EOS 7D and resolution

Discussion in 'Digital SLR' started by Charles, Sep 25, 2009.

  1. Charles

    Charles Guest

    18 MP is perhaps a stretch ... about 4 microns per pixel versus about 6
    microns, which was recently the low-end limit for a reasonable tradeoff
    between SNR and resolution. I like the gapless micro-lenses and the
    shrinking of the surrounding support electronics at each photo diode site.
    But, is the smaller photo-site size really a good idea?

    Then, there are the lenses that we can buy. With 18 MP crammed onto an
    APS-C sensor, are available lenses up to the task? Which ones?

    As to being feature-rich, the 7D is way up there. However, the issue for
    many is dynamic range, low-light performance, and what one can capture with
    available lenses.

    Any thoughts on this?

    The MP race will only be over when it is over.
     
    Charles, Sep 25, 2009
    #1
    1. Advertisements

  2. Charles

    Robert Coe Guest

    :
    : > 18 MP is perhaps a stretch ... about 4 microns per pixel versus about 6
    : > microns, which was recently the low-end limit for a reasonable tradeoff
    : > between SNR and resolution. I like the gapless micro-lenses and the
    : > shrinking of the surrounding support electronics at each photo diode site.
    : > But, is the smaller photo-site size really a good idea?
    : >
    : > Then, there are the lenses that we can buy. With 18 MP crammed onto an
    : > APS-C sensor, are available lenses up to the task? Which ones?
    :
    : Almost all the better FF lenses will be fine, although maybe only when
    : stopped down a bit. I'd guess a lot of APS-C only lenses would be iffy.
    :
    : > As to being feature-rich, the 7D is way up there.
    :
    : Lots of very nice features. Two-axis in-viewfinder level! On-demand grid
    : lines. Better AF, live view, video.

    I've always wondered why we couldn't have an in-viewfinder level. I have the
    misfortune of being able to immediately see a half-degree horizon error, while
    routinely making errors of a degree or more, even with a viewfinder crammed
    with grid lines. That feature alone makes me lust for a 7D.

    It seems to me that if it lives up to its specs, the 7D raises serious
    questions about the future of FF. It halves the MP gap between the 50D and the
    5D2 at a price $1000 below the 5D2, and it preserves a user's investment in
    APS-C lenses. Unless the price of FF cameras comes down quickly and
    dramatically, I suspect that many serious amateurs who might have gone FF
    sooner rather than later may simply rethink the whole concept.

    How about you, Bret? Would you have still bought a 5D2 if the 7D had been an
    option at the time?

    Bob
     
    Robert Coe, Sep 26, 2009
    #2
    1. Advertisements

  3. Really? Such as "is MPix everything?" and "If you need more
    MPix, why not upscale your image?"?
    I'm using a 20D. I don't feel much of a gap. Sure, many P&S
    currently out-pixel me. They don't out-image me.

    Limits? Of course a 20D has limits. Some of them are clearly
    felt, but they mostly relate to low light capability for me.
    Too few pixels? Never felt that as a problem.

    Yes, I lust after a 5D2, because the pixel size is the same but
    the sensor has been much improved. For me, give me larger pixels,
    not more.
    Kit lenses and their ilk aren't much of an investment, and those
    who made investments other than the 10-22mm *knew* they would
    stay with crop frames. Me, I like to keep my way to FF open.
    Pixel fixation speaking?

    Building a FF with the 7D pixel size would yield 46MPix.
    Always remember that the tricks you can do with crop sensors
    work just as well on FF.

    -Wolfgang
     
    Wolfgang Weisselberg, Sep 26, 2009
    #3
  4. Charles

    Robert Coe Guest

    : > It seems to me that if it lives up to its specs, the 7D raises serious
    : > questions about the future of FF.
    :
    : Really? Such as "is MPix everything?" and "If you need more
    : MPix, why not upscale your image?"?
    :
    : > It halves the MP gap between the 50D and the
    : > 5D2 at a price $1000 below the 5D2,
    :
    : I'm using a 20D. I don't feel much of a gap. Sure, many P&S
    : currently out-pixel me. They don't out-image me.
    :
    : Limits? Of course a 20D has limits. Some of them are clearly
    : felt, but they mostly relate to low light capability for me.
    : Too few pixels? Never felt that as a problem.
    :
    : Yes, I lust after a 5D2, because the pixel size is the same but
    : the sensor has been much improved. For me, give me larger pixels,
    : not more.
    :
    : > and it preserves a user's investment in
    : > APS-C lenses.
    :
    : Kit lenses and their ilk aren't much of an investment, and those
    : who made investments other than the 10-22mm *knew* they would
    : stay with crop frames. Me, I like to keep my way to FF open.
    :
    : > Unless the price of FF cameras comes down quickly and
    : > dramatically, I suspect that many serious amateurs who might have gone FF
    : > sooner rather than later may simply rethink the whole concept.
    :
    : Pixel fixation speaking?
    :
    : Building a FF with the 7D pixel size would yield 46MPix.
    : Always remember that the tricks you can do with crop sensors
    : work just as well on FF.

    My 50D is less than a year old, so I have no dog in this hunt. But if I were
    in your shoes, with a once excellent APS-C camera that's gone a bit long in
    the tooth, I think I might be more likely to replace it with a 7D than with a
    5D2.

    But that wasn't my point. My point was that a lot of people may feel that way.
    And that Canon may have concluded that that's where the future is, at least in
    the near term. Your vigorously defensive reaction syggests to me that you may
    secretly suspect that I'm right. ;^)

    Bob
     
    Robert Coe, Sep 27, 2009
    #4
  5. Charles

    John Sheehy Guest

    At high ISO, certainly, based on the RAWs available. For base ISO, I've
    seen at least one alarmingly poor sample, with an 8-column periodic
    blackpoint offset pattern, with an amplitude of almost 20 14-bit ADUs.
    Hope that was just a lemon, or an uncalibrated pre-production unit.
     
    John Sheehy, Sep 27, 2009
    #5
  6. The camera is about as good as it was on the day it was made.
    It's even better in my hands than it was there in the beginning
    since I have learned a lot about the dos and don'ts with that body.
    :)

    So, I have an excellent APS-C camera that is eclipsed by some even
    more excellent offerings. Life is good whether I upgrade or not.
    Isn't that great? :)
    I don't think I represent a majority or important minority.
    It seems a lot of people are, uhm, a bit megapixel crazy.

    A chain being only as strong as the weakest link, I feel that
    resolution is already being limited by most lenses outside
    their best apertures at lower than the present megapixels.
    (I don't think you'll get a noticeably better image from a wide
    open EF 50mm f/1.4 at 18MPix than at 8MPix + sensible upscaling
    (e.g. Lanczos scaling) where needed, for a drastic example.)
    Canon has concluded that they need to sell cameras and to sell
    cameras they need to compete in features and marketing numbers.
    Megapixel record numbers belong there.

    If suddenly a serious part of the camera buyers(!) would decide
    cameras needed to look like puppys and smell of burned toast ---
    well, Canon would be stupid to not offer them exactly that.

    As for *me*, that's not where *my* future lies, yes.

    -Wolfgang
     
    Wolfgang Weisselberg, Sep 29, 2009
    #6
  7. Charles

    DRS Guest

    [...]
    Yes and no. They dropped the megapixel count on the G11.
     
    DRS, Sep 29, 2009
    #7
  8. Charles

    ColinD Guest

    -------- Original Message --------
    Subject: Re: EOS 7D and resolution
    From: Wolfgang Weisselberg <>
    To:
    Date: 29/09/2009 12:33 p.m.
    Even if the lens cannot equal the sensor for resolution, the image will
    still be better than with fewer pixels. The resultant resolution is
    always a function of both lens and sensor. I've pointed this out
    several times in the past.

    The maximum resolution obtainable in practical photographic work is
    limited both by the camera lens and by the film/sensor. The formula
    often used to predict the resolution of a camera original is:

    1/Rt2 = 1/Rs2 + 1/RL2 (Higgins, G.C.Appl. Opt. 3, v.1, 9, Jan 1964)

    Rt = Resolution of the system (lens + sensor)

    Rs = Resolution of the sensor

    RL = Resolution of the lens

    As you can see, system resolution is not just lens resolution alone.
     
    ColinD, Sep 29, 2009
    #8
  9. Charles

    ColinD Guest

    -------- Original Message --------
    Subject: Re: EOS 7D and resolution
    From: Wolfgang Weisselberg <>
    To:
    Date: 29/09/2009 12:33 p.m.
    Even if the lens cannot equal the sensor for resolution, the image will
    still be better than with fewer pixels. The resultant resolution is
    always a function of both lens and sensor. I've pointed this out
    several times in the past.

    The maximum resolution obtainable in practical photographic work is
    limited both by the camera lens and by the film/sensor. The formula
    often used to predict the resolution of a camera original is:

    1/Rt2 = 1/Rs2 + 1/RL2 (Higgins, G.C.Appl. Opt. 3, v.1, 9, Jan 1964)

    Rt = Resolution of the system (lens + sensor)

    Rs = Resolution of the sensor

    RL = Resolution of the lens

    As you can see, system resolution is not just lens resolution alone.
     
    ColinD, Sep 29, 2009
    #9
  10. Charles

    ColinD Guest

    Oops. Sorry for the double post. Damn news client said it failed to
    send, but it didn't, obviously.

    Colin D.
     
    ColinD, Sep 30, 2009
    #10
  11. Sure, if you use arbitrary wrong definitions for 'better'.
    Point out as much as you want. Doesn't mean it's true or relevant.
    See how you went from "better image" to a "*prediction* of
    resolution"?

    By your logic infinitely small sensor sizes, unable to capture
    even a single photon, would produce 'better' images than any
    other sensor with the same lens. Even you should be able to
    see several catches with that.

    Your claim, *as it stands*, is obviously wrong.

    And misses a very important word to be relevant to practical use,
    "noticeable" or "visible".
    Ah, yes, a formula from the good old days a decade before the
    first digital image sensor, back when higher resolution meant
    slower films with finer, less notable grain.
    I can see that you probably don't understand 'better' or
    'noticeable' --- or what tinier pixels do unless you raise the
    technology in between.

    Maybe you should look at the independent lens test sites and
    see if your formula comes up with the right answers as you test
    the same lenses at the same format size against different sensor
    resolutions ...

    -Wolfgang
     
    Wolfgang Weisselberg, Sep 30, 2009
    #11
  12. Charles

    Wally Guest

    What would that show?

    Wally

    Wally
     
    Wally, Sep 30, 2009
    #12
  13. Charles

    ColinD Guest

    Well, if one removes the sarcasm from your reply, there's practically
    nothing left of it. *Of course* resolution is a function of both
    film/sensor and the lens. If it were not so, then every lens would
    deliver an image dependent only on lens resolution, regardless of sensor
    resolution, a conclusion that is patently incorrect. Whether you can
    'notice' a 'visible' difference depends on the manner in which you view
    the image, but that is irrelevant; a difference will exist whether you
    can see it with naked eye or not.



    And the age of the formula is totally irrelevant also - or has E=MC^2
    passed it's use-by date as well?
     
    ColinD, Sep 30, 2009
    #13
  14. I see you still misuse other peoples' domains and email addresses.
    Sure, if you relabel valid criticism --- like pointing out your
    asocial misuse of the email address you knowingly falsely claim
    as yours --- as sarcasm ...
    Obviously. Even
    R = min(R_S, R_L)
    is a function of both sensor and lens.
    Now go away with your strawmen and *prove* your 'equation'.
    You can find some data at photozone.de --- you should be able
    to correlate the observed resolutions for the same lens/focal
    length/aperture settings at different sensor resolutions.

    But of course that would be work and it would turn out that your
    claim is neither as clear-cut nor as correct as you'd like it
    to be.
    A difference that has no impact has no relevance.
    Differences that cannot be seen are thus irrelevant to visual arts.

    Of course, that truth will be labled 'sarcasm' by you.
    Not yet(!). It may yet end as phlogiston did, or be outdated Newton's
    laws of gravitation --- the future will tell.

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 1, 2009
    #14
  15. It would show if the formula was correct. Spot checks say
    it's off.

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 1, 2009
    #15
  16. Charles

    Wally Guest

    You are not very convincing!

    Wally
     
    Wally, Oct 1, 2009
    #16
  17. Tell me, what would convince you?

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 5, 2009
    #17
  18. Yeah, 10.000.000MPix --- the best way to fill memory cards
    and hard drives with useless information. And absolutely
    needed for showing them on a monitor or printing 4x6. You
    need more MPix for 8x12.
    So you are often printing posters of house size?

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 9, 2009
    #18
  19. | And they should raise it again, when they figure out how to get the same or
    | better image-level read noise with the higher densities.
    ^^^^^^^^^^^^^^^^^^^^^^
    You get the same noise at image level, so cropping like crazy
    (what's that? 3% of the length and 3% of the height of the image?
    And why didn't you frame the shot correctly in the first place?)
    is enlarging like crazy, and you get all the noise blown up.
    On pixel level --- where your crop argument counts --- the noise
    is much worse.
    The "obscene 21Mpixels" aren't any more than the 20D's 8 MPix!
    It's just that the old, old 20D does the cropping already in
    the hardware, saving you a lot of unneeded data to be
    stored, transfered and handled. And you don't get all that
    visible vignetting when shooting wide open.

    I'll admit that the 5D2 pixels can do higher ISO better,
    though.

    Now compare that to 15 or 18 MPix on the same (or even slightly
    smaller) area as the 8 MPix of the 20D --- you ask that much more
    from your lens, you get less light per pixel ...
    I don't know your chips ... (probably high end athlon CPUs)

    16GB cards from good brands (and I wouldn't consider any
    other type of card) are not exactly cheap especially if you want
    to buy a *fast* 16GB card. Which is a really good idea, unless
    you shoot slowly, never filling the buffer, and don't mind waiting
    while transferring the data from the card to the computer.
    Not too bad, but not a huge amount either.
    A RAW on disk costs a lot more than just the space it occupies.

    Do you have a reliable "the house gets burned down, swamped
    with water, burgled, earthquarked" backup method? Have you
    implemented it? Do you exercise it regularly?

    Do you have an extra fast computer? One that is overpowered
    by a factor of at least 2 or 3 (or even much more for obscene
    megapixels) for any and all tasks except RAW conversion? Do you
    keep it up to speed, upgrading at least yearly?

    Do you check your shots at the same screen magnification as
    before --- so more pixels mean more time spend checking the
    image?

    Can you afford more than 24 hours a day?

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 12, 2009
    #19
  20. Megapixels don't equal resolution, your question is illogical.
    So with a perfect optics you need infinite pixels for a 4x6
    print (in terms of proper sampling) --- if your claim was true.
    Never mind you cannot put even 4 MPix on a 4x6 inch print
    even with 400 PPI.

    How many cycles/mm does your optics do, in terms of MTF?
    Just to irritate you, I'll mention old digital cameras that took
    4 and more pixels to transition between black and white ... at
    lowly 4 MPix. According to you, they have more than enough
    pixels (transition takes more than 3 pixels) and at the same
    time too few (pixel structure).

    Maybe you should decide one day what you want.
    Finer? Are you sure it's not stronger, since it has not been
    averaged over a larger area (the amount of light captured is the
    same, but the per-pixel amount is less than half) --- and you
    have added read noise 18 million times instead of 8 million times?
    You seem to forget what an AA filter is.
    And that is different with "High-MP images" in what way? Do
    they capture all of the detail at the highest image
    frequencies they can capture or what?

    Did you know that the higher per-pixel noise of "High-MP
    images" overlays the very same high frequencies they offer as
    an advantage over lower MPix images?
    So you say.
    Let us assume it was true that the 7D had the best IQ (however
    you want to measure that, there are so many ways) despite having
    the highest pixel density, what do you think a 20D manufactured
    with the knowledge and techniques used on the 7D would get as IQ?
    You ... expect. The proof of the pudding is in the eating,
    not in the expecting of cooking one.
    Do your tests, produce your results and we'll see.
    Very often the populus is wrong by design, which is the case,
    IMHO, with pixel resolution.

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 12, 2009
    #20
    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.