Sensor noise: Nikon D200 versus Canon S70

Discussion in 'Digital Photography' started by Roger N. Clark (change username to rnclark), Sep 18, 2006.

  1. I've completed two new digital camera analyses:

    The Nikon D200 Digital Camera:
    Sensor Noise, Dynamic Range, and Full Well Analysis
    http://www.clarkvision.com/imagedetail/evaluation-nikon-d200

    Compare those results to a small pixel P&S camera:

    The Canon S70 Digital Camera:
    Sensor Noise, Dynamic Range, and Full Well Analysis
    http://www.clarkvision.com/imagedetail/evaluation-canon-s70

    I've included comparisons of the above results in
    some new figures (3b, 6 and 7) and discussion on this page:

    Digital Cameras: Does Pixel Size Matter?
    Factors in Choosing a Digital Camera
    http://www.clarkvision.com/imagedetail/does.pixel.size.matter

    Roger
     
    Roger N. Clark (change username to rnclark), Sep 18, 2006
    #1
    1. Advertising

  2. Roger N. Clark (change username to rnclark)

    Paul Rubin Guest

    "Roger N. Clark (change username to rnclark)" <> writes:
    > I've completed two new digital camera analyses:
    > The Nikon D200 Digital Camera:...
    > The Canon S70 Digital Camera:...


    Cool. Do you want similar data from other cameras?
     
    Paul Rubin, Sep 18, 2006
    #2
    1. Advertising

  3. Roger N. Clark (change username to rnclark)

    acl Guest

    Roger N. Clark (change username to rnclark) wrote:
    > I've completed two new digital camera analyses:
    >
    > The Nikon D200 Digital Camera:
    > Sensor Noise, Dynamic Range, and Full Well Analysis
    > http://www.clarkvision.com/imagedetail/evaluation-nikon-d200
    >
    > Compare those results to a small pixel P&S camera:
    >
    > The Canon S70 Digital Camera:
    > Sensor Noise, Dynamic Range, and Full Well Analysis
    > http://www.clarkvision.com/imagedetail/evaluation-canon-s70
    >
    > I've included comparisons of the above results in
    > some new figures (3b, 6 and 7) and discussion on this page:
    >
    > Digital Cameras: Does Pixel Size Matter?
    > Factors in Choosing a Digital Camera
    > http://www.clarkvision.com/imagedetail/does.pixel.size.matter
    >
    > Roger



    Roger,
    How do all the Canon cameras manage such low read noises? I did ask once
    before, and you said you also found it strange that they managed to
    reduce read noise to such low levels for their CMOS sensors. Well, it's
    equally low for this compact, which has a CCD. Any idea how this can
    happen (except by some very effective image processing before doing
    anything else to the raw data)?

    Cheers.
     
    acl, Sep 18, 2006
    #3
  4. "acl" <> wrote in message
    news:...
    SNIP
    > How do all the Canon cameras manage such low read noises?


    <http://web.canon.jp/Imaging/cmos/technology-e/noise_reduction.html>

    SNIP
    > Any idea how this can happen (except by some very effective image
    > processing before doing anything else to the raw data)?


    There may be some specific noise suppression in the DIGIC processor as
    well, but at least it is not obvious (apart from the absence of
    noise).

    --
    Bart
     
    Bart van der Wolf, Sep 19, 2006
    #4
  5. Roger N. Clark (change username to rnclark)

    acl Guest

    Hello Bart,


    Bart van der Wolf wrote:
    > "acl" <> wrote in message
    > news:...
    > SNIP
    > > How do all the Canon cameras manage such low read noises?

    >
    > <http://web.canon.jp/Imaging/cmos/technology-e/noise_reduction.html>
    >


    Right, but the S70 has a CCD, so it's not what is described there. The
    S70 has extremely low read noise (as measured by Roger) for a CCD. It's
    lower than most CCDs for scientific applications (I asked people at the
    place where I work who deal with these things about usual read noise
    levels, as I personally am clueless in practical matters).

    > SNIP
    > > Any idea how this can happen (except by some very effective image
    > > processing before doing anything else to the raw data)?

    >
    > There may be some specific noise suppression in the DIGIC processor as
    > well, but at least it is not obvious (apart from the absence of
    > noise).


    I didn't mean the same sort of thing that Neat Image (say) does. But I
    always had trouble understanding how else they get the read noise so
    low (I didn't have quantitative data, but looking at high ISO images in
    the shadows it looked quite incredible). I think it is clear that some
    very clever and effective processing is taking place.

    At least, I can't think of any other explanation. But this is still
    strange, how is it possible that Canon come up with such an algorithm
    but nobody else?
     
    acl, Sep 19, 2006
    #5
  6. Paul Rubin wrote:

    > "Roger N. Clark (change username to rnclark)" <> writes:
    >
    >>I've completed two new digital camera analyses:
    >> The Nikon D200 Digital Camera:...
    >> The Canon S70 Digital Camera:...

    >
    > Cool. Do you want similar data from other cameras?


    Yes, I would. Others are doing these analyses too, and I've
    collected the information and summarized it in Tables 1-4
    here:
    http://www.clarkvision.com/imagedetail/digital.signal.to.noise

    But most efforts are by amateur astronomers trying for
    cameras more suited to low light work.
    I would be interested in filling in more data
    between the 2.3 and 6 micron pixel range, and if any camera
    comes out much below 2.3 microns.

    A requirement is that the camera must put out raw data.
    Taking the data is pretty easy. Another requirement
    is that there is software to get the linear raw data
    from the raw file. Then it takes me several days
    of work to analyze and write it up.

    Roger
     
    Roger N. Clark (change username to rnclark), Sep 19, 2006
    #6
  7. acl wrote:

    > Hello Bart,
    >
    > Bart van der Wolf wrote:
    >
    >>"acl" <> wrote in message
    >>news:...
    >>SNIP
    >>
    >>>How do all the Canon cameras manage such low read noises?

    >>
    >><http://web.canon.jp/Imaging/cmos/technology-e/noise_reduction.html>
    >>

    > Right, but the S70 has a CCD, so it's not what is described there. The
    > S70 has extremely low read noise (as measured by Roger) for a CCD. It's
    > lower than most CCDs for scientific applications (I asked people at the
    > place where I work who deal with these things about usual read noise
    > levels, as I personally am clueless in practical matters).


    Canon's S60 camera has much worse read noise, so they
    have certainly done something different on the S70. One trick
    used in CCDs is to read the signal multiple times effectively
    averaging the read noise toward zero. I don't know if
    Canon does some trick like that (the web page
    you reference does imply multiple reads).

    >>>Any idea how this can happen (except by some very effective image
    >>>processing before doing anything else to the raw data)?

    >>
    >>There may be some specific noise suppression in the DIGIC processor as
    >>well, but at least it is not obvious (apart from the absence of
    >>noise).

    >
    > I didn't mean the same sort of thing that Neat Image (say) does. But I
    > always had trouble understanding how else they get the read noise so
    > low (I didn't have quantitative data, but looking at high ISO images in
    > the shadows it looked quite incredible). I think it is clear that some
    > very clever and effective processing is taking place.
    >
    > At least, I can't think of any other explanation. But this is still
    > strange, how is it possible that Canon come up with such an algorithm
    > but nobody else?


    I don't think they are averaging pixels or doing post processing
    on the raw data. That would show in the spatial resolution,
    and it would be especially bad on stars. Nikon's median
    filter shows such effects, for example.

    Roger
     
    Roger N. Clark (change username to rnclark), Sep 19, 2006
    #7
  8. Roger N. Clark (change username to rnclark)

    Guest

    "Bart van der Wolf" <> writes:

    > "acl" <> wrote in message
    > news:...

    ....

    >> Any idea how this can happen (except by some very effective image
    >> processing before doing anything else to the raw data)?


    > There may be some specific noise suppression in the DIGIC processor
    > as well, but at least it is not obvious (apart from the absence of
    > noise).


    Most of it is in the much inproved photo-diode resetting in CMOS
    units, and using double correlation read-outs.

    --
    Paul Repacholi 1 Crescent Rd.,
    +61 (08) 9257-1001 Kalamunda.
    West Australia 6076
    comp.os.vms,- The Older, Grumpier Slashdot
    Raw, Cooked or Well-done, it's all half baked.
    EPIC, The Architecture of the future, always has been, always will be.
     
    , Sep 24, 2006
    #8
  9. Roger N. Clark (change username to rnclark)

    acl Guest

    wrote:
    > "Bart van der Wolf" <> writes:
    >
    >>"acl" <> wrote in message
    >>news:...

    > ...
    >
    >>>Any idea how this can happen (except by some very effective image
    >>>processing before doing anything else to the raw data)?

    >
    >>There may be some specific noise suppression in the DIGIC processor
    >>as well, but at least it is not obvious (apart from the absence of
    >>noise).

    >
    > Most of it is in the much inproved photo-diode resetting in CMOS
    > units, and using double correlation read-outs.
    >


    However, the S70, which has a CCD, also has extremely low readout noise.
    This was the point of the question.
     
    acl, Sep 25, 2006
    #9
    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. Iris Nadine Kartasasmita

    Canon S70: aggressive noise reduction?

    Iris Nadine Kartasasmita, Mar 22, 2005, in forum: Digital Photography
    Replies:
    6
    Views:
    465
    Iris Nadine Kartasasmita
    Mar 24, 2005
  2. Rich

    Ugly noise versus "good" noise

    Rich, Oct 4, 2005, in forum: Digital Photography
    Replies:
    3
    Views:
    378
  3. ThomasH
    Replies:
    33
    Views:
    854
    Philip Homburg
    Nov 2, 2005
  4. wayne

    Nikon D200 and Canon 350D/Rebel XT Noise Tests

    wayne, Apr 4, 2006, in forum: Digital Photography
    Replies:
    2
    Views:
    339
    wayne
    Apr 5, 2006
  5. Peter Potamus the Purple Hippo

    Re: Mozilla versus IE versus Opera versus Safari

    Peter Potamus the Purple Hippo, May 8, 2008, in forum: Firefox
    Replies:
    0
    Views:
    892
    Peter Potamus the Purple Hippo
    May 8, 2008
Loading...

Share This Page