Kit
Senior Member
Originally posted by Fundee
everybody ????
including you ??????
Nope, I prefer to stay in the REAL world.
Originally posted by Fundee
everybody ????
including you ??????
Originally posted by Kit
Nope, I prefer to stay in the REAL world.
Originally posted by Fundee
real world ????...where.???? when.????
Originally posted by Kit
Here...... now. Realistic reasonable expectations. Something many people struggle to come to terms with.
Originally posted by Fundee
need to struggle ??????..wow !!!
Originally posted by Fundee
me using s2..the only thing that i not quite like is why the 2 battery system...izzit no time for R&D ?????....raw mode only with 12 mgpixel not 6....other than that all ok.....nice colour, with custom setting...i dunno what is the limit, there's so many thing to try....if only this camera can do without the 1.5 multiplier and a better focussing module like the F5..... ;p
Originally posted by Klause
In terms of handling, I find the s2 a bit weird in shape(not insulting it ;p). I heard that there's no battery grip made for it but you can make on urself from a grip of other model. Is that right?
Originally posted by Bedpan
Raw Files are about 12.5mb each
3024x2016 = 6096384 pixels
6096384 x 16bits = 97542144bits
97542144bits / 8 /1024 /1024 = 11.63mb
Add in the Camera info and a Jpg and you got your 12.5mb file.
As you can see the Raw data is just that 6mp of raw data!
Originally posted by Jed
Very, very fancy maths. I'm impressed, someone must have bought you a calculator for Christmas that you opened early :bsmilie: :bsmilie:
Seriously though, the only problem with your maths is that it's a lovely case of getting the formula to fit the result. Last I heard, the Fuji, and all other digital cameras, have slightly more than 16 bits in total, since they have at least 8 bits per channel. If I'm not wrong the Fuji records 12 per channel in RAW, meaning an RGB output of 36 bits, not 16. Which leaves your mathematical formula just slightly in need of recomputing
In fact, your formula actually IS right. But not for RAW files, but for TIFF files. Bearing in mind that they record as either 8 bits per channel (in which case, substitute 24 for your 16) or 16 bits per channel (sub 48 for 16).
RAW files depend largely on how the camera manufacturers decide to organise data, but involves the raw information from the CCDs without any processing.
Originally posted by Bedpan
The RAW files are 16bits per pixel... only 12bits of used data, 4 bits of padding are included for each pixel. In Raw each pixel is only a single colour. During conversion these extra bits are dumped and the pixel colour is extrapolated using the surrounding colour.. This is also why comparable 6mp camera Raw files are smaller (d100, d60); They only save the 12bit data without the padding.
At least this is how I understand it from the Designer of Qimage, SharpRaw, and Bibble. Apparently it made it easier for them to make the conversion software as data is easier to break in 2 byte blocks as compared to 12 bit blocks...
Originally posted by roygoh
So the RAW data is before Bayer filtering, thus straight from the CCD, and the RAW image viewer has to apply the Bayer filtering, and that's why RAW files from different camera manufacturers are not compatible, am I right?
Thanks.
ROy
Originally posted by ckiang
AFAIU, the RAW data is after the filtering and anti-aliasing filter, but before the interpolation that creates the resultant image. By right it should be smaller than the TIFF it creates, right? The RAW convertor will then do the necessary interpolation, exposure compensation, white balancing, sharpening etc which is normally done in-camera.
Regards
CK
You said that each 12-bits color pixel in RAW format is represented by 16 bits, due to 4 bits padding. In that case, the resultant file should not be smaller than the 48-bit TIFF.Originally posted by Bedpan
Each pixel is recorded as just a single colour. In the case of the S2 12bits of information is recorded per pixel (then padded with 4 extra bits).
Originally posted by tsdh
You said that each 12-bits color pixel in RAW format is represented by 16 bits, due to 4 bits padding. In that case, the resultant file should not be smaller than the 48-bit TIFF.
Is that true?
OK, I know what you mean. I forgot the RGB filter on each pixel of the sensor.Originally posted by Bedpan
Ummmm.....
RAW (in the case of the S2)
Each pixel = 1 colour
Each Pixel = 12bits actual data (16bits including padding)
8 bit TIFF
Each Pixel = 3 colours
Each pixel = 8bits x 3 colours = 24bits per pixel
16 bit TIFF
Each Pixel = 3 colours
Each pixel = 16bits x 3 colours = 48bits per pixel
So in the case of 16 bit TIFF's vs. RAW, you can see each pixel in 16 bit TIFF contains 48bits, RAW is only 16bits (12bits actual data). So you can see the Tiff should be 3 times as large as the RAW. This will not be exact as each contains differnt headers and other embedded data...
Hope this helps... Having a hard time explaining it![]()
Originally posted by Goondu
No Sigma SD9?;p
Originally posted by Galaa
D60 seems to be winning...and i thot its suppost to be the least popular one, acording to dpreview forums...they prefered D100 and S2 Pro...