Few noteworthy jaggies here, either

Discussion in 'Digital Photography' started by K2, Dec 1, 2003.

  1. K2

    K2 Guest

    This is another example of an image you'd expect to have many jaggies
    per Foveon critics, but I only see minor ones on the rail behind the
    blue boards and some similar planes. Yes, they are there, but you have
    to look for them. They certainly don't "LOOK SHIT" (quoting our
    nameless trailer park poster).

    http://img.dpreview.com/news/0209/sigmasd9_sunny/SD9IMG04837.jpg

    Above, I describe what I see at 1024x768 on a 17" monitor with 0.25mm
    pitch. Switching to 1280x1024, the jaggies almost entirely vanish.
    That's why anti-aliasing is a relic of older monitors and older sensor
    designs. It doesn't bother me that I see minor jaggies with my
    practical monitor setting. I know the monitor is the limiting factor,
    not the camera.

    However, if I took that same image, applied some sharpening to
    overdramatize the jaggies, uploaded it to www.pbase.com and called it
    an "original," people would be claiming it's a "typical" Foveon image.

    I am beginning to understand the mindset around here, and why reasoned
    technical debate is impossible with some posters.

    K2
     
    K2, Dec 1, 2003
    #1
    1. Advertising

  2. K2

    Rafe B. Guest

    On Mon, 01 Dec 2003 03:10:58 GMT, K2 <> wrote:


    >I am beginning to understand the mindset around here, and why reasoned
    >technical debate is impossible with some posters.



    Given that you utterly fail to comprehend the basic
    concept of aliasing, you ought not to be speaking of
    "reasoned technical debate."


    rafe b.
    http://www.terrapinphoto.com
     
    Rafe B., Dec 1, 2003
    #2
    1. Advertising

  3. K2

    K2 Guest

    On Mon, 01 Dec 2003 04:52:33 GMT, Rafe B. <>
    wrote:

    >On Mon, 01 Dec 2003 03:10:58 GMT, K2 <> wrote:
    >
    >
    >>I am beginning to understand the mindset around here, and why reasoned
    >>technical debate is impossible with some posters.

    >
    >
    >Given that you utterly fail to comprehend the basic
    >concept of aliasing, you ought not to be speaking of
    >"reasoned technical debate."


    Given that you keep repeating the same smug statements over and over,
    I'll have to start ignoring you (I don't believe in killfiles).

    No comments about those benign jaggies, I notice. All that matters is
    whether an image fits your rigid parameters. Whether or not it looks
    good seems to be beside the point.

    K2
     
    K2, Dec 1, 2003
    #3
  4. K2

    K2 Guest

    On Mon, 01 Dec 2003 04:52:33 GMT, Rafe B. <>
    wrote:

    >On Mon, 01 Dec 2003 03:10:58 GMT, K2 <> wrote:
    >
    >
    >>I am beginning to understand the mindset around here, and why reasoned
    >>technical debate is impossible with some posters.

    >
    >
    >Given that you utterly fail to comprehend the basic
    >concept of aliasing, you ought not to be speaking of
    >"reasoned technical debate."


    This article explains why anti-aliasing is needed after interpolation
    has been done in a traditional Bayer sensor.

    http://www.linear-systems.com/support/anti_aliasing.htm

    "When the information is color interpolated, under-sampled data is
    created and results in unwanted color, such as color aliasing,
    spectral highlights, and moiré patterns. To remove color-aliasing
    artifacts, an anti-aliasing filter can be installed in the camera. The
    filter is an optical assembly that fits in front of the CCD and behind
    the camera lens. It eliminates aliasing by lowering the
    high-frequency/high-contrast areas in an image that would normally
    exceed the sampling capability of the CCD."

    The article says little or nothing about jaggies, or that
    anti-aliasing is desirable with full-color sensors that don't NEED
    color interpolation. Anti-aliasing is basically a patch for the older
    mosaic sensor design. Do you think someone at the level of Carver
    Meade doesn't get those concepts? Don't call me undereducated!

    Sure, you could apply anti-aliasing to a Foveon image if you had zero
    tolerance for any jaggies (at cruder monitor resolutions), but there
    is no law mandating it. If you are that fearful of the slightest
    jaggy, nobody's forcing you to give up your current camera, so relax!

    But don't approach this as if you have infinitely deeper knowledge
    than anyone who likes Foveon images.

    K2
     
    K2, Dec 1, 2003
    #4
  5. K2

    Mike Guest

    "Rafe B." <> wrote in message
    news:...
    > On Mon, 01 Dec 2003 03:10:58 GMT, K2 <> wrote:
    >
    >
    > >I am beginning to understand the mindset around here, and why reasoned
    > >technical debate is impossible with some posters.

    >
    >
    > Given that you utterly fail to comprehend the basic
    > concept of aliasing, you ought not to be speaking of
    > "reasoned technical debate."
    >
    >
    > rafe b.
    > http://www.terrapinphoto.com

    Dear All of the foveon war alliance,
    I have found that with any given image the Monitor setting is important some
    may give a good non jaggy image at 100% Experiment whilst looking at the
    various resolution charts for all digital cameras in DP Review using 50% to
    200% and you will se what I mean. Your monitor has resolution of 72 to 90
    dpi which can have an have detrimental effect on viewing a given image
    depending on its resolution etc. The only way to really be sure of a given
    images quality is to print it on a quality printer at 300 dpi at the maximum
    size you may need for your photography. For example I am in a camera club
    where I enter competitions using A4 or A3 size prints, so the camera will
    have to give good quality a A3.

    If your only going to exhibit on the Internet as a lot of people do you
    don't need such a high resolution camera to enter Internet camera club
    competitions because the resolution for the net is only 70 to 90 dpi. as
    against the printer resolution of 300 dpi.
    MikeS
     
    Mike, Dec 1, 2003
    #5
  6. "K2" <> wrote in message
    news:...
    SNIP
    > Above, I describe what I see at 1024x768 on a 17" monitor with 0.25mm
    > pitch. Switching to 1280x1024, the jaggies almost entirely vanish.


    Of course. You cannot display more than 1240 horizontal pixels on an approx.
    310 mm wide 0.25mm dot pitch CRT, without hiding a large number of them with
    the shadow mask. You are not looking at the pixels themselves.
    It's like closing your eyes, and saying: "I don't see it, you must be
    wrong".

    > That's why anti-aliasing is a relic of older monitors and older sensor
    > designs. It doesn't bother me that I see minor jaggies with my
    > practical monitor setting. I know the monitor is the limiting factor,
    > not the camera.


    It has been pointed out to you many times already, Aliasing is not just
    about jaggies, and how can you comment on pixel quality if you can't see
    them?

    Bart
     
    Bart van der Wolf, Dec 1, 2003
    #6
  7. "K2" <> wrote in message
    news:...
    SNIP
    > This article explains why anti-aliasing is needed after interpolation
    > has been done in a traditional Bayer sensor.


    Why do you feel a need to demonstrate your ignorance, over and over again?
    You have been told numerous times, aliasing has nothing to do with Bayer
    sensors. Aliasing has everything to do with discrete sampling and the lack
    of low-pass filtering, PRIOR to the light reaching the sensor.

    > The article says little or nothing about jaggies, or that
    > anti-aliasing is desirable with full-color sensors that don't NEED
    > color interpolation.


    Don't you understand? All sensors with a regular sampling pitch will show
    aliasing artifacts, unless filtered.

    > Anti-aliasing is basically a patch for the older
    > mosaic sensor design. Do you think someone at the level of Carver
    > Meade doesn't get those concepts? Don't call me undereducated!


    It has (probably) nothing to do with Meade, but with the camera design.
    Sigma failed to prevent the erroneous results of an unfiltered sensor,
    hoping to (apparently successfully) delude people like you.

    > Sure, you could apply anti-aliasing to a Foveon image


    No you can't. You can only apply an AA-filter to the sensor, to prevent it.

    SNIP
    > But don't approach this as if you have infinitely deeper knowledge
    > than anyone who likes Foveon images.


    You are insulting knowledgeable Foveon users that recognize it's
    limitations, but can live with them for their particular use.

    Bart
     
    Bart van der Wolf, Dec 1, 2003
    #7
  8. Re: Few noteworthy photos from theSD-9 either

    Few noteworthy photos from the preddy gallery either...



    "K2" <> wrote in message
    news:...
    > This is another example of an image you'd expect to have many jaggies
    > per Foveon critics, but I only see minor ones on the rail behind the
    > blue boards and some similar planes. Yes, they are there, but you have
    > to look for them. They certainly don't "LOOK SHIT" (quoting our
    > nameless trailer park poster).
    >
    > http://img.dpreview.com/news/0209/sigmasd9_sunny/SD9IMG04837.jpg
    >
    > Above, I describe what I see at 1024x768 on a 17" monitor with 0.25mm
    > pitch. Switching to 1280x1024, the jaggies almost entirely vanish.
    > That's why anti-aliasing is a relic of older monitors and older sensor
    > designs. It doesn't bother me that I see minor jaggies with my
    > practical monitor setting. I know the monitor is the limiting factor,
    > not the camera.
    >
    > However, if I took that same image, applied some sharpening to
    > overdramatize the jaggies, uploaded it to www.pbase.com and called it
    > an "original," people would be claiming it's a "typical" Foveon image.
    >
    > I am beginning to understand the mindset around here, and why reasoned
    > technical debate is impossible with some posters.
    >
    > K2
     
    Betty Swallocks, Dec 1, 2003
    #8
  9. "K2" <> wrote in message
    news:...
    SNIP
    > This article explains why anti-aliasing is needed after interpolation
    > has been done in a traditional Bayer sensor.


    Why do you feel a need to demonstrate your ignorance, over and over again?
    You have been told numerous times, aliasing has nothing to do with Bayer
    sensors. Aliasing has everything to do with discrete sampling and the lack
    of low-pass filtering, PRIOR to the light reaching the sensor.

    > The article says little or nothing about jaggies, or that
    > anti-aliasing is desirable with full-color sensors that don't NEED
    > color interpolation.


    Don't you understand? All sensors with a regular sampling pitch will show
    aliasing artifacts, unless filtered.

    > Anti-aliasing is basically a patch for the older
    > mosaic sensor design. Do you think someone at the level of Carver
    > Meade doesn't get those concepts? Don't call me undereducated!


    It has (probably) nothing to do with Meade, but with the camera design.
    Sigma failed to prevent the erroneous results of an unfiltered sensor,
    hoping to (apparently successfully) delude people like you.

    > Sure, you could apply anti-aliasing to a Foveon image


    No you can't. You can only apply an AA-filter to the sensor, to prevent it.

    SNIP
    > But don't approach this as if you have infinitely deeper knowledge
    > than anyone who likes Foveon images.


    You are insulting knowledgeable Foveon users that recognize it's
    limitations, but can live with them for their particular use.

    Bart
     
    Bart van der Wolf, Dec 1, 2003
    #9
  10. "Rafe B." <> wrote in message
    news:...
    > On Mon, 01 Dec 2003 03:10:58 GMT, K2 <> wrote:
    >
    >
    > >I am beginning to understand the mindset around here, and why reasoned
    > >technical debate is impossible with some posters.

    >
    >
    > Given that you utterly fail to comprehend the basic
    > concept of aliasing, you ought not to be speaking of
    > "reasoned technical debate."


    "During our tests, we photographed all kinds of subjects - architectural as
    well as "natural" ones - and out of 2,000-3,000 pictures, only one showed
    any signs of aliasing. It was one of four otherwise identical pictures taken
    at the same time and place. The only difference between them was the angle
    at which the hand-held camera was positioned, and/or the focal length of the
    zoom lens. So although the tendency to alias is theoretically greater in
    SD9-made pictures than in pictures from cameras using an anti-aliasing
    filter, the reality is that it was extremely rare in our tests.

    This situation raises a provocative question that each photographer must
    answer on his own: which is worse, a camera that shows maximum sharpness 100
    percent of the time while aliasing once in 2,000-3,000 shots like the SD9,
    or a camera that never aliases but blurs, albeit ever so slightly, 100
    percent of its pictures, like all the others?"

    http://www.digitaloutput.net/back edit/may03/ss3.html
     
    George Preddy, Dec 1, 2003
    #10
  11. K2

    Azzz1588 Guest

    In article <>, K2 <>
    writes:

    >why reasoned
    >technical debate is impossible with some posters.
    >


    And you are a classic example of someone impossible to
    reason with.

    Boy, the Sigma sockpuppetts are just as bad as the Mac ones...............











    "Only a Gentleman can insult me, and a true Gentleman never will..."
     
    Azzz1588, Dec 1, 2003
    #11
  12. K2

    Azzz1588 Guest

    Re: Few noteworthy photos from theSD-9 either

    In article <bqfim2$p83$>, "Betty Swallocks"
    <bettystrailerhomeofthe1Ds7awhip&> writes:

    >Few noteworthy photos from the preddy gallery either...


    Dont you mean NO noteworthy pictures ???


    I havnt seen one yet........................
    (unless you like yellow people, and lot's of jaggies in pic's)


















    "Only a Gentleman can insult me, and a true Gentleman never will..."
     
    Azzz1588, Dec 1, 2003
    #12
  13. K2 wrote:

    > This is another example of an image you'd expect to have many jaggies
    > per Foveon critics, but I only see minor ones on the rail behind the
    > blue boards and some similar planes. Yes, they are there, but you have
    > to look for them. They certainly don't "LOOK SHIT" (quoting our
    > nameless trailer park poster).
    >
    > http://img.dpreview.com/news/0209/sigmasd9_sunny/SD9IMG04837.jpg


    Dude, I'm not going to comment on the jaggie factor. I see the whole
    'jaggie issue' as a trade off. Sigma's implementation certainly gets
    the most apparent shrpness from each pixel. The downside is that jaggie
    WILL show up if you magnify the image. Looking at your picture at 100%
    mag, on a 19" monitor, running @ 1280x1024, I cannot see any
    objectionable jaggies.

    What I do see are some hellacious color artifacts. Indeed, I don't
    think you could have chosen a 'better' picture to demonstrate the color
    accuracy shortcomings that the Sigma suffers. At 100%, I can see a
    faint red/blue halo surrounding each diagonal in the corrugated hatch
    roof. The worst color artifacting occurs in the square plate(?) with
    all the holes in it. ( http://members.cox.net/geonerd/Wacky_color.jpg )
    I dare say that Sigma needs to do a 'bit' more firmware R&D.

    -Greg
     
    Greg Campbell, Dec 1, 2003
    #13
  14. K2

    Azzz1588 Guest

    In article <r5Myb.464$yf.258@fed1read01>, Greg Campbell <>
    writes:

    >What I do see are some hellacious color artifacts.


    You could have just stopped at hellacious color !!!!!

    Look at the kid's, houses, and the violin, and cat in
    his galleries.

    Damn if the colors dont have a *cartoon* effect to
    them. Dont know how else to describe it quickly, but none seem to
    show consistant color rendition as well..........




    "Only a Gentleman can insult me, and a true Gentleman never will..."
     
    Azzz1588, Dec 1, 2003
    #14
  15. Hello K2,

    K2 wrote:

    > This is another example of an image you'd expect to have many jaggies
    > per Foveon critics, but I only see minor ones on the rail behind the
    > blue boards and some similar planes. Yes, they are there, but you have
    > to look for them. They certainly don't "LOOK SHIT" (quoting our
    > nameless trailer park poster).
    >
    > http://img.dpreview.com/news/0209/sigmasd9_sunny/SD9IMG04837.jpg
    >
    > Above, I describe what I see at 1024x768 on a 17" monitor with 0.25mm
    > pitch. Switching to 1280x1024, the jaggies almost entirely vanish.
    > That's why anti-aliasing is a relic of older monitors and older sensor
    > designs. It doesn't bother me that I see minor jaggies with my
    > practical monitor setting. I know the monitor is the limiting factor,
    > not the camera.
    >
    > However, if I took that same image, applied some sharpening to
    > overdramatize the jaggies, uploaded it to www.pbase.com and called it
    > an "original," people would be claiming it's a "typical" Foveon image.
    >
    > I am beginning to understand the mindset around here, and why reasoned
    > technical debate is impossible with some posters.
    >


    I just looked through the sd9 samples in www.pbase.com.
    And I couldnt see a jaggie problem, that would affect me.

    I saw another problem: 50% of the images have strange colors.

    I know, that all sd9 images are raw and must be postprocessed. So the
    question is: why are there so much wrong-color-images?

    Are there users, which do the postprocessing incorrectly?
    Will the new Sigma Photo Pro 2 give better results?

    On the other side, there are some pictures with impressing sharpness and
    luminousity.
    My explanation for this:
    I understand well, that a screen or a printer cannot reproduce natural
    light dynamics and color, because in nature the light dynamic is some 1000
    times higher than in an image.
    So landscape photos will need some corrections on gamma factor and color,
    when the picture should give the illusion of natural luminousity.

    When these corrections are manually well done then a outstanding picture
    will result from this, when the corrections are done automatic then the
    result will be average, and when the correction is made manually and
    wrong, then the result is bad.

    I don't care about jaggies.
    SD9 jaggies are neglectible, when the pictures are viewed with reasonable
    scaling.
    The new SD10 has an inbuilt antialias filter anyway, because it uses
    microlenses. Maybe it will have less jaggies.

    My concern is about the effort which it would need to achieve good color.
    I like it, to treat color and luminace manually. But I dont want to do
    this for every image.

    regards,

    peter
     
    Peter Heckert, Dec 1, 2003
    #15
  16. "Peter Heckert" <> wrote in message
    news:p...
    SNIP
    > The new SD10 has an inbuilt antialias filter anyway, because it uses
    > microlenses. Maybe it will have less jaggies.


    Not trying to nitpick, but the microlenses are not the same as an
    anti-aliasing filter. The microlenses do reduce the tendency for aliasing
    because the worst type of discrete sampling for imaging is point sampling
    that should represent an area. An area to represent an area gives a much
    more accurate response, but photosites (especially on CMOS devices) have
    supporting electronic components per photosite which reduces the so-called
    fill factor (% of light sensitive area). Using microlenses to concentrate
    the incident radiation on the light sensitive part of the photosite, mimics
    a near to 100% fill factor.

    An anti-aliasing filter is an additional optical component placed in front
    of the sensor, and is often combined with an InfraRed filter.

    Bart
     
    Bart van der Wolf, Dec 1, 2003
    #16
  17. "Bart van der Wolf" <> wrote in message
    news:3fcb4f81$0$1494$4all.nl...
    >
    > "K2" <> wrote in message
    > news:...
    > SNIP
    > > This article explains why anti-aliasing is needed after interpolation
    > > has been done in a traditional Bayer sensor.

    >
    > Why do you feel a need to demonstrate your ignorance, over and over again?
    > You have been told numerous times, aliasing has nothing to do with Bayer
    > sensors. Aliasing has everything to do with discrete sampling and the lack
    > of low-pass filtering, PRIOR to the light reaching the sensor.
    >
    > > The article says little or nothing about jaggies, or that
    > > anti-aliasing is desirable with full-color sensors that don't NEED
    > > color interpolation.

    >
    > Don't you understand? All sensors with a regular sampling pitch will show
    > aliasing artifacts, unless filtered.


    You mean blurred. Anyway...

    "During our tests, we photographed all kinds of subjects - architectural as
    well as "natural" ones - and out of 2,000-3,000 pictures, only one showed
    any signs of aliasing. It was one of four otherwise identical pictures taken
    at the same time and place. The only difference between them was the angle
    at which the hand-held camera was positioned, and/or the focal length of the
    zoom lens. So although the tendency to alias is theoretically greater in
    SD9-made pictures than in pictures from cameras using an anti-aliasing
    filter, the reality is that it was extremely rare in our tests.

    This situation raises a provocative question that each photographer must
    answer on his own: which is worse, a camera that shows maximum sharpness 100
    percent of the time while aliasing once in 2,000-3,000 shots like the SD9,
    or a camera that never aliases but blurs, albeit ever so slightly, 100
    percent of its pictures, like all the others?"

    http://www.digitaloutput.net/back edit/may03/ss3.html
     
    George Preddy, Dec 2, 2003
    #17
  18. K2

    Guest Guest

    Extract from the highly respsected photography journal....

    Digital Output - The Business Giude For Electronic Publishers



    > You mean blurred. Anyway...
    >
    > "During our tests, we photographed all kinds of subjects - architectural

    as
    > well as "natural" ones - and out of 2,000-3,000 pictures, only one showed
    > any signs of aliasing. It was one of four otherwise identical pictures

    taken
    > at the same time and place. The only difference between them was the angle
    > at which the hand-held camera was positioned, and/or the focal length of

    the
    > zoom lens. So although the tendency to alias is theoretically greater in
    > SD9-made pictures than in pictures from cameras using an anti-aliasing
    > filter, the reality is that it was extremely rare in our tests.
    >
    > This situation raises a provocative question that each photographer must
    > answer on his own: which is worse, a camera that shows maximum sharpness

    100
    > percent of the time while aliasing once in 2,000-3,000 shots like the SD9,
    > or a camera that never aliases but blurs, albeit ever so slightly, 100
    > percent of its pictures, like all the others?"
    >
    > http://www.digitaloutput.net/back edit/may03/ss3.html
    >
    >
    >
     
    Guest, Dec 2, 2003
    #18
  19. "George Preddy" <> wrote in message
    news:bqh06b$job$...
    >
    > "Bart van der Wolf" <> wrote in message
    > news:3fcb4f81$0$1494$4all.nl...
    > >
    > > "K2" <> wrote in message
    > > news:...
    > > SNIP
    > > > This article explains why anti-aliasing is needed after interpolation
    > > > has been done in a traditional Bayer sensor.

    > >
    > > Why do you feel a need to demonstrate your ignorance, over and over

    again?
    > > You have been told numerous times, aliasing has nothing to do with Bayer
    > > sensors. Aliasing has everything to do with discrete sampling and the

    lack
    > > of low-pass filtering, PRIOR to the light reaching the sensor.
    > >
    > > > The article says little or nothing about jaggies, or that
    > > > anti-aliasing is desirable with full-color sensors that don't NEED
    > > > color interpolation.

    > >
    > > Don't you understand? All sensors with a regular sampling pitch will

    show
    > > aliasing artifacts, unless filtered.

    >
    > You mean blurred. Anyway...


    No, I mean low-pass spatial frequency filtered. Blurring would also affect
    low frequencies, whereas real anti-aliasing affects primarily high spatial
    frequencies (only a brick-wall filter can exclusively filter out high
    frequencies, which is not possible in normal optics).

    Bart
     
    Bart van der Wolf, Dec 2, 2003
    #19
  20. "Bart van der Wolf" <> wrote in message
    news:3fcc9599$1$1496$4all.nl...

    > > > Don't you understand? All sensors with a regular sampling pitch will

    > show
    > > > aliasing artifacts, unless filtered.

    > >
    > > You mean blurred. Anyway...

    >
    > No, I mean low-pass spatial frequency filtered. Blurring would also affect
    > low frequencies, whereas real anti-aliasing affects primarily high spatial
    > frequencies (only a brick-wall filter can exclusively filter out high
    > frequencies, which is not possible in normal optics).


    Everything in a Bayer image is blurry.
     
    George Preddy, Dec 2, 2003
    #20
    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. Paul H.

    Re: G3 jaggies

    Paul H., Jul 28, 2003, in forum: Digital Photography
    Replies:
    0
    Views:
    425
    Paul H.
    Jul 28, 2003
  2. Allegro

    Re: G3 jaggies

    Allegro, Jul 28, 2003, in forum: Digital Photography
    Replies:
    0
    Views:
    412
    Allegro
    Jul 28, 2003
  3. George Preddy

    Look ma, no jaggies...

    George Preddy, Nov 30, 2003, in forum: Digital Photography
    Replies:
    66
    Views:
    1,795
  4. Petey

    Jaggies

    Petey, Feb 22, 2004, in forum: Digital Photography
    Replies:
    15
    Views:
    952
    Bart van der Wolf
    Feb 22, 2004
  5. =?Utf-8?B?VmluY2VudCBMZXVuZw==?=

    Anyone here got a specific cert after passing either 70-622 or 70-

    =?Utf-8?B?VmluY2VudCBMZXVuZw==?=, Jul 9, 2007, in forum: Microsoft Certification
    Replies:
    2
    Views:
    473
    Montreal MCT
    Jul 14, 2007
Loading...

Share This Page