Light fall off on dSLRs - an experiment

Discussion in 'Digital SLR' started by Kennedy McEwen, Mar 16, 2006.

  1. David,
    The problem with making the lenses so they improve wide angle
    lens response, is they would hurt longer focal length lens response.

    Roger
     
    Roger N. Clark (change username to rnclark), Mar 21, 2006
    1. Advertisements

  2. Kennedy McEwen

    JPS Guest

    In message <>,
    How could you *NOT* want to know what the gradients look like? That
    would certainly indicate to you the best places to measure values, or be
    an instant clue if something is, in fact, wrong. Your brain is not just
    for examining numbers; it has excellent pattern-recognition algorithms.
    Your statistics are too global to tell the entire story.
    --
     
    JPS, Mar 21, 2006
    1. Advertisements

  3. Yes, I can see that.

    David
     
    David J Taylor, Mar 21, 2006
  4. Kennedy McEwen

    Rich Guest

    How is it possible to accurately determine if "angle of incidence" of
    incoming light makes that much of a difference (vis a vis, legacy
    glass versus "digital" glass) when there is no such thing as
    legacy glass that was made for sub-35mm frame sizes (to match
    current sensors) and no such thing as "digital glass" designed for FF
    sensors? Unless Canon has created some FF lenses for digital sensors
    that I haven't seen. My initial reaction to a question about this
    would be, "Legacy glass simply because of its much larger area of
    coverage should beat digital glass designed to illuminate only sensors
    with areas smaller than FF." (This of course disregards the angle of
    incidence of the light exiting the back end of the lens of the legacy
    glass) Using the centre portion of any lens automatically reduces
    aberration levels, it's the principal of stopping down lenses except
    that in this case, the sub-FF sensor functions as the stop-down.
    -Rich
     
    Rich, Mar 21, 2006
  5. Kennedy McEwen

    Rich Guest

    Philips does make some good televisions. I wanted to buy a model but
    couldn't get it in North America. Perhaps B&O does with Philips what
    Sony does (did) for their XBRs, select the best (most accurate) 10% of
    the Trinitron line to use in their products?
    No, Sony does not make a camera with the qualilty of the D200 or the
    D2x. They may never do this because:
    1. No pro is going to buy a pro or semi-pro (enthusiasts) camera
    based on using Minolta glass.
    2. There may be too small a market to interest them.

    The only problems I can see is acceptance by traditional Canon or
    Nikon buyers and Sony's idiotic "corporate edict" that they have to
    use as much Sony in-house product prior to using products from
    other companies. Maybe this is changing?

    The wildcard is Zeiss. If Sony decided to market a higher end DSLR
    (say on the level of the D200) as well as entry-level things (as
    Samsung has done by rebranding Pentax) then perhaps they could
    use the Zeiss moniker on lenses they produce? The lens on the R1
    appears to be somewhere on the level of above run of the mill "kit"
    lenses and Canon's L-glass, in terms of quality.
    Zeiss has already decided (though why I do not know, Canon needed more
    "help" with its lens offerings than did Nikon) to market glass for
    Nikon. Perhaps it was the perception that Nikon users would be more
    likely to "jump ship" for higher end glass than Canon users?
    -Rich
    -Rich
     
    Rich, Mar 21, 2006
  6. Although not impossible, that is technically very difficult to do. The
    way image chips are designed it to create a "unit cell" of a small
    number of pixels, say 4, and then simply repeat that automatically
    across the frame. Correcting each pixel based on its radius from the
    centre would require another level of design process.

    Also, it would show up with normal (perpendicular) incidence and there
    doesn't appear to be any increase in response at the edges or corners
    with that orientation.
    I think I understand what you are getting at, but I don't see any
    evidence of it and know, from my own experience, how difficult such a
    graded variation across the frame would be.
     
    Kennedy McEwen, Mar 22, 2006
  7. Actually, I do want to know the gradient, but not for this particular
    purpose. The fact that the shadows are dark enough to have no influence
    on the results is enough for this particular test.

    What I am interested in the gradient for is something else entirely.
    Having gone to the trouble of creating a pretty good collimated source,
    albeit unnecessarily to satisfy concerns about light scattering in the
    original test, I can now create a shadow on the focal plane which is
    only limited by diffraction effects and the MTF of the detector and the
    AA filter. Consequently, I can use the gradient across a sloping edge
    shadow to calculate the MTF of the sensor alone in the Canon 5D
    completely independent of the lens. Combined with MTF measurements of
    the lens and camera as a system this should provide a relatively simple
    method of calculating the MTF of the lenses I have for comparison
    against published data.
     
    Kennedy McEwen, Mar 22, 2006
  8. Read Stacey's post again - he reckons he has compared his *OM* lenses on
    both film and digital and, as referenced in another thread, found the
    digital images to show more light fall-off! So legacy lenses do exist,
    in fact Olympus were pressured to rescind their original refusal to
    provide a 4/3rds adapter for those legacy OM lenses as a result of
    pressure from the OM owners group.

    I have to say, I doubt Stacey's remarks on this topic because I have
    checked most of my OM lenses on the full frame Canon 5D and find no
    discernible difference that can be assigned to a difference in angular
    response between film and digital sensor. Gamma and toe response
    certainly, but not angular response.

    So, either Stacey is mistaken or Panasonic don't know how to make half
    frame sensors capable of accepting off-axis light. If the schematics on
    the Oly website are even remotely representative of real detectors they
    are using then this is quite likely. This latter situation would, of
    course, partially explain Olympus' obsession with off-axis light fall
    off, which my recent measurements show is not a problem with the Canon
    FF sensors until you get to angles which are impossible for Canon lenses
    to produce.
     
    Kennedy McEwen, Mar 22, 2006
  9. Kennedy McEwen

    Skip M Guest


    Commercial. I was thinking more in line with TVs, though. Sony has always
    been a player in broadcast video cameras, of course. But in consumer
    electronics, Sony has never tried to compete with the likes of B&O, Marc
    Levinson, Thiel, etc.
     
    Skip M, Mar 22, 2006
  10. Kennedy McEwen wrote:
    []
    Thanks, Kennedy.

    I had been thinking that the micro-lenses were manufactured as a separate
    layer, and subsequently bonded. By scaling down the micro-lens step-and
    repeat just a fraction of a percent, you could make the lenses a fraction
    of a pixel out at the edges.
    But of, course, Roger picked me up on the point that it would degrade the
    performance for more telephoto lenses!

    Cheers,
    David
     
    David J Taylor, Mar 22, 2006
  11. Seems a rather fine distinction. Aren't news cameramen the video
    equivalent of newspaper photographers? One uses Sony the other
    Nikon/Canon.
    I would put these products in line with Leica and Minox cameras.
    Expensive and niche, but hardly standard professional workhorses.
     
    Kennedy McEwen, Mar 22, 2006
  12. *If* I was the chief sensor designer for Canon, and therefore had
    access to the light ray bundle angles for all of the commonly used
    lenses, I reckon i would be highly tempted to progressively angle those
    outer edge sensors *slightly* inwards. Surely you could do this to
    improve the problem without adversely affecting telephoto/telecentric
    lenses simply by keeping the angle within the 'sweet spot of about
    10-15 degrees. After all, at those outer edges, you can pretty much
    ignore lightrays coming from beyond 90 degrees... Get my drift?

    I don't know how difficult it would be to angle arrays of microlenses
    progressively.. do they grow them, press them, grind them? Little
    leprechauns with super glue? Just give them some tiny little wedges...

    (O:
     
    mark.thomas.7, Mar 22, 2006
  13. Wonderful thought! <G>
     
    David J Taylor, Mar 22, 2006
  14. Kennedy McEwen

    Skip M Guest


    Thus the "high end" distinction, rather than commercial. You left out this
    part:
    "I was thinking more in line with TVs, though. Sony has always been a
    player in broadcast video cameras, of course."
     
    Skip M, Mar 22, 2006
  15. It would be pretty difficult, but not impossible. The lenses are made
    by spinning a layer of etch resist polymer (single drop in the centre of
    the wafer which then smears out as it is spun at speed), exposing it to
    a grid pattern of light and then developing it to remove the exposed
    areas (or the inverse if positive etch resist). This leaves little
    squares of etch resist on the pixel, all isolated from each other, their
    thickness depending on the viscosity of the original material and the
    spin rate. The wafer is then heated and the etch resist melts, forming
    little near spherical surfaces that act as microlenses. They certainly
    don't grow them or grind them.
     
    Kennedy McEwen, Mar 22, 2006
  16. Kennedy McEwen

    J. Clarke Guest

    The big problem I see with Sony is that they would have to establish that
    they were in the market for the long pull and that they understood
    standards. I have one of the best PDAs ever made, a Sony. They decided
    after a while to pull out of that market. They've also changed their
    "memory stick" standard twice at least since that unit shipped and some of
    the larger ones just plain won't work in it.

    So, even if Sony makes the best DSLR in the world, I can't see any pro
    buying one for anything other than a special project of some kind that only
    it would do, simply because five years down the road Sony may decide that
    they aren't making enough money in that market and pull out of it leaving
    the camera orphaned, or they may continue to change their storage standards
    every time the wind shifts until a point is reached where whatever type of
    storage that unit needs is either unavailable or priced through the roof
    due to becoming a niche product. And that leaves aside the possibility
    that they'll decide to change the lens mount every now and again.

    If Sony's been in the SLR market for ten years and done a good job of
    maintaining quality and compatibility, _then_ one of their cameras might be
    a good choice for professional use.
     
    J. Clarke, Apr 4, 2006
  17. Kennedy McEwen

    ian lincoln Guest

    They will just have to have some form of the memory stick AND compact flash.
     
    ian lincoln, Apr 8, 2006
  18. Kennedy McEwen

    J. Clarke Guest

    They tried that on some Clie NX models. The CF slot was supposed to be just
    for a wifi card but it wasn't long before someone came up with drivers that
    allowed it to be used for memory. Trouble is that Sony buggered the
    hardware somehow so that if you leave a CF in the slot the battery goes
    flat right quick.
     
    J. Clarke, Apr 10, 2006
    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.