Panasonic Lumix DMC-LX5 announced - looks good!
in [Digital Cameras]
|
Prev: The horror of the small sensor P&S image
Next: Scott Linstead's high-speed photographs capture creatures frozen in time
From: Peter on 24 Jul 2010 20:07 "Chris Malcolm" <cam(a)holyrood.ed.ac.uk> wrote in message news:8b1863FeeuU2(a)mid.individual.net... > Peter <peternew(a)nospamoptonline.net> wrote: >> "David J Taylor" <david-taylor(a)blueyonder.co.uk.invalid> wrote in message >> news:i2brsa$kjt$1(a)news.eternal-september.org... >>> "Bruce" <docnews2011(a)gmail.com> wrote in message >>> news:hmah46d44eql5al69cktajudm79givupd1(a)4ax.com... >>> [] >>>> Olympus offers an 8mm f/3.5 fisheye lens for Four Thirds. You can >>>> remove the fisheye "distortion" in post processing, just as you can >>>> with the AF Nikkor 10.5mm f/2.8G DX. >>> >>> True, but not what I was thinking of as a compact, rectilinear lens. > >> did you look at the Sigma 8-16. It is not compact, but is rectilinear. I >> played with one last week and published my short review in one of the >> groups. >> Summary: I am seriously considering getting that lens > > Got one, like it very much! Here's an "all four walls" interior shot, > plus some other examples. > > http://www.flickr.com/photos/chris_malcolm/4815223713/in/set-72157624398875799/ > Forgot to mention, I really like your lily shot. -- Peter
From: David Ruether on 25 Jul 2010 10:03 "Chris Malcolm" <cam(a)holyrood.ed.ac.uk> wrote in message news:8b18gnFeeuU3(a)mid.individual.net... > David Ruether <d_ruether(a)thotmail.com> wrote: >> [...] But, if you meant, fisheye lenses make images >> that are more like the way we see", I think you are right! ;-) >> --DR > Not the way I see. I can't detect any barrel distortion of straight > lines in my eyesight, although there might be a little. > -- > Chris Malcolm OK, here are a few "exercises" for you...;-). First, learn to see "wide angle" (as in being able to become aware of the visual field beyond the center of your vision, without directly looking at it). This does require some practice to widen your "field of attention" when you want (fun while driving! ;-). Once this is possible for you, if you are (for instance) at a computer desk that is facing a wall, and you are directly facing that wall, expanding your field of view (without moving your eyes, which should remain aimed straight ahead), you should be able easily to see the curvature of the wall/ceiling intersection as it passes overhead in front of you (or at least see the left and right simultaneous convergence of parallel lines - which could not happen in rectangular perspective without abrupt bends or breaks in the lines). This may be easier to see at first by standing against one wall midway along a long hallway with doors (as in a hotel) and looking fixedly directly across the hallway and then observing the door tops and wall/ceiling and floor/wall intersections as they move away from you (they clearly curve to move toward intersection in the distance, something that could not happen with rectangular perspective, by definition). Also, next time you are at a beach with a clearly defined sky and water intersection, try looking somewhat up. Guess what? The horizon line clearly bends into an upward-facing bowl-curve! Now you are ready to prove to your friends (and otherwise ;-) that they don't see the way they think they do! 8^) Remember when I pointed out earlier in this thread that within a narrow angle of view, all the perspective types look about the same? It is when you "can go wide" with your vision (or with photography, with wide-angle lenses) that you can see the differences in perspective types. Unfortunately, most people seem satisfied with "telephoto sight", and must move their eyes around wildly to construct an understanding of their surroundings. How limiting! 8^) --David Ruether www.donferrario.com/ruether d_ruether....@....hotmail.com
From: David Ruether on 25 Jul 2010 11:26 "David Ruether" <d_ruether(a)thotmail.com> wrote in message news:i2hg79$kon$1(a)ruby.cit.cornell.edu... > "Chris Malcolm" <cam(a)holyrood.ed.ac.uk> wrote in message > news:8b18gnFeeuU3(a)mid.individual.net... >> David Ruether <d_ruether(a)thotmail.com> wrote: >>> [...] But, if you meant, fisheye lenses make images >>> that are more like the way we see", I think you are right! ;-) >>> --DR >> Not the way I see. I can't detect any barrel distortion of straight >> lines in my eyesight, although there might be a little. >> -- >> Chris Malcolm > OK, here are a few "exercises" for you...;-). First, learn to see > "wide angle" (as in being able to become aware of the visual > field beyond the center of your vision, without directly looking > at it). This does require some practice to widen your "field of > attention" when you want (fun while driving! ;-). Once this is > possible for you, if you are (for instance) at a computer desk > that is facing a wall, and you are directly facing that wall, > expanding your field of view (without moving your eyes, which > should remain aimed straight ahead), you should be able easily > to see the curvature of the wall/ceiling intersection as it passes > overhead in front of you (or at least see the left and right > simultaneous convergence of parallel lines - which could not > happen in rectangular perspective without abrupt bends or > breaks in the lines). This may be easier to see at first by standing > against one wall midway along a long hallway with doors (as in > a hotel) and looking fixedly directly across the hallway and > then observing the door tops and wall/ceiling and floor/wall > intersections as they move away from you (they clearly curve > to move toward intersection in the distance, something that > could not happen with rectangular perspective, by definition). > Also, next time you are at a beach with a clearly defined sky > and water intersection, try looking somewhat up. Guess what? > The horizon line clearly bends into an upward-facing bowl-curve! > Now you are ready to prove to your friends (and otherwise ;-) > that they don't see the way they think they do! 8^) Remember > when I pointed out earlier in this thread that within a narrow angle > of view, all the perspective types look about the same? It is when > you "can go wide" with your vision (or with photography, with > wide-angle lenses) that you can see the differences in perspective > types. Unfortunately, most people seem satisfied with "telephoto > sight", and must move their eyes around wildly to construct an > understanding of their surroundings. How limiting! 8^) > --David Ruether [For an example of a "perspective-free" view, look here -- http://www.donferrario.com/ruether/500mm-Nikkor.htm ] To add -- one of the advantages of fisheye perspective is that as the point of view is changed, the peripheral "events" are relatively undisturbed compared with extreme WA rectangular-perspective views in motion (proportions remain relatively unchanged, as do angles) - for an example, look at this image at -- http://www.donferrario.com/ruether/web_photos/aht_photos/ba09.jpg and note that the building vertical parallel lines remain nearly vertical and parallel, unlike what a super-wide rectangular lens would show when tipped. Also note the "natural" front-to-back subject element proportioning as opposed to the same that would be exaggerated in a similarly wide rectangular-perspective view. For an example of that, see -- http://www.donferrario.com/ruether/web_photos/phun_fotoz/people/ba17.jpg or, for a more extreme example, see -- http://www.donferrario.com/ruether/12mm-Voightlander.htm For an attempt at approximating what we see (including an angle of view greater than 180 degrees horizontally), and with facial structural obstructions, see -- http://www.donferrario.com/ruether/eyes-view.htm (note that within narrower angles, the view is not extremely unlike a moderate rectangular perspective type, so long as the straight parallel lines do not occupy a relatively major part of the image). But, again, for more on this (instead of making uninformed assertions...;-), it may be worth one's while to look over these sites -- http://www.donferrario.com/ruether/lens_perspective_types.htm http://www.donferrario.com/ruether/lens-angle-of-view-and-perspective.htm http://www.donferrario.com/ruether/seeing_and_perspective.htm http://www.fullscreen360.com/st-helens.htm --David Ruether www.donferrario.com/ruether d_ruether....@....hotmail.com
From: Chris Malcolm on 26 Jul 2010 03:35 David Ruether <d_ruether(a)thotmail.com> wrote: > "Chris Malcolm" <cam(a)holyrood.ed.ac.uk> wrote in message > news:8b18gnFeeuU3(a)mid.individual.net... >> David Ruether <d_ruether(a)thotmail.com> wrote: >>> [...] But, if you meant, fisheye lenses make images >>> that are more like the way we see", I think you are right! ;-) >>> --DR >> Not the way I see. I can't detect any barrel distortion of straight >> lines in my eyesight, although there might be a little. >> -- >> Chris Malcolm > OK, here are a few "exercises" for you...;-). First, learn to see > "wide angle" (as in being able to become aware of the visual > field beyond the center of your vision, without directly looking > at it). This does require some practice to widen your "field of > attention" when you want (fun while driving! ;-). Once this is > possible for you, if you are (for instance) at a computer desk > that is facing a wall, and you are directly facing that wall, > expanding your field of view (without moving your eyes, which > should remain aimed straight ahead), you should be able easily > to see the curvature of the wall/ceiling intersection as it passes > overhead in front of you (or at least see the left and right > simultaneous convergence of parallel lines - which could not > happen in rectangular perspective without abrupt bends or > breaks in the lines). This may be easier to see at first by standing > against one wall midway along a long hallway with doors (as in > a hotel) and looking fixedly directly across the hallway and > then observing the door tops and wall/ceiling and floor/wall > intersections as they move away from you (they clearly curve > to move toward intersection in the distance, something that > could not happen with rectangular perspective, by definition). > Also, next time you are at a beach with a clearly defined sky > and water intersection, try looking somewhat up. Guess what? > The horizon line clearly bends into an upward-facing bowl-curve! > Now you are ready to prove to your friends (and otherwise ;-) > that they don't see the way they think they do! 8^) Remember > when I pointed out earlier in this thread that within a narrow angle > of view, all the perspective types look about the same? It is when > you "can go wide" with your vision (or with photography, with > wide-angle lenses) that you can see the differences in perspective > types. Unfortunately, most people seem satisfied with "telephoto > sight", and must move their eyes around wildly to construct an > understanding of their surroundings. How limiting! 8^) > --David Ruether > www.donferrario.com/ruether > d_ruether....@....hotmail.com Yes, I agree with that. You're correct that the optics of the eye demand that kind of result. But we don't see with the eye, what we see is the results of the brain's processing of te RAW data from the eye. That's why we don't see the blind spot, or indeed any of the twiggery of the blood vessels etc. that pass across the front of the retina. The brain edits them out. But it doesn't just do what you might call noise reduction, dud "pixels" removal, etc., it stitches together the latest input from the retina with stuff from previous views as the eye flicks around in its flickering saccadic jumps. That's why we think we can see a fairly wide fairly sharp area, such as the page of a book, instead of only the few sharp words we're looking at and the rest a blur. Attention and expectation also play a big part in this visual construction of the brain. There are tricks by which you can fool the brain into letting you see the blind spot, the twiggery, just as the tricks you've described which let you see the inevitable barrel distortion of a fixed view. If you like these tricks let us get closer to the RAW image from a stationary retina. But that's not how the eye normally works. We normally see a highly post-processed stitch up by the brain. That's what we see when we look at a view, and what we see when we look at a photograph of the view. My point is that my brain works quite hard to present me with a rectilinear representation of the world, despite not having rectilinear optics. -- Chris Malcolm Warning: none of the above is indisputable fact.
From: Chris Malcolm on 26 Jul 2010 04:06
Peter <peternew(a)nospamoptonline.net> wrote: > "Chris Malcolm" <cam(a)holyrood.ed.ac.uk> wrote in message > news:8b1863FeeuU2(a)mid.individual.net... >> Peter <peternew(a)nospamoptonline.net> wrote: >>> "David J Taylor" <david-taylor(a)blueyonder.co.uk.invalid> wrote in message >>> news:i2brsa$kjt$1(a)news.eternal-september.org... >>>> "Bruce" <docnews2011(a)gmail.com> wrote in message >>>> news:hmah46d44eql5al69cktajudm79givupd1(a)4ax.com... >>>> [] >>>>> Olympus offers an 8mm f/3.5 fisheye lens for Four Thirds. You can >>>>> remove the fisheye "distortion" in post processing, just as you can >>>>> with the AF Nikkor 10.5mm f/2.8G DX. >>>> >>>> True, but not what I was thinking of as a compact, rectilinear lens. >> >>> did you look at the Sigma 8-16. It is not compact, but is rectilinear. I >>> played with one last week and published my short review in one of the >>> groups. >>> Summary: I am seriously considering getting that lens >> >> Got one, like it very much! Here's an "all four walls" interior shot, >> plus some other examples. >> >> http://www.flickr.com/photos/chris_malcolm/4815223713/in/set-72157624398875799/ > I found some minor CA in the one I tried. I have been using the Nikon 12-24 > and the 10.5 semi fisheye and correcting the linear distortion in > CaptureNX2. The additional 2 makes a world of difference. What vices have > you found in the lens, besides some CA and the expected barrel distortion? I've found small amounts of CA near the edges, but only if there are extreme contrasts involved, such as twigs or building edges against sky. Quite a lot less than my copy of Sigma's 10-20mm, which unfortunately I won't be able to direct direct comparisons with until it's been repaired. But I have been taking it around to locations where I've previously got good shots with my 10-20mm, and in general it seems to be at least a slightly better lens all round, which performs better at 8mm than my 10-20mm did at 10mm, and also holds its performance better when the aperture is opened up. I haven't done enough testing to confirm this, but I suspect that optimal sharpness in the central area comes at lower apertures than f8, which if true is impressive. It does suffer from what I suspect may be a generic vice of ultra wide angle lenses, at least with some DSLR AF systems, and that is unreliable autofocus. While trying to hand hold the camera steadily pointing at some large central thing with central spot focus it can jump a lot in repeated focuses between what it thinks is the correct focus. Even when it's stable and doesn't jump around, it will quite often indicate a distance on the lens barrel which is clearly silly. That this is in fact silly is confirmed by careful manual focusing which always produces more sensible focus distances. My camera can do live view manual focusing with the LCD image magnified as much as I like, which makes accurate manual focusing very easy, in fact much easier than on any other camera I've ever owned, including film SLRs in the days before autofocus. Is this autofocus uncertainty just due to the huge DoF of such wide lenses? Perhaps a contributor, but certainly not the whole story, because I find when using autofocus with both the 8-16mm and the 10-20mm that I'll often get a rather soft image, and occasionally a sharper one, indicating that the problem isn't the lens. Using manual focus gets me a higher proportion of sharp shots. So I usually either focus it manually or leave it preset at a handy hyperfocal setting for the kind of distance ranges I'm encountering. -- Chris Malcolm Warning: none of the above is indisputable fact. |