Re: 12 bit images
- From: clay@xxxxxxxxxxxxxxx
- Date: Wed, 30 Apr 2008 13:07:06 -0700 (PDT)
On Apr 30, 3:08 pm, Jerry Avins <j...@xxxxxxxx> wrote:
c...@xxxxxxxxxxxxxxx wrote:
On Apr 30, 10:03 am, Jerry Avins <j...@xxxxxxxx> wrote:
rajesh wrote:
...
Surprising to see that google hasnt yielded a single image.I'm not surprised. Most images are intended to be looked at with eyes.
Eight bits produces all the monochrome shading that most people can
appreciate or that most printing processes can reproduce.
Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Hello Jerry et al,
8 bits usually works well enough to hold an image in its final
presentation if being presented by the usual means like paper or LCD.
Ultrahigh dynamic range displays are a different matter altogether.
Most digital SLR cameras record data at 12 bits per channel in their
raw modes. One (Nikon D3) is now offered that has 14 bits per channel!
Since with this camera the dynamic range was not increased 4 fold, the
resulting steps are closer together. This proves helpful in some
cases.
If the OP wants a 12 bit image, he should just go over to a
photography forum like the ones onwww.dpreview.comand ask. I'm sure
many there will offer up an image or two. The question now is what
format does he want. I doubt he wants a proprietary format which the
raw files from DSLRs are offered. Someone could convert the raw to a
16 bit tif and he can go from there.
I was just about to see what Gimp could do by way of diddling the raw
data from my camera. I guess the OP could do that as well as I. It
needn't be a panchromatic conversion if he isn't concerned with color
fidelity; we made good pictures with blue-sensitive film, and with
orthochromatic film after that.
Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯- Hide quoted text -
- Show quoted text -
Hello Jerry,
If you are getting into trying to read the raw data, there are some
links to info on the web. For instance I know some people have reverse
engineered most of Nikon NEF format. But down underneath pretty much
all raw files contain the actual A/D values for each of the
photosites. Of course each of photosites only detects one color, so
the three primary colored arrays need to be interpolated and overlaid.
Recall the Bayer sensor pattern uses
Row A GRGBGRGBGRGBGRGBGRGB
Row B RGBGRGBGRGBGRGBGRGBG
Row C GRGBGRGBGRGBGRGBGRGB
etc.
Bayer's method uses a color filter array in front of the array of
wideband photodetectors. And the green is used twice as often as the
red and blue.
IHTH,
Clay
.
- References:
- 12 bit images
- From: rajesh_d27
- Re: 12 bit images
- From: aaac111
- Re: 12 bit images
- From: rajesh
- Re: 12 bit images
- From: Andre
- Re: 12 bit images
- From: rajesh
- Re: 12 bit images
- From: Jerry Avins
- Re: 12 bit images
- From: clay
- Re: 12 bit images
- From: Jerry Avins
- 12 bit images
- Prev by Date: Re: jitter calculation for ADC
- Next by Date: Re: CIC filter
- Previous by thread: Re: 12 bit images
- Next by thread: Newton-Rapshon method over a multiplicative or modulated noise process
- Index(es):
Relevant Pages
|