Hot pixels on a Canon 300D

Discussion in 'Digital Photography' started by Nick, Nov 17, 2003.

  1. Nick

    Nick Guest

    I just bought a Canon EOS 300D and, being the paranoid type, immediately ran
    deadpixeltest.exe on it. I took the pictures using the RAW setting with the
    lens cover on. And at 5 seconds or more, it finds this group of pixels:

    http://members.chello.se/nscho/bad_pixels.gif

    It seems strange that there should be a whole bunch of hot pixels so close
    to each other. And no, it's no compression artifact, they really look like
    this on the RAW image.

    I'm not very likely to take too many pictures using more than 5 seconds, but
    the fact that they are all clustered together bothers me. So what do you
    think, is it a bad sensor, or can something else cause this?
    Nick, Nov 17, 2003
    #1
    1. Advertising

  2. Nick

    jriegle Guest

    No, that is not a compression artifact. Even in RAW, the camera still has to
    "blend the pixels" or thin lines will get color artifacts since it has a
    Bayer sensor.. You are simply seeing the effect of "hot" red sensing cell.
    The software that blends the pixels together and softens the image somewhat
    causes this in the outpu image.

    It is often missunderstood that RAW means "right off the sensor". That is
    not the case.
    John

    "Nick" <> wrote in message
    news:Lmcub.3692$%W3.19960@amstwist00...
    > I just bought a Canon EOS 300D and, being the paranoid type, immediately

    ran
    > deadpixeltest.exe on it. I took the pictures using the RAW setting with

    the
    > lens cover on. And at 5 seconds or more, it finds this group of pixels:
    >
    > http://members.chello.se/nscho/bad_pixels.gif
    >
    > It seems strange that there should be a whole bunch of hot pixels so close
    > to each other. And no, it's no compression artifact, they really look like
    > this on the RAW image.
    >
    > I'm not very likely to take too many pictures using more than 5 seconds,

    but
    > the fact that they are all clustered together bothers me. So what do you
    > think, is it a bad sensor, or can something else cause this?
    >
    >
    >
    jriegle, Nov 17, 2003
    #2
    1. Advertising

  3. Nick

    Mr Blobby Guest

    Is that what's causing the red dots in this 15-20 Min exposure ?
    http://www.users.bigpond.com/andrewc/Astronomy/pages/South Pole.htm
    I originally thought they were Geostationary satellites !

    Andrew


    "jriegle" <> wrote in message
    news:s6dub.278492$...
    > No, that is not a compression artifact. Even in RAW, the camera still has

    to
    > "blend the pixels" or thin lines will get color artifacts since it has a
    > Bayer sensor.. You are simply seeing the effect of "hot" red sensing cell.
    > The software that blends the pixels together and softens the image

    somewhat
    > causes this in the outpu image.
    >
    > It is often missunderstood that RAW means "right off the sensor". That is
    > not the case.
    > John
    >
    > "Nick" <> wrote in message
    > news:Lmcub.3692$%W3.19960@amstwist00...
    > > I just bought a Canon EOS 300D and, being the paranoid type, immediately

    > ran
    > > deadpixeltest.exe on it. I took the pictures using the RAW setting with

    > the
    > > lens cover on. And at 5 seconds or more, it finds this group of pixels:
    > >
    > > http://members.chello.se/nscho/bad_pixels.gif
    > >
    > > It seems strange that there should be a whole bunch of hot pixels so

    close
    > > to each other. And no, it's no compression artifact, they really look

    like
    > > this on the RAW image.
    > >
    > > I'm not very likely to take too many pictures using more than 5 seconds,

    > but
    > > the fact that they are all clustered together bothers me. So what do you
    > > think, is it a bad sensor, or can something else cause this?
    > >
    > >
    > >

    >
    >
    Mr Blobby, Nov 18, 2003
    #3
  4. Nick

    chibitul Guest

    In article <Lzdub.15004$>,
    "Mr Blobby" <> wrote:

    > Is that what's causing the red dots in this 15-20 Min exposure ?
    > http://www.users.bigpond.com/andrewc/Astronomy/pages/South Pole.htm
    > I originally thought they were Geostationary satellites !
    >
    > Andrew


    Geostationary satellites are all in the equatorial plane of the Earth,
    since they *have to* rotate at the same speed with the Earth to appear
    geostationarry to us... thus they have to be in the equator plane, not
    north or south of this unique plane. Your hot pixels are all ober the
    place!
    chibitul, Nov 18, 2003
    #4
  5. Nick

    Vadim Guest

    It won't answer your question exactly, but you can find very useful
    information on hot pixels here:
    http://webpages.charter.net/bbiggers/DCExperiments/html/hot_pixels.html


    "jriegle" <> wrote in message news:<s6dub.278492$>...
    > No, that is not a compression artifact. Even in RAW, the camera still has to
    > "blend the pixels" or thin lines will get color artifacts since it has a
    > Bayer sensor.. You are simply seeing the effect of "hot" red sensing cell.
    > The software that blends the pixels together and softens the image somewhat
    > causes this in the outpu image.
    >
    > It is often missunderstood that RAW means "right off the sensor". That is
    > not the case.
    > John
    >
    > "Nick" <> wrote in message
    > news:Lmcub.3692$%W3.19960@amstwist00...
    > > I just bought a Canon EOS 300D and, being the paranoid type, immediately

    > ran
    > > deadpixeltest.exe on it. I took the pictures using the RAW setting with

    > the
    > > lens cover on. And at 5 seconds or more, it finds this group of pixels:
    > >
    > > http://members.chello.se/nscho/bad_pixels.gif
    > >
    > > It seems strange that there should be a whole bunch of hot pixels so close
    > > to each other. And no, it's no compression artifact, they really look like
    > > this on the RAW image.
    > >
    > > I'm not very likely to take too many pictures using more than 5 seconds,

    > but
    > > the fact that they are all clustered together bothers me. So what do you
    > > think, is it a bad sensor, or can something else cause this?
    > >
    > >
    > >
    Vadim, Nov 18, 2003
    #5
  6. In article <Lmcub.3692$%W3.19960@amstwist00>, "Nick" <>
    wrote:

    > I just bought a Canon EOS 300D and, being the paranoid type, immediately ran
    > deadpixeltest.exe on it. I took the pictures using the RAW setting with the
    > lens cover on. And at 5 seconds or more, it finds this group of pixels:
    >
    > http://members.chello.se/nscho/bad_pixels.gif
    >
    > It seems strange that there should be a whole bunch of hot pixels so close
    > to each other. And no, it's no compression artifact, they really look like
    > this on the RAW image.
    >
    > I'm not very likely to take too many pictures using more than 5 seconds, but
    > the fact that they are all clustered together bothers me. So what do you
    > think, is it a bad sensor, or can something else cause this?
    >
    >
    >


    That's a hot pixel. The width is probably due to some software blurring
    to avoid Bayer pattern aliasing.

    Here's a 19.3 minute exposure:

    http://www.pixelmemory.us/Photos/Outdoors/Mt. Hamilton/Oct 22 2003%
    20Star%20Spin/CRW_0142.CRW

    Or it's reduced JPG:
    http://www.pixelmemory.us/Photos/Outdoors/Mt. Hamilton/Oct 22 2003%
    20Star%20Spin/spin.jpg

    (The red line is a laser from the Lick Observitory)


    That's really not too many hot pixels for 19 minutes. They could even
    be masked with a threshold based filter. Yeah, it's touch blurry. It's
    windy up there and I haven't gotten around to upgrading my tripod yet.
    Kevin McMurtrie, Nov 18, 2003
    #6
  7. Nick

    Mr Blobby Guest

    Thanks for that, does that mean they follow the planets line ?

    "chibitul" <> wrote in message
    news:...
    > In article <Lzdub.15004$>,
    > "Mr Blobby" <> wrote:
    >
    > > Is that what's causing the red dots in this 15-20 Min exposure ?
    > > http://www.users.bigpond.com/andrewc/Astronomy/pages/South Pole.htm
    > > I originally thought they were Geostationary satellites !
    > >
    > > Andrew

    >
    > Geostationary satellites are all in the equatorial plane of the Earth,
    > since they *have to* rotate at the same speed with the Earth to appear
    > geostationarry to us... thus they have to be in the equator plane, not
    > north or south of this unique plane. Your hot pixels are all ober the
    > place!
    Mr Blobby, Nov 18, 2003
    #7
  8. Nick

    Phil Guest

    Kevin McMurtrie wrote:

    > In article <Lmcub.3692$%W3.19960@amstwist00>, "Nick" <>
    > wrote:
    >
    >
    >>I just bought a Canon EOS 300D and, being the paranoid type, immediately ran
    >>deadpixeltest.exe on it. I took the pictures using the RAW setting with the
    >>lens cover on. And at 5 seconds or more, it finds this group of pixels:
    >>
    >>http://members.chello.se/nscho/bad_pixels.gif
    >>
    >>It seems strange that there should be a whole bunch of hot pixels so close
    >>to each other. And no, it's no compression artifact, they really look like
    >>this on the RAW image.
    >>
    >>I'm not very likely to take too many pictures using more than 5 seconds, but
    >>the fact that they are all clustered together bothers me. So what do you
    >>think, is it a bad sensor, or can something else cause this?
    >>
    >>
    >>

    >
    >
    > That's a hot pixel. The width is probably due to some software blurring
    > to avoid Bayer pattern aliasing.


    I'm not sure I agree. My first 300D had a true, undeniable hot pixel --
    and it showed up at all exposure speeds, even 1/1000 sec. And it was
    visible as a bright point in all images where it was in an area with a
    darkish background (e.g., grass).

    A nunber of factors MAY result in apparent hot pixels with long exposures:

    o How warm was the camera/CMOS when the test was done?

    o Was there light leakage from the viewfinder (i.e., was the VF cover
    supplied with the camera in place)?

    I was also curious at what file was tested in DeadPixelTest. The
    versions I've used deal with jpeg or tiff, but not RAW. What file was
    really tested.

    So I'm not quite ready to conclude that this is a "hot pixel" situation
    ... vs. noise or light leakage or a conversion artifact.

    Phil
    Phil, Nov 18, 2003
    #8
  9. "Mr Blobby" <> writes:
    >Thanks for that, does that mean they follow the planets line ?


    I don't think so. The geostationary satellites all have to be in the
    Earth's equatorial plane. The planets are (mostly) found in the solar
    system's plane. But the earth's axis of rotation is tilted 23 degrees
    with respect to the solar system's plane, so the 2 planes are separated
    by that much angle.

    Dave
    Dave Martindale, Nov 18, 2003
    #9
  10. Nick

    Mr Blobby Guest

    Doh !


    Of course !

    Thanks.


    "Dave Martindale" <> wrote in message
    news:bpcmm8$1pt$...
    > "Mr Blobby" <> writes:
    > >Thanks for that, does that mean they follow the planets line ?

    >
    > I don't think so. The geostationary satellites all have to be in the
    > Earth's equatorial plane. The planets are (mostly) found in the solar
    > system's plane. But the earth's axis of rotation is tilted 23 degrees
    > with respect to the solar system's plane, so the 2 planes are separated
    > by that much angle.
    >
    > Dave
    Mr Blobby, Nov 18, 2003
    #10
  11. Nick

    Nick Guest

    "Phil" <> wrote in message
    news:5Mkub.86374$...
    > I'm not sure I agree. My first 300D had a true, undeniable hot pixel --
    > and it showed up at all exposure speeds, even 1/1000 sec. And it was
    > visible as a bright point in all images where it was in an area with a
    > darkish background (e.g., grass).
    >
    > A nunber of factors MAY result in apparent hot pixels with long exposures:
    >
    > o How warm was the camera/CMOS when the test was done?
    >
    > o Was there light leakage from the viewfinder (i.e., was the VF cover
    > supplied with the camera in place)?
    >
    > I was also curious at what file was tested in DeadPixelTest. The
    > versions I've used deal with jpeg or tiff, but not RAW. What file was
    > really tested.
    >
    > So I'm not quite ready to conclude that this is a "hot pixel" situation
    > .. vs. noise or light leakage or a conversion artifact.


    The camera was room temperature, and I converted a RAW image taken with ISO
    100 to TIFF for use in DeadPixelText. So there should be no conversion
    problem, but the software in the 300D may be responsible for the adjacent
    pixels (that is, there may be only one hot pixel).

    I did some more tests today, to make sure there were no light leakage, but
    the pixels are still there. If I take a picture of my livingroom wall using
    f/22 and 10 seconds, the red spot is there. So it's definitely something
    wrong with the sensor.

    Looks like it's going back to the shop. :-(
    Nick, Nov 18, 2003
    #11
  12. "Mr Blobby" <> skrev i meddelandet
    news:FClub.15731$...
    > Doh !
    >
    >
    > Of course !
    >
    > Thanks.
    >
    >
    > "Dave Martindale" <> wrote in message
    > news:bpcmm8$1pt$...
    > > "Mr Blobby" <> writes:
    > > >Thanks for that, does that mean they follow the planets line ?

    > >
    > > I don't think so. The geostationary satellites all have to be in the
    > > Earth's equatorial plane. The planets are (mostly) found in the solar
    > > system's plane. But the earth's axis of rotation is tilted 23 degrees
    > > with respect to the solar system's plane, so the 2 planes are separated
    > > by that much angle.


    What is deadpixeltest.exe ?

    Where is it found?


    Morgan O.
    Morgan Ohlson, Nov 18, 2003
    #12
  13. Nick

    Nick Guest

    Nick, Nov 18, 2003
    #13
  14. Nick

    Phil Guest

    Nick wrote:

    > I did some more tests today, to make sure there were no light leakage, but
    > the pixels are still there. If I take a picture of my livingroom wall using
    > f/22 and 10 seconds, the red spot is there. So it's definitely something
    > wrong with the sensor.
    >
    > Looks like it's going back to the shop. :-(
    >


    Sounds like the right decision, Nick.

    Phil
    Phil, Nov 18, 2003
    #14
  15. In article <8Pmub.3749$%W3.20100@amstwist00>, "Nick" <>
    wrote:

    > "Phil" <> wrote in message
    > news:5Mkub.86374$...
    > > I'm not sure I agree. My first 300D had a true, undeniable hot pixel --
    > > and it showed up at all exposure speeds, even 1/1000 sec. And it was
    > > visible as a bright point in all images where it was in an area with a
    > > darkish background (e.g., grass).
    > >
    > > A nunber of factors MAY result in apparent hot pixels with long exposures:
    > >
    > > o How warm was the camera/CMOS when the test was done?
    > >
    > > o Was there light leakage from the viewfinder (i.e., was the VF cover
    > > supplied with the camera in place)?
    > >
    > > I was also curious at what file was tested in DeadPixelTest. The
    > > versions I've used deal with jpeg or tiff, but not RAW. What file was
    > > really tested.
    > >
    > > So I'm not quite ready to conclude that this is a "hot pixel" situation
    > > .. vs. noise or light leakage or a conversion artifact.

    >
    > The camera was room temperature, and I converted a RAW image taken with ISO
    > 100 to TIFF for use in DeadPixelText. So there should be no conversion
    > problem, but the software in the 300D may be responsible for the adjacent
    > pixels (that is, there may be only one hot pixel).
    >
    > I did some more tests today, to make sure there were no light leakage, but
    > the pixels are still there. If I take a picture of my livingroom wall using
    > f/22 and 10 seconds, the red spot is there. So it's definitely something
    > wrong with the sensor.
    >
    > Looks like it's going back to the shop. :-(
    >
    >
    >


    It sounds like a stuck pixel if it shows in such short exposures.
    That's a good reason for returning it.
    Kevin McMurtrie, Nov 19, 2003
    #15
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. John Wilson

    Sigh... Canon 10D and hot pixels. Again.

    John Wilson, Aug 14, 2003, in forum: Digital Photography
    Replies:
    6
    Views:
    504
    Jim Townsend
    Aug 15, 2003
  2. PeterH

    How to check for dead pixels on Canon 300D??

    PeterH, Apr 22, 2004, in forum: Digital Photography
    Replies:
    3
    Views:
    2,691
    PeterH
    Apr 23, 2004
  3. mike

    Canon S60 - hot pixels

    mike, Jul 29, 2004, in forum: Digital Photography
    Replies:
    2
    Views:
    604
  4. BD

    Question: Canon Rebel and hot pixels...

    BD, Jul 25, 2005, in forum: Digital Photography
    Replies:
    4
    Views:
    338
    Dave R knows who
    Jul 25, 2005
  5. kl_tom

    Can hot pixels become dead pixels?

    kl_tom, Oct 4, 2006, in forum: Digital Photography
    Replies:
    4
    Views:
    523
    Bill Funk
    Oct 5, 2006
Loading...

Share This Page