otto.f Posted November 16, 2014 Share #81 Â Posted November 16, 2014 Advertisement (gone after registration) Ok captain Link to post Share on other sites More sharing options...
Advertisement Posted November 16, 2014 Posted November 16, 2014 Hi otto.f, Take a look here New article on overgaard.dk - "Leica 75mm Summilux-M f/1.4". I'm sure you'll find what you were looking for!
CheshireCat Posted November 17, 2014 Share #82 Â Posted November 17, 2014 A picture taken with a 50 mm lens at f/1.4 has approx. the same depth-of-field as a picture taken with a 90 mm lens at f/2.5 when the former gets cropped to the same field-of-view as the latter. Â This sentence is not technically correct. The 90 at f/2.5 will give the same look. The 90 at f/4.5 will give the same DoF (CoC < pixelsize). Link to post Share on other sites More sharing options...
01af Posted November 17, 2014 Share #83 Â Posted November 17, 2014 A picture taken with a 50 mm lens at f/1.4 has approx. the same depth-of-field as a picture taken with a 90 mm lens at f/2.5 when the former gets cropped to the same field-of-view as the latter.This sentence is not technically correct. It is. Â Â The 90 at f/2.5 will give the same look.The 90 at f/4.5 will give the same DoF ... Now this is a statement that doesn't make sense. Â Â ... (CoC < pixelsize). Pixel size has absolutely nothing to do with this. Link to post Share on other sites More sharing options...
jaapv Posted November 17, 2014 Share #84 Â Posted November 17, 2014 Â Â Â Pixel size has absolutely nothing to do with this. Sensu strictu it has. Effectively the CoC can never be smaller than the pixel size. Not that that has any relationship to the realworld situation. Link to post Share on other sites More sharing options...
M9reno Posted November 17, 2014 Share #85 Â Posted November 17, 2014 Sensu stricto, just to add to the pedantry! Link to post Share on other sites More sharing options...
01af Posted November 17, 2014 Share #86 Â Posted November 17, 2014 Sensu strictu it has. No, it hasn't. Â Â Effectively the CoC can never be smaller than the pixel size. Yes, it can. Link to post Share on other sites More sharing options...
jaapv Posted November 17, 2014 Share #87  Posted November 17, 2014 Advertisement (gone after registration) No, it hasn't.   Yes, it can. Yes it can, but it won't have any effect any more.The DOF remaining will be determined by the pixel size. Link to post Share on other sites More sharing options...
TonyField Posted November 17, 2014 Share #88 Â Posted November 17, 2014 This sentence is not technically correct.The 90 at f/2.5 will give the same look. The 90 at f/4.5 will give the same DoF (CoC < pixelsize). Â Olaf is definitely correct when you look at final image magnification and viewing distance using one of our trusty Leica full frame cameras. His use of DOF assumes that the desire is to have the 90 mm image at F2.5 enlarged full frame to display (for example, a 24x36 cm print image). The corresponding image from the 50 mm lens at F1.4 taken from the same shooting position and enlarged to display the SAME FIELD OF VIEW (i.e. a strong crop from the frame) with the same 24x36 cm final image size will have very similar visual DOF when both final images are viewed from the same distance. Â Your suggestion of 90mm at F4.5 is certainly true only if the full frame of both images are enlarged to the same spatial dimensions without crop. Â In all cases, the computation of the DOF must have a circle of confusion value that is correct for final image magnification and final viewing distance. With these two lenses, the COC must be adjusted by a factor of 50/90 = 0.555 if you want the same final image field of view. Link to post Share on other sites More sharing options...
01af Posted November 17, 2014 Share #89  Posted November 17, 2014 Yes it can, but it won't have any effect any more.The DOF remaining will be determined by the pixel size. This is a common misconception.  Seriously—pixel size totally doesn't matter in all things concerning depth-of-field. Link to post Share on other sites More sharing options...
jaapv Posted November 17, 2014 Share #90 Â Posted November 17, 2014 As DOF is linked to the resolving power of the eye and the pixel size is the maximum you can increase resolution to, so there is a limit. Obviously the COC can be made smaller, but it will have no effect any more. I am talking about this as a limit. As soon as the CoC is larger than the pixel size the pixel size has no influence -obviously. Link to post Share on other sites More sharing options...
pop Posted November 17, 2014 Share #91 Â Posted November 17, 2014 Obviously the COC can be made smaller, but it will have no effect any more. Â I think this would hold true if the COC (the blurred image of a point shaped object) was smack in the middle of a pixel. Move it close to the border and it will taint the adjacent pixel. Link to post Share on other sites More sharing options...
01af Posted November 17, 2014 Share #92  Posted November 17, 2014 ... the pixel size is the maximum you can increase resolution to, so there is a limit. Obviously the COC can be made smaller, but it will have no effect any more. As I said—this is a common misconception.  Obviously you are unwilling to believe what I'm saying ... but you don't have to. Simply try it for yourself in real life. Compute the hyperfocal distance for a lens and an aperture of your choice when the maximum diameter for the circle of confusion matches your camera's pixel pitch. Set the said lens to said aperture and focus at the computed hyperfocal distance. Shoot something detailed at a far-away distance (beyond the hyperfocal distance) on a clear day, then shoot the same scene with the lens properly focused at infinity.  According to your theory, there shouldn't be any difference in the pictures' sharpness of far-away objects. But in fact, there will be. Because circles of confusion smaller than the pixel pitch will yield a sharper image than circles of confusion equal to the pixel pitch. Link to post Share on other sites More sharing options...
jaapv Posted November 17, 2014 Share #93 Â Posted November 17, 2014 @ POP: Ok, in that case we might assume half pixel size. The argument holds, however. Link to post Share on other sites More sharing options...
TonyField Posted November 17, 2014 Share #94 Â Posted November 17, 2014 Of course we have an additional concept of "pixel". With the exception of the Leica MM, an image pixel is a square of four pixels (GRGB) which further complicates the issue. Link to post Share on other sites More sharing options...
jaapv Posted November 17, 2014 Share #95 Â Posted November 17, 2014 In this case we are arguing angels on the head of a pin - which is much larger than a pixel Link to post Share on other sites More sharing options...
pop Posted November 17, 2014 Share #96 Â Posted November 17, 2014 Actually, the things under the Bayer filter should not be called "pixels" at all. Â A pixel is an "atomic" picture element holding the complete information about one location within the image. The pit on the sensor and the corresponding digit in the raw image file holds the brightness value for one color channel only. The pixels in later steps of the processing chain are quite artificial, being reconstructed by some computing tricks. Link to post Share on other sites More sharing options...
01af Posted November 17, 2014 Share #97  Posted November 17, 2014 ... a square of four pixels (GRGB) which further complicates the issue. No, it doesn't. In this context, pixel size doesn't matter anyway, so it also doesn't matter what exactly a pixel—or a sensel—is. Link to post Share on other sites More sharing options...
TonyField Posted November 17, 2014 Share #98  Posted November 17, 2014 No, it doesn't. In this context, pixel size doesn't matter anyway, so it also doesn't matter what exactly a pixel—or a sensel—is.  I think you are going to have to explain this.  The only pixels that are "perfectly sharp" (assuming perfect optics, etc) are those at the exact plane of focus. This is, of course, independent of the pixel size. Other parts of the image are "very slightly" or possibly "very much" out of focus ... depending on the focus deviation from the image plane. This is a continuous function Any degree of out-of-focus will affect some or all adjacent pixels to some extent.  As Jaap implies, there is some probably some size of CoC relative to pixel size at which the decrease of CoC value can not change the apparent sharpness detectable by the human eye.  If you disagree with this assessment, please explain why. Link to post Share on other sites More sharing options...
Exodies Posted November 18, 2014 Share #99 Â Posted November 18, 2014 Pixel size sets an upper limit on the number of line pairs per mm which can be resolved by the sensor. If, in practice, this is much larger than the customary value (5) used to calculate acceptable coc diameters then 01af would be correct in real situations. Otherwise, it just sounds mysterious. I too would like to read an explanation. Link to post Share on other sites More sharing options...
TonyField Posted November 18, 2014 Share #100  Posted November 18, 2014 ... Compute the hyperfocal distance for a lens and an aperture of your choice when the maximum diameter for the circle of confusion matches your camera's pixel pitch. Set the said lens to said aperture and focus at the computed hyperfocal distance. Shoot something detailed at a far-away distance (beyond the hyperfocal distance) on a clear day, then shoot the same scene with the lens properly focused at infinity. According to your theory, there shouldn't be any difference in the pictures' sharpness of far-away objects. But in fact, there will be. Because circles of confusion smaller than the pixel pitch will yield a sharper image than circles of confusion equal to the pixel pitch.  If the CoC matches the pixel pitch and you shoot at the computed hyperfocal distance, then we should see sharpness gains by focusing on infinity. This is obvious since focusing on infinity yields "perfect sharpness" ... closer subject material is, of course, "out of focus" to some degree. However, there will be some detail smearing between pixels (detail that is partly in two adjacent pixel).  If the CoC is much smaller than the pixel pitch (lets assume 1/4 the size or smaller), you are pushing the hyperfocal focus plane "closer to infinity". There should be less smearing of the sharp detail and you should in theory see the sharpness gain.  I do not know if a CoC in the vicinity of the pixel pitch (and smaller) is practical. I wonder how good new or old Leica lenses really are when dealing with detail at this very high resolution (I only use ancient film day lenses that I collected since the late 1960's and these lenses are not up to modern sharpness and contrast standards). This type of critical sharpness has never been of concern to me  However, no matter how I think of it, pixel / sensel size does matter Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.