NOISE

Ok here is the blue channel, it does not look that odd to me, what in
this do you see that looks processed?
http://www.pbase.com/konascott/image...4/original.jpg

Scott

Patrick Boch wrote:
Why do we kick noise?....Film has noise...Damn boring if it didn't...Perhaps
its what the noise looks like? Another words; many times-I see noise in
digital images...that's more like- spotted artifacts...With film...its more
grain like...Digital needs to get past this-plastic like-sheen..and
incorporate...film noise...into there cameras....

Why should digital photography have the good features of film photography, and also the bad features? Each medium has some
advantages that can be used creatively, and some disadvantages. The same can be said of any art medium.

secheese wrote:

On Sat, 12 Feb 2005 22:38:27 GMT, "Patrick Boch"
wrote:


Why do we kick noise?....Film has noise.


There is no noise in the world of film photography; film grain is not
noise. In the CCD/CMOS world, noise is pixels that have some value,
that wasn't recorded by photons entering the camera. Film is a
chemical media and each and every grain of emulsion are only affected
by the light striking it.


Assume somehow I option a truly uniform background, and photograph it
with film. I then scan the film with a microdensitometer with a very
small aperture. If I then look at a plot of density vs horizontal
position in frame, I will see some noise. Now, admittedly some of that
will be electronic noise in the microdensitometer. However, for small
apertures, some will be due to structure of film itself, and statistics
of photon stream and photochemical process itself.

Sure, larger densitometer apertures average/reduce noise, but we could
do same exact thing with digital, though on a coarser basis.

There is virtually NO measurement we can make in our universe that has
zero noise (and photography basically involves measurement).

On 13 Feb 2005 08:05:38 -0800, "Scott W" wrote:

Ok here is the blue channel, it does not look that odd to me, what in
this do you see that looks processed?
http://www.pbase.com/konascott/image...4/original.jpg

Scott


If you look at the this image -- and even
more so in leaves.jpg -- you'll see nice
evenly centered histograms in Red and Green,
but the Blue channel has bottomed out.

It's not a color balance issue, because
the fencepost is quite neutral.

There's something about the fencepost
image that looks distinctly two dimensional.

I'm not sure why but I've got a theory or
two. One is that the green grass in the
background is oversaturated. In a typical
3D scene, as objects recede, they lose
saturation; that's a fundamental visual cue.

I'm not trying to make a broad blanket
condemnation of 10D images or digital
captures in general. It's just that,
to a long time film user, occasionally
the digicam captures do take some
getting used to.


rafe b.
http://www.terrapinphoto.com

rafe bustin wrote:
On 13 Feb 2005 08:05:38 -0800, "Scott W" wrote:

Ok here is the blue channel, it does not look that odd to me, what
in
this do you see that looks processed?
http://www.pbase.com/konascott/image...4/original.jpg

Scott


If you look at the this image -- and even
more so in leaves.jpg -- you'll see nice
evenly centered histograms in Red and Green,
but the Blue channel has bottomed out.

It's not a color balance issue, because
the fencepost is quite neutral.

There's something about the fencepost
image that looks distinctly two dimensional.

I'm not sure why but I've got a theory or
two. One is that the green grass in the
background is oversaturated. In a typical
3D scene, as objects recede, they lose
saturation; that's a fundamental visual cue.

I'm not trying to make a broad blanket
condemnation of 10D images or digital
captures in general. It's just that,
to a long time film user, occasionally
the digicam captures do take some
getting used to.


rafe b.
http://www.terrapinphoto.com

The histogram for the blue channel of the leaves looks just like I
would think it would have to. You have an image with a lot of blue and
yellow. As for the blue being bottomed out this is not too much of a
surpise, yellows often have close to zero blue in them and it is likely
that the raw converter pushed the black level up just a bit off zero.

Do you have the raw file from this image, I would love to have a look
at it.

Scott

rafe bustin wrote:
On 13 Feb 2005 09:06:23 -0800, "Scott W" wrote:


The histogram for the blue channel of the leaves looks just like I
would think it would have to. You have an image with a lot of blue
and
yellow. As for the blue being bottomed out this is not too much of a
surpise, yellows often have close to zero blue in them and it is
likely
that the raw converter pushed the black level up just a bit off
zero.

Do you have the raw file from this image, I would love to have a
look
at it.


Yes, somewhere... it'll take a moment to find.
I'll email it to you privately -- if you can
handle a 6M attachment or so...

Even so... I have lots of roughly similar images
taken with C41 (color negative) film and have not
seen a similar bottoming-out of the blue.


rafe b.
http://www.terrapinphoto.com
No problem with large emails
send them to

On 13 Feb 2005 09:06:23 -0800, "Scott W" wrote:


The histogram for the blue channel of the leaves looks just like I
would think it would have to. You have an image with a lot of blue and
yellow. As for the blue being bottomed out this is not too much of a
surpise, yellows often have close to zero blue in them and it is likely
that the raw converter pushed the black level up just a bit off zero.

Do you have the raw file from this image, I would love to have a look
at it.


Yes, somewhere... it'll take a moment to find.
I'll email it to you privately -- if you can
handle a 6M attachment or so...

Even so... I have lots of roughly similar images
taken with C41 (color negative) film and have not
seen a similar bottoming-out of the blue.


rafe b.
http://www.terrapinphoto.com

In message ,
Kevin McMurtrie wrote:

1600 - Noise and noise reduction badly damages fine details

Does anyone know of any links to a RAW 300D ISO1600 file? I'd like to
look at one and see if it is really ISO 800 pushed, like the 10D.
--


John P Sheehy

In message ,
"David J. Littleboy" wrote:

Neither has anyone else. Nor with any other Canon dSLR. Stacey's simply
wrong on this.

It happens when people over-noise-reduce. This is done by some P&S
cameras, but not DSLRs, as far as I know.
--


John P Sheehy

secheese wrote:

There is no noise in the world of film photography; film grain is not
noise. In the CCD/CMOS world, noise is pixels that have some value,
that wasn't recorded by photons entering the camera. Film is a
chemical media and each and every grain of emulsion are only affected
by the light striking it.


I don't think there is no noise but you do have a point that grain isn't
the same as digital noise, it's more like an overlay of speckles but for
all I know the actual grains are the correct color, just an introduced
texture. The noise comparisons I've seen shoot a grey card and the high
ISO images have colored speckles. I'm not sure film would have these
color noise effects though it certainly gets an obvious dimpled texture
that may not be introducing color aberations.

I might guess film has a soft blurry color noise behind the grain
whereas digital has a per-pixel noise.