On Fri, 29 Jul 2005 19:55:55 +0200, Waldemar Zwierzchlejski wrote:

Lee napisa(a):
http://newsimg.bbc.co.uk/media/image...m_nasa_203.jpg

Anybody know where we can download a hi-res photo of that?

Cheers

Lee


http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Thanks!
Rich

Rich Grise wrote:

http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Next time, they should replace the camera, because of all the stuck
pixels.

Nick
--
Motormodelle / Engine Models:
http://www.motor-manufaktur.de
Ellwe 2FB * VTM 87 * DLM-S3a * cubic
more to come ...

On Fri, 29 Jul 2005 18:18:06 -0500, Nick Müller wrote
(in article ):

Rich Grise wrote:

http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Next time, they should replace the camera, because of all the stuck
pixels.

Nick


They might be hot pixels for a reason - solar or cosmic radiation, for
instance. I'd like to see multiple exposures at the same settings to
see if the same pixels are still fubar'd.

--
"Fame may be fleeting but obscurity is forever." ~Anonymous
"I believe as little as possible and know as much as I can."
~Todd Stuart Phillips
www.angryherb.net

"Herb Schaltegger" wrote in
message .com...
On Fri, 29 Jul 2005 18:18:06 -0500, Nick M|ller wrote
(in article ):

Rich Grise wrote:

http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph
I've
ever seen in my whole life.

Next time, they should replace the camera, because of all the stuck
pixels.

Nick


They might be hot pixels for a reason - solar or cosmic radiation, for
instance. I'd like to see multiple exposures at the same settings to
see if the same pixels are still fubar'd.


Or it could be the dings and pits in the windows due to micrometeors
hitting them since the station was put in orbit.. Take a look at the
pics taken through the shuttle windows and you see a LOT more pits. Just
got the press message from NASA about the new planet they found beyond
Pluto. Guess it's time to make a new solar model...



----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

that IS a pretty awesome photo. thanks for the link. somehow it seems i've
never seen the shuttle that way before, usually it's not so close up, i
don't think i've ever seen it so "lumpy" before. usually it looks much more
sleek and smooth. really looks funny. wrapped in refractory fire bricks.
like they hired a brick mason to assemble a spacecraft. or a ceramicist.

i wish they had photographed the exterior last time it flew. :-(

yay the shuttle flys again! safe trip home!

b.w.




Rich Grise wrote:

http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph
I've
ever seen in my whole life.

Heres another

http://www.nasa.gov/images/content/1...76_ys_full.jpg


"Rich Grise" wrote in message
news
On Fri, 29 Jul 2005 19:55:55 +0200, Waldemar Zwierzchlejski wrote:

Lee napisa(a):
http://newsimg.bbc.co.uk/media/image...m_nasa_203.jpg

Anybody know where we can download a hi-res photo of that?

Cheers

Lee


http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Thanks!
Rich

"Rich Grise" wrote in message
news
On Fri, 29 Jul 2005 19:55:55 +0200, Waldemar Zwierzchlejski wrote:

Lee napisa(a):

http://newsimg.bbc.co.uk/media/image...m_nasa_203.jpg

Anybody know where we can download a hi-res photo of that?

Cheers

Lee


http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Thanks!
Rich

Here is another cool (or maybe "hot" is a better term) one
from a different perspective ...
http://antwrp.gsfc.nasa.gov/apod/ap050729.html

mikey

Herb Schaltegger wrote:

On Fri, 29 Jul 2005 18:18:06 -0500, Nick Müller wrote
(in article ):


Rich Grise wrote:


http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Next time, they should replace the camera, because of all the stuck
pixels.

Nick



They might be hot pixels for a reason - solar or cosmic radiation, for
instance. I'd like to see multiple exposures at the same settings to
see if the same pixels are still fubar'd.

They might even be defects in the CCD. Often the super large ones are that way
and are just used with programs that know the effects. Also - merge overlaps and
shifted pictures that provide data or verify good and sometime/all time flaky.

Mostly good memory is the term. The picture is 3032 x 2008 in size, JPG.

Martin



--
Martin Eastburn
@ home at Lion's Lair with our computer lionslair at consolidated dot net
NRA LOH, NRA Life
NRA Second Amendment Task Force Charter Founder

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

That one looks rotated -

Dimension : 2012 x 3032
Prior one : 3032 x 2008

Thanks - Wish I had them on Skylab - long time since I 'played' with it!

Martin

Grady wrote:

Heres another

http://www.nasa.gov/images/content/1...76_ys_full.jpg


"Rich Grise" wrote in message
news
On Fri, 29 Jul 2005 19:55:55 +0200, Waldemar Zwierzchlejski wrote:


Lee napisa(a):

http://newsimg.bbc.co.uk/media/image...m_nasa_203.jpg

Anybody know where we can download a hi-res photo of that?

Cheers

Lee



http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Thanks!
Rich






--
Martin Eastburn
@ home at Lion's Lair with our computer lionslair at consolidated dot net
NRA LOH, NRA Life
NRA Second Amendment Task Force Charter Founder

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

On Fri, 29 Jul 2005 20:56:56 GMT, Rich Grise
wrote:

On Fri, 29 Jul 2005 19:55:55 +0200, Waldemar Zwierzchlejski wrote:

Lee napisa(a):
http://newsimg.bbc.co.uk/media/image...m_nasa_203.jpg

Anybody know where we can download a hi-res photo of that?

Cheers

Lee


http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Thanks!
Rich

Here's another good one.
http://antwrp.gsfc.nasa.gov/apod/ap050729.html

"Steve W." wrote in message
...

"Herb Schaltegger" wrote in
message .com...
On Fri, 29 Jul 2005 18:18:06 -0500, Nick M|ller wrote
(in article ):

Rich Grise wrote:

http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph
I've
ever seen in my whole life.

Next time, they should replace the camera, because of all the stuck
pixels.

Nick


They might be hot pixels for a reason - solar or cosmic radiation, for
instance. I'd like to see multiple exposures at the same settings to
see if the same pixels are still fubar'd.


Or it could be the dings and pits in the windows due to micrometeors
hitting them since the station was put in orbit.. Take a look at the
pics taken through the shuttle windows and you see a LOT more pits. Just
got the press message from NASA about the new planet they found beyond
Pluto. Guess it's time to make a new solar model...



One way to find out would be to subtract a black image from the image in
question and see if the hot pixels go away. Of course, to do this
correctly, one would need to create the black image from the camera in
question. There is a program that creates black images, but I'm not sure
how reliable it is.

"Mike Fields" wrote in message
...

"Rich Grise" wrote in message
news
On Fri, 29 Jul 2005 19:55:55 +0200, Waldemar Zwierzchlejski wrote:

Lee napisa(a):

http://newsimg.bbc.co.uk/media/image...m_nasa_203.jpg

Anybody know where we can download a hi-res photo of that?

Cheers

Lee


http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Thanks!
Rich

Here is another cool (or maybe "hot" is a better term) one
from a different perspective ...
http://antwrp.gsfc.nasa.gov/apod/ap050729.html

mikey


Now that deserves a little respect!

George wrote:

There is a program that creates black images, but I'm not sure
how reliable it is.

Oh, I guess that "program" is the black plastic disk that is frequently
forgotten to be removed when taking photos.
No, that disk doesn't store any information beside the print
"Pentax/Olympus/Sony/..."


Nick
--
Motormodelle / Engine Models:
http://www.motor-manufaktur.de
Ellwe 2FB * VTM 87 * DLM-S3a * cubic
more to come ...

" George" wrote in message
news:gSIGe.197523$x96.104937@attbi_s72...

"Steve W." wrote in message
...

"Herb Schaltegger" wrote
in
message .com...
On Fri, 29 Jul 2005 18:18:06 -0500, Nick M|ller wrote
(in article ):

Rich Grise wrote:


http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest
photograph
I've
ever seen in my whole life.

Next time, they should replace the camera, because of all the
stuck
pixels.

Nick


They might be hot pixels for a reason - solar or cosmic radiation,
for
instance. I'd like to see multiple exposures at the same settings
to
see if the same pixels are still fubar'd.


Or it could be the dings and pits in the windows due to micrometeors
hitting them since the station was put in orbit.. Take a look at the
pics taken through the shuttle windows and you see a LOT more pits.
Just
got the press message from NASA about the new planet they found
beyond
Pluto. Guess it's time to make a new solar model...



One way to find out would be to subtract a black image from the image
in
question and see if the hot pixels go away. Of course, to do this
correctly, one would need to create the black image from the camera in
question. There is a program that creates black images, but I'm not
sure
how reliable it is.


After looking it over really close I noticed trails behind many of the
bright points. Looks like many are cosmic dust flying by since if it was
bad pixels or damaged glass it would remain fixed in position.



----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

On Fri, 29 Jul 2005 23:44:10 -0500, the opaque "Martin H. Eastburn"
clearly wrote:

That one looks rotated -

Dimension : 2012 x 3032
Prior one : 3032 x 2008

Thanks - Wish I had them on Skylab - long time since I 'played' with it!

Martin

Grady wrote:

Heres another

http://www.nasa.gov/images/content/1...76_ys_full.jpg

I let Firefox expand that one and scrolled down the left side.
When I got to the bottom, I thought
"MAN, look at the size of those woofers!"
That was fun.


--
Guns don't kill people. Rappers do!
-----------------------------------
www.diversify.com Rap-free Website Development

"Martin H. Eastburn" wrote in message
...
Herb Schaltegger wrote:

On Fri, 29 Jul 2005 18:18:06 -0500, Nick Müller wrote
(in article ):


Rich Grise wrote:


http://www.nasa.gov/images/content/1...2673_hires.jpg

There's a non-zero probability that that's the coolest photograph I've
ever seen in my whole life.

Next time, they should replace the camera, because of all the stuck
pixels.

Nick



They might be hot pixels for a reason - solar or cosmic radiation, for
instance. I'd like to see multiple exposures at the same settings to
see if the same pixels are still fubar'd.

They might even be defects in the CCD. Often the super large ones are
that way
and are just used with programs that know the effects. Also - merge
overlaps and
shifted pictures that provide data or verify good and sometime/all time
flaky.

Mostly good memory is the term. The picture is 3032 x 2008 in size, JPG.

yes thats truly an awesome picture, make the hairs on my neck tingle.

intersting some of the 1 pixel spots are blue or red, but many of them are
white, I gues a defect may affect more than one colour of the pixel at a
time, the tails of the spots are so perfectly vertical they must be cemera
artifact.

Colin =^.^=

""Nick Müller"" wrote in message
...
George wrote:

There is a program that creates black images, but I'm not sure
how reliable it is.

Oh, I guess that "program" is the black plastic disk that is frequently
forgotten to be removed when taking photos.
No, that disk doesn't store any information beside the print
"Pentax/Olympus/Sony/..."


Nick

No. BlackFrame NR 1.0 BlackFrame NR eliminates noise from CCD of your
digital camera during long exposure night shots. Try again.

colin wrote:

intersting some of the 1 pixel spots are blue or red, but many of them are
white, I gues a defect may affect more than one colour of the pixel at a
time, the tails of the spots are so perfectly vertical they must be cemera
artifact.

Pixels can either be dead (=black) stuck (=white) or have partial
defects (red/green/blue stuck or dead).
A closer look also shows strictly vertical lines that are
dimmer/brighter than the neighboring pixels. This are defects in the
sensor. Also note, that the spots don't have the slightes halo around
them, which also shows that this are defects in the sensor.

Maybe that the defects are from the radiation in space.

Most people don't know that their digital camera also has defects,
because these are mapped out in the camera's SW before you can see them.
Better cameras have an option to initiate a new pixle defect mapping.

Nick
--
Motormodelle / Engine Models:
http://www.motor-manufaktur.de
Ellwe 2FB * VTM 87 * DLM-S3a * cubic
more to come ...

George wrote:

No. BlackFrame NR 1.0 BlackFrame NR eliminates noise from CCD of your
digital camera during long exposure night shots. Try again.

No.
It does _not_ make black photos, because these would be simply ...
black. It does make noise reduction, adjustment of contrast and
brightness and shifts/adjusts the histogram.

Nick
--
Motormodelle / Engine Models:
http://www.motor-manufaktur.de
Ellwe 2FB * VTM 87 * DLM-S3a * cubic
more to come ...

(Nick Müller) wrote:

colin wrote:

intersting some of the 1 pixel spots are blue or red, but many of them are
white, I gues a defect may affect more than one colour of the pixel at a
time, the tails of the spots are so perfectly vertical they must be cemera
artifact.

Pixels can either be dead (=black) stuck (=white) or have partial
defects (red/green/blue stuck or dead).
A closer look also shows strictly vertical lines that are
dimmer/brighter than the neighboring pixels. This are defects in the
sensor. Also note, that the spots don't have the slightes halo around
them, which also shows that this are defects in the sensor.

Maybe that the defects are from the radiation in space.

Most people don't know that their digital camera also has defects,
because these are mapped out in the camera's SW before you can see them.
Better cameras have an option to initiate a new pixle defect mapping.

This doesn't tell the whole story.

Firstly, modern camera CCDs don't have three sensors per pixel, so they can't
have a pixel stuck full on, ie white. (with rare exceptions, ie there is a
3-layer CCD on the market with three sensors per pixel in three layers. But a
white stuck pixel in these would be extremely rare as three separate sensors
would have to fail all at once)

The sensors in almost all modern digital cameras are layed out as follows:
rgrg
gbgb
rgrg
gbgb

ie every second sensor is green, and blue/red each account for every fourth
sensor, layed out in alternating rows.

The above grid represents a 16 pixel CCD. The two missing colours on each pixel
are simply interpolated from adjacent sensors of the relevant colour.

So a stuck blue sensor will result in not just a bright blue pixel, but a cross
shape, as the four orthogonally adjacent pixels will be assigned 50% of full
blue, and the diagonally adjacent pixels will get 25%. (They will also get a
contribution from other nearby blue sensors, we assume this is negligible
compared to the contribution from the stuck sensor)

Likewise for red.

This we can see in the picture, as bright red and bright blue cross-shaped spots
occur frequently.

Stuck green pixels on the other hand will result in orthogonally adjacent pixels
being assigned 25% green and have no effect on diagonally adjacent pixels.

Still, we would expect bright green spots with a less pronounced cross, but in
fact the only anomalous green spots are much fainter and decidedly not
cross-shaped.

While antialiasing no doubt is affecting the spots, it doesn't explain the
absence of bright green spots, nor the presence of white dots which have no
cross-shape at all, ie which look like single-pixel white spots.

The vertical 'trails' associated with some of the bright stuck pixels I would
attribute to some property of the circuitry not handling a 100% colour signal
from a pixel correctly, ie a design flaw which results in all pixels vertically
above the stuck pixel being assigned incorrect values. In fact this tells us
that the CCD is probably being scanned in vertical columns, from the bottom of
the picture to the top, and whatever mechanism carries forward the residual
'extra charge' from a stuck sensor is reset at the end of the column.

Can anyone explain the lack of bright green spots (the green sensors never fail
'full on'?), or the presence of single-pixel white spots?
--
Patrick Hamlyn posting from Perth, Western Australia
Windsurfing capital of the Southern Hemisphere
Moderator: polyforms group )

Larry Jaques wrote:

On Fri, 29 Jul 2005 23:44:10 -0500, the opaque "Martin H. Eastburn"
clearly wrote:


That one looks rotated -

Dimension : 2012 x 3032
Prior one : 3032 x 2008

Thanks - Wish I had them on Skylab - long time since I 'played' with it!

Martin

Grady wrote:


Heres another

http://www.nasa.gov/images/content/1...76_ys_full.jpg


I let Firefox expand that one and scrolled down the left side.
When I got to the bottom, I thought
"MAN, look at the size of those woofers!"
That was fun.


O rings look like the rings around the woofer! - Agree.

Like one on "back to the future" ? Where the kid hits it once and blows him across the room.

--
Martin Eastburn
@ home at Lion's Lair with our computer lionslair at consolidated dot net
NRA LOH, NRA Life
NRA Second Amendment Task Force Charter Founder

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

This doesn't tell the whole story.

The sensors in almost all modern digital cameras are layed out as follows:
rgrg
gbgb

Can anyone explain the lack of bright green spots (the green sensors never fail
'full on'?), or the presence of single-pixel white spots?

This doesn't tell the whole story either. The sensors in almost all
modern digital cameras are a uniform array of sensor cells with a
filter layer above them that implements the color pattern you
described.

There is no electronic difference between a "red" sensor or a "green"
sensor, thus no reason why particular colors would exhibit different
failure modes.

""Nick Müller"" wrote in message
...
George wrote:

No. BlackFrame NR 1.0 BlackFrame NR eliminates noise from CCD of your
digital camera during long exposure night shots. Try again.

No.
It does _not_ make black photos, because these would be simply ...
black. It does make noise reduction, adjustment of contrast and
brightness and shifts/adjusts the histogram.

Nick

Hot pixels is a known problem of all CCD cameras during long exposures (1s
and up)
A hot pixel is created by an element with a higher rate of charge leakage
than its neighbors.
On long exposure the leakage may cross the exposed value treshold thus
producing a white dot on the image. BlackFrame NR uses a black-frame
method to remove such noise with preserving all details, so the makers say.
I haven't used this program, although I do have a copy. I have used black
frames many times in astrophotography to remove hot pixels.

wrote:

The sensors in almost all modern digital cameras are a uniform array of
sensor cells with a filter layer above them that implements the color
pattern you described.

As Patrick described, there are 4 cells for each pixel. 2 green, 1 red 1
blue. I only would have a different layout:

rgrg
gbgb
rgrg
gbgb But I don't bother if Patrick insists on his version.

The thing with filters is an outdated technology for photographic
cameras.


Nick
--
Motormodelle / Engine Models:
http://www.motor-manufaktur.de
Ellwe 2FB * VTM 87 * DLM-S3a * cubic
more to come ...

George wrote:
""Nick Müller"" wrote in message
...


Hot pixels is a known problem of all CCD cameras during long exposures (1s
and up)
A hot pixel is created by an element with a higher rate of charge leakage
than its neighbors.
On long exposure the leakage may cross the exposed value treshold thus
producing a white dot on the image. BlackFrame NR uses a black-frame
method to remove such noise with preserving all details, so the makers say.
I haven't used this program, although I do have a copy. I have used black
frames many times in astrophotography to remove hot pixels.



Recently, in group s.e.d, there was a link to a site on spatial
filtering. It shows some examples of hot spot removal.

http://www2.polito.it/ricerca/cgvg/teaching/ComputerVision/SpatialFiltering.pdf

On Sat, 30 Jul 2005 12:53:18 -0500, the opaque "Martin H. Eastburn"
clearly wrote:

Larry Jaques wrote:

On Fri, 29 Jul 2005 23:44:10 -0500, the opaque "Martin H. Eastburn"
clearly wrote:

Heres another

http://www.nasa.gov/images/content/1...76_ys_full.jpg


I let Firefox expand that one and scrolled down the left side.
When I got to the bottom, I thought
"MAN, look at the size of those woofers!"
That was fun.

O rings look like the rings around the woofer! - Agree.

The curved triangular section plus the ring wer the 2 things
visible in the picture as I scrolled down. Very evocative.


Like one on "back to the future" ? Where the kid hits it once and blows him across the room.

Or like the stereo setup the guy (Seth Green) gets in ending of The
Italian Connection? He puts the ladies in front of it and it blows
their clothes right off? I simply MUST get wunna them! Of course,
I'd be perfectly happy settling as the guy who runs off with Charlize
Theron. SCHWING!


--
Guns don't kill people. Rappers do!
-----------------------------------
www.diversify.com Rap-free Website Development

"Aubrey McIntosh, Ph.D." wrote in
message ...


George wrote:
""Nick Müller"" wrote in message
...


Hot pixels is a known problem of all CCD cameras during long exposures
(1s and up)
A hot pixel is created by an element with a higher rate of charge
leakage than its neighbors.
On long exposure the leakage may cross the exposed value treshold thus
producing a white dot on the image. BlackFrame NR uses a black-frame
method to remove such noise with preserving all details, so the makers
say. I haven't used this program, although I do have a copy. I have
used black frames many times in astrophotography to remove hot pixels.

Recently, in group s.e.d, there was a link to a site on spatial
filtering. It shows some examples of hot spot removal.

http://www2.polito.it/ricerca/cgvg/teaching/ComputerVision/SpatialFiltering.pdf

Thanks for the link. Always looking for enhancement technics.

(Nick Müller) wrote:

wrote:

The sensors in almost all modern digital cameras are a uniform array of
sensor cells with a filter layer above them that implements the color
pattern you described.

As Patrick described, there are 4 cells for each pixel. 2 green, 1 red 1
blue.

That's not what I said, and it's wrong. Each pixel has only one actual sensor
associated with it - A pixel doesn't have 4 'cells' (ie sensors). By your
definition my diagrammed CCD would actually be a four-pixel sensor. The missing
colour information for each pixel is simply interpolated from nearby sensors of
the correct colour.

When they say 'our camera has a 12-bit (or 8-bit, or 16-, or 24- etc) 6
megapixel sensor' they're using marketing hype (ie an untruth), since more than
half of the information is interpolated. However the actual full-colour
resolution of such a sensor is not immediately obvious, and it looks similar to
a full-resolution sensor if you don't zoom too far. At that level the pixel info
is usually slightly 'fudged' anyway due to antialiasing etc.

... I only would have a different layout:

rgrg
gbgb
rgrg
gbgb But I don't bother if Patrick insists on his version.

Hard to know what you mean here, since that was the same layout I used.

The thing with filters is an outdated technology for photographic
cameras.

Say what??

All modern CCDs use filters in front of the CCD to control what light falls on
which sensor, including the '3 sensors per pixel' CCDs.
--
Patrick Hamlyn posting from Perth, Western Australia
Windsurfing capital of the Southern Hemisphere
Moderator: polyforms group )

In
sci.space.shuttle,sci.electronics.design,rec.puzzl es,rec.crafts.metalworking,
On Sat, 30 Jul 2005 23:18:39 GMT, " George"
wrote:


"Aubrey McIntosh, Ph.D." wrote in
message ...


George wrote:
""Nick Müller"" wrote in message
...


Hot pixels is a known problem of all CCD cameras during long exposures
(1s and up)
A hot pixel is created by an element with a higher rate of charge
leakage than its neighbors.
On long exposure the leakage may cross the exposed value treshold thus
producing a white dot on the image. BlackFrame NR uses a black-frame
method to remove such noise with preserving all details, so the makers
say. I haven't used this program, although I do have a copy. I have
used black frames many times in astrophotography to remove hot pixels.

Recently, in group s.e.d, there was a link to a site on spatial
filtering. It shows some examples of hot spot removal.

http://www2.polito.it/ricerca/cgvg/teaching/ComputerVision/SpatialFiltering.pdf

Thanks for the link. Always looking for enhancement technics.

For more, check out some of the software and techniques discussed
in sci.astro.amateur where they do a lot of this kind of stuff for
digital astrophotography. I've seen pics from posters there as good as
any in astronomy textbooks.

-----
http://www.mindspring.com/~benbradley

Aubrey McIntosh, Ph.D. wrote:



George wrote:

""Nick Müller"" wrote in message
...


Hot pixels is a known problem of all CCD cameras during long exposures
(1s and up)
A hot pixel is created by an element with a higher rate of charge
leakage than its neighbors.
On long exposure the leakage may cross the exposed value treshold thus
producing a white dot on the image. BlackFrame NR uses a black-frame
method to remove such noise with preserving all details, so the makers
say. I haven't used this program, although I do have a copy. I have
used black frames many times in astrophotography to remove hot pixels.


Recently, in group s.e.d, there was a link to a site on spatial
filtering. It shows some examples of hot spot removal.

http://www2.polito.it/ricerca/cgvg/teaching/ComputerVision/SpatialFiltering.pdf


That was interesting. I do some of that transforming pictures through scanners
into an outline of certain for conversion in my Plasma software. So much to learn
on graphic translation.

Martin

--
Martin Eastburn
@ home at Lion's Lair with our computer lionslair at consolidated dot net
NRA LOH, NRA Life
NRA Second Amendment Task Force Charter Founder

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

In article ,
Nick Müller wrote:
George wrote:

There is a program that creates black images, but I'm not sure
how reliable it is.

Oh, I guess that "program" is the black plastic disk that is frequently
forgotten to be removed when taking photos.
No, that disk doesn't store any information beside the print
"Pentax/Olympus/Sony/..."

The Nikon D70 (among other digital cameras) has a mode for long
exposures when, immediately after the exposure is taken, another image
is taken, with the shutter closed and the same exposure time. This is
then subtracted from the original exposure, to remove pattern noise from
long exposures. The disadvantage is that it means that if you take a
five second exposure, you will have to wait another five seconds before
you can take another shot. (As a result, I turned this mode off when
taking shots of the 4th of July fireworks in town. Most of my exposures
were short enough so there was no problem -- the long ones were usually
when there was an unexpected pause between firings, and I eventually let
the shutter close and waited for the sound of the next launch.

Enjoy,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

Patrick Hamlyn wrote:

Hard to know what you mean here, since that was the same layout I used.

I'm sorry. Yes, it's the same. Somehow I garbled your pattern when I
answered.

Nick
--
Motormodelle / Engine Models:
http://www.motor-manufaktur.de
Ellwe 2FB * VTM 87 * DLM-S3a * cubic
more to come ...

"Ben Bradley" wrote in message
...
In
sci.space.shuttle,sci.electronics.design,rec.puzzl es,rec.crafts.metalworking,
On Sat, 30 Jul 2005 23:18:39 GMT, " George"
wrote:


"Aubrey McIntosh, Ph.D." wrote in
message ...


George wrote:
""Nick Müller"" wrote in message
...


Hot pixels is a known problem of all CCD cameras during long exposures
(1s and up)
A hot pixel is created by an element with a higher rate of charge
leakage than its neighbors.
On long exposure the leakage may cross the exposed value treshold thus
producing a white dot on the image. BlackFrame NR uses a black-frame
method to remove such noise with preserving all details, so the makers
say. I haven't used this program, although I do have a copy. I have
used black frames many times in astrophotography to remove hot pixels.

Recently, in group s.e.d, there was a link to a site on spatial
filtering. It shows some examples of hot spot removal.

http://www2.polito.it/ricerca/cgvg/teaching/ComputerVision/SpatialFiltering.pdf

Thanks for the link. Always looking for enhancement technics.

For more, check out some of the software and techniques discussed
in sci.astro.amateur where they do a lot of this kind of stuff for
digital astrophotography. I've seen pics from posters there as good as
any in astronomy textbooks.

-----
http://www.mindspring.com/~benbradley

I frequent alt.binaries.pictures.astro. They too post amazing photos and
discuss imaging techniques. Thanks for the suggestion.

Hi Pat,

My goodness, a real intelligent person talking about something he is
expert on - on sss.

Please email me. My email is real.

Thanks,

Rick



Patrick Hamlyn wrote:
(Nick Müller) wrote:


colin wrote:


intersting some of the 1 pixel spots are blue or red, but many of them are
white, I gues a defect may affect more than one colour of the pixel at a
time, the tails of the spots are so perfectly vertical they must be cemera
artifact.

Pixels can either be dead (=black) stuck (=white) or have partial
defects (red/green/blue stuck or dead).
A closer look also shows strictly vertical lines that are
dimmer/brighter than the neighboring pixels. This are defects in the
sensor. Also note, that the spots don't have the slightes halo around
them, which also shows that this are defects in the sensor.

Maybe that the defects are from the radiation in space.

Most people don't know that their digital camera also has defects,
because these are mapped out in the camera's SW before you can see them.
Better cameras have an option to initiate a new pixle defect mapping.


This doesn't tell the whole story.

Firstly, modern camera CCDs don't have three sensors per pixel, so they can't
have a pixel stuck full on, ie white. (with rare exceptions, ie there is a
3-layer CCD on the market with three sensors per pixel in three layers. But a
white stuck pixel in these would be extremely rare as three separate sensors
would have to fail all at once)

The sensors in almost all modern digital cameras are layed out as follows:
rgrg
gbgb
rgrg
gbgb

ie every second sensor is green, and blue/red each account for every fourth
sensor, layed out in alternating rows.

The above grid represents a 16 pixel CCD. The two missing colours on each pixel
are simply interpolated from adjacent sensors of the relevant colour.

So a stuck blue sensor will result in not just a bright blue pixel, but a cross
shape, as the four orthogonally adjacent pixels will be assigned 50% of full
blue, and the diagonally adjacent pixels will get 25%. (They will also get a
contribution from other nearby blue sensors, we assume this is negligible
compared to the contribution from the stuck sensor)

Likewise for red.

This we can see in the picture, as bright red and bright blue cross-shaped spots
occur frequently.

Stuck green pixels on the other hand will result in orthogonally adjacent pixels
being assigned 25% green and have no effect on diagonally adjacent pixels.

Still, we would expect bright green spots with a less pronounced cross, but in
fact the only anomalous green spots are much fainter and decidedly not
cross-shaped.

While antialiasing no doubt is affecting the spots, it doesn't explain the
absence of bright green spots, nor the presence of white dots which have no
cross-shape at all, ie which look like single-pixel white spots.

The vertical 'trails' associated with some of the bright stuck pixels I would
attribute to some property of the circuitry not handling a 100% colour signal
from a pixel correctly, ie a design flaw which results in all pixels vertically
above the stuck pixel being assigned incorrect values. In fact this tells us
that the CCD is probably being scanned in vertical columns, from the bottom of
the picture to the top, and whatever mechanism carries forward the residual
'extra charge' from a stuck sensor is reset at the end of the column.

Can anyone explain the lack of bright green spots (the green sensors never fail
'full on'?), or the presence of single-pixel white spots?

Huh?

Nevermind..



Patrick Hamlyn wrote:
(Nick Müller) wrote:


wrote:


The sensors in almost all modern digital cameras are a uniform array of
sensor cells with a filter layer above them that implements the color
pattern you described.

As Patrick described, there are 4 cells for each pixel. 2 green, 1 red 1
blue.


That's not what I said, and it's wrong. Each pixel has only one actual sensor
associated with it - A pixel doesn't have 4 'cells' (ie sensors). By your
definition my diagrammed CCD would actually be a four-pixel sensor. The missing
colour information for each pixel is simply interpolated from nearby sensors of
the correct colour.

When they say 'our camera has a 12-bit (or 8-bit, or 16-, or 24- etc) 6
megapixel sensor' they're using marketing hype (ie an untruth), since more than
half of the information is interpolated. However the actual full-colour
resolution of such a sensor is not immediately obvious, and it looks similar to
a full-resolution sensor if you don't zoom too far. At that level the pixel info
is usually slightly 'fudged' anyway due to antialiasing etc.


... I only would have a different layout:

rgrg
gbgb
rgrg
gbgb But I don't bother if Patrick insists on his version.


Hard to know what you mean here, since that was the same layout I used.


The thing with filters is an outdated technology for photographic
cameras.


Say what??

All modern CCDs use filters in front of the CCD to control what light falls on
which sensor, including the '3 sensors per pixel' CCDs.

This depends somewhat on the definition of sensor and detector. We
fought this when talking to others all the time. To those of us in EO
when the first mosaiced sensors came out, a single chip is one sensor.
Each sensor could have many detectors on it. Before the mosaics, when
scanning was done with mirrors, then indeed sensor and detector meant
the same thing. But when folks developed mosaics, such as CCDs,
addressable diode arrays, and such, then we still said a single
processed piece was the sensor, but each sensor could have multiple
detectors, with detector and "pixel" being the same thing in a
non-scanned system.



That's not what I said, and it's wrong. Each pixel has only one actual
sensor
associated with it - A pixel doesn't have 4 'cells' (ie sensors). By your
definition my diagrammed CCD would actually be a four-pixel sensor.
The missing
..

In article ,
Patrick Hamlyn wrote:

Can anyone explain the lack of bright green spots (the green sensors never fail
'full on'?), or the presence of single-pixel white spots?

Perhaps a nonstandard layout, consisting of an luminance pixel and two
chrominance pixels.

--
There's no such thing as a free lunch, but certain accounting practices can
result in a fully-depreciated one.