Sigma SD9 Photos Reviewed

Discussion in 'Digital Cameras' started by Frank Weston, Dec 27, 2003.

  1. Frank Weston

    Mike Engles Guest


    Hello

    I don't own anything, except a LS 40 scanner.
    With the exception of you, I get the impression that no user other than
    Canon post in these conversations.

    It is a religious war, where each side is intolerant of the preferences
    of the other.
    As far as I can see there is no right or wrong. It is really a matter of
    taste. My eyes prefer images made by the Sigma to images made by Canons.
    My eyes do not like the blurriness of Canons, or the colour artifacts of
    the Sigmas.

    On balance I prefer Sigma, but I also think that images from Canon10ds
    can be very impressive.

    Mike Engles

    Mike Engles
     
    Mike Engles, Jan 14, 2004
    1. Advertisements

  2. Frank Weston

    JPS Guest

    In message <>,
    As a recording of the scene, it is flat out *WRONG*. It pretends to
    show detail, but you can not really surmise the actual shape of any
    small details from the image. All the edges are in the wrong place, and
    many of them are missing. It is wrong.
    --
     
    JPS, Jan 14, 2004
    1. Advertisements

  3. I wouldn't mention their Pornaroid colors in one sentence
    with value. One has to be pretty preddy to do so.
     
    Michael Quack, Jan 15, 2004
  4. So does the Foveon.

    But in comparison to Bayer sensors the manufacturing
    problems resulting in varying layer thickness of X3
    chips, (varying over the area of each sensor!) make
    it impossible to really calibrate the X3.

    Which explains many of their color problems. Bayer has
    much better defined basis for their interpolation than
    Sigma has.
     
    Michael Quack, Jan 15, 2004
  5. Frank Weston

    Mike Engles Guest

    Hello

    I have had another look ath the image in Photoshop.
    It is excessively oversharpened. I cannot see how you are so intimate
    with the scene that you KNOW what is right and wrong. You are making a
    guess at best. There might be some false detail as you say,but it is
    impossible to say what it is. I would prefer some generally unkownable
    false detail in certain parts of a image, than a sense of no detail at
    all. That of course is my taste. A great deal of photography is more
    about mood, than absolute truth. It is the mood created, that gives the
    WOW factor. Some of the best photographs have no relation to any
    reality, but they make you say WOW!.

    Mike Engles
     
    Mike Engles, Jan 15, 2004
  6. Frank Weston

    JPS Guest

    In message <>,
    It doesn't matter; the sharpness is inconsistent, and that comes right
    from the original capture.
    I can't see how you can be so blind as not to see all the long, straight
    lines of the same color that don't exist in real, complex scenes like
    this one.
    What do you mean impossible? How many plants have you seen that look
    like Legos? How many walls? How many flowers?

    The bricks and shingles have extremely well defined edges that are
    uncannily aligned with the pixel grid, and in other areas, you can
    barely see the mortar and overlapping shingles. There's one spot in the
    picture where a line of solid color runs between several rows of stones.
    How could that be real?
    Maybe you have a cognitive or perceptive defect; I don't know. It is
    blaringly obvious that the picture has anti-aliasing everywhere, and I
    can tell more about what's really there with the softer default 10D
    images, especially if sharpened with a tiny radius. If I see two pixels
    that are whiter than their neigbors, and one is brighter than the other;
    I can interpolate exactly where the center of light "mass" is, with a
    filtered image. With the unfiltered image, most likely, one or both
    will be white, and everything else black. The pixel grid distorts the
    actual location and borders of objects when antialiasing occurs.
    If you shot that same scene with a Canon 1Ds, and used nearest neighbor
    to downsize to 2268*1512, it would probably look as good or better in
    terms of the qualities you are looking for. Personally, I would use
    Lanczos, and then sharpen it a tad.

    You're talking about art, not recording a scene. This is a technology
    newsgroup, rec.photo.DIGITAL. It is not an art forum, per se. A person
    with a camera can take artistic photos, or just use it for making an
    accurate record of a scene. It is the recording quality that I am
    addressing here; some great photos have been taken with vaseline on a
    filter; that doesn't make the camera that did it a good recording
    camera.

    Just say you like the effect in some cases, and admit that it is
    technically a poor reproduction of the scene.
    --
     
    JPS, Jan 15, 2004
  7. Frank Weston

    Steve m... Guest

    I love my Digital Rebel. Takes the best pictures I've ever had. I
    considered the SD-9 for about an instant then I saw that you just saying how
    "great it is". Wrong. I bought a bayer sensor. I like it better. Canon
    is a good name when you are spending $1000. I wouldn't consider a Sigma
    because of that.

    Steve m...
     
    Steve m..., Jan 16, 2004
  8. Frank Weston

    Lisa Horton Guest

    Right. Except for all those who post that they are very happy with
    their Canon DSLR's. Which would be virtually all of them. Except for
    virtually all Canon users, yes, the "group" is disappointed.

    Lisa
     
    Lisa Horton, Jan 16, 2004
  9. Frank Weston

    Mike Engles Guest


    Hello

    Oversharpening and adding contrast will cause pretty much all the
    defects. That is what has happened. The USP of the sustem is to have the
    lego effect,which is not noticable or printable at 100%. I have printed
    a Sigma image at native resolution and posted it. It prints extremely
    well. What the print has also rather too well is the green fringing in
    ceratin parts. This I do not like. As far as I can see most other
    digital camers images have a soft plastic effect which is no where near
    the truth either.

    We will have to agree to differ.

    Mike Engles
     
    Mike Engles, Jan 16, 2004
  10. Frank Weston

    Mike Engles Guest


    Hello

    By the way the image must have aliasing, because if it had antialiasing,
    it would not have the effect you claim. That is the point if the Sigma,
    it does not have anti-alising.


    Mike Engles
     
    Mike Engles, Jan 16, 2004
  11. It is. We've been over this many times. See more means seeing more lens
    defects.

    The 15-30 EX is obviously the best lens in its class, and by a long shot.
    That doesn't mean it is perfect. An SD9 is not the camera to use to make a
    lens look good, you'll see every tiny defect in stunning detail, to include
    single pixel wide CA which is absolutely impossible to see with a blurry 4
    pixel resolution Bayer DSLR.

    Here's a nice explanation, written by a digtial imaging giant...
    http://forums.dpreview.com/forums/read.asp?forum=1000&message=6554320
    It can't, it doesn't interpolate, so no made up data is inserted. All the
    data is pure optical.
     
    George Preddy, Jan 17, 2004
  12. Easily viewed on pbase, since almost all Canon DSLR images are radically
    downsized. 1Ds images often appear around 2-3MP (as opposed to 10D/300D/D60
    images, which generally yield less than 1MP usable, blowing the
    Canon-troll-theory that 640-480 is all pros ever want). Shouldn't be
    surprising, 2.76MP is the 1Ds's non-interpolated optical resolution,
    afterall.

    Few downsized 1Ds images equal the sharpness and clarity Foveon's full size
    images, but it is clear a closer fight than the diminutive 10D sensor can
    muster. Although, even downized, the SD9 generally kills the (way outdated
    CMOS) Canon 1Ds, try it and see...

    http://www.pbase.com/image/24208071
    http://www.pbase.com/image/24323811

    Man, just look at that Sigma 15-30 EX sharpness at 15mm, corner to corner.
    Incredible lens. Granted, the 1Ds spreads roughly the same number of
    sensors over a full frame, so all glass performs relatively poorly on it.
     
    George Preddy, Jan 17, 2004
  13. Frank Weston

    Crownfield Guest

    nice sharpening halo.
     
    Crownfield, Jan 17, 2004
  14. That's X3FL.
    it.
     
    George Preddy, Jan 17, 2004
  15. Absolutely, a Foveon can see lens defects a Bayer has no hope of resolving.
    Seeing single pixel wide features is a pipe dream with a blurry 4 pixel
    resolution Bayer.
    The Nikkor isn't nearly as sharp, not that it matters all that much on a
    Bayer.
    Post a full size D100 image at 15mm, here's an SD9 shot. Corner to corner,
    the Sigma 15-30 EX is King Kong at 1/3rd the price of the just-ok Nikkor...

    http://www.pbase.com/image/24323811/original

    I'm sure you won't post a shot, because Nikon/Nikkor can't compete in this
    league. Not even remotely competitive with Sigma/Foveon. Anything less
    than you posting a full size shot is an admission that this is correct. And
    so obviously, it is.

    <Snipped the rest of your post, it was just sour grapes/jealousy>
     
    George Preddy, Jan 17, 2004
  16. SNIP
    Written by someone that won't admit that he is the person he suggests he is.

    Asked several times over whether he can state his relationship with Foveon,
    is answered by deafening silence. His "examples" have disappeared from
    Phase.
    He starts out almost superficially believable (leaving out stuff that could
    undermine his position of course), but gets less coherent as the discussion
    focuses on flaws in his reasoning, a lot like your MO. Hmm.

    Bart
     
    Bart van der Wolf, Jan 17, 2004
  17. is.

    He signed it. You are losing self control.
     
    George Preddy, Jan 17, 2004
  18. ?? You mean someone typed a name, what does that proof?

    Bart
     
    Bart van der Wolf, Jan 17, 2004
  19. George Preddy, Jan 17, 2004
  20. You are not making sense again. Time to take you medicine...

    Bart
     
    Bart van der Wolf, Jan 17, 2004
    1. Advertisements

Ask a Question

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

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.