Picture taken with the lens caps on, ISO = 400, exposure = 1", has red dark specks of 3x3px size

Hi All!

Two weeks ago I get my new CANAON EOS DIGITAL REBEL XT 350D.

Accidentally a made a shot with the lens caps on.
The last picture should be black, but, to my surprise, there was a
group of dark red specks in the image.

I did more images with different ISO/exposure settings (RAW format),
but the speck group didn't disappeared.
Its appox. size is 3x3px.

The non-black pixels are arranged in the following way:
1-2-1
2-3-2
1-2-1

where "3" means the brightest pixel, and "1" means the dart-dark red
one.

I was thinking it could be defective sensor, so I contacted the
support. They recommended me to send the camera
to one of theirs support centers "...to be examined...". .
Now I want them to look on the pictures because don't want to send the
camera unless I'm sure the problem is
with the sensor and they agree on it.

And the reason I'm writing to the conference is, that maybe someone has
had such issues and he/she could
share the experience... It also will be interesting if someone will
test his/her camera in the way I did.

I have my images uploaded and they are available on Inet:

1. Picture taken with the lens caps on, ISO = 400, exposure time =
1"
http://www.mysharefile.com/v/2957648/img_2911.cr2.html


2. Picture taken with the lens caps on, ISO = 400, exposure time =
10"
http://www.mysharefile.com/v/8143079/img_2910.cr2.html


3. Picture taken with the lens caps on, ISO = 1600, exposure time
= 1"
http://www.mysharefile.com/v/7246556/img_2912.cr2.html

I also can post my correspondence with the support if interested.


Thanks,
Sergey Adamenko
Houston, TX

wrote:

Hi All!

Two weeks ago I get my new CANAON EOS DIGITAL REBEL XT 350D.

Accidentally a made a shot with the lens caps on.
The last picture should be black, but, to my surprise, there was a
group of dark red specks in the image.

Sounds like a hot pixel. Search using that term, there are always a few
hot pixels, it's not a big deal. It may be possible to map it out, I'm
not sure.

I did more images with different ISO/exposure settings (RAW format),
but the speck group didn't disappeared.
Its appox. size is 3x3px.

The non-black pixels are arranged in the following way:
1-2-1
2-3-2
1-2-1

That's the bayer filter interpolation spreading one bad pixel around.


where "3" means the brightest pixel, and "1" means the dart-dark red
one.

I was thinking it could be defective sensor, so I contacted the
support. They recommended me to send the camera
to one of theirs support centers "...to be examined...". .
Now I want them to look on the pictures because don't want to send the
camera unless I'm sure the problem is
with the sensor and they agree on it.

And the reason I'm writing to the conference is, that maybe someone has
had such issues and he/she could
share the experience... It also will be interesting if someone will
test his/her camera in the way I did.

I have my images uploaded and they are available on Inet:

1. Picture taken with the lens caps on, ISO = 400, exposure time =
1"
http://www.mysharefile.com/v/2957648/img_2911.cr2.html


2. Picture taken with the lens caps on, ISO = 400, exposure time =
10"
http://www.mysharefile.com/v/8143079/img_2910.cr2.html


3. Picture taken with the lens caps on, ISO = 1600, exposure time
= 1"
http://www.mysharefile.com/v/7246556/img_2912.cr2.html

I also can post my correspondence with the support if interested.


Thanks,
Sergey Adamenko
Houston, TX

In message .com,
wrote:

And the reason I'm writing to the conference is, that maybe someone has
had such issues and he/she could
share the experience... It also will be interesting if someone will
test his/her camera in the way I did.

I looked at 2910 ... you have four obvious hot pixels. There are three
near the top and right of center, in a small constellation. Black in
that RAW file is about 256. The hot pixel at the lower right of the
cluster is 510 in the RAW data; the pixels surrounding it are not hot at
all. They are all within a few units of 256. The next hot pixel to the
upper left is like 459, or something, and the top-left one is in the
three hundreds. The really hot pixel is about 40% from the left and 70%
down from the top; it is 3705 in the RAW data, and the pixels
surrounding it are not hot at all. They are also all within a few units
of 256.

So, there is no tight cluster of hot pixels; these are single hot
pixels, and they look fat to you because they have been demosaiced.
--


John P Sheehy

Thanks the answer!

I've read about mapping.
Wish Canon provided a tool in to mark hot pixels, so that clients
can resolve such issues yourselves.

Sergey.