE buying the camera, not after.
The "vast majority" of digicam users don't know or care about such subtleties, just as the vast majority of film camera users don't know or care about how film or cameras work. Of much greater concern is that many digicam users don't know anything at all about the need to archive their images, and will be in for a rude surprise somewhere down the road. Those who know the difference between RAW, TIF and JPG are already ahead of the game, at least they're asking the right questions and know what a file is. rafe b. http://www.terrapinphoto.com
Had a look at your website.. nice stuff. bably agree as to vast majority.. though my feelings on the subject are probably less helpfull than yours. I tend to feel if they dont investigate before spending money, then they dont qualify for free help. I guess Im just a grumpy old curmedgeon.
Another problem is that there are so many variations on TIFF - even a simply freeware app like IrfanView gives you the option of saving with no compression, or using LZW, Packbits, JPEG, or ZIP compression... PMView Pro also offers the option to apply horizontal differencing for LZW compression, the choice of "White is zero" or "Black is zero" photometrics, selectable encoding strip sizes (4k, 8k, 16, 32k and single-strip), and the use of either Intel (LSB-first, PC-standard) or Motorola (MSB-first, Mac-standard) byte ordering... All of these have different uses and different benefits and drawbacks depending on specifics of the image. You can'e expect the camera to always choose the right one. And some software doesn't support some forms of compression (I tried loading a TIFF into my Avery label software and it complained that it didn't support LZW compression - resaving the file without compression allowed it to load).
As long as we're wishing for features... What I'd really like to see is a camera that can talk to a USB storage device. The cameras already do FAT file systems and already do USB. If you could have the camera transfer pictures from the flash card to a USB hard disk then people wouldn't have to spend hundreds of dollars on devices that do the same.
There's not much difference... there are portable devices for well under $200, including 20G or 40G hard drive, which will read a CF card and copy it to the hard drive. Several examples here: <http://www.mydigitaldiscount.com/> rafe b. http://www.terrapinphoto.com
Kibo informs me that (Dr. Joel M. Hoffman) stated that: No. TIFF is far less space efficient than RAW. Try converting a RAW file to 16 bit TIFF yourself, & you'll find that the TIFF file is much, *much* larger than the original RAW file.