Professional Web Applications Themes

determining native iso of sensor - Photography

Is there a way to determine the native ISO of the sensor in my Canon 20D? From what I've heard, it isn't necessarily the lowest ISO setting. Someone said the sensor in the 10D was best at 200 and that 100 was actually a slight interpolation....

  1. #1

    Default determining native iso of sensor

    Is there a way to determine the native ISO of the sensor in my Canon
    20D? From what I've heard, it isn't necessarily the lowest ISO
    setting. Someone said the sensor in the 10D was best at 200 and that
    100 was actually a slight interpolation.
    drs@canby.com Guest

  2. #2

    Default Re: determining native iso of sensor

    com wrote in news:com:
     

    As I understand it - the ISO number of a CCD sensor is determined
    by the white clipping point, just as it is done for slide film.

    For low contrast subjects you can then use a lower ISO number.


    /Roland
    Roland Guest

  3. #3

    Default Re: determining native iso of sensor

    On Sun, 13 Feb 2005 08:51:10 -0800, com wrote:
     


    That's a bit like asking "what is the native volume setting on my stereo?"

    ISO in a digital camera is just the calibration of the system as compared with
    film... there is no 'native' ISO.

    For example, if an ISO 100 film has a perfect gray scale image with a certain
    light and an F stop and a shutter speed and lens... then that same photo when
    taken with a digital camera (and all the same settings) will mean the camera is
    set to ISO 100... in other words, the engineer will adjust the gain of the
    prototype camera to get the same picture and then gain is said to be set to ISO
    100...

    Normally the lowest ISO number means the best signal to noise ratio and the best
    picture, but I imagine that a camera could be rated to be better at a higher
    number, due to various compromises, you need to know exactly why someone said
    200 is better...

    Bob Guest

  4. #4

    Default Re: determining native iso of sensor

    In message <133.1.4>,
    Roland Karlsson <com> wrote:
     
    >
    >As I understand it - the ISO number of a CCD sensor is determined
    >by the white clipping point, just as it is done for slide film.
    >
    >For low contrast subjects you can then use a lower ISO number.[/ref]

    I think that drs' question is actually something like, "is the 20D's ISO
    100 actually its ISO 200, pulled one stop", implying a loss of one stop
    of highlights in the RAW data, to which the answer is "no". It's "no"
    for the 10D, also; both have full dynamic range at ISO 100.

    --

    <>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
    John P Sheehy <komm> 
    JPS@no.komm Guest

  5. #5

    Default Re: determining native iso of sensor

    com wrote:
     

    Native? It's a sensor that 'aculates' a 'charge' of light over time.

    For example, it does not care if the exposure time is 1/4000 or 40 seconds. At
    some point there isn't enough light ("zero" aclation) and at some point there
    is too much (saturated -> burt highlights).

    What you might be referring to is "at what ISO setting is the gain set to 1" (0
    dB). Even then I doubt the answer is that simple.

    Best ISO? I believe if you go to dpreview and look at the noise graphs you'll
    find the D20 is least noisy at ISO 100.
    http://www.dpreview.com/reviews/konicaminolta7d/page18.asp

    Cheers,
    Alan

    --
    -- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
    -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
    -- [SI] gallery & rulz: http://www.pbase.com/shootin
    -- e-meil: there's no such thing as a FreeLunch.
    Alan Guest

  6. #6

    Default Re: determining native iso of sensor

    Alan Browne wrote: 
    >
    >
    > Native? It's a sensor that 'aculates' a 'charge' of light over time.
    >
    > For example, it does not care if the exposure time is 1/4000 or 40
    > seconds. At some point there isn't enough light ("zero" aclation)
    > and at some point there is too much (saturated -> burt highlights).
    >
    > What you might be referring to is "at what ISO setting is the gain set
    > to 1" (0 dB). Even then I doubt the answer is that simple.
    >
    > Best ISO? I believe if you go to dpreview and look at the noise graphs
    > you'll find the D20 is least noisy at ISO 100.
    > http://www.dpreview.com/reviews/konicaminolta7d/page18.asp[/ref]

    Then using the term "native" in this context is quite clear to most. See
    JP Sheehy's reply also.
    --
    John McWilliams
    John Guest

  7. #7

    Default Re: determining native iso of sensor

    On Sun, 13 Feb 2005 21:01:51 GMT, komm wrote:
     

    Thanks for the info. I know so little about this that I can't word the
    question accurately. But you're right, I was referring to a suggestion
    that ISO 200 might be less noisy than 100 on the 10D and 20D, and the
    person used the term native to describe that best setting.

    Another question: Do sensors show a lot of variation between
    individual units? i.e. might the one in my 20D be quite different from
    another 20D?
    drs@canby.com Guest

  8. #8

    Default Re: determining native iso of sensor


     

    As a general rule, the lowest ISO will yield the lowest noise and the best
    possible signal to noise ratio. The inherent signal to noise ratio of any
    sensor is only degraded by adding gain after the sensor. It's a basic
    engineering principle. There can be an improvement in signal to noise ratio
    by adding gain before a sensor, but never after the sensor. Boosting ISO on
    a digital camera simply adds gain after the sensor.
     

    On average, the difference from sensor to sensor would be minimal. However,
    there could be local problems with any given sensor. These include dead
    pixels, stuck pixels, etc.


    Charles Guest

  9. #9

    Default Re: determining native iso of sensor

    In message <com>,
    com wrote:
     

    I don't know, but if it did, the issue would be obscured by calibration
    of the amplifiers. We can't see what the actual levels are in the
    sensor, as all we can see is the RAW data that is digitized from it.
    Some people's 20Ds seem prone to banding with horizontal bands about 9
    pixels aparts (a 9-pixel cycle), but this may have more to do with the
    read-out circuitry than the sensor itself.
    --

    <>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
    John P Sheehy <komm> 
    JPS@no.komm Guest

  10. #10

    Default Re: determining native iso of sensor

    John McWilliams wrote:
     
    >>
    >>
    >>
    >> Native? It's a sensor that 'aculates' a 'charge' of light over time.
    >>
    >> For example, it does not care if the exposure time is 1/4000 or 40
    >> seconds. At some point there isn't enough light ("zero" aclation)
    >> and at some point there is too much (saturated -> burt highlights).
    >>
    >> What you might be referring to is "at what ISO setting is the gain set
    >> to 1" (0 dB). Even then I doubt the answer is that simple.
    >>
    >> Best ISO? I believe if you go to dpreview and look at the noise
    >> graphs you'll find the D20 is least noisy at ISO 100.
    >> http://www.dpreview.com/reviews/konicaminolta7d/page18.asp[/ref]
    >
    >
    > Then using the term "native" in this context is quite clear to most. See
    > JP Sheehy's reply also.[/ref]

    I did, and JPS' posts are always worth reading attentively.

    Unless you have the engineering data and know at what point _all_ the gains are
    set to 1, then you don't know what the "native" sensitivity is. And I would bet
    strongly that it is not a nicely rounded number like 100 or 200 or 50.

    Since the D20 is least noisy at ISO 100 (see link above) that is the sensitivity
    one should aim to use, light, lens and time allowing.

    In short, the term 'native' just doesn't and can't apply (IMO).

    Cheers,
    Alan

    --
    -- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
    -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
    -- [SI] gallery & rulz: http://www.pbase.com/shootin
    -- e-meil: there's no such thing as a FreeLunch.
    Alan Guest

  11. #11

    Default Re: determining native iso of sensor

    Alan Browne <ca> wrote in news:cuqldi$o2v$1
    inews.gazeta.pl:
     

    Yes - it can. The native ISO number is determined by the white
    clipping level and this is the lowest ISO you can use at
    normal contrast pictures without clipping. You can then increase
    the ISO by amplifying the signal before A/D conversion until
    you no longer can stand the noise.


    /Roland
    Roland Guest

  12. #12

    Default Re: determining native iso of sensor

    Roland Karlsson wrote:
     
    >
    >
    > Yes - it can. The native ISO number is determined by the white
    > clipping level and this is the lowest ISO you can use at
    > normal contrast pictures without clipping. You can then increase
    > the ISO by amplifying the signal before A/D conversion until
    > you no longer can stand the noise.[/ref]

    The likelyhood that the native sensitivity in ISO is user selectable is
    unlikely, and that seems to be what the poster is looking for. As the lowest
    ISO yields the lowest noise, that would appear to be the best setting for him
    for lowest noise images providing he has the aperture and time to do it.

    --
    -- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
    -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
    -- [SI] gallery & rulz: http://www.pbase.com/shootin
    -- e-meil: there's no such thing as a FreeLunch.
    Alan Guest

  13. #13

    Default Re: determining native iso of sensor

    In message <cuqldi$o2v$gazeta.pl>,
    Alan Browne <ca> wrote:
     

    I doubt that it's as straightforward as "there's a voltage is the sensor
    itself, and this voltage is used directly at ISO 100, and all other ISOs
    have to amplify the signal by 2x, 4x, 8x, etc". They most likely all go
    throught the same process, with the 1 of 1x just being a point of
    reference, and may actually be 0.13 or 57 or 1311.
    --

    <>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
    John P Sheehy <komm> 
    JPS@no.komm Guest

  14. #14

    Default Re: determining native iso of sensor

    In message <133.1.4>,
    Roland Karlsson <com> wrote:
     
    >
    >Yes - it can. The native ISO number is determined by the white
    >clipping level and this is the lowest ISO you can use at
    >normal contrast pictures without clipping.[/ref]

    .... but why not just call this the minimum full-DR ISO? Native sounds
    like everything else is artificial, and a host of other connotations,
    and so much is this is purely arbitrary. My 20D meters the same as my
    Sekonic 558 meter, and I get dark images from it, unless I set the EC to
    +1. My 10D meters like the Sekonic at ISO 64, at "ISO 100". Both give
    dark images that waste dynamic range in JPEG, and a stop more in RAW.
    The 10D can handle it's "ISO 100" shots with a light meter set to ISO 32
    with slide-like headroom (with a little extra in the red channel). You
    can call any level in the 0-4095 (often only 0-4007 in some ISO modes in
    some cameras) "average grey", and render your gamma-corrected image from
    there. In fact, that's what all digital cameras do with separate color
    channels.

    If there were some ISO value that could be ascribed to a white object at
    the clipping point of the lowest full-DR ISO, I doubt it would be 100
    times a power of 2.
     

    Do you actually think that the sensor data goes straight to the ADC
    circuit at "the native ISO", and that the other ISOs go through a
    different route with an amplifier that the "native" doesn't pass
    through?
    --

    <>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
    John P Sheehy <komm> 
    JPS@no.komm Guest

  15. #15

    Default Re: determining native iso of sensor

    In message <cur4rp$6ji$gazeta.pl>,
    Alan Browne <ca> wrote:
     

    Of course, many cameras have extensions to the ISO range which are
    compromised. Most "ISO 3200" modes on DLRS are just the camera pushing
    amplified ISO 1600 to 3200 for you automatically. ISO 50 on the Canon
    1D mk II and 1Ds mk II are just ISO 100 pulled for you by the camera,
    with a loss of a stop of headroom (I wonder if they posterize the data,
    or contain a tag in the RAW file to expose what would otherwise be ISO
    100 data a stop darker). I shoot like this on my 10D and 20D all the
    time, even though the LCD doesn't say, "ISO 50". It says "ISO 100" and
    "+1 EC" instead. On the other end of the spectrum, I don't set my 20D
    to ISO 3200 anymore. I set it to ISO 1600 and -1 EC instead, which
    gives the same quality in the mutual dynamic range, plus an extra stop
    for specular highlights and light sources. The only exception is that
    bad pixels are interpolated with higher resolution at "ISO 3200", which
    is trivial (all 20D ISO 3200 RAW values are even, except some of the
    interpolated bad pixels).

    If I am drifting a bit off topic, I just woke up from an evening nap and
    don't have my bearings straight yet.
    --

    <>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
    John P Sheehy <komm> 
    JPS@no.komm Guest

  16. #16

    Default Re: determining native iso of sensor

    komm wrote: 
    >
    >
    > I doubt that it's as straightforward as "there's a voltage is the sensor
    > itself, and this voltage is used directly at ISO 100, and all other ISOs
    > have to amplify the signal by 2x, 4x, 8x, etc". They most likely all go
    > throught the same process, with the 1 of 1x just being a point of
    > reference, and may actually be 0.13 or 57 or 1311.[/ref]

    One thing that is being overlooked in the ISO rating is that 20Ds (just
    the same as 10Ds before them) lose definition at higher ISO settings
    until it is impossible to see detail in an otherwise sharply focused
    image at 800 and higher ISO when you look at the actual pixel size on a
    monitor. Sometimes the image will print 'looking' sharper but not
    anywhere near as sharp as a 100 ISO similar shot. Just another reason to
    use fast glass with digitals.

    JD

    --
    EOS my GOD,
    Give me ISO for I have not yet seen the light.
    Take away my grain, give me colour and you
    shall have given me the edge!
    Deciple Guest

  17. #17

    Default Re: determining native iso of sensor

    In message <42118600$tpgi.com.au>,
    Deciple of EOS <com> wrote:
     

    Who is overlooking that? It's not mentioned often, because it is taken
    for granted; all other things being equal, higher ISOs are noisier and
    therefore harder to see pixel-to-pixel detail in.
     

    True, but not all fast lenses are sharp wide open, nor is the depth of
    field sufficient for all uses.

    People who avoid high ISOs at all costs often ruin their pictures. In
    DPReview, just last night, someone started a thread asking why his ISO
    100 shot of a gull on a beach was so "noisy". I looked at the image,
    and it looked more like over-quantization than sensor noise. I then
    looked at the EXIF data, and it was shot at 1/8000 f/2.8. That's
    sunny-f/25, and I know that you need sunny-f/11 for a decent exposure on
    the 10D or 20D (and it was obviously taken uner a hazy sky, so sunny-f/8
    would probably be better). The image was under-exposed by about 5
    stops, probably using no more than 4 to 6 of the 12 bits in the RAW
    data. Maybe he thought that ISO 100 would give him the best picture,
    just a bit too hard.
    --

    <>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
    John P Sheehy <komm> 
    JPS@no.komm Guest

  18. #18

    Default Re: determining native iso of sensor

    komm wrote:
     
    >
    >
    > I doubt that it's as straightforward as "there's a voltage is the sensor
    > itself, and this voltage is used directly at ISO 100, and all other ISOs
    > have to amplify the signal by 2x, 4x, 8x, etc". They most likely all go
    > throught the same process, with the 1 of 1x just being a point of
    > reference, and may actually be 0.13 or 57 or 1311.[/ref]

    That's really what I was getting to. There may be a point within ISO 100 to
    3200 where the gain is 1 (pre A/D), it may be that the unity gain is at ISO
    39.07... who knows, and it ain't important in the end.

    Cheers,
    Alan.


    --
    -- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
    -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
    -- [SI] gallery & rulz: http://www.pbase.com/shootin
    -- e-meil: there's no such thing as a FreeLunch.
    Alan Guest

  19. #19

    Default Re: determining native iso of sensor

    komm wrote:
     
    >
    >
    > Of course, many cameras have extensions to the ISO range which are
    > compromised. Most "ISO 3200" modes on DLRS are just the camera pushing
    > amplified ISO 1600 to 3200 for you automatically. ISO 50 on the Canon
    > 1D mk II and 1Ds mk II are just ISO 100 pulled for you by the camera,
    > with a loss of a stop of headroom (I wonder if they posterize the data,
    > or contain a tag in the RAW file to expose what would otherwise be ISO
    > 100 data a stop darker). I shoot like this on my 10D and 20D all the
    > time, even though the LCD doesn't say, "ISO 50". It says "ISO 100" and
    > "+1 EC" instead. On the other end of the spectrum, I don't set my 20D[/ref]

    in "100 +1 EC" is the shutter period increased ("A" mode) or is the image sensor
    'gained'? If it just increases the exposure time, then it has nothing to do
    with the sensor gain.
     

    This suggests that you're simply over/underexposing, not actually changing the
    gain of the sensor. Eg, you're managing the limit cases optimally knowing what
    you're likely to get during RAW conversion.
     

    Oil your bearings. You're drifting a bit, but hey, it's usenet.

    Cheers,
    Alan
    --
    -- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
    -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
    -- [SI] gallery & rulz: http://www.pbase.com/shootin
    -- e-meil: there's no such thing as a FreeLunch.
    Alan Guest

  20. #20

    Default Re: determining native iso of sensor

    On Tue, 15 Feb 2005 14:59:18 GMT, komm wrote:
     

    After reading all these posts, I realize that the notion of a "native"
    ISO is at least far more complicated than I thought. And possibly a
    misnomer. The idea that ISO 200 on the 20D might produce less noise
    than any other setting is incorrect. So I'm back to keeping it simple:
    for reduced noise, shoot at as low an ISO as possible, which is what I
    thought before I heard about native ISO. But now I have to look up
    "over-quantization."
    drs@canby.com Guest

Page 1 of 2 12 LastLast

Similar Threads

  1. Dust on sensor
    By Lars in forum Photography
    Replies: 25
    Last Post: March 5th, 10:33 PM
  2. Replies: 45
    Last Post: February 13th, 12:41 AM
  3. First exprience with dust on sensor
    By Sheldon in forum Photography
    Replies: 4
    Last Post: February 12th, 05:37 AM
  4. Sensor cleaning
    By Musty in forum Photography
    Replies: 53
    Last Post: January 15th, 07:00 AM
  5. Nikon's new image sensor
    By J. in forum Photography
    Replies: 3
    Last Post: July 25th, 03:39 PM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139