Sharpest "Reasonably priced" Canon lens

Discussion in 'Canon' started by Mike, Nov 21, 2007.

  1. Mike

    Mike Guest

    Yes, I know that "sharpest" is a rather generic description, but bear
    with me....
    I have a Canon 20D with the 17 - 85mm zone. It is a very nice "walk
    around" lens and is fine for most casual use.. but... Sometimes I
    really want to take the high resolution (sharp) pictures that I'm sure
    the 20D is capable of. Even stopping down to f11 and, keeping the
    shutter speed up and using a tripod still does not product the sharpness
    and detail I'm looking for. So, I want to do a bit of a test....
    I want to get a lens that would at least match the capabilities of the
    20D's 8 MP sensor resolution, but I can't really justify the cost of the
    nice L lenses. I've seen comments about the 50mm lens being very sharp,
    and I figure that would be a reasonable test lens... So..
    What (single focal length) Canon lens is the sharpest while staying
    within a reasonable price point. For example, Canon offers 50mm 1.8,
    1.4 and even 1.2 (L) or 1.0 (L). Without looking them up, I suspect the
    1.2 and 1.0 L lenses will be a bit pricey. Would the 1.8 (or 1.4) lens
    be my best bet, or should I consider dropping down to the 35mm F2 or
    even 28mm f2.8?? Since this is primarily an 'experiment', I'm not real
    concerned about focal length; I can use about anything within the
    moderate wide to moderate tele range. If this experiment works out, I
    may then see about some other focal length lenses that provide the
    'sharpness' I'm looking for.
    Thanks
     
    Mike, Nov 21, 2007
    #1
    1. Advertisements

  2. Don't be so sure about L-glass meaning anything these days. Some recent tests
    last month prove that L-glass on even a tripod mounted Mk-II can't beat the lens
    and resolution in a hand-held $400 P&S superzoom camera. Seen the comparison
    photos with my own eyes. Dem's da fac's.
     
    TrevorAndersen, Nov 21, 2007
    #2
    1. Advertisements

  3. <snip>

    Take a look at

    "http://www.bobatkins.com/photography/reviews/best_canon_eos_lenses.html"

    and the lens reviews at

    "http://www.bobatkins.com/photography/reviews/index.html" (under
    "Equipment Reviews" scroll down to "Lenses."

    Especially look at
    "http://www.bobatkins.com/photography/reviews/ef-s_17-85_review_4.html"
    where he compares the sharpness of the lens you have against several
    other lenses, at different settings.

    So the short answer is that the 50mm primes will be sharper than the
    zoom you're using.

    The real gem in the EF-s line is the 10-22, which has L quality optics,
    but at $625 it's expensive.
     
    SMS 斯蒂文• å¤, Nov 21, 2007
    #3
  4. Mike

    /dev/null Guest

    Produce full sized comparisions with a proper resolution chart. Seat of the
    pants puffery doesn't count.
     
    /dev/null, Nov 21, 2007
    #4
  5. Mike

    Eatmorepies Guest

    The 70-200mm f4 L is reasonably priced. I can get one in the UK for £350.
    They are very sharp. Also - you can ignore the price because you can re-sell
    it for 90% (or more) of what you pay for it (so long as you keep it clean
    and retain the packaging).

    I have a 50mm f1.4 and it's not as sharp as the L lenses I have. I only use
    it if I want a physically small lens or for photography in cities at night
    because I want to use a large apeture. The lack of sharpness of this lens
    does't matter so much if you're producing grainy monochrome pictures.

    John
     
    Eatmorepies, Nov 21, 2007
    #5
  6. Yeah, ain't that the kicker? They were full-res images of the moon. An easily
    available, reproducible, easily shared by all, high-resolution, high-contrast,
    test target. Even fractal in nature, because you can always use crater sizes on
    it as small as you want to go to test resolution limits. Photos of the moon are
    great for comparing cameras in the world. Everyone is posting an image of the
    moon from their camera at some time or another. They were 100% pixel crops. No
    seat of the pants puffery there. As I said, "Dem's da fac's." Can't you read?

    The even bigger kicker? The images coming from the Mk-II with L-glass, locked on
    tripod, mirror locked out of the way to prevent any shake at all, were done by
    someone who prides their self on their astro-photography (for some reason). He
    was trying to use his equipment as best as he knew how. One of the DSLR maniacs
    who is always trying to justify why he spent so much, $13,000 in this instance.
    He couldn't even blame atmospheric "seeing" conditions on the differences. The
    DSLR+L-glass taken with the moon high in the sky, the P&S guy taking his image
    of the moon near the horizon through denser and more turbulent atmosphere. The
    $400 P&S camera guy was pointing his hand-held camera at the moon shortly after
    purchase just to see how it would do just for the hell of it, not even trying to
    get the best shot he could. The $400 P&S camera still won.

    Dem's da fac's, Mam. Nuttin' but da fac's.

    Deal with it.
     
    TrevorAndersen, Nov 21, 2007
    #6
  7. Minor correction in "dem fac's" ... now I remember why the P&S moon image was
    more orange. It wasn't due to being lower to the horizon, that was never stated
    by the photographer (but could be true). He was taking the image of the moon
    through all the smoke and haze of the huge fires in California USA last month,
    even more atmospheric turbulence than average on a bad "seeing" night. Like
    trying to shoot through heat-waves, only ones fill with smoke too. And it
    _still_ beat the Mk-II + L-glass for resolution on a clear night of seeing. This
    is why I remembered being certain that there was more atmospheric turbulence.
    One more minor kicker, the L-glass was a prime lens, not even a zoom. The P&S
    camera was a super-zoom. A Panasonic FZ18 if I recall now.

    Again, deal with it.
     
    TrevorAndersen, Nov 21, 2007
    #7
  8. Mike

    Chris W Guest

    I'm not sure the moon is a good universal test subject for comparison.
    The distance from you to the moon depends on how high above the horizon
    the moon is. While that probably isn't significant, the amount of
    atmosphere between you and the moon and the relative pollution level
    where you are could significantly effect the sharpness of the image.



    --
    Chris W
    KE5GIX

    "Protect your digital freedom and privacy, eliminate DRM,
    learn more at http://www.defectivebydesign.org/what_is_drm"

    Ham Radio Repeater Database.
    http://hrrdb.com
     
    Chris W, Nov 21, 2007
    #8
  9. I guess paying 32X's more for "subtle" sure is worth it. $13,000/$400=32.5 Yes,
    you could buy 32 cameras that had about twice the resolution of the Mk-II for
    the same price as the Mk-II, with prime L-glass no less. You'd have enough left
    over for $200 worth of accessories for those P&S cameras too. Even if they were
    close in comparison, and they were not, when a $400 camera comes close to the
    best DSLR and best L-glass available there's something drastically wrong with
    that DSLR and L-glass, and even more wrong with the mentality of someone that
    would buy it. Plus I don't call twice the resolution shown in the P&S image as
    exactly subtle. That's like going from a 5MPX resolution to a 20MPX resolution
    (it's a function of area, remember? not linear). That's how drastic the
    difference was. Where on earth (or the moon) do you get an absurd evaluation
    that "the larger sensor produced the better quality image" when it only produced
    half the resolution, I have no idea. Are all people who buy DSLRs this blind?
    You must be. You keep proving it.

    Let me guess, you're another one of those DSLR pushers that just talks out of
    your ass and you never even saw those comparison photos. It's the only possible
    explanation.
     
    TrevorAndersen, Nov 21, 2007
    #9
  10. Mike

    Chris W Guest

    I guess I am missing something be cause I can't see how is what you said
    is relevant to the point I presented. I also don't know what 2 images
    you are talking about... I didn't see any links to photos in this thread.

    To be more clear my point is this. If someone in East LA takes a photo
    of the moon while it is just over the horizon in the western sky with
    camera X, and someone else on top of a mountain in Montana takes a photo
    of the moon while it is directly over head with camera Y. A comparison
    of the 2 images would be of little usefulness in determining the
    relative quality of the cameras/lenses that took the photos.


    --
    Chris W
    KE5GIX

    "Protect your digital freedom and privacy, eliminate DRM,
    learn more at http://www.defectivebydesign.org/what_is_drm"

    Ham Radio Repeater Database.
    http://hrrdb.com
     
    Chris W, Nov 21, 2007
    #10
  11. Sure, find the same "versatility" in 18x worth of zoom in prime-lenses of
    L-glass. :) Then lug it all around. Got a freight container handy? You get more
    versatility for $400, at longer focal-lengths with more aperture to play with
    than is even availble in L-glass, meaning (drum roll) you don't need those high
    ISOs for most of the range of that lens. Keep trying to justify that DSLR, it
    just gets more humorous the more you try.


    They just can't stand it, can they.
     
    TrevorAndersen, Nov 21, 2007
    #11
  12. Mike

    Chris W Guest

    I think this whole DSLR vs P&S is pointless. It's like arguing over
    what's the best vehicle, a Toyota Prius or a Ford F350 one ton diesel.
    Obviously it depends on what you want to do with it. Just for the
    nitpickers, I don't mean to imply that the comparison between the two
    types of cameras is at all similar to the comparison between the two
    vehicles, other than to state that such comparisons are pointless.

    --
    Chris W
    KE5GIX

    "Protect your digital freedom and privacy, eliminate DRM,
    learn more at http://www.defectivebydesign.org/what_is_drm"

    Ham Radio Repeater Database.
    http://hrrdb.com
     
    Chris W, Nov 21, 2007
    #12
  13. Mike

    Mike Guest

    (from the OP)

    First, thanks for all the comments (on subject and off.. :) )
    It sounds like the Canon 50mm 1.8 is probably the best bet, although
    there have been some comments about durability..
    Just a couple of additional comments, in case they trigger more ideas:
    - I was using various f-stops, generally between 5.6 and 11. Some
    difference from 5.6 to 8, but not much change after that.
    - I was collecting both raw and the 'finest' jpgs for my initial
    test/comparison. Although the raw files certainly allow more
    'processing' options. I couldn't detect any difference in 'sharpness' or
    resolution between the file types.
    - My test pictures were not of a scientific 'chart' or such, just some
    images that (potentially) contained a lot of detail.
    The 17-85 mm is certainly a fine lens for most uses, I just want to see
    how much better I might be able to get with better optics...

    Thanks for the input...
     
    Mike, Nov 21, 2007
    #13

  14. These two images were taken at different moon phases. So you can't use any
    high-contrast areas along the terminator in the half-moon phase (DSLR) compared
    to the full-moon image (P&S) for comparison. Instead, for a FAIR comparison,
    you'll have to look at the detail available on the right sides in equivalent
    evenly lit Lunar Sea areas to compare them properly. Look at the details in the
    smaller craters, and structures at the edges of the darker seas. Study them VERY
    carefully. Try to not let your biases and prejudices sway you. I detect almost
    2x's the resolution in the P&S image.

    Below image taken with a $400 Panasonic FZ18, hand-held, partial digital-zoom
    used. The photographer was playing with new features. So yes, even that nasty
    digital zoom was used too. Shot through the haze and heat from dense brush and
    forest fires that wiped out several small towns and thousands of homes--read
    HUGE fires = turbulent atmosphere for days. Original JPG file, 100% crop right
    from the camera. Oh yes, I even forgot to mention this too, prepared to be
    surprised that not even an expensive DSLRs RAW file could beat it.

    http://lh4.google.com/aniramca/RvnV1xjv2_I/AAAAAAAAFzg/H9Y5LgnsCxk/P1010370_1.jpg

    If you can't open it from that link, then go here
    http://picasaweb.google.com/aniramca/Moon_shots It's the 4th image from the
    left. Try to open the larger size in a way where you can see it at its full
    1024x772 resolution. You may have to save it to hard-drive and view it in your
    image viewer to bypass google resizing your browser's view of it.


    Below image taken with a Canon 1D Mark II, 300 mm f/4 L IS lens (390mm eq.),
    ISO 100, manual exposure 1/125, f/5.6, mounted on a tripod, mirror lock-up, IS
    disabled, RAW file converted to TIFF, AND -- I quote from the photographer's
    page -- (can you believe he tried this many post-processing tricks and STILL
    couldn't beat a $400 P&S camera?)

    "The image was enlarged in Photoshop by another 1.141x to give 2x the original
    size. Next, the shadow/highlight tool was used to decrease the intensity of the
    brightest parts of the image. The Photoshop converter produced excessive noise
    in the sky, so the sky was selected and the curves tool was used to reduce
    noise. Next, InagesPlus was used with a Richardson-Lucy image restoration run,
    5x5 box, 29 iterations. Finally a jpeg created with highest quality setting
    using save for web. This image shows what can be done with a high
    signal-to-noise DSLR image."

    (For those of you who don't know, the Richardson-Lucy algorithm is a supposedly
    advanced upsampling method to try to regain even more resolution, inventing new
    details it as it goes along, even that didn't help.)

    http://www.clarkvision.com/imagedetail/moon-test1/moon-300mm.JZ3F9763.phtshp6144-rlnb.jpg


    Right. Enjoy your $10,000-$12,000 DSLRs and $1000-$3000 L-glass prime lenses!

    Thanks Roger, for showing everyone "what can be done with a high signal-to-noise
    DSLR image." You couldn't have made it more clear that today's P&S cameras are
    better no matter how hard you tried.

    For a P&S to beat a DSLR of this supposedly great quality, with all the
    drawbacks going for the P&S no less, or even if you don't see that as clearly as
    I do, you have to be honest that they at least match in resolution and clarity
    if nothing else. It's just simply total madness, complete and utter insanity, to
    even consider throwing that much money into any DSLR camera these days. To top
    it off? The FZ18 isn't even one with the sharpest optics available in P&S
    cameras. It's good, it's damn good, but there are P&S cameras with even slightly
    better glass than it has (just shorter zoom reach).
     
    TrevorAndersen, Nov 21, 2007
    #14

  15. Correction: Richardson-Lucy is an algorithm for refocusing soft edges without
    the haloing typically associated with unsharp mask. I was confusing this with
    another discussion I had recently. Still ... even THAT didn't help restore
    enough data to their proper places.
     
    TrevorAndersen, Nov 21, 2007
    #15
  16. Mike

    Mike Guest

    I tried viewing at several different settings, from 'full screen' to
    magnified so that I could see and 'study' apparent individual sensor
    pixels.
    Yes, resolution is only one aspect of 'sharpness'. I'm sure many people
    could add their way of judging 'sharpness'. I think that it is, to some
    degree, like pornography, I can't define it, but I know it when I see
    it. A really 'sharp' picture almost jumps out at you, it feels much
    more alive and vibrant. While color may also influence the perception
    of sharpness, it is a different attribute of a picture. I've taken some
    really sharp pictures in the past, with with good equipment (I still
    have my Canon F1 with it's 1.4 lens and associated equipment) or good
    luck. I just want to increase the chances of such successes in the future.
    Thanks....
     
    Mike, Nov 22, 2007
    #16
  17. Mike

    Douglas Guest

    160mm Focal length is hardly a usable portrait lens Frank. That's the field
    of view a 100mm lens ends up having when fitted to a 1.6 crop camera. It
    gets worse too. The field of view equal to 160mm and the depth of field of
    100mm. Hardly the ideal situation.

    This is why Canon make a 60mm Macro lens specifically to "equal" a 100mm
    macro lens on a 1.6 crop camera's field of view. It does not however provide
    any of the traditional benefits obtained by using a true "portrait lens" for
    35mm cameras of 75mm FL.

    You often hear about the perspective of a portrait. In fact a year or so ago
    some jokers in this group tried to tell me that because the perspective of a
    lens is unchanged at any focal length, using a correct lens for portraits
    had nothing to do with getting the perspective right.

    Here's my input to the OP. All Canon lenses are sharp, you just have be
    careful not to cut yourself with them!

    Douglas
     
    Douglas, Nov 22, 2007
    #17
  18. My apologies to the OP, I just thought it important to mention that L-glass
    isn't all what they make it out to be, an important consideration for his
    question. I'm just answering questions from others since then. This is my last
    post on the subject in your thread.

    Good luck in your search.


    Not my shot. I would have taken much more time on it to make sure it was even
    better than this.
    I guess we'll have to allow the FZ18 owner the same option of using RAW and 33
    different post-processing steps then, as were used on the Mk-II image. Then it
    would be fairer, wouldn't it? The FZ18 is right from the camera's JPG file, the
    Mk-II image is all the processing that can possibly be done to reclaim the best
    image available from the RAW data, even using noise reduction on it. Can you
    think of any other ways that you can falsely and blindly justify the DSLR?
    I find it interesting how people will subconsciously try to save the value of
    the DSLR kit in any way that they can. That's 390mm (35mm eq.) to 504mm (35mm
    eq.), not 300mm. A 22% difference in magnification is not going to account for
    nearly 200% difference in resolution. (might I remind you that this is native
    JPG vs time-consuming and grossly post-processed RAW too?)

    Doesn't the fact that a super-zoom lens on a $400 P&S camera can outdo the
    performance of a $2000 lens on a $10,000-$12,000 DSLR mean anything to you? That
    its JPG image is challenging the post-processed RAW file? I don't even care that
    it beat the DSLR, even if it came close it would still mean that there was
    something wrong with the DSLR and lens. The very fact that they are even
    comparable means that the DSLR lost. The P&S camera could have an image 1/2 the
    resolution as the DSLR and it would have still won in my book when you take
    everything else into account.
    I'm sorry, how does this matter? Do explain it so that everyone can understand
    your thought processes behind this. It must be fascinating.
    Yes it will. This is why the physically larger diameter DSLR lens can't have as
    sharp of optics wide-open as a smaller lens. It's impossible to configure larger
    glass to the same precision. At least not at consumer prices. This is why every
    DSLR lens can only be used at one or two f/stops for optimum resolution. Not
    true with a smaller better lens where its performance is about equal at all
    f/stops. They only have to figure DSLR glass good enough to define one large
    sensor area to get it past your inspection. You'll never notice that it's never
    any sharper than that. A smaller sensor lens has to define sensor areas 1/4th
    the size or smaller, meaning optics 4X's the precision, at least.
    Interesting. Still no blow-your-socks-off difference in resolution nor image
    quality. Side by side the FZ18 wins again for resolution. I hope you found the
    purchase worth it. I'd much rather spend my money on a lens that goes from 28mm
    f//2.8 to 508mm f/4.2 with the same or more image quality. How is that 504mm
    f/4.2 lens working out for you on that DSLR of yours? Does it provide the same
    image resolution and aperture as this one does? How much did it cost you? Now
    how much did it cost for all of your lenses combined to attain those
    focal-lengths? Don't forget that little bit of important "do the math" info when
    you start adding things up. How much do you have to haul around in weight? What
    happens when you are taking an image of a distant bird in flight with that 500mm
    lens and two unusual animals jump out of the woods, fighting just 15 feet away
    from you, the perfect chance to get those shots of some once-in-a-lifetime
    animal behavior (or video of it). Or a meteor or jet in trouble streaks across
    the whole sky, you need to capture the full trail of its passage before the
    trail dissipates. Luckily you zoomed out to 28mm in time. You even captured the
    explosion at the end of its flight. Did you change your DSLR lens in time? You
    all pride yourselves on how much better they are for fast-action needs. Surely
    your DSLR must have been able to capture that when you go from 500mm to 28mm in
    less than a second. It has to if it's that good for fast action uses. I would
    expect nothing less for spending that much money on a camera.

    None of this matters though. What does reality have to do with any DSLR owner's
    decisions. I'm sure you'll still find ways to justify your DSLR. You all do.
     
    TrevorAndersen, Nov 22, 2007
    #18
  19. The optical zoom on this camera is apparently equivalent to 504 mm on
    full-frame 35. That sets the image magnification; the digital zoom is
    irrelevant because it adds no extra detail.
    See the "390 mm equivalent" there? So the lens on the FZ18 was
    effectively about 1.3 times longer, capturing about 77% of the FOV
    compared to the lens used on the Canon. With that much extra
    magnification, the FZ18 is *expected* to capture more detail of the
    subject in the first place.

    The comparison would be approximately fair if the Canon had used a 400
    mm lens, but it didn't. So it makes no sense to compare these two
    photos for the amount of detail present.

    Having said that, I don't see any extra detail in the FZ18 image.

    Overall, it looks a bunch worse than the Canon one - but that's mostly
    because of the flat lighting and lack of any contrast enhancement, and
    we can't blame that on any difference between cameras.
    Um, you're forgetting that the P&S had the advantage of 30% more lens
    magnification. That's not a drawback, it's a major advantage.
    Since long zoom reach appears to be what gave the FZ18 whatever
    advantage it has here, there may be no better P&S camera for this
    particular test. However, that advantage could be wiped out by the
    DSLR camera owner simply using a 400 mm lens. Anything longer than
    that gives the advantage to the DSLR. 500 or 600 mm or 800 mm should
    do nicely, and are available as camera lenses. Or the DSLR owner could
    couple their camera to an astronomical telescope to get a lens with an
    effective focal length of 1000-2000 mm, which would capture *far* more
    detail under good conditions than either of the sample images discussed.

    (The moon is only 0.5 degree of arc in size, so a 2000 mm lens makes its
    image fill nearly the full height of the 1D sensor, and that's about the
    optimum focal length for this camera and subject).

    The advantage of the DSLR is that you *can* use the longer camera lens,
    and you can couple it to the astro telescope. I don't know why Roger
    used only a 300 mm lens in his test, unless it's the longest he happens
    to have - it's clearly not the best the camera can do for this subject.

    Dave
     
    Dave Martindale, Nov 22, 2007
    #19
  20. No P&S camera is ever at the "limit of its range" if you give the same freedom
    to them as the DSLR owner is using, which they all have. Put on an accessory
    lens to extend the range. From microscope to telescope, it matters not. Your
    argument sinks faster than a lead brick when reality is applied to it.

    The only limits are the ones that you planted in your mind.
     
    Howard Freine, Nov 22, 2007
    #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.