BillBrittain Posted March 14, 2007 Share #1 Posted March 14, 2007 Advertisement (gone after registration) No, I have not been drinking. Having read and responded to Chris Tribble's helpful survey thread, I found my self thinking today about Leica's next generation digital camera. I've always like the 6x6 crop as it comes from some medium format cameras. Recently, on another forum, a Canon "full frame" zealot was going on about how othr digital's "wasted" the len's full image area. These things got me to thinking. Why can't the digital sensor be round? Setting aside the monumental R&D costs for a new chip, it seems to me that offering a round RAW file would allow the photographer to crop in traditional 35mm framing, square, or any other format, whilst maximizing the use of the len's projected image. The rangefinder could still produce traditional 35mm framelines, or Leica could produce special custom framelines in whatever format the buyer desires. Talk about Full-Frame! What am I missing here? Be gentle please Link to post Share on other sites More sharing options...
Advertisement Posted March 14, 2007 Posted March 14, 2007 Hi BillBrittain, Take a look here Why not a ROUND sensor?. I'm sure you'll find what you were looking for!
Guest guy_mancuso Posted March 14, 2007 Share #2 Posted March 14, 2007 Square box in a round hole , or round peg in a square box. Camera size would have to increase to do it for one. But on the other hand the lens projects a image circle. Then i would have to change all my picture frames in the house to round ones. LOL Actually this has come up before and in theory is not a bad idea at all. You could take that 24 x36 and make it more like 30 X 36 but the mirror box would have to change and everything else Link to post Share on other sites More sharing options...
dugby Posted March 14, 2007 Share #3 Posted March 14, 2007 In a full frame, careful examination of the corners of the image will reveal lens based spherical aberrations. Put that 36mmx24mm rectangle into a circle. Most of the image on the outsides of the corner, following the circle will be blurred, out of focus and the colours changed. So in an existing FF lens, the best image is actual a square, without doing the exact mathematics, about 30mmx30mm giving an area of 900mm~2, whilst a FF is 864mm~2 Link to post Share on other sites More sharing options...
dist Posted March 14, 2007 Share #4 Posted March 14, 2007 Maybe the fact that when you cut circles from a plane area you waste 22.5% of the material enters the equation as well. Just guessing... Link to post Share on other sites More sharing options...
BillBrittain Posted March 14, 2007 Author Share #5 Posted March 14, 2007 In a full frame, careful examination of the corners of the image will reveal lens based spherical aberrations. Put that 36mmx24mm rectangle into a circle. Most of the image on the outsides of the corner, following the circle will be blurred, out of focus and the colours changed. So in an existing FF lens, the best image is actual a square, without doing the exact mathematics, about 30mmx30mm giving an area of 900mm~2, whilst a FF is 864mm~2 This is what entered into my thinking. A square would be the most efficient rectangular crop, but starting from a round sensor one would have the freedom to use the maximum area of the sensor regardless of the crop selected. Guy, if they make a round sensor it sounds like we'll have to start manufacturing round frames! Link to post Share on other sites More sharing options...
george + Posted March 14, 2007 Share #6 Posted March 14, 2007 A square sensor makes a lot of sense from a lens optics point of view. But a round sensor would create quite a few electronic - and manufacturing - challenges. Then come the round frames. Link to post Share on other sites More sharing options...
Guest guy_mancuso Posted March 14, 2007 Share #7 Posted March 14, 2007 Advertisement (gone after registration) Let's not forget round photo paper . If you really think about it when 35mm was born. It was more probably out of camera size than anything else where they came up with 24x36 because if it was taller say 30mm it would be a much taller camera ,. Sort of like the V hassy it would have been like looking. Plus i guess some of that decision may have been to be like stationary letter size. Interesting to see WHY 24 x36 was born in the first place Link to post Share on other sites More sharing options...
ennjott Posted March 14, 2007 Share #8 Posted March 14, 2007 It was born by using a double cine frame (24x18).. now the question is, where did THAT come from? =) Link to post Share on other sites More sharing options...
jaapv Posted March 14, 2007 Share #9 Posted March 14, 2007 The only reason they came up with 24x36 was because that was the standard movie film size back in the early years of the 20th century. (well, actually 24x18, because one wants the film to run through a moviecamera from top to bottom.) Barnack started out with the idea of building an exposure meter for movie film and thought of pulling the film through horizontally, enabling a doubling of the format. And that was how it all started.... Link to post Share on other sites More sharing options...
jaapv Posted March 14, 2007 Share #10 Posted March 14, 2007 It was born by using a double cine frame (24x18).. now the question is, where did THAT come from? =) 35 mm wide film was introduced in 1889 by Edison and George Eastman as a compromise between cost and quality.It turned out to be unwieldy in use and was reduced to 24 mm film and later to other formats, like the 16 mm film and the 9.5mm film by Pathe which had quite a run of popularity in the 1900rds. Later, as emulsions improved, 8 mm came in use for amateur film, ending, of course, with the post-WWII Super 8 film. And then video killed it all.... Link to post Share on other sites More sharing options...
dugby Posted March 14, 2007 Share #11 Posted March 14, 2007 I seem to recall something about Greek mathematics, architecture and Pythagorus..... the ancient Greek architects and designers found the 'magic' ratio of 3x4x5 for buildings, and artwork was visually pleasant on the eye. This may have influenced Leica to pursue to 4 x 3 image ratio lately, in contrast to the other vendors pursuit of 2 x 3 Link to post Share on other sites More sharing options...
stephengilbert Posted March 15, 2007 Share #12 Posted March 15, 2007 Other than the added cost of producing chips some of whose area would not often be used, why not? They have round image film cameras (Bostick-Sullivan :: Hobo Cameras, and they produce very interesting images. The frames might be costly, though. Link to post Share on other sites More sharing options...
rceres Posted March 15, 2007 Share #13 Posted March 15, 2007 Seems to me like the next logical step would be to consider sensor pixel placement and shape options other than a stacked rectillinear array. I'm thinking that perhaps hexagonal pixels might provide better sensor performance and very different moire sensetivity. Link to post Share on other sites More sharing options...
andit Posted March 15, 2007 Share #14 Posted March 15, 2007 While not exactly round, Fuji Film has been working in this direction for some time. Where most companies use square pixels on there sensors, Fuji got thinking that a octagnol pixel would work better. That is basically Fuji's Super CCD technology. What they do at this stage is use two different size octagnol pixels on there sensors (the sensor is square at the end however and if you enlarged it enough would look like a honey comb). The large pixels are for reducing noise in lowlight and the small ones for highlight situations. That is why you will see there CCD's have something like a total of 6MP with an effective 12MP resolution. They are also working on different layer CCD's - much like film that had 3 or 4 emulsion layers. Is this the future? I don't know, I have not worked with Fuji equipment, so I can not say how the image quality is. In their defence however, they are working in new directions and doing a lot of R&D. One day this will lead to something. Andreas Link to post Share on other sites More sharing options...
BillBrittain Posted March 15, 2007 Author Share #15 Posted March 15, 2007 Having thought about this overnight, it seems to me that the technical advantages of a round sensor would be great. Even with a slightly increased body size, I would enjoy the ability to crop as I see fit and to use the maximum number pof pixels available from the lens. Perhaps the biggest problem (again, apart from R&D) would be marketing. Explaining the availabilty of different aspect ratios would be tricky for the average marketing department. Then again, Leica could do it and say something like "We were the first to get film right, now we're the first to get digital right". Also, Leica might be the only small format brand that has enough user sophistication to understand the benefits of variable aspect ratios and to integrate cropping into their RAW workflow. OK Leica, if use this idea I will test a camera for free Link to post Share on other sites More sharing options...
larry Posted March 15, 2007 Share #16 Posted March 15, 2007 It was born by using a double cine frame (24x18).. now the question is, where did THAT come from? =) The Golden Mean (The Golden Mean) Link to post Share on other sites More sharing options...
gennfiks Posted March 15, 2007 Share #17 Posted March 15, 2007 The next logical step is a curved (spherical) sensor to guarantee that all the rays fall at 90º. Also one should be able to change the curvature of the sensor when one changes the lens from 28 to 90mm type, for example. Ones that is acheved we could come back to use thinner bodies as the older 35mm cameras and FF sensors. Wishfull thinking, but.... Cheers, Gennady Link to post Share on other sites More sharing options...
george + Posted March 15, 2007 Share #18 Posted March 15, 2007 A curved sensor is an interesting thought too. How you get a curved piece of silicon is a question. But more importantly, you may not be able to use current lenses which, I think, are designed to project to flat surfaces. Link to post Share on other sites More sharing options...
gennfiks Posted March 15, 2007 Share #19 Posted March 15, 2007 I would like to address the last question, before going into the first one. Why are we using microlenses on the ccd sesnors then? Is it not because the rays are not hitting the sensor at 90º and are not being "registeres" by a photodetector, another words being reflected? At least my understanding of the issue was that. Please correct me if I'm wrong. All this noise about the film being capable of absorbing light coming from different angles and sesnor (glass) not. That is why so much work was done on "digital lenses" to be able to deliver light at 90º to the sensor. The curved sensor - one "stupid" solution that I could imagine is based on MEMS (machined micromirros) used extensively nowdays which can rotate if the voltage is applied to it. Well, create an array of these where each micromirror has 1 or various pixels on it. Each mirror can be rotated individually and the whole array could be made to assume a spherical shape of various radiai. G Link to post Share on other sites More sharing options...
rceres Posted March 15, 2007 Share #20 Posted March 15, 2007 The next logical step is a curved (spherical) sensor to guarantee that all the rays fall at 90º. Also one should be able to change the curvature of the sensor when one changes the lens from 28 to 90mm type, for example. Ones that is acheved we could come back to use thinner bodies as the older 35mm cameras and FF sensors. Wishfull thinking, but.... Cheers, Gennady This would require a complete redesign of all the lenes since they currently are designed to focus on a flat plane. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.