Sigma Digital SLR Cameras.

Discussion in 'Digital Photography' started by Biff Mullins, Jun 18, 2004.

  1. Biff Mullins

    Biff Mullins Guest

    Thinking about buying the Sigma DSLR. Either the SD9 or the SD10.
    Is the SD10 worth the extra money? What would be the best lens to start out with.

    Oh yeah, does anyone know when that Foveon point and shoot will be available?


    Biff Mullins
    Biff Mullins, Jun 18, 2004
    1. Advertisements

  2. available?

    If you need higher ISO settings, the SD10 is the better choice. If
    you shoot at lower ISO settings as I do (I'm just a causal
    photographer) than the SD9 is just fine. IMO< the 50mm EX lens is a
    great starting point with either camera.
    Peter A. Stavrakoglou, Jun 18, 2004
    1. Advertisements

  3. Biff Mullins

    D.R. Guest

    Don't waste your money.
    D.R., Jun 18, 2004
  4. Biff Mullins

    Zebedee Guest

    Before you look for a Sigma DSLR, have a look at the scathing review on about it.

    Personally, I feel that the Sigma is a capable camera let down by the poor
    Sigma lenses and the restrictive image file format.



    (Claiming asylum in an attempt
    to escape paying his debts to
    Dougal and Florence)
    Zebedee, Jun 18, 2004
  5. Biff Mullins

    TP Guest

    There is nothing scathing about that review. It is a highly
    complimentary review that expresses some reservations about the
    quality of Sigma lenses. But we knew that anyway.

    Some excerpts will show that the review is far from scathing:




    You will, however, find that this 3.4 Megapixel image blows away
    anything you've seen before in terms of image quality. I think you'll
    agree after viewing the sample photos attached to this review.


    Included with the SD10 is the excellent Sigma Photo Pro 2.0 software.
    This software should serve as an example to other camera
    manufacturers, with its great user interface and robust performance.
    You'll be spending a lot of time in Photo Pro, since the camera only
    saves photos in RAW (X3F) format. At first glance, Photo Pro looks
    like just another photo browser (see above photo). But select an image
    and you'll find a lot more functionality.


    In general, I was pleased with the results I saw using the auto
    correction feature. If you're not satisfied with how that looks, you
    can tweak the settings until the image looks like what you were

    The manual included with the SD10 is pretty good for a camera manual.
    I especially liked the lengthy explanations, and minimal fine print.


    The first thing you'll notice while looking at the back of the camera
    is just how simple everything is. There are just a few buttons and no
    dials, making it a welcome change from more complex D-SLRs.


    If you clicked on those images above, or if you know anything about
    the X3 sensor, you won't be surprised to hear my assessment of the
    image quality: it's awesome, especially after a trip through that auto
    adjustment feature. The sharpness and resolution are truly stunning


    While the Sigma SD10 would be a nice D-SLR if it used a traditional
    CCD or CMOS sensor, it's the SD10's Foveon X3 image sensor which
    really makes it stand out from the crowd. Image quality is, in a word,


    Another standout feature of the camera is the Sigma Photo Pro
    software, which you'll learn to love. That's because you must
    post-process every image, as they're all saved in RAW format.
    Thankfully, the Photo Pro software is both capable and responsive. You
    can let the camera make automatic adjustments to the image, or you can
    tweak it yourself.


    What I liked:

    Stunning image quality
    Robust performance
    FireWire and USB ports
    Uses AA batteries (instead of lithium-ion)
    Easy to use
    All the benefits of a D-SLR: lenses, flashes, and full manual controls
    Excellent Photo Pro software

    What I didn't care for:

    Poor low light focusing without external flash
    Dust seems to be a problem
    Some of Sigma's lenses are not so great
    All images must be post-processed (this may be a good thing, depending
    on your point of view)
    Above average noise at high ISOs
    TP, Jun 18, 2004
  6. Please ignore this crossposting troll behind an anonimizer.
    Bart van der Wolf, Jun 18, 2004
  7. out with.

    The SD10 is definitely worth the extra money. The best single lens to start
    with would be the 24-70 EX

    Be aware of course, that the Nikon D70 and the Canon 10D, are not only less
    expensive, but far more capable. The Sigma D-SLRs are very low resolution
    (3.4 megapixel), and have a host of issues. Furthermore, with the Nikon and
    Canon D-SLRs the amateur lenses are sufficient, while with Sigma you'll want
    to use only EX lenses.

    The acknowledged best source for information on the SD10 is
    http:/ .

    I got an e-mail from World Wide Resources that didn't answer that exact
    question. But their web site recently was updated to say September of 2004,
    a delay of three months.

    This is a 1.5 megapixel camera, so don't pay more than $100 for it.
    You're welcome GP.
    Steven M. Scharf, Jun 18, 2004
  8. Steven has an axe to grind with Sigma cameras, take his advice with a
    grain of salt.
    Peter A. Stavrakoglou, Jun 19, 2004
  9. An axe would be the perfect tool to grind a Sigma camera with.
    Randall Ainsworth, Jun 19, 2004
  10. Steven,

    Your horse manure about 3.34 MP of resolution of course invites a
    comparison. Got a 3+ MP camera from your side you would like to

    Can't wait to meet you on the other side when the 6 MP X3 Foveon chip
    is released. That should make an interesting comparison with the
    full-feature 6 MP class CFA!
    Laurence Matson, Jun 19, 2004
  11. Biff Mullins

    JPS Guest

    In message <>,
    Unless they've come up with a whole new technology, that will mean even
    worse color discrimination by the layers, or the same as current, if the
    chip is larger and the pixel pitch remains the same.
    JPS, Jun 19, 2004
  12. Biff Mullins

    Skip M Guest

    When, or if? I've heard nothing about any continued development of the
    Foveon chip, except from you. Can you point me in the direction of any
    published accounts of this?
    By the way, what does CFA stand for?
    Skip M, Jun 19, 2004
  13. Here is a nice one, there is even a 3.4MP Bayer image that shows the
    cameras, it's obviously a very high res machine. The Sigma SD9 beats
    the Canon 10D by at least 10:6 (the sensor counts).

    When shooting in Medium Resolution, the SD9 roughly equals the 10D's
    optical resolution. That shouldn't be surprising, since the SD9 is
    using only 1.5M RGB triplets in its Medium Resolution, same as the 10D
    at its max resolution.
    Georgette Preddy, Jun 20, 2004
  14. Oops, here's the link...
    Georgette Preddy, Jun 20, 2004
  15. SNIP
    Why do you propose to compare a properly Low-Pass filtered CCD with a
    non-filtered Foveon? Because you hope the non-observant viewer doesn't make
    a decent enlargement (8x10 inches or larger) of the images, but gazes at
    pixels on a low resolution CRT display?
    Yes, that's what many here have said from the beginning. The sensor's image
    enlargement potential, and non-aliasing Luminance resolution pales in
    comparison to some of it's anti-aliasing filtered competitors, and it's
    lagging further behind as the months pass. The low pixel count has made
    people wonder if the manufacturing complexity prohibits an larger pixel
    count version for a competitive price. It would still not address color
    accuracy and noise issues, but those could perhaps be improved by
    significant profiling and post-processing.

    Also the fact that the latest announced Foveon sensor array is pointed at a
    consumer market (nothing new as DSLRs are concerned) doesn't inspire many
    people to invest in a DSLR growth path. A 5 or 6MP Foveon sensor would at
    least make a few people reconsider, although locking people into Sigma
    lenses, where they may already have invested in quality lenses of another
    brand, seems a short term strategy.

    It's nice to philosophize, but let's wait for some *solid* evidence from
    Foveon before getting excited. Vaporware is only too well known by the
    regulars in this forum, so it'll take more than an unconfirmed suggestion
    from a sales rep to make people anxious. And the reputation of Sigma with
    misleading advertizing (e.g. switching from 3.4MP-->10.3MP, and a
    paintbrushed(?) advertisement 'demonstrating' the DSLR image quality)
    doesn't help either.

    Bart van der Wolf, Jun 20, 2004
  16. I wonder why the SD9 only outrespoves the Canon 10D by 2X in the blue
    channel, while it outresolves it by more than 3X in the Green channel?
    Anyone have an explanation for the chart results as shown?

    I suppose it makes sense since the 10D has no better green capture
    than red or blue, since 2 greens have to be averaged and weighted the
    same as red and blue when the 10D spatially interpolates and simulates
    color. This means that the phenomen exists on the Foveon sensor
    itself. It has most difficulty resolving blue for some reason with
    only double the blue resolution of a 6MP Bayer. its lowest.

    Bayer does marginally ok if all you care about is photographing
    2-color binary black and white objects--but Foveon is still clearly
    substantially better than 6MP Bayer at that too.

    Funnny how all of the big review sites refuse to publish color
    resolution results, except

    Sites like and are all about making
    money at the expense of their readership. Sponsorhip sponsorship
    sponsorship--that's what they think as they type their reviews--how
    can I get them to like me more than another review site? Disgraceful
    "professional" incompentence.
    Georgette Preddy, Jun 20, 2004
  17. Biff Mullins

    JPS Guest

    In message <>,
    The "chart results" are in your head. You can't count the Sigma as
    resolving lines in any area where there are hints of lines other than
    the ones on the prints. Anything within that is aliasing garbage, not
    resolution. The reason there is so much aliasing garbage is that you
    made the charts wrong; they're supposed to be sinusoidal as they sweep
    around the image; not two-tone on and off.
    JPS, Jun 20, 2004
  18. SNIP
    Despite the lens differences and different focal length and aperture
    settings used, let's evaluate the results. Measurements were based on the
    grayscale element because obviously the RGB derivates are not proper
    sinusoidal signals due to wrong processing .

    ** 10D ** (30338215.ResChart_10D.jpg)
    Sensor Nyquist frequency is 67.8 cycles/mm
    Limiting sensor resolution Hor/Ver 49/49 px --> 52.8 cy/mm
    Diagonal sensor resolution approx. 54 px --> 47.9 cy/mm. Photoshop's ACR as
    used, was unable to avoid some aliasing and subsequent color moiré not
    present in other Raw converter results. Also, a strange (Raw processor?)
    effect reduced diagonal resolution below hor/ver resolution, which is
    uncommon for a rectangular grid (which has 41% higher diagonal resolution by

    ** SD-9 ** (30338216.ResChart_SD9.jpg)
    Sensor Nyquist frequency is 54.8 cycles/mm
    Limiting sensor resolution Hor/Ver 38/40 px --> 55.1 / 52.4 cy/mm
    Diagonal sensor resolution apparently approx. 36 px --> 58.2 cy/mm, which
    exceeds the Nyquist limit and that leads to aliasing, as witnessed by the
    hyperbolic modulation deviations from straight lines.
    For equal full sensor output size versus 10D, multiply cy/mm by 1512/2048=

    So to summarize, based on the samples used;
    hor/ver *normalized equal output size* limiting resolution:
    10D: 52.8 cy/mm + some aliasing,
    SD-9: 40.7-38.7 cy/mm + some aliasing, and
    diagonal *normalized equal output size* limiting resolution:
    10D: 47.9 cy/mm + some aliasing (a strange Raw processor? effect reduced
    diagonal resolution),
    SD-9: 42.9 cy/mm + significant aliasing.

    Thanks for proving the points many have already made. Output resolution of
    the SD-9 without a Low-pass filter approaches (but is still some 33% lower
    than) the 10D's output resolution with a Low-pass filter. Aliasing artifacts
    will deteriorate image quality/accuracy, but on some subjects one might get
    away with that. Also the choice of Raw converter can influence the accuracy
    of the conversion.

    Bart van der Wolf, Jun 20, 2004
  19. SNIP
    Because that is not a resolution test.

    This is when using a same Field-of-View comparison:
    , which is what you get after correcting this difference in target
    <> .

    Also consider that the original comparison was designed to "fool" the 1Ds
    with an unlikely color combination, and that the 1Ds image was low-pass
    filtered to almost half the sampling resolution, whereas the Sigma wasn't,
    resulting in horrible aliasing artifacts in Sigma images, especially
    noticeable when output is 5x7in or larger.

    And for a more meaningful resolution comparison, compare:
    which starts to misrepresent the finer lines above the 12 marker (even
    counts only 4 or 5 black lines at the 20 marker instead of 9), with: which has
    little trouble resolving the finest detail at the 20 marker (we need more
    detailed test cards for this camera), and the Fuji S2 at 12MP output setting
    on the same page also
    reaches a reasonable 18+ mark.

    Bart van der Wolf, Jun 20, 2004
  20. Biff Mullins

    tekfull Guest

    Still no pics
    tekfull, Jun 20, 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.