Compression in JPEG files in digital cameras

I need some help in explaining the JPEG compression feature in digital
cameras. My camera (which likely is similar to most others) has the
feature to compress the photo JPEG files in the storage card. It also
has the choice to have different pixel sizes (example: 3000x2250,
2000x1500, 1024x768, etc). What is the difference of the above two
features? If you store a 3000x2250 pixel data in compressed mode,
does it loose its quality? Can it be re-instated to full uncompressed
size without loosing photo quality?. When I compressed the data, it
will fit more pictures in a single storage card. But, is it the same
if I choose 2000x1500 pixel and no compression instead?
Thanks for info.

wrote in message
ups.com...
I need some help in explaining the JPEG compression feature in
digital
cameras. My camera (which likely is similar to most others) has the
feature to compress the photo JPEG files in the storage card. It
also
has the choice to have different pixel sizes (example: 3000x2250,
2000x1500, 1024x768, etc). What is the difference of the above two
features? If you store a 3000x2250 pixel data in compressed mode,
does it loose its quality? Can it be re-instated to full
uncompressed
size without loosing photo quality?. When I compressed the data, it
will fit more pictures in a single storage card. But, is it the same
if I choose 2000x1500 pixel and no compression instead?
Thanks for info.



Don't ask much, do you?

Jpg is a lossy form of saving the picture - Google on 'explanation of
jpg' and it will give you the detail - as distinct from compressed per
se. As you have lost picture info you cannot get it back, so there is
no way to regenerate the original intact. If you want to save in a
non-compressed format you need to use TIFF (OK it is slightly
compressed) or better still RAW which is just the raw data off the
sensor without adjustment.

The differences in resolution affect file size and quality of
reproduction. 3000x2250 is about 6.7Mp, 2000x1500 is 3Mp, etc. For
'normal' printing at full frame (i.e. without cropping) 3Mp will
produce an acceptable colour picture at A4 or thereabouts. For most
day-to-day use 5Mp is enough, 6Mp is really the limit for a compact.
If you go much higher than that you start to get digital noise in
saturated colours due mainly to thermal effects in the sensor. Another
point that is often missed is that the lenses on many compacts -
particularly at the cheaper end of the market - often do not have the
resolution in themselves to match the resolution of the sensor.

In a nutshell, use RAW or TIFF if your camera has it, otherwise go for
the best quality that the camera can provide - memory cards are dirt
cheap these days.


--
Woody

harrogate3 at ntlworld dot com

In rec.photo.digital harrogate3 wrote:

JPEG is a lossy form of saving the picture -

The Bayer sensor is already lossy, having only one of RGGB at
any of four pixel locations, so one could make the argument that
JPEG imposes no further loss.

Google on 'explanation of jpg' and it will give you the detail

or read
http://photo.net/jpeg/learn/

Bill Tuthill wrote:
In rec.photo.digital harrogate3 wrote:

JPEG is a lossy form of saving the picture -

The Bayer sensor is already lossy, having only one of RGGB at
any of four pixel locations, so one could make the argument that
JPEG imposes no further loss.

You could argue that until you are blue in the face, and it
still wouldn't be anything near correct though! :-)

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

In rec.photo.digital Floyd L. Davidson wrote:

JPEG is a lossy form of saving the picture -

The Bayer sensor is already lossy, having only one of RGGB at
any of four pixel locations, so one could make the argument that
JPEG imposes no further loss.

You could argue that until you are blue in the face, and it
still wouldn't be anything near correct though! :-)

I'm not going to argue it that long,
but I would say that nobody has quantified the loss that comes from
the Bayer sensor versus the loss that comes from JPEG.

The real problem with JPEG is that it is not edit-safe,
rather than loss of information at creation time.

In rec.photo.digital Bill Tuthill wrote:

I'm not going to argue it that long,
but I would say that nobody has quantified the loss that comes from
the Bayer sensor versus the loss that comes from JPEG.

The real problem with JPEG is that it is not edit-safe,
rather than loss of information at creation time.


Is it not obvious to you that the "loss" from using a bayer sensor (which
isn't really a loss in the same sense) and the loss due to JPEG compression
are going to be cummulative in nature. Thus, the JPEG losses will only add to
any losses due to using a bayer sensor.

--
Thomas T. Veldhouse

We have more to fear from the bungling of the incompetent than from the
machinations of the wicked.

Floyd L. Davidson wrote:

JPEG is a lossy form of saving the picture -

The Bayer sensor is already lossy, having only one of RGGB at
any of four pixel locations, so one could make the argument that
JPEG imposes no further loss.

You could argue that until you are blue in the face

If you shot using raw you'd have a better chance of correcting it.

Daniele

Bill Tuthill wrote:
In rec.photo.digital harrogate3 wrote:

JPEG is a lossy form of saving the picture -

The Bayer sensor is already lossy, having only one of RGGB at
any of four pixel locations, so one could make the argument that
JPEG imposes no further loss.

However, JPEG imposes an additional loss, because of the approximations it
makes. You can control the degree of approximation by the quality setting
on cameras, or by the compression or quality settings in your software.

David

David J Taylor added these comments in the current discussion du
jour ...

JPEG is a lossy form of saving the picture -

The Bayer sensor is already lossy, having only one of RGGB at
any of four pixel locations, so one could make the argument
that JPEG imposes no further loss.

However, JPEG imposes an additional loss, because of the
approximations it makes. You can control the degree of
approximation by the quality setting on cameras, or by the
compression or quality settings in your software.

Here we go again! David, you and I get along OK. You're level
headed and not any kind of elitist, while at the same time, you
have an excellent foundation in the theory of all of this. I think
the OP is a pretty rank novice and has probably gone unconscious by
now in this thread. You have such a great way to express your ideas
in words that people of all technical levels can understand that I
bet you can do far better than my feeble attempt at helping the OP.

--
HP, aka Jerry

Bill Tuthill added these comments in the current discussion du jour
....

In rec.photo.digital harrogate3 wrote:

JPEG is a lossy form of saving the picture -

The Bayer sensor is already lossy, having only one of RGGB at
any of four pixel locations, so one could make the argument that
JPEG imposes no further loss.

Google on 'explanation of jpg' and it will give you the detail

or read
http://photo.net/jpeg/learn/

Bill, I'm not nearly enough of a mathematicion to understand this,
but it is WAY over the OP's head. Might it be better to describe
the issue in more qualitative than theoretical ways he can
understand?

--
HP, aka Jerry