11MP digital or medium format film?

Discussion in 'Digital Photography' started by Beowulf, Aug 22, 2004.

  1. Beowulf

    Crownfield Guest

    Georgette Preddy wrote:
    >
    > Roland Karlsson <> wrote in message news:<Xns9552C35F5CB6klotjohan@130.133.1.4>...
    > > "dylan" <> wrote in news:GDCXc.33$:
    > >
    > > > From the tests I've seen 11MP is as good as, if not slightly better than
    > > > 35mm (start debate here), but not equivalent to Medium Format.

    > >
    > > High quality 11MP digital is much smoother than 35 mm film and it
    > > is in the same league as medium format.

    >
    > I agree with that, but there are no 11MP Bayers camera due to doubling
    > green by reducing red and blue, their highly inefficient 2D design
    > leaves no choice--4 slots for 3 primaries. Even the 14n/c is only
    > 3.3MP optical before digital interpolation, that is less than the
    > 10.3MP SD10.


    actually, 3 mp. and you call it 10, 13, what next, 22?
    Crownfield, Aug 29, 2004
    #41
    1. Advertising

  2. Crownfield <> wrote in message news:<>...
    > Georgette Preddy wrote:
    > >
    > > Roland Karlsson <> wrote in message news:<Xns9552C35F5CB6klotjohan@130.133.1.4>...
    > > > "dylan" <> wrote in news:GDCXc.33$:
    > > >
    > > > > From the tests I've seen 11MP is as good as, if not slightly better than
    > > > > 35mm (start debate here), but not equivalent to Medium Format.
    > > >
    > > > High quality 11MP digital is much smoother than 35 mm film and it
    > > > is in the same league as medium format.

    > >
    > > I agree with that, but there are no 11MP Bayers camera due to doubling
    > > green by reducing red and blue, their highly inefficient 2D design
    > > leaves no choice--4 slots for 3 primaries. Even the 14n/c is only
    > > 3.3MP optical before digital interpolation, that is less than the
    > > 10.3MP SD10.

    >
    > actually, 3 mp. and you call it 10, 13, what next, 22?


    The SD10 is exactly 13.72MP as output by the camera, interpoloted to
    the precise degree as all Bayer images are interpolated by default.
    The 1Ds is precisely 2.74MPs, before interpolation.
    George Preddy, Aug 31, 2004
    #42
    1. Advertising

  3. Roland Karlsson <> wrote in message news:<Xns9554BDEFEEDFEklotjohan@130.133.1.4>...
    > (Georgette Preddy) wrote in
    > news::
    >
    > > I agree with that, but there are no 11MP Bayers camera due to doubling
    > > green by reducing red and blue, their highly inefficient 2D design
    > > leaves no choice--4 slots for 3 primaries. Even the 14n/c is only
    > > 3.3MP optical before digital interpolation, that is less than the
    > > 10.3MP SD10.
    > >

    >
    > How nice of you to tell us. Unfortunately, for you, no one
    > believes in your math.


    You don't need to "believe in" math unless you use an interpolated
    camera. \

    The math is simple, a 6MP-monochrome sensor doesn't have anywhere near
    the full color resolution of a 3.43MP-full-color sensor. A has 230%
    of the RGB triples of a tiny 10D sensor. It has 170% of the RGB
    triples of the outdated 1D Mk II sensor. And the SD9 has a whopping
    130% of the RGB triples of the smallish $8000 Canon 1Ds sensor.
    George Preddy, Aug 31, 2004
    #43
  4. Beowulf

    Annika1980 Guest

    >From: (George Preddy)

    >The math is simple, a 6MP-monochrome sensor doesn't have anywhere near
    >the full color resolution of a 3.43MP-full-color sensor.


    Go look up the definition of "resolution."
    Then link us to those 100's of full-size images you've posted. Tons of them.
    Is that interpolated? More like extrapolated.
    Annika1980, Aug 31, 2004
    #44
  5. Beowulf

    Crownfield Guest

    George Preddy wrote:
    >
    > Crownfield <> wrote in message news:<>...
    > > Georgette Preddy wrote:
    > > >
    > > > Roland Karlsson <> wrote in message news:<Xns9552C35F5CB6klotjohan@130.133.1.4>...
    > > > > "dylan" <> wrote in news:GDCXc.33$:
    > > > >
    > > > > > From the tests I've seen 11MP is as good as, if not slightly better than
    > > > > > 35mm (start debate here), but not equivalent to Medium Format.
    > > > >
    > > > > High quality 11MP digital is much smoother than 35 mm film and it
    > > > > is in the same league as medium format.
    > > >
    > > > I agree with that, but there are no 11MP Bayers camera due to doubling
    > > > green by reducing red and blue, their highly inefficient 2D design
    > > > leaves no choice--4 slots for 3 primaries. Even the 14n/c is only
    > > > 3.3MP optical before digital interpolation, that is less than the
    > > > 10.3MP SD10.

    > >
    > > actually, 3 mp. and you call it 10, 13, what next, 22?

    >
    > The SD10 is exactly 13.72MP as output by the camera, interpoloted to
    > the precise degree as all Bayer images are interpolated by default.


    actually still 3 mp. nothing you do or say will change that.

    > The 1Ds is precisely 2.74MPs, before interpolation.


    actually, no. but you know that also.
    you just seem to lack any sense of honesty.
    Crownfield, Aug 31, 2004
    #45
  6. The thread goes the wrong way:

    11 Mpix is 11 Mpix in luminance (black and white). The different colors have different sensitivity to luminance, but this
    is calibrated by the image processing. Only the colors are interpolated. So, a 11 Mpix sensor gives only about 4 Mpix
    color, which is interpolated and added to the black and white image. That is no harm because the human eye has also a
    reduced resolution capacity for colors. Also: 11 Mpix = 11 Mpix. Look at your pictures and see!

    AFH

    George Preddy schrieb:

    > Crownfield <> wrote in message news:<>...
    > > Georgette Preddy wrote:
    > > >
    > > > Roland Karlsson <> wrote in message news:<Xns9552C35F5CB6klotjohan@130.133.1.4>...
    > > > > "dylan" <> wrote in news:GDCXc.33$:
    > > > >
    > > > > > From the tests I've seen 11MP is as good as, if not slightly better than
    > > > > > 35mm (start debate here), but not equivalent to Medium Format.
    > > > >
    > > > > High quality 11MP digital is much smoother than 35 mm film and it
    > > > > is in the same league as medium format.
    > > >
    > > > I agree with that, but there are no 11MP Bayers camera due to doubling
    > > > green by reducing red and blue, their highly inefficient 2D design
    > > > leaves no choice--4 slots for 3 primaries. Even the 14n/c is only
    > > > 3.3MP optical before digital interpolation, that is less than the
    > > > 10.3MP SD10.

    > >
    > > actually, 3 mp. and you call it 10, 13, what next, 22?

    >
    > The SD10 is exactly 13.72MP as output by the camera, interpoloted to
    > the precise degree as all Bayer images are interpolated by default.
    > The 1Ds is precisely 2.74MPs, before interpolation.
    A.F. Hobbacher, Aug 31, 2004
    #46
  7. (George Preddy) wrote in
    news::

    > You don't need to "believe in" math unless you use an interpolated
    > camera. \


    I do believe in math - but I have not the slightest idea what an
    interpolated camera is. Hmmmm ... an interpolated car? Hmmm ...
    an interpolated boat? Nope ... I don't get it.

    > The math is simple, a 6MP-monochrome sensor doesn't have anywhere near
    > the full color resolution of a 3.43MP-full-color sensor. A has 230%
    > of the RGB triples of a tiny 10D sensor. It has 170% of the RGB
    > triples of the outdated 1D Mk II sensor. And the SD9 has a whopping
    > 130% of the RGB triples of the smallish $8000 Canon 1Ds sensor.


    This is not real math - it is superstition. The Church O´ Math calls
    this talking blasphemy.


    /Roland
    Roland Karlsson, Aug 31, 2004
    #47
  8. "A.F. Hobbacher" <> wrote in
    news::

    > 11 Mpix is 11 Mpix in luminance (black and white). The different
    > colors have different sensitivity to luminance, but this is calibrated
    > by the image processing. Only the colors are interpolated. So, a 11
    > Mpix sensor gives only about 4 Mpix color, which is interpolated and
    > added to the black and white image. That is no harm because the human
    > eye has also a reduced resolution capacity for colors. Also: 11 Mpix =
    > 11 Mpix. Look at your pictures and see!
    >


    We all know that. Even Mr Preddy does. But Mr Preddy like to play games.


    /Roland
    Roland Karlsson, Aug 31, 2004
    #48
  9. "A.F. Hobbacher" <> wrote in message news:<>...
    > The thread goes the wrong way:
    >
    > 11 Mpix is 11 Mpix in luminance (black and white). The different colors have different sensitivity to luminance, but this


    11M monochrome sensors only yields 11M 1/3rd luminance readings--full
    luminance requires full color. Of which 25% are discarded during the
    interpolation process which treats each RGGB set as a single full
    color pixel, plus some added influences from even farther neighbors
    (blurrier).

    The SD10 has 10.3M 1/3rd luminance readings for about $7000 less.
    More importantly it uses all of them, since it has 25% more RGB
    triples than the old $8000 1Ds.

    > is calibrated by the image processing. Only the colors are interpolated. So, a 11 Mpix sensor gives only about 4 Mpix
    > color, which is interpolated and added to the black and white image. That is no harm because the human eye has also a
    > reduced resolution capacity for colors. Also: 11 Mpix = 11 Mpix. Look at your pictures and see!
    >
    > AFH
    Georgette Preddy, Sep 2, 2004
    #49
  10. "Georgette Preddy" <> wrote in message
    news:...
    > "A.F. Hobbacher" <> wrote in message

    news:<>...
    > > The thread goes the wrong way:
    > >
    > > 11 Mpix is 11 Mpix in luminance (black and white). The different

    colors have different sensitivity to luminance, but this
    >
    > 11M monochrome sensors only yields 11M 1/3rd luminance

    readings--full
    > luminance requires full color.


    Which is an example of more of the usual nonsensical misinformation by
    the preddiot.

    The sensors are not monochrome, they are sensitive (after filtering)
    for roughly 1/3rd of the visual spectrum. The native luminance
    sensitivity of silicon based sensors is not uniform over the full
    sensitivity range from 350 nm to 1000 nm. Filters can be used to
    adjust the amount of energy from different wavelengths that reach the
    sensor, which defines it luminance sensitivity for incident light.

    Given the known absorption characteristics of the filters, multiplied
    by the sensitivity and quantum efficiency (which is wavelength and
    design dependent) of the sensor, the exact scene luminance can be
    quite accurately determined, as is witnessed by a Bayer CFA
    reconstruction:
    http://www.xs4all.nl/~bvdwolf/main/foto/bayer/bayer_cfa.htm (first
    example).
    The result may be influenced further by the use of an adequate
    Anti-aliasing filter, the absence of which will lead to aliasing
    artifacts that are lower resolution than the spatial frequencies that
    caused those artifacts.

    Bart
    Bart van der Wolf, Sep 2, 2004
    #50
  11. Beowulf

    Crownfield Guest

    the preddiot prattled:


    > The SD10 has 10.3M 1/3rd luminance readings for about $7000 less.
    > More importantly it uses all of them, since it has 25% more RGB
    > triples than the old $8000 1Ds.


    but the sd10 is really useless,
    because it has only 3 mp sensors distributed spatially in the image.
    thats a whopping ** 72% ** less.

    3 mp is not a pro camera,
    its a mini consumer toy.


    >
    Crownfield, Sep 2, 2004
    #51
  12. Beowulf

    Summitar Guest

    >George Preddy (Steve Giovanella) wrote:

    >The SD10 has 10.3M 1/3rd luminance readings for about $7000 less.
    >More importantly it uses all of them, since it has 25% more RGB
    >triples than the old $8000 1Ds.



    Shut up, Preddy.
    Summitar, Sep 5, 2004
    #52
  13. Beowulf

    Gadgets Guest

    I once heard Kodachrome 25, 35mm quoted as equivalent to 21 million pixels
    as far as resolving fine detail. Colour accuracy is a whole different
    field...

    If you've seen a poster sized Velvia print on a high gloss positive paper,
    you'd see why high end work is still film... digital has many strengths, but
    large output isn't one of them!

    Cheers, Jason (remove ... to reply)
    Video & Gaming: http://gadgetaus.com
    Gadgets, Sep 5, 2004
    #53
  14. "Gadgets" <info@gadgetaus...com> wrote in
    news:413aca86$:

    > I once heard Kodachrome 25, 35mm quoted as equivalent to 21 million
    > pixels as far as resolving fine detail. Colour accuracy is a whole
    > different field...


    Yes, you need to scan at 20 Mpixels or so to catch everything in
    a K25, but ... that does nort mean that it is equal to 20 Mpixels.
    Those pixels are extremely low quality. I think you shall make a
    realty check. Look at some high quality pixels and then at the
    pixels that a 20 Mpixel scan of K25 produces. It is a huge difference.

    > If you've seen a poster sized Velvia print on a high gloss positive
    > paper, you'd see why high end work is still film... digital has many
    > strengths, but large output isn't one of them!


    If a poster from a 35 mm Velvia looks good is it because it is a
    stunning picture and you think that film noise is beautiful. It is
    actually - and that is the main strength of film - beautiful noise.

    BTW - a poster from 35 mm film is scaled 25-50 times or so. You
    cannot scale up film that much without seeing the the noise and/or
    get non optimal sharpness.


    /Roland
    Roland Karlsson, Sep 5, 2004
    #54
  15. "Roland Karlsson" <> wrote:
    > "Gadgets" <info@gadgetaus...com> wrote:
    >
    > > I once heard Kodachrome 25, 35mm quoted as equivalent to 21 million
    > > pixels as far as resolving fine detail. Colour accuracy is a whole
    > > different field...

    >
    > Yes, you need to scan at 20 Mpixels or so to catch everything in
    > a K25, but ... that does nort mean that it is equal to 20 Mpixels.
    > Those pixels are extremely low quality. I think you shall make a
    > realty check. Look at some high quality pixels and then at the
    > pixels that a 20 Mpixel scan of K25 produces. It is a huge difference.


    Really. The sensible pages I've seen call 35mm film as being around 8MP.
    I've just started playing with a 300D, and one stupid snapshot, a mere JPEG
    with blown highlights, has some of the most beautiful pixels I've ever seen.
    I usually downsample my film scans to 1/3 the original pixel count (to
    create files that look very good printed at 300 dpi), but even at that size
    (2400 dpi), I've never seen pixels as clear and clean. (2400 dpi from 645 is
    19MP, 12 x 17 @ 300 dpi.)

    What's even more hilarious is the 35mm film types squawk about K25 and Ektar
    25 and Tech Pan, all of which have been discontinued. They talk about image
    quality, but they refused to buy enough high-res film to make the stuff
    commercially viable. To the point that there are no high-res films being
    manufactured today. None. Zilch. Zero. Sheesh.

    > > If you've seen a poster sized Velvia print on a high gloss positive
    > > paper, you'd see why high end work is still film... digital has many
    > > strengths, but large output isn't one of them!

    >
    > If a poster from a 35 mm Velvia looks good is it because it is a
    > stunning picture and you think that film noise is beautiful. It is
    > actually - and that is the main strength of film - beautiful noise.


    Grain noise looks gross to me. I think it's just that most people became
    used to it that some people think it looks good.

    However, posters from 4x5 Velvia do look good. If you can stand the
    over-the-top reds and generally dizzy color rendition.

    > BTW - a poster from 35 mm film is scaled 25-50 times or so. You
    > cannot scale up film that much without seeing the the noise and/or
    > get non optimal sharpness.


    Really. To get back to the subject, from playing with downloaded 1Ds
    samples, it looks as though I'm getting slightly more detail per frame from
    645 + Nikon 8000. But not a lot. And scanning's a pain.

    David J. Littleboy
    Tokyo, Japan
    David J. Littleboy, Sep 5, 2004
    #55
  16. "Gadgets" <info@gadgetaus...com> wrote in message
    news:413aca86$...
    > I once heard Kodachrome 25, 35mm quoted as equivalent to 21 million pixels
    > as far as resolving fine detail. Colour accuracy is a whole different
    > field...
    >
    > If you've seen a poster sized Velvia print on a high gloss positive paper,
    > you'd see why high end work is still film... digital has many strengths,

    but
    > large output isn't one of them!
    >

    Kodachrome isn't a really super fine grain, or high resolution film. Fuji
    Astia is one of the highest resolution "regular" films. The specs say 150
    lpmm. Many however will do the math of 150x24 and 150 x36 and come up with
    (3600x5400 lines) and get 19,440,000. But a grid isn't a pixel, besides many
    lenses don't resolve that high. So I would cut that by 50% so you end up
    with only around 9.7 Megapixels. So a sharp Astia slide perfectly exposed
    will edge out a digital image in their own native formats. Figure most
    regular films are lower resolution than Astia (including Velvia) you would
    end up with 6 to 8 Megapixels, assuming a lens that can resolve 92 lpmm on
    film... your milage may vary! Scanning a film means you are comparing dSLR
    with a scanner, and there will be some losses.

    Darrell Larose
    Pentax *istD & DA 14mm f:2.8
    Pentax LX
    Darrell Larose, Sep 5, 2004
    #56
  17. "David J. Littleboy" <> wrote in message
    news:cherrr$gdb$...
    >
    > "Roland Karlsson" <> wrote:
    > > "Gadgets" <info@gadgetaus...com> wrote:
    > >
    > > > I once heard Kodachrome 25, 35mm quoted as equivalent to 21 million
    > > > pixels as far as resolving fine detail. Colour accuracy is a whole
    > > > different field...

    > >
    > > Yes, you need to scan at 20 Mpixels or so to catch everything in
    > > a K25, but ... that does nort mean that it is equal to 20 Mpixels.
    > > Those pixels are extremely low quality. I think you shall make a
    > > realty check. Look at some high quality pixels and then at the
    > > pixels that a 20 Mpixel scan of K25 produces. It is a huge difference.

    >
    > Really. The sensible pages I've seen call 35mm film as being around 8MP.
    > I've just started playing with a 300D, and one stupid snapshot, a mere

    JPEG
    > with blown highlights, has some of the most beautiful pixels I've ever

    seen.
    > I usually downsample my film scans to 1/3 the original pixel count (to
    > create files that look very good printed at 300 dpi), but even at that

    size
    > (2400 dpi), I've never seen pixels as clear and clean. (2400 dpi from 645

    is
    > 19MP, 12 x 17 @ 300 dpi.)
    >
    > What's even more hilarious is the 35mm film types squawk about K25 and

    Ektar
    > 25 and Tech Pan, all of which have been discontinued. They talk about

    image
    > quality, but they refused to buy enough high-res film to make the stuff
    > commercially viable. To the point that there are no high-res films being
    > manufactured today. None. Zilch. Zero. Sheesh.
    >
    > > > If you've seen a poster sized Velvia print on a high gloss positive
    > > > paper, you'd see why high end work is still film... digital has many
    > > > strengths, but large output isn't one of them!

    > >
    > > If a poster from a 35 mm Velvia looks good is it because it is a
    > > stunning picture and you think that film noise is beautiful. It is
    > > actually - and that is the main strength of film - beautiful noise.

    >
    > Grain noise looks gross to me. I think it's just that most people became
    > used to it that some people think it looks good.
    >
    > However, posters from 4x5 Velvia do look good. If you can stand the
    > over-the-top reds and generally dizzy color rendition.
    >
    > > BTW - a poster from 35 mm film is scaled 25-50 times or so. You
    > > cannot scale up film that much without seeing the the noise and/or
    > > get non optimal sharpness.

    >
    > Really. To get back to the subject, from playing with downloaded 1Ds
    > samples, it looks as though I'm getting slightly more detail per frame

    from
    > 645 + Nikon 8000. But not a lot. And scanning's a pain.
    >
    > David J. Littleboy
    > Tokyo, Japan
    >

    Kodachrome isn't a really super fine grain, or high resolution film. Fuji
    Astia is one of the highest resolution "regular" films. The specs say 150
    lpmm. Many however will do the math of 150x24 and 150 x36 and come up with
    (3600x5400 lines) and get 19,440,000. But a grid isn't a pixel, besides many
    lenses don't resolve that high. So I would cut that by 50% so you end up
    with only around 9.7 Megapixels. So a sharp Astia slide perfectly exposed
    will edge out a digital image in their own native formats. Figure most
    regular films are lower resolution than Astia (including Velvia) you would
    end up with 6 to 8 Megapixels, assuming a lens that can resolve 92 lpmm on
    film... your milage may vary! Scanning a film means you are comparing dSLR
    with a scanner, and there will be some losses.

    So David is correct, IMHO resolution numbers of the film is useless, as the
    lens resolution and camera/subject motion will effect the end result. I
    recall seeing the resolution figures of K64 vs. K64. They were the same. K25
    has slightly finer grain 9-11 RMS vs. 10-12 RMS. The only shortcoming I have
    found in my dSLR is projecting, a well exposed chrome on a screen is still a
    delight, the "digital projectors" still have a Power Point reality, and a
    digital photo doesn't look as good projected in comparison.
    Darrell Larose, Sep 5, 2004
    #57
    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. Guest
    Replies:
    4
    Views:
    497
    Mxsmanic
    May 23, 2004
  2. Guest
    Replies:
    0
    Views:
    422
    Guest
    May 23, 2004
  3. spassig

    Epson Perfection 2400 - scanning medium format film?

    spassig, Oct 13, 2004, in forum: Digital Photography
    Replies:
    1
    Views:
    709
    Gene Palmiter
    Oct 13, 2004
  4. Rick Baker
    Replies:
    6
    Views:
    1,152
    Scott W
    Mar 17, 2006
  5. Progressiveabsolution
    Replies:
    15
    Views:
    704
    Alfred Molon
    Sep 17, 2006
Loading...

Share This Page