D70 vs P&S cameras: in-camera sharpening.

Newbie question follows, so keep the flame throwers sheathed

I watched a consumer TV programme last night where they shot a few
indoor party scenes with a range of consumer P&S cameras, printed out
at A4 size, and compared results with those from a Nikon D70. A
Fuji P&S was found to be "sharper" than the D70.

Now from what I have read here about "soft" results reported by new
users of DSLRs, I presume this was due to more in-camera sharpening
being applied in the consumer P&S than in the D70.

So, here is the question: does this mean that the D70 is more
accurately outputting the detail information that is hitting the
sensor, and that if it looks soft, that is because that is all the
information that can be resolved? This would also mean that any
sharpening up being applied later in the digital darkroom is
"inventing" information that is not in the original, would it not?

Looking forward to getting some more education about sharpening

Depends, I have just read that Raw files are a little less sharp than an 'in
camera cooked' Jpeg, so as to allow for post digital manipulation. So if the
reviewers were comparing on that basis thinking that the Raw is necessarily
superior to a Jpeg, as usual they don't know there arse from their elbow.
The in camera sharpening on a P&S's Jpeg is perhaps tweaked to appeal to the
average snapshooter but the slight softness of a raw file is perfect as it
is the raw data, unassembled, allowing for post manipulation which is what
digital is all about really, the art of light wave/rule bending,
My 2p's worth

"Bryn James" wrote in message
...
Newbie question follows, so keep the flame throwers sheathed

I watched a consumer TV programme last night where they shot a few
indoor party scenes with a range of consumer P&S cameras, printed out
at A4 size, and compared results with those from a Nikon D70. A
Fuji P&S was found to be "sharper" than the D70.

Now from what I have read here about "soft" results reported by new
users of DSLRs, I presume this was due to more in-camera sharpening
being applied in the consumer P&S than in the D70.

So, here is the question: does this mean that the D70 is more
accurately outputting the detail information that is hitting the
sensor, and that if it looks soft, that is because that is all the
information that can be resolved? This would also mean that any
sharpening up being applied later in the digital darkroom is
"inventing" information that is not in the original, would it not?

Looking forward to getting some more education about sharpening

Depends, I have just read that Raw files are a little less sharp than an 'in
camera cooked' Jpeg, so as to allow for post digital manipulation. So if the
reviewers were comparing on that basis thinking that the Raw is necessarily
superior to a Jpeg, as usual they don't know there arse from their elbow.
The in camera sharpening on a P&S's Jpeg is perhaps tweaked to appeal to the
average snapshooter but the slight softness of a raw file is perfect as it
is the raw data, unassembled, allowing for post manipulation which is what
digital is all about really, the art of light wave/rule bending,
My 2p's worth

"Bryn James" wrote in message
...
Newbie question follows, so keep the flame throwers sheathed

I watched a consumer TV programme last night where they shot a few
indoor party scenes with a range of consumer P&S cameras, printed out
at A4 size, and compared results with those from a Nikon D70. A
Fuji P&S was found to be "sharper" than the D70.

Now from what I have read here about "soft" results reported by new
users of DSLRs, I presume this was due to more in-camera sharpening
being applied in the consumer P&S than in the D70.

So, here is the question: does this mean that the D70 is more
accurately outputting the detail information that is hitting the
sensor, and that if it looks soft, that is because that is all the
information that can be resolved? This would also mean that any
sharpening up being applied later in the digital darkroom is
"inventing" information that is not in the original, would it not?

Looking forward to getting some more education about sharpening

"Bryn James" wrote in message
...

So, here is the question: does this mean that the D70 is more
accurately outputting the detail information that is hitting the
sensor, and that if it looks soft, that is because that is all the
information that can be resolved? This would also mean that any
sharpening up being applied later in the digital darkroom is
"inventing" information that is not in the original, would it not?

Yes indeed. The sharpening process actually throws away information, so it
should be done only after all other adjustments.

"Bryn James" wrote in message
...

So, here is the question: does this mean that the D70 is more
accurately outputting the detail information that is hitting the
sensor, and that if it looks soft, that is because that is all the
information that can be resolved? This would also mean that any
sharpening up being applied later in the digital darkroom is
"inventing" information that is not in the original, would it not?

Yes indeed. The sharpening process actually throws away information, so it
should be done only after all other adjustments.

"Bryn James" wrote in message
...
Newbie question follows, so keep the flame throwers sheathed

I watched a consumer TV programme last night where they shot a few
indoor party scenes with a range of consumer P&S cameras, printed out
at A4 size, and compared results with those from a Nikon D70. A
Fuji P&S was found to be "sharper" than the D70.

Now from what I have read here about "soft" results reported by new
users of DSLRs, I presume this was due to more in-camera sharpening
being applied in the consumer P&S than in the D70.

Bingo! The in-camera sharpening of the D70 is adjustable, and by default,
isn't much.

So, here is the question: does this mean that the D70 is more
accurately outputting the detail information that is hitting the
sensor, and that if it looks soft, that is because that is all the
information that can be resolved? This would also mean that any
sharpening up being applied later in the digital darkroom is
"inventing" information that is not in the original, would it not?

Sharpening doesn't "invent" detail, it brings out what is already there.
The idea is that when two adjacent pixels (or actually pixels within a
certain distance of each other) are different, you increase the difference.
For example, suppose you had a row of black-and-white pixels that went like
this:

10 10 10 10 13 17 20 20 20 20

This is most likely a slightly blurred edge between an area of brightness 10
and an area of brightness 20. The sharpening algorithm will do something
like this:

10 10 10 10 11 19 20 20 20 20

(increasing the difference between the 13 and the 17) or maybe even this:

10 10 10 09 08 22 21 20 20 20

(producing a very exaggerated edge).

This counteracts the effects of blur from any source, within reason (the
actual theory is quite complicated and depends on the point spread
function). HOWEVER, it can also make edges look sharper than they were in
the original subject, and it can produce very unnatural effects.

Low-end digital cameras do a fixed amount of in-camera sharpening with the
assumption that you are taking pictures under typical conditions and want a
print between 4x6 and 8x10 inches in size. Higher-end DSLRs have adjustable
in-camera sharpening and are not very heavy-handed; they expect you'll do
your _real_ processing outside the camera.

So... Comparing the "sharpness" of the D70 to a low-end digital camera is
not a fair test. The D70 definitely captures enough detail that people can
use it to test Nikon lenses. It just doesn't do a lot of sharpening by
itself.

Also, the D70 has a quite aggressive noise-reduction algorithm. If you make
a 1- or 5-second exposure with a D70, it won't be speckled the way a long
exposure with a cheaper camera will be.

--
Clear skies,

Michael A. Covington
Author, Astrophotography for the Amateur
www.covingtoninnovations.com/astromenu.html

"Bryn James" wrote in message
...
Newbie question follows, so keep the flame throwers sheathed

I watched a consumer TV programme last night where they shot a few
indoor party scenes with a range of consumer P&S cameras, printed out
at A4 size, and compared results with those from a Nikon D70. A
Fuji P&S was found to be "sharper" than the D70.

Now from what I have read here about "soft" results reported by new
users of DSLRs, I presume this was due to more in-camera sharpening
being applied in the consumer P&S than in the D70.

Bingo! The in-camera sharpening of the D70 is adjustable, and by default,
isn't much.

So, here is the question: does this mean that the D70 is more
accurately outputting the detail information that is hitting the
sensor, and that if it looks soft, that is because that is all the
information that can be resolved? This would also mean that any
sharpening up being applied later in the digital darkroom is
"inventing" information that is not in the original, would it not?

Sharpening doesn't "invent" detail, it brings out what is already there.
The idea is that when two adjacent pixels (or actually pixels within a
certain distance of each other) are different, you increase the difference.
For example, suppose you had a row of black-and-white pixels that went like
this:

10 10 10 10 13 17 20 20 20 20

This is most likely a slightly blurred edge between an area of brightness 10
and an area of brightness 20. The sharpening algorithm will do something
like this:

10 10 10 10 11 19 20 20 20 20

(increasing the difference between the 13 and the 17) or maybe even this:

10 10 10 09 08 22 21 20 20 20

(producing a very exaggerated edge).

This counteracts the effects of blur from any source, within reason (the
actual theory is quite complicated and depends on the point spread
function). HOWEVER, it can also make edges look sharper than they were in
the original subject, and it can produce very unnatural effects.

Low-end digital cameras do a fixed amount of in-camera sharpening with the
assumption that you are taking pictures under typical conditions and want a
print between 4x6 and 8x10 inches in size. Higher-end DSLRs have adjustable
in-camera sharpening and are not very heavy-handed; they expect you'll do
your _real_ processing outside the camera.

So... Comparing the "sharpness" of the D70 to a low-end digital camera is
not a fair test. The D70 definitely captures enough detail that people can
use it to test Nikon lenses. It just doesn't do a lot of sharpening by
itself.

Also, the D70 has a quite aggressive noise-reduction algorithm. If you make
a 1- or 5-second exposure with a D70, it won't be speckled the way a long
exposure with a cheaper camera will be.

--
Clear skies,

Michael A. Covington
Author, Astrophotography for the Amateur
www.covingtoninnovations.com/astromenu.html

On Tue, 07 Dec 2004 13:22:55 GMT, Bryn James
wrote:

Newbie question follows, so keep the flame throwers sheathed

I watched a consumer TV programme last night where they shot a few
indoor party scenes with a range of consumer P&S cameras, printed out
at A4 size, and compared results with those from a Nikon D70. A
Fuji P&S was found to be "sharper" than the D70.

Now from what I have read here about "soft" results reported by new
users of DSLRs, I presume this was due to more in-camera sharpening
being applied in the consumer P&S than in the D70.

So, here is the question: does this mean that the D70 is more
accurately outputting the detail information that is hitting the
sensor, and that if it looks soft, that is because that is all the
information that can be resolved? This would also mean that any
sharpening up being applied later in the digital darkroom is
"inventing" information that is not in the original, would it not?

Looking forward to getting some more education about sharpening

Well spotted, this is typical bull**** you expect from a consumer
program. I've seen them (being general here) get it *completely* wrong
in other industries and in one particular experience I had about 10
years ago, they were verging on being corrupt, selecting an obviously
inferior brand by applying questionable and irrelevant logic.

For a Granny, the inability to modify the sharpness setting in the D70
'sharpening' menu might mean this program's conclusion was helpful,
but for the rest of the world, it's trash.

Here's a general rule on in-camera vs out of camera processing:

** If the same source data is made available to both algorithms, the
out-of-camera (read Photoshop) will do a better job. It has MUCH more
processing power, MANY more sliders for fine control, MUCH more R&D
and is in it's 9th release, compared to version 2 of the camera
firmware.

Some things (and I believe digital zoom on non-DSLRs could fall into
this category) are best done in-camera because they are done before
JPEG conversion, the (zoom for example) is done on the RAW and then
converted to JPEG. But, this really only matters if JPEG is being
used.


For D70, I would shoot RAW with sharpening to NONE (BTW, I am not sure
if it would actually sharpen the RAW, or just use it as a
recommendation for the RAW importer). Don't do any sharpening on
import (annoying RAW thing in photoshop has this to 25% by default)
and use a third-party sharpener plugin such as nik sharpener pro which
asks you target output size, print rez and viewing distance and even
scans the image to determine the correct amount and method of
sharpening.

Even if you want small internet-only images, if card capacity isn't an
issue, I'd still shoot RAW and resize them in Photoshop, the result
has less step-artifacts than doing it in-camera.

--
Owamanga!

On Tue, 07 Dec 2004 13:22:55 GMT, Bryn James
wrote:

Newbie question follows, so keep the flame throwers sheathed

I watched a consumer TV programme last night where they shot a few
indoor party scenes with a range of consumer P&S cameras, printed out
at A4 size, and compared results with those from a Nikon D70. A
Fuji P&S was found to be "sharper" than the D70.

Now from what I have read here about "soft" results reported by new
users of DSLRs, I presume this was due to more in-camera sharpening
being applied in the consumer P&S than in the D70.

So, here is the question: does this mean that the D70 is more
accurately outputting the detail information that is hitting the
sensor, and that if it looks soft, that is because that is all the
information that can be resolved? This would also mean that any
sharpening up being applied later in the digital darkroom is
"inventing" information that is not in the original, would it not?

Looking forward to getting some more education about sharpening

Well spotted, this is typical bull**** you expect from a consumer
program. I've seen them (being general here) get it *completely* wrong
in other industries and in one particular experience I had about 10
years ago, they were verging on being corrupt, selecting an obviously
inferior brand by applying questionable and irrelevant logic.

For a Granny, the inability to modify the sharpness setting in the D70
'sharpening' menu might mean this program's conclusion was helpful,
but for the rest of the world, it's trash.

Here's a general rule on in-camera vs out of camera processing:

** If the same source data is made available to both algorithms, the
out-of-camera (read Photoshop) will do a better job. It has MUCH more
processing power, MANY more sliders for fine control, MUCH more R&D
and is in it's 9th release, compared to version 2 of the camera
firmware.

Some things (and I believe digital zoom on non-DSLRs could fall into
this category) are best done in-camera because they are done before
JPEG conversion, the (zoom for example) is done on the RAW and then
converted to JPEG. But, this really only matters if JPEG is being
used.


For D70, I would shoot RAW with sharpening to NONE (BTW, I am not sure
if it would actually sharpen the RAW, or just use it as a
recommendation for the RAW importer). Don't do any sharpening on
import (annoying RAW thing in photoshop has this to 25% by default)
and use a third-party sharpener plugin such as nik sharpener pro which
asks you target output size, print rez and viewing distance and even
scans the image to determine the correct amount and method of
sharpening.

Even if you want small internet-only images, if card capacity isn't an
issue, I'd still shoot RAW and resize them in Photoshop, the result
has less step-artifacts than doing it in-camera.

--
Owamanga!

Owamanga wrote in
:

Some things (and I believe digital zoom on non-DSLRs could fall into
this category) are best done in-camera because they are done before
JPEG conversion, the (zoom for example) is done on the RAW and then


Would not all operations be performed on the RAW data? I don't usually
shoot RAW on my camera because it takes so long to write to the card.

Bob