Lens fov-coverage comparison chart

Discussion in 'Education' started by robert100, Jun 26, 2012.

  1. Years ago in film photography I had a handy visual-aid "chart", giving a quick visual view of what you'd get in a photo,
    using a 35 mm camera, with various length lenses. IE a photo of an elk, taken at 500 yards or something, showing what
    you'd get in the photo with a 50mm, with a 100, with 135, etc etc, all the "common" lens lengths. This is a supremely
    handy reference, yet I have hunted the web using every term I can think of as a search-word, and can find nothing.,
    Anyone know of where to download such a beast ?
     
  2. thanks, most appreciated. Just took a flip into it, it doesn't actually have what I'm looking for exactly, it has 'some sample images of some shorter lengths', and it has the FOV charts of pretty well every length you could think of (from which I could, using math and taking a 10 meg image i shot at a length i know, then "build" what I'm after, but that is a few days work, and i'm not kidding, it's a long process). However,,,,even tho it may not have what i'm after right now, i really sincerely appreciate being sent to it because it looks like a superb site. I'll explain the reasoning behind what I want for a quick-visual. My present equipment is beginning to age and likely no longer cost effective to repair. So, I'm in the process of debating the "where do i go from here". IE I may switch from Canon and go with Nikon. I may, either instead of or in addition to one item, buy a super-zoom (either fuji hs30 or the Fuji XS1). In the dslr debate, i may pickup a nikon d5100 and/or a d3200... or....i may now go, finally, with the 'rebel' series of canon.....in either nikon or canon, once i do choose, although it is only wise to buy new bodies, used lenses are okay to buy....so...right now, for example, in canon, right now there is a used L-series 70-200 for sale....non IS....and i'm trying to give myself a visual peek-a-boo at "what i get in view, from a 200, versus a 300"....because there is also a 300 mm IS for sale used, good lens but not an "L"....so.....I want to play with "am i better off with the L series clarity and then crop up out of a 200 mm image? or does the xtra length and IS, make it wiser to go with the 300 'lesser' quality lens" ? I have built myself, in about an hour and a half last nite (using a photo of a bighorn sheep photo'd with an 85 canon at 10 meg as the starting point), a visual showing myself what I would have as a 'croppable image' of the actual bighorn itself "if" I had photo'd the exact same photo, with that exact 85 mm lens, using a 12 meg camera, a 14 meg, a 15, a 16, an 18 (ie canon now has that sensor in 3 bodies), and, last but not least, the 24 meg sensor in the nikon d3200 (and Sony, but sony is out for other reasons). This was interesting, in that regardless of the "mental math" you do in your head, it was really revealing to build the visual and "see" the difference. It really let me know, that moving from ten meg to anything prior to and less than 18, would not really have been worthwhile...but....that moving to 18 is worthwhile (which makes the canon a great choice) "and"....when I looked at what 24 meg does to it, I was "astounded" at what that can mean. So...."extend that", to taking into consideration what happens if you step up to either 18 meg, or to the 24 meg, "then, how much lens do you need ?" IE, a 200 mil L-series, hooked to an 18 meg image, is a whole lot different than my "experienced feel" has me thinking/feeling about because my "internal visuals" are based on years of my eyes looking at 10 meg images. This may sound dis-jointed. Does what I am saying make sense ? You combine a high- resolution 200 mil lens, to a 24 meg sensor, and nothing in my visual 'experience', in years of working at 10 meg, gives me a "physical sensation" of how enormous a gain that is...and while obviously, going to 300 mil is 'more', i'm only facing the question because a 200 mil L is there at a decent used price. All this may begin to sound "anal" and overly analytical, which in fact I'm actually not, And there are people out there, for whom 'one more' camera is no big deal, whose approach is "hey....go buy it". But, I'm an artist and photographer, and I work out of a mobile set-up. I have limited space, I do not like clutter, and I am trying to have a "small footprint". Trying to learn to buy the right tool, and only the right tool. So thanks for the advice. Assuming it loads, i'll post the composite photo i built as reference of what the bighorn photod at 10 meg would be in other meg-sensors using the same lens from the same spot.
    00aY7L-477431684.jpg
     
  3. okay, the image did go up. So now, using the reference image I built, my opinion is that to jump from 10
    meg (my 40d's) to anything less than 16 meg would have been a wasted move regardless. Go buy a longer
    lens. At 16 meg, it's "borderline", longer lens versus maybe time to go to a newer body and more megs.
    Now that 18 is here and good quality, easy decision, 18 is a justified jump from 10. And 24 ? WOW. 24 is
    definitely worth the move from 10. "But"....is Nikon's 24 worth skipping past Canon's 18 ? A heck of a lot
    more Interesting as a visual, than it is as a 'mental math' exercise.
     
  4. Craig: 11 minutes into the a.m. on wednesday....found it...the book you sent me to, page 125, has an
    exact visual of 200 versus 300 i can use to convert.....on my first flip thru i'd seen it jump from 100 directly
    to 400 and couldn't see the value in it...they showed the other, later....thanks again, great site....
     
  5. If all pixels were equal you might be on to something, but they are not, so you are not. If you need 300mm then cropping
    a 200mm is a poor substitute, a better thing to do is get a 1.4 TC to give you a one stop slower 280mm.
     
  6. thanks for the input....if you're referring back to the original query for a fov chart to compare image size
    gain from a 300 versus 200, then pixels are irrelevant to that portion of the topic (ie in film days, it
    wouldn't matter if i was shooting 64 asa chrome or 800 asa black and white, the fov of the lens doesn't
    change).... nor is the question of lens quality a factor in it......., the quality of the lens elemsnts will
    impact on any images produced, but have nothing to do with fov coverage.....if you are referring to the
    little "image chart" i tossed together using a photo of a bighorn sheep as a reference, then number of
    pixels is the only relevant data involved "for that specific purpose", as again, the chart has nothing to do
    with comparing "quality" of pixels coming off different sensors, its only application is to show the
    number (regardless of quality) of pixels you'd have to begin working with for cropping purposes..........in
    terms of quality though, cropping from a 200 mm L series Canon will "usually" produce a more
    satisfactory finished image than anything taken out of a 300 mm canon 75-300 kit lens even though it
    had longer reach and gave a "more pixels" image to work with, and the crop from a 200L f4 is going to
    give a much better starting-point to work with something out of a lesser quality 200 with a tele attached,

    another factor, from experience, one of the interesting things which impacts on lens choices etc, is that
    it is definitely easier to gain accurate focus on something 200 yards off if you use a 300 lens, than it is
    to have that same object in accurate focus in a 200 mm crop.....experience with each lens' own DOF
    factors become involved... etc etc....(and so does the viewfinder you're looking through).....

    i "may not be onto something" as you state.....but my experience tells me i probably are......IE, i get
    more "basic image" to work with using a 200 mm lens on an 18 meg sensor in a rebel t2i, than i do with
    an equivalent quality 300 mm lens on my 10 meg 40d......., and it definitely impacts on decisions as to
    "how many of what to pack around, at six pounds apiece".....going from a 10 meg sensor in an xti to the
    18 meg sensor in the t2i added nothing in weight to the rebel cameras....but from a 200 to a 300 lens in
    the same brand etc usually adds weight and bulk ....

    and whether i are or aren't onto something, you taking time to read it all and provide input is truly
    appreciated......have a great day
     
  7. This page might help.
    http://www.mylensdb.com/guide-field-of-view-calculator-visualizer/
     
  8. It isn't difficult to work out fov calculations from images if you know the starting point. Or just copy one of these images to give you a feel out in the field.
    With regards pixel density and cropping, if you up-sample the smaller mp to give you the same number of pixels and then display them next to each other there is often virtually no difference in detail. This is one example, if you download the 10mp and the 18mp files and upsize the smaller then the detail is virtually identical, even in perfect studio conditions where the differences should be maximised. Just move the sample square to the yellow feather and you will see the 80% more pixels give virtually no more resolved detail.
    And before you think this is all hypothesis, here are two images of mine shot from the same place with the same lens with different cameras. Both are well over 100% crops but one image is composed of over twice the pixels of the other, common wisdom dictates it should hold much more detail, but it doesn't. It does have a little more detail, but not appreciably more and, again, these were shot in optimal conditions that should favour the resolution advantage of the higher MP camera.
    00adFF-483389584.jpg
     
  9. Bob S, that fov calculator visualization tool is exactly what i was looking for. Most appreciated.

    Scott, of the two images above, the one on the right has much better detail, so my automatic guess is
    that all other things being constant, it was from the higher pixel sensor. Re the feather sample at
    dpreview, I took a quick glance at it but have to go back and play with it before commenting.

    Without using the exact numbers from the bighorn sheep example I set up (I can go back and calculate
    them but it'll take time and i was originally aiming at theory not exact numbers), here is my theory
    behind what I've suggested:

    If an object in the initial frame - ie the bighorn sheep - occupies, say, 10 percent (as an easy-round-
    number) of the total image....and if the overall image is a ten megapixel image....then that object, the
    bighorn in this case, is a one megapixel image......

    okay,,,,you take exactly the same photo from the same spot, same lens, but with an 18 megapixel
    sensor......the bighorn still, because of the fov of the lens being constant, still occupies 10 percent of
    the overall image.......and, in this case, ten percent of the overall image, is now one point eight
    megapixels in size.......thus when you take the bighorn you into digital manipulation, you are working
    with a starting image which already has...it must have......one point eight times as many individual
    points of data with which you are commencing work......

    Right ? I am commencing work on a 1 point 8 megapixel data bighorn, versus previously starting work
    with only a 1 megapixel data bighorn ......which either should make a difference, or we could all go back
    to using the first big one-point-one megapixel "pro" camera from 1990.....

    this incidentally is an issue "on top of" my original sample, but re your mention that not all pixels are
    equal, that is very true.....as an added thought-comparison, take my original sample using the 10 meg
    image from the 40d (which was actually in jpeg not raw, i was in the mountains for three days and
    shooting fine jpegs to make sure i had card-capacity) and "pretend" it is a RAW image crop at one meg
    sampled in a 12 bit processor, versus a one-point-eight meg sized crop from a sensor with a 14 bit
    RAW processor......it "should" add even more usable quality to the mix....

    the intriguing thing about the whole pixel-race of course, is that when we were using the original one-
    point-one meg beasts and then they brought out two-meg beasts, they then published charts saying
    "one meg is okay up to about 20x30 inch but over that you gotta go two meg"....and now they put out
    charts claiming 12 meg is no good for anything larger than a 4x6.....ya gotta love 'em....

    thanks for the feedback, both of you, most beneficial
     
  10. Robert,

    Don't confuse noise with resolution, the two images were unprocessed apart from the resizing. The version with more
    than twice the pixels has more than twice the noise, but very little additional resolution, it should have 50% more at the
    least.

    I understood your initial premise, that is why I commented, the numbers do not add up, pixels are very far from equal, and
    nothing beats sensor real estate. The logical extension of this is that a cropped high density 200mm image will not
    compare favorably with a lower density 300mm image, well that has been my finding.
     
  11. Scott, your findings, that the 200 high density crop is beaten by a lower density 300 non-crop, is vitally
    important because determining that is my objective of this beginning this entire exercise. You haven't
    mentioned in that however, "how much" denser an image your findings are based on. IE, there has never
    been any doubt in my mind that moving from 10 to 12 meg was pointless, or 10 to 14 or 15.....at 16 it
    struck me it was starting to enter a realm where it may though....and at 18 and up, it struck me the
    math may start to make it appear to become sensible....all based on equivalent sensor size ie
    apsc/apsh as a constant (with minor dimensional variance by mfr), and ignoring the differences in q
    between nikon canon etc etc.....(that becomes a different factor in any discusssion).

    What i started out to do, never found the fov chart i wanted and thus did not do initially, was to do the
    math to determine, as i did re sensor-percentage-size of a given object, the percent size of that same
    given object of the overall image, if photod with a 300 versus with a 200 as the only variable. The
    number this produces will be the "reverse effect" of what my little bighorn-sample showed happens with
    pixels.....in that study, more pixels in the original identical fov, the more pixels in a given object within
    it.....now, in moving from a 300 image and "backing-off" to a 200 mm fov, the same number of pixels
    have to cover a much much larger field....thus, the singular object, the bighorn, will consist once again
    of a smaller number of the overall pixels....a smaller percentage of the total. And it is entirely possible,
    that this math may show that in the end result, that object is back to being 10 percent or less of the
    overall image, thus negating the potential advantage. This "may" or "may not" be.....I haven't used the
    fov site to work it out yet.

    Re the two images you posted, just so there is no confusion, I am calling the one on "my" right as I look
    at my monitor, the "right hand image". (if you look at it from a performing-stage view they'd state it
    oppositely, so i wanted to clarify that). As I sit at my keyboard, the image on my monitor which appears
    above the "8-9-0" side I am calling the right hand image, and the one above the "1-2-3" I am calling the
    left hand image. And on my monitor, the right hand image is head-and-shoulders above the left hand one
    in quality, the resolution of the details of edges of lines, and of clarity of shapes etc, is significantly
    better than in the left hand one. IE if those two images were posted and claimed to be a "lens
    comparison", I would be prepared to pay far more for the lens producing the right hand one vs the left.
     
  12. Scott: re my comments about your two images, these are three of the objects I based my statement
    on. In looking at the loop of wire, the wire itself is substantially crisper and clearer in the right hand, and
    you really see it in the two highlights on the wire. In the left hand image, those two highlights "bulge"
    out, as though they are balloons of highlight...in the right hand image those two highlights are very
    obviously "part of" the image of the wire, they are captured "crisply", they don't "bulge out" beyond the
    shape of the wire itself. Then, looking at the "scratch" on the metal surface running from the bottom up
    to the "m"...on the left hand image that scratch is "blurry", while in the right hand image it is well-
    defined, with crisp edge-lines....and the same thing when looking at the white scracthed areas of high
    between the vertical marking lines above the "m"....in the left hand image those highs are "furry", in the
    right hand image, they have clearly defined shapes and edges. The right hand image is far superior in
    quality.

    Getting to view that has been very, very, appreciated and helpful. It would be interesting to know the two
    sensors involved.
     
  13. robert,
    Yes the right hand image is the one with twice the pixels, but I would urge you to download them and do a simple contrast adjust to them, there is very little resolution difference when the images are optimised. The right hand image should contain considerably more detail than it does.
    I don't recall the magnification, but it is considerably over 100%, but the right hand image contains well over twice the number of pixels. It is a comparison, effectively, between a 7mp camera and an 18mp camera, the higher number is also younger generation so should be even better.
    But one comment of yours is very relevant, lens price. A 200 f2.8 non IS is $820, a 300 f4 with IS is $1,359, not a small difference. If you are going to start comparing 200mm primes against similar priced 300mm zooms then my findings might be reversed.
    You have a lot more research to do to make your own conclusions, for my uses I discovered that pixel density is vastly overblown as a focal length "multiplier" and if I wanted a 300mm lens fov, using a 200mm lens on a higher pixel density body did not compare well to actually buying the 300mm.
     
  14. Scott, thanks for permission to download and play with the images. Actually, it downloads as a single
    image, which is no sweat, as I'm able to, and did, work on each half of it independently in Elements.
    Bottom line ? No matter what I did to the left hand, the one you indicate is 7 meg (? who made a 7 meg
    camera who also now makes an 18 meg one ?) I could not get the lines, ie the little loop of wire, to be
    as "crisp and accurate a line" as it was in the right hand image with no additional reworking at all.

    I could add chroma saturation density which improved the overall image, I could adjust the levels adding
    shadow density which improved it, I adjusted the contrast which improved the image a little as a single,
    and I worked on it using 'sharpening' and with the 'unsharp mask'....all of which improved it "but" never
    did manage to make the lines as crisp and certain as they were in the right hand as an untouched
    image.

    The only way I could have (and did not bother) added additional "crispness" to the wire-line highlights was to take the image into the 'distortion' tool and use a small, tight, brush to "push" the little bulgey balloon flares into being a crisp edge line. (it can be done that way if you only have one image and are forced to make it usable, and i have on a couple of occasions done so...and it is unbelievably slogging, time consuming work, make lottsa coffee before you start)

    Then, I applied the exact same work to the right hand image, and the differences in what was available
    to work with became hugely apparent, The saturation and density increases had significantly more
    "goodness" to them, each color had far more quality-depth to it than the left one ...and when I played
    with sharpness and an unsharp mask, I couldn't even apply the same amounts as I had with the left
    hand image....the right hand was already sharp enough to begin with, that when i started "increasing"
    that sharpness I was only able to even apply a minimal amount before it went "over the edge", into the
    realm where it started to look "obviously overworked".

    Bottom line: that right hand image is far, far, superior to the left one. Despite both coming down as
    jpegs, the right hand one has considerably more information-per-pixel already inherent in it to work with
    once you take it into post-processing.

    You allowing me to work with them was most, most, appreciated. Everything you've been adding to the topic has made a huge jump in what I'm gaining from this.
     

Share This Page