CCD vs interpolated

What is the difference? If it is advertized a certain MP, I have always
assumed CCD is more accurate, but don't know what the difference is.
Thanks

Richard Bornstein wrote:
: What is the difference? If it is advertized a certain MP, I have always
: assumed CCD is more accurate, but don't know what the difference is.
: Thanks

I may be wrong, but I think you are talking two different things here. A
CCD is a device that captures light intensity levels at specific
locations. Interpolated is not a device but a process. I believe it is
generally a way to double the number of pixels of a photo file by
compairing the output of two adjacent pixels (picture elements), which
could be captured by a CCD element, and creating an interpreted pixel to
go inbetween. So one side of your equasion is a device and the other is a
process.

Now as to which is best, IMHO, a direct measuring of the image pixels
would seem to be more accurate than a computer process "imagining" a
proportion of the image elements. Most of the time when I have seen
"Interpolated" it has been in relation to a scanner. By only having half
as many image sensors as the final image would require, the cost is
reduced. Sometimes with film scanners, some form of interpolation may be
present as the miniaturization of the optics can create a limit on the
number and size of image elements sensed. Eventually the cost of
increasing the number of sampled points vs the cost of that increase
becomes prohibitive.

Randy

==========
Randy Berbaum
Champaign, IL

Richard Bornstein wrote:
: What is the difference? If it is advertized a certain MP, I have always
: assumed CCD is more accurate, but don't know what the difference is.
: Thanks

I may be wrong, but I think you are talking two different things here. A
CCD is a device that captures light intensity levels at specific
locations. Interpolated is not a device but a process. I believe it is
generally a way to double the number of pixels of a photo file by
compairing the output of two adjacent pixels (picture elements), which
could be captured by a CCD element, and creating an interpreted pixel to
go inbetween. So one side of your equasion is a device and the other is a
process.

Now as to which is best, IMHO, a direct measuring of the image pixels
would seem to be more accurate than a computer process "imagining" a
proportion of the image elements. Most of the time when I have seen
"Interpolated" it has been in relation to a scanner. By only having half
as many image sensors as the final image would require, the cost is
reduced. Sometimes with film scanners, some form of interpolation may be
present as the miniaturization of the optics can create a limit on the
number and size of image elements sensed. Eventually the cost of
increasing the number of sampled points vs the cost of that increase
becomes prohibitive.

Randy

==========
Randy Berbaum
Champaign, IL

In article ,
"Richard Bornstein" wrote:

What is the difference? If it is advertized a certain MP, I have always
assumed CCD is more accurate, but don't know what the difference is.
Thanks

Is this regarding Fuji's magic sensor shapes that defy physics and
double the resolution?

Look for "effective pixels" of the sensor. That's how many sensor
points are actually in use. The lens quality matters a lot too because
plenty of 8MP cameras are so blurry that they offer nothing more than
their 4MP predecessors.

In article ,
"Richard Bornstein" wrote:

What is the difference? If it is advertized a certain MP, I have always
assumed CCD is more accurate, but don't know what the difference is.
Thanks

Is this regarding Fuji's magic sensor shapes that defy physics and
double the resolution?

Look for "effective pixels" of the sensor. That's how many sensor
points are actually in use. The lens quality matters a lot too because
plenty of 8MP cameras are so blurry that they offer nothing more than
their 4MP predecessors.

Richard Bornstein wrote:
What is the difference? If it is advertized a certain MP, I have always
assumed CCD is more accurate, but don't know what the difference is.
Thanks

Its the native resolution of the CCD that counts, interpolation just gives
you big files - and you can do that yourself in photoshop if you really
have to.

Think of it in the same way as optical vs digital zoom.

--

Rick

Digital Printing
www.intelligence-direct.com - 01270 215550

Richard Bornstein wrote:
What is the difference? If it is advertized a certain MP, I have always
assumed CCD is more accurate, but don't know what the difference is.
Thanks

Its the native resolution of the CCD that counts, interpolation just gives
you big files - and you can do that yourself in photoshop if you really
have to.

Think of it in the same way as optical vs digital zoom.

--

Rick

Digital Printing
www.intelligence-direct.com - 01270 215550

Right now there are three kinds of sensors used in digital cameras.
CCD devices capture photons and read them out as a long string of
buckets as in a bucket brigade. To get color, the individual cells
are covered with red, green and blue (transmitting) filters.
CMOS devices capture photons and read them out like SRAM, one
storage location at a time. To get color, the individual cells
are covered with red, green and blue (transmitting) filters.
Foveon devices capture photons and read them out like SRAM,
one storage location at a time. To get color, the individual cells
have three different wells under the photon capture area. The lowest
energy photons (red) are captured in the top layer, the mid-energy
photons are captured in the middle layer, and the high energy
photons are captured deeper in the 3rd layer. Each layer can be
accessed individually.

CCD and CMOS color sensors need to be processed by an algorithm
that performs demoasicing (AKA interpolation). There are at least
22 different algorithms that perform this Bayer interpolation. Some
are better than others on one kind of imaging, others better on other
kinds of imaging. Almost everyone agrees that the pictures comming
out of Bayer sensors are top notch.

The Foveon sensors do not need this demosaicing processing, but they
need a different color-discrimination process to obtain nice RGB
colors.
Most people can agree that well exposed Foveon images are top notch.

Where the disagreement lies is with shadow level detail where many
see that CCD and CMOS sensors have better shadow level detail
due in part to the spectral sensitivity differences between the filter
model of CCD/CMOS verusus the energy level discriminatioon of the
Foveons.

Right now there are three kinds of sensors used in digital cameras.
CCD devices capture photons and read them out as a long string of
buckets as in a bucket brigade. To get color, the individual cells
are covered with red, green and blue (transmitting) filters.
CMOS devices capture photons and read them out like SRAM, one
storage location at a time. To get color, the individual cells
are covered with red, green and blue (transmitting) filters.
Foveon devices capture photons and read them out like SRAM,
one storage location at a time. To get color, the individual cells
have three different wells under the photon capture area. The lowest
energy photons (red) are captured in the top layer, the mid-energy
photons are captured in the middle layer, and the high energy
photons are captured deeper in the 3rd layer. Each layer can be
accessed individually.

CCD and CMOS color sensors need to be processed by an algorithm
that performs demoasicing (AKA interpolation). There are at least
22 different algorithms that perform this Bayer interpolation. Some
are better than others on one kind of imaging, others better on other
kinds of imaging. Almost everyone agrees that the pictures comming
out of Bayer sensors are top notch.

The Foveon sensors do not need this demosaicing processing, but they
need a different color-discrimination process to obtain nice RGB
colors.
Most people can agree that well exposed Foveon images are top notch.

Where the disagreement lies is with shadow level detail where many
see that CCD and CMOS sensors have better shadow level detail
due in part to the spectral sensitivity differences between the filter
model of CCD/CMOS verusus the energy level discriminatioon of the
Foveons.