 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
| PhotoKB HomeDiscussion Groups Digital Photography Digital PhotoDSLR CamerasZLR CamerasPoint & Shoot CamerasFilm Photography 35 mmLarge FormatMedium formatDarkroomFilm and LabsOther EquipmentPhoto Technique Nature PhotographyPeople PhotographyTechnique GeneralGeneral Photo Topics General TopicsAustralian PhotographyUK PhotographyDirectoryPhoto Clubs |
Photo Forum / Digital Photography / DSLR Cameras / September 2007Nikon D3 hints at a way "out" for Olympus
Olympus needs a strategy to exit the confines of the 4/3rds sensor without losing too much face. Nikon's D3 is compatible with DX lenses via an aperture stop that uses only part of the sensor (5 megapixels) so people owning DX lenses won't necessarily have to dump them. Olympus needs to go the other way. The Olympus digital lenses produce an image circle just slightly smaller than an 35mm sensor. You can see for yourself if you put a digital lens on a Konica SLR, which shares a similar bayonette to the digital Olympus cameras. It vignettes the outer edge, but not the top and bottom, at least not too much. Olympus needs to go to a larger sensor (hey, why not a 4/3rds in the same physical area as an APS sensor? It make FAR more sense than continuing to support the ancient 3:2 format) to progress against increasingly sophisticated and strong competition. So, implementing an APS sized sensor with a 4/3rd configuration, at 14 megapixels with a built-in mask function (optionally able to be turned off when using current digital lenses, for those who want to live with Canon WA edge aberrations) to allow the use of their current digital lenses would help here. Since the image circle covered by the Olympus digitals just about supports an APS sensor size, there is NO downside. Reason I say this is simple: The DR issue is important, at least as much as noise. So why would someone, barring a heavy investment in Olympus glass, choose the upcoming E-3 over a (for e.g.) Nikon D300, from what we know at this point? Or even the Canon 40D which is likely $400 less? In other words, knowing that no matter how much you spent on glass, you'd never quite reach the image quality of a rival, why would you not BUY the rival? Isn't this in-part what cost Nikon the high end against Canon, having a $5000 camera that couldn't keep up with the competition's $5000 camera? We do not know what sensor the E-3 is going to use, but if it's the same one as in the entry-level E-410 and 510, someone dropping $4000 (the cost of the E-3 plus it's new 12-60mm kit lens and the 50-200mm lens) might think it was time to consider alternate system paths. <snip> > much. Olympus needs to go to a larger sensor (hey, why not a 4/3rds > in the same physical area as an APS sensor? It make FAR more sense [quoted text clipped - 6 lines] > help here. Since the image circle covered by the Olympus digitals just > about supports an APS sensor size, there is NO downside. <snip> The problem with doing that for Olympus though, is the small distance between the back of the lens, and the sensor in the 4/3 system. Increasing sensor size means a larger mirror, and a larger mirror won't fit in the distance they have available. They physically can't go to a bigger sensor system unless they... A) go to a complete new lens mount with greater sensor/flange distance B) dispense with the reflex system, and rely solely on liveview - basically a bigger-sensor, interchangeable lens version of most of today's big-zoom compacts. C) come up with a completely different reflex system, such as having the viewfinder in the film-plane, swinging it away and swinging the sensor into place. Option B is probably the only really viable one, but with screen resolutions still a long way short of viewfinder screens, I can't see it being a very popular one. Option A is the only option that completely addresses other issues like autofocus too, but it would see them having to support 2 different lens mounts. Of course if designed properly, it would be fairly elementary to create an adapter allowing the new lenses to be mounted to a 4/3 body. Canon are in a similar situation with 2 physically different lens mounts in EF and EF-S. The difference is that EF will fit onto an EF-S body, but not vice-versa. I guess this is kind of like the hypothetical olympus situation above. Interesting to note that Pentax/Samsung and Sony/Minolta could go to full-frame using the Nikon system if they wished. Both manufacturers have maintained 100% compatibility between their small-sensor and large-sensor lenses/bodies, with only the image circle changing. > <snip> >> much. Olympus needs to go to a larger sensor (hey, why not a 4/3rds [quoted text clipped - 10 lines] > The problem with doing that for Olympus though, is the small distance between the back of the lens, and the sensor in the 4/3 > system. Increasing sensor size means a larger mirror, Not necessarily, see below. > and a larger mirror won't fit in the distance they have available. They physically can't go to a bigger sensor system unless > they... [quoted text clipped - 3 lines] > C) come up with a completely different reflex system, such as having the viewfinder in the film-plane, swinging it away and > swinging the sensor into place. or D: Put in the larger sensor, but stick with a mirror about the same size as at present and accept they'll have much less of the FoV showing in the viewfinder. Cheers  Signaturecmyk > <snip>> much. Olympus needs to go to a larger sensor (hey, why not a 4/3rds > > in the same physical area as an APS sensor? It make FAR more sense [quoted text clipped - 14 lines] > system unless they... > A) go to a complete new lens mount with greater sensor/flange distance The distance between the lens and sensor in digitals makes the cameras much "thicker" than most of the old film cameras, which is unfortunate in some ways. But, the lens mount doesn't determine the lens-sensor gap, and I think the Olympus lens mount will support a larger sensor format. The bodies will definitely need a re-design. > B) dispense with the reflex system, and rely solely on liveview - basically > a bigger-sensor, interchangeable lens version of most of today's big-zoom > compacts. I doubt they'd do that but you never know. > Interesting to note that Pentax/Samsung and Sony/Minolta could go to > full-frame using the Nikon system if they wished. Both manufacturers have > maintained 100% compatibility between their small-sensor and large-sensor > lenses/bodies, with only the image circle changing. Problem is that Olympus never really got into the AF game in a big way until they switched to digital sensors. Their AF stuff from the 1990s won't work with anything except film cameras of that era. A poor strategy on their part I as far as consumers go, I guess, but then it does sell more lenses now because of it. The old OM lenses (manual) will work with an adapter to the digital bodies. > But, the lens mount doesn't determine the lens-sensor gap, and I think > the Olympus lens mount will support a larger sensor format. The > bodies will definitely need a re-design. Doesn't the lens determine the lens/sensor gap? Isn't the lens closest to the sensor when focused at infinity? (*) So it cannot be moved further from the sensor without giving up focus at infinity, no? (*) That's oversimplified, ignoring modern lens design where groups move independently of each other. But in general, don't they tend to be closest to the sensor when focussed at infinity? -hank >>But, the lens mount doesn't determine the lens-sensor gap, and I think >>the Olympus lens mount will support a larger sensor format. The [quoted text clipped - 7 lines] > move independently of each other. But in general, don't they tend to > be closest to the sensor when focussed at infinity? Moving the lens away from the sensor does cause the loss of infinity focus, at least that much holds true. The distance from sensor to lens is generally equal to the diagonal of the sensor which for 24x36 is about 43mm and that's also the 'normal' lens focal length. APS DSLRs could come closer but they are designed to accomodate the old lenses. It's easier to design fast wide angle lenses if you can get them closer to the sensor so there is some disadvantage to that compromise. I believe the Canon 'digital' lenses do mount closer to the sensor and that's why those lenses won't mount on their full frame cameras where Nikons will... but Nikon doesn't have a 10-22mm lens either. Olympus has an 11-22mm f/2.8-3.5 which is faster than the Canon 10-22mm f/3.5-4.5 which is faster & better than the Sigma 10-20mm f/4-5.6 that fits the Nikon mount. There is a Nikon 10.5mm f/2.8 fisheye but not rectalinear. SignaturePaul Furman Photography http://edgehill.net Bay Natives Nursery http://www.baynatives.com >> A) go to a complete new lens mount with greater sensor/flange distance > > The distance between the lens and sensor in digitals makes the cameras > much "thicker" than most of the old film cameras, which is unfortunate > in some ways. Not quite. The sensor/lens distance has stayed the same for all digitals except Olympus. It is not this that makes digitals thicker than film cameras. It is the stuff that has to go behind the sensor that makes digitals thicker. The sensor & screen, both have supporting circuitry that needs to go behind them and that increases the distance between the sensor and the rear of the camera. A film camera only needed a back door and a pressure plate to push on the quite thin film. > But, the lens mount doesn't determine the lens-sensor gap, and I think > the Olympus lens mount will support a larger sensor format. The > bodies will definitely need a re-design. While it would be possible to make a lens using the existing bayonet but with a greater distance to the sensor, you have to consider that all current lenses are designed to have the sensor plane a certain distance from the flange. If you move the sensor plane further back, you lose infinity focus. This is why you can't have a simple adaptor to allow Canon FD lenses to be used on EOS bodies. The other consideration is not just the distance from the flange to the sensor, but the distance from the rear of the lens to the sensor. It is this for example that stops Canon EF-S lenses mounting to film/full-frame bodies. The rear of the lens protrudes further into the body, so a full-frame mirror hits it. >> B) dispense with the reflex system, and rely solely on liveview - >> basically >> a bigger-sensor, interchangeable lens version of most of today's big-zoom >> compacts. > > I doubt they'd do that but you never know. Considering they were the first to introduce live-view, it wouldn't surprise me. It would be the easiest way out for them. > Olympus needs a strategy to exit the confines of the 4/3rds sensor > without losing too much face. There is no easy way out of the box Olympus is in. However they have been innovative in the past, so perhaps they will be again. If they have to make a change, and I believe they will eventually, what limits them to a particular sensor size or format? I see the Canon/Nikon/Pentax/Sony industry slowly arriving at the full frame sensor for consumer DSLRs. When they get there, what will they do next? Olympus could be the leader of the pack if they would go with a still larger sensor. Perhaps 645 format? There must be a format that will allow a reasonably small body plus lens, with more pixels with less density than currently available. There have been compact full-frame 35mm cameras for years, and it is no longer necessary to provide extra space for two spools of film! Fred >There is no easy way out of the box Olympus is in. However they have >been innovative in the past, so perhaps they will be again. If they >have to make a change, and I believe they will eventually, what limits >them to a particular sensor size or format? I think the fact that Olympus has invested heavily in Four Thirds, but hasn't achieved anything like the rate of return hoped for, means that Olympus are not in a any position to invest in another format. I have been a little starled by some of the ideas that have been suggested about larger sensor sizes. For the avoidance of doubt, all the Olympus Four Thirds lenses have an image cicle that is just big enough for the Four Thirds sensors, and no more. There is no question of using any of the Olympus Four Thirds lenses with a larger sensor. > >There is no easy way out of the box Olympus is in. However they have > >been innovative in the past, so perhaps they will be again. If they [quoted text clipped - 10 lines] > enough for the Four Thirds sensors, and no more. There is no question > of using any of the Olympus Four Thirds lenses with a larger sensor. This depends entirely on what you mean by "just big enough." Olympus users have become used to getting near perfect corner performance because unlike the APS and FF cameras, their lenses more than cover the 4/3rds format. While Canon FF users put up with terrible edge aberrations in their FFs, and Olympus user would be shocked by it. This would seem to suggest that the Olympus digital lenses cover a large frame area relative to their design than do the DX and the FF film lenses now used on Nikons and Canons. But, if anyone is curious, I can take a shot through a film camera with an Olympus digital lens attached and show you the coverage. >> >There is no easy way out of the box Olympus is in. However they have >> >been innovative in the past, so perhaps they will be again. If they [quoted text clipped - 22 lines] > shot through a film camera with an Olympus digital lens attached and > show you the coverage. The question I have is, Is the poor corner performance of Canon FF caused because the image circle is _just_ big enough, or is it because of the angle the light hits the digital sensor? Bear in mind, light fall off isn't as bad on 35mm film as it is in FF digital. > >> >There is no easy way out of the box Olympus is in. However they have > >> >been innovative in the past, so perhaps they will be again. If they [quoted text clipped - 28 lines] > Bear in mind, light fall off isn't as bad on 35mm film as it is in FF > digital. Well, Canon until recently made ZERO effort to address the differences between lenses and how they work with film and sensors, so it could very well be the angle. The fact some Canon users seem to be getting reasonable performance from higher-end old film lenses with different designs seems to indicate this could be the case. Until Canon gets off their a-- and makes reasonably telecentric wide angle lenses for FF, then you'd have to examine the designs of other lenses, try them on the FF Canon or Nikon and see how they perform. I'm not sure how angle of incidence effects vignetting with a sensor, but some of the Canon WAs have terrible vignetting issues. >>>>> There is no easy way out of the box Olympus is in. However they have >>>>> been innovative in the past, so perhaps they will be again. If they [quoted text clipped - 35 lines] > angle of incidence effects vignetting with a sensor, but some of the > Canon WAs have terrible vignetting issues. They offset microlenses to compensate to some degree, but biased to a particular lens. There's some hint that Nikon's D3 (and D300?) take a different approach, using a gapless dual layer of microlenses, with the second layer refocusing light from the edges back down to the sensor diode. Don't ask me for a link - I read and saw some diagrams from Nikon, but didn't save the URL. Too early to say if it works, but the samples from D3 look very good. >Well, Canon until recently made ZERO effort to address the differences >between lenses and how they work with film and sensors, so it could [quoted text clipped - 4 lines] >FF, then you'd have to examine the designs of other lenses, try them >on the FF Canon or Nikon and see how they perform. You will find that most of Canon's tele lenses are just as telecentric as any lens in the entire Olympus line-up. >I'm not sure how >angle of incidence effects vignetting with a sensor, In which case the advice that it is better to shut up and be thought an idiot than to speak up and remove all doubt seems most appropriate! > but some of the >Canon WAs have terrible vignetting issues. Only at wide apertures - and that is clearly a lens issue, not a sensor issue. Put the camera on a tripod, shine a flashlight in the eyepiece and look at how that circular full aperture appears lie a cat's eye when you stand near the edge or corner of the field - THAT is why the lens vignettes, and THAT is why the issue disappears when the lens stops down. SignatureKennedy Yes, Socrates himself is particularly missed; A lovely little thinker, but a bugger when he's pissed. Python Philosophers (replace 'nospam' with 'kennedym' when replying) > In article <1190680850.457340.41...@r29g2000hsg.googlegroups.com>, RichA > <rander3...@gmail.com> writes [quoted text clipped - 10 lines] > You will find that most of Canon's tele lenses are just as telecentric > as any lens in the entire Olympus line-up. And yet they still have rotten edge definition. Compare Canon, either a 1.5 or full frame with one of their WA zooms to an Olympus using their 11-22mm. The difference at the edge is pretty shocking. > >I'm not sure how > >angle of incidence effects vignetting with a sensor, > > In which case the advice that it is better to shut up and be thought an > idiot than to speak up and remove all doubt seems most appropriate! The reason I qualified it was that I had heard Canon was working on new microlens arrays for it's sensors. Using them in a certain way can alleviate some of the problems with light fall off of sensors due to photons not hitting the bottom of the "well." It's possible by doing that they can keep using their poor WA lenses. > > but some of the > >Canon WAs have terrible vignetting issues. [quoted text clipped - 5 lines] > vignettes, and THAT is why the issue disappears when the lens stops > down. In one sense, that is actually good. It means there is little use in spending four figures for an f2.8 zoom when an f4 will function similarly, since you MUST stop down the f2.8 to get a decent image. Even Canon's 17-40L f4 lens stopped down to f5.6 showed marked edge distortion. >> You will find that most of Canon's tele lenses are just as telecentric >> as any lens in the entire Olympus line-up. > >And yet they still have rotten edge definition. Compare Canon, either >a 1.5 or full frame with one of their WA zooms to an Olympus using >their 11-22mm. The difference at the edge is pretty shocking. Which proves, does it not, that the "telecentric argument" is complete hogwash! >> >I'm not sure how >> >angle of incidence effects vignetting with a sensor, [quoted text clipped - 7 lines] >to photons not hitting the bottom of the "well." It's possible by >doing that they can keep using their poor WA lenses. Why would Canon need to be working on a new means of reducing light fall-off when measurements of light fall off on the two year old 5D show it is better than film already? Hint: higher fill factor is not the same thing, and that is what Canon have brought to market. >>Put the camera on a tripod, shine a flashlight in the eyepiece >> and look at how that circular full aperture appears lie a cat's eye when [quoted text clipped - 5 lines] >spending four figures for an f2.8 zoom when an f4 will function >similarly, Cobblers - that is NOT what it means at all! >since you MUST stop down the f2.8 to get a decent image. And you will have to stop down an f/4 lens even further! This is a lens design issue and if a lens designer considers a cost/performance compromise acceptable in an f/2.8 design then he is likely to consider, at the very least, that the same cost/performance compromise is acceptable in the lower priced f/4 equivalent. Life has always been this way. You can't afford a lens without compromise. The only question is which compromises you can afford to live with. >Even Canon's 17-40L f4 lens stopped down to f5.6 showed marked edge >distortion. Edge distortion is something completely different from what is being discussed. SignatureKennedy Yes, Socrates himself is particularly missed; A lovely little thinker, but a bugger when he's pissed. Python Philosophers (replace 'nospam' with 'kennedym' when replying) >>> >There is no easy way out of the box Olympus is in. However they have >>> >been innovative in the past, so perhaps they will be again. If they [quoted text clipped - 22 lines] >> shot through a film camera with an Olympus digital lens attached and >> show you the coverage. Absolute nonsense, Rich. You cannot fit a Four Thirds lens to a film camera with the correct lens flange to film plane distance. All 35mm film cameras/lenses have a greater lens flange to film plane distance than Four Thirds, which will of course make the lens appear to offer greater coverage than it really has. So it would prove nothing. >The question I have is, Is the poor corner performance of Canon FF caused >because the image circle is _just_ big enough, or is it because of the angle >the light hits the digital sensor? >Bear in mind, light fall off isn't as bad on 35mm film as it is in FF >digital. The angle of incidence is to blame. The light fall-off is worse with some lenses than others. > >>> >There is no easy way out of the box Olympus is in. However they have > >>> >been innovative in the past, so perhaps they will be again. If they [quoted text clipped - 28 lines] > than Four Thirds, which will of course make the lens appear to offer > greater coverage than it really has. So it would prove nothing. We aren't talking about a huge difference in lens-flange distances. Besides, this is just a side-issue, a curiosity. My main point is that Olympus does not need a new bayonette mount to create a larger sensored camera, which would at least put it on an equal footing with Nikon, Canon, Pentax and Sony. >> >>> >There is no easy way out of the box Olympus is in. However they >> >>> >have [quoted text clipped - 39 lines] > sensored camera, which would at least put it on an equal footing with > Nikon, Canon, Pentax and Sony. I guess the shape of the bayonet doesn't need to change, but other specs about the lens mount do need to change. For example the position, size and location of the electronic pickups, the position of the rear of the lens etc. These all interfere with it's ability to be used on larger sensors. Yes it would indeed be possible for them to make a camera with a 35mm or APS sized sensor and use the same bayonet. But it would be necessary to relocate these other items which would stop the lens from working. It would be better for them to create a new mount, and then provide an adaptor back to 4/3 for the owners of 4/3 cameras, and use 4/3 as an in-between system. Perhaps 4/3 could be likened to the 110 format SLR film cameras - good enough quality for many, smaller than 35mm, but always playing 2nd fiddle to the larger formats for the people who demand quality. There is no reason whatsoever, that the things they have learnt in doing 4/3 can't be adapted to larger formats - eg telecentric lenses, so in that sense the technology wouldn't be wasted. But, I can't see any of these scenarios happening. Olympus have committed to the 4/3 system, and so as all the other manufacturers move to larger formats, olympus will stay megapixel/noise limited. Probably their best course of action is their current one - cameras like the E410 that take advantage of the smaller size, so they become a bridge between P&S and larger SLR. What I would like to see however is the 4/3 sensors being used in compact cameras. While we complain about their noise, lack of DR etc, they are still a lot better than the ultra-mini sensors in P&S. 4/3 is a good in-between size that delivers good results, but with smaller lenses than APS/35mm. > >> >>> >There is no easy way out of the box Olympus is in. However they > >> >>> >have [quoted text clipped - 51 lines] > adaptor back to 4/3 for the owners of 4/3 cameras, and use 4/3 as an > in-between system. Just as my original point, the Nikon D3 does just that. All they need to do is keep people from ending up with useless lenses. > But, I can't see any of these scenarios happening. Olympus have committed to > the 4/3 system, and so as all the other manufacturers move to larger > formats, olympus will stay megapixel/noise limited. Probably their best > course of action is their current one - cameras like the E410 that take > advantage of the smaller size, so they become a bridge between P&S and > larger SLR. So far, the only on announced (will it ever be released?) is the Sigma Foveon. Olympus could offer a pocketable P&S with a fixed focal length pancake or very short zoom. > What I would like to see however is the 4/3 sensors being used in compact > cameras. While we complain about their noise, lack of DR etc, they are still > a lot better than the ultra-mini sensors in P&S. 4/3 is a good in-between > size that delivers good results, but with smaller lenses than APS/35mm. They had the 2/3rds sensor, they didn't even keep that. P&S is aimed straight at the DSLR-ignorant, those who have some need for 400mm equivalent zooms and lots of noise. I think what happened was that they got sick of the higher-end P&S's being money-losers (can't charge $900 for one now like some used to cost) and they decided to cost-cut and the first place was the sensor and the lenses. >> >> >>> >There is no easy way out of the box Olympus is in. However they >> >> >>> >have [quoted text clipped - 63 lines] > Just as my original point, the Nikon D3 does just that. All they need > to do is keep people from ending up with useless lenses. Nope not the same. Because Nikon, Pentax, Sony/Minolta, and to a lesser extent Canon, started with a 35mm system, and kept the dimensions of that, except with an APS sensor, the APS lenses use the same flange-sensor distance, same electronic linkages etc as the 35mm system. Therefore you can put an APS lens on a 35mm camera, and it will work as per normal except with the smaller image circle. On the D3, Nikon simply crop away the black bits. I actually use my 18-55 Pentax lens with a 35mm body - vignetting is terrible at 18mm, but by 24mm is completely gone. If Olympus created an APS or 35mm camera, they wouldn't be able to use the existing lens mount because of flange-sensor distance, position of electronic linkages etc. If done right, they would be able to use an adaptor to put the larger-format lenses onto a 4/3 body, but wouldn't be able to go back the other way. A real-world example of this type of scenario would be the Pentax SLR family. You could mount 67/645 lenses on a 35mm body with an adapter but you couldn't mount 35mm lenses on 67/645 bodies. Likewise you could mount 35mm lenses on a 110 body with an adaptor, but you couldn't mount 110 lenses on 35mm. Not being able to mount smaller-format lenses on larger-format bodies had nothing to do with image circle, but had more to do with flange-film distances, size of the opening, etc. \> [] > They had the 2/3rds sensor, they didn't even keep that. P&S is aimed > straight at the DSLR-ignorant, those who have some need for 400mm [quoted text clipped - 3 lines] > cost) and they decided to cost-cut and the first place was the sensor > and the lenses. P&S has its place alongside the DSLR. Most people don't want or need the size weight and complexity of a DSLR, it's not a case that they are ignorant, simply that they want a lightweight and compact camera for their photos. Even some DSLR owners take along a compact camera for just those reasons. It would certainly be interesting to see what could be done with the 4/3 sensor in a compact camera. The Sony "APS-C" camera failed perhaps because of its bulk - would a 4/3 camera fare any better? A "low-light" or even "party" camera? I suspect not. Cheers, David On Sep 26, 4:04 am, "David J Taylor" <david-tay...@blueyonder.not-this- bit.nor-this-bit.co.uk> wrote: > [] > [quoted text clipped - 11 lines] > photos. Even some DSLR owners take along a compact camera for just those > reasons. Yes, but I wouldn't describe long zoom P&S cameras as compact. > It would certainly be interesting to see what could be done with the 4/3 > sensor in a compact camera. The Sony "APS-C" camera failed perhaps > because of its bulk - would a 4/3 camera fare any better? A "low-light" > or even "party" camera? > > I suspect not. No, because the market has conditioned the buyers to look for zoom range and megapixels. A 6 megapixel 4/3 size sensor in a compact camera would be nice. > On Sep 26, 4:04 am, "David J Taylor" [] >> P&S has its place alongside the DSLR. Most people don't want or >> need the size weight and complexity of a DSLR, it's not a case that [quoted text clipped - 3 lines] > > Yes, but I wouldn't describe long zoom P&S cameras as compact. Well, I would certainly describe the Panasonic TZ3 as compact: http://www.dpreview.com/reviews/panasonictz3/ Lens: 28-280mm (35mm equiv) Dimensions: 105 x 59.2 x 36.7 mm (4.20 x 2.37 x 1.47 in) Weight (excl batt): 232 g (0.51lb) Cheers, David In article <46f83f55$0$4550$5a62ac22@per-qv1-newsreader-01.iinet.net.au>, Doug Jewell <noone@nowhere.com> writes >Bear in mind, light fall off isn't as bad on 35mm film as it is in FF >digital. Bear in mind that it does! I initially measured the SENSOR light fall off almost two years ago and reported the results in this forum at the time - slightly less fall-off than film with the full frame Canon 5D. Since then, a couple of others have confirmed similar results with the same camera. Light fall-off with full frame sensors is just one of the many misconceptions that pushed Olympus into the corner they now find themselves in. SignatureKennedy Yes, Socrates himself is particularly missed; A lovely little thinker, but a bugger when he's pissed. Python Philosophers (replace 'nospam' with 'kennedym' when replying) > Doug Jewell writes > [quoted text clipped - 6 lines] > Since then, a couple of others have confirmed similar results with the > same camera. The issue is sharpness and/or chromatic aberration, yes? Largely corrected by microlenses but caused by the angle of light hitting pixels with depth versus flat film for wide angle lenses. If that's the case, Canon's closer sensor to lens design mount is a bit more problematic but a Nikon lens on a Canon would retain it's telecentricity. I'm not saying it's a huge problem but this is what we are discussing. > Light fall-off with full frame sensors is just one of the > many misconceptions that pushed Olympus into the corner they now find > themselves in. Olympus should actually have more of an issue with this since their mount is even closer and the lenses need to be wider... although the sensor is smaller so that decreases angle right there. SignaturePaul Furman Photography http://edgehill.net Bay Natives Nursery http://www.baynatives.com >> Doug Jewell writes >> [quoted text clipped - 8 lines] > >The issue is sharpness and/or chromatic aberration, yes? No - the issue is the statement quoted in the top line above: light fall-off. >Largely corrected by microlenses but caused by the angle of light >hitting pixels with depth versus flat film for wide angle lenses. Silicon sensors have LESS absorption depth than film! The Olympus misconception was that the microlenses, added to increase the pixel fill factor, caused reduced absorption with angle of incidence. At least in some cases that has certainly been proven to be false. >If that's the case, Canon's closer sensor to lens design mount is a bit >more problematic but a Nikon lens on a Canon would retain it's >telecentricity. I'm not saying it's a huge problem but this is what we >are discussing. It isn't a problem at all within the constraints of the Canon mount on a full frame sensor. Quite the opposite actually. >> Light fall-off with full frame sensors is just one of the many >>misconceptions that pushed Olympus into the corner they now find [quoted text clipped - 3 lines] >mount is even closer and the lenses need to be wider... although the >sensor is smaller so that decreases angle right there. Precisely, but that is completely the opposite of what Olympus were peddling - they used telecentric lenses because of light fall off concerns, not because they chose an excessively short backworking distance. SignatureKennedy Yes, Socrates himself is particularly missed; A lovely little thinker, but a bugger when he's pissed. Python Philosophers (replace 'nospam' with 'kennedym' when replying) >hey, why not a 4/3rds >in the same physical area as an APS sensor? Because that is physically impossible - like fitting a quart into a pint pot! > It make FAR more sense >than continuing to support the ancient 3:2 format) to progress against >increasingly sophisticated and strong competition. So, implementing >an APS sized sensor with a 4/3rd configuration, at 14 megapixels 4/3 format is NOT the aspect ratio of the sensor! The 4/3 consortium DEFINES a 22mm image diagonal for the image. The consortium name is derived from that fundamental definition. It comes from the usual sensor naming conventions of 1/3", 2/3" etc. 1/3" sensors have image diagonals of 5.5mm, 2/3" sensors have image diagonals of 11mm and 4/3" sensors have image diagonals of 22mm. You can, and the original 4/3 consortium announcement made this clear, have 4/3 sensors with 4:3, 3:2 and even 1:1 aspect ratios, even circular formats, and ALL 4/3 system lenses must be compatible with all possible formats - ie. they must produce an image which can be inscribed by a 22mm diagonal circle. Please explain how you intend to fit a sensor of approximately 22.5mm x 15mm into a 22mm diagonal. SignatureKennedy Yes, Socrates himself is particularly missed; A lovely little thinker, but a bugger when he's pissed. Python Philosophers (replace 'nospam' with 'kennedym' when replying) > In article <1190602810.351847.299...@19g2000hsx.googlegroups.com>, RichA > <rander3...@gmail.com> writes [quoted text clipped - 4 lines] > Because that is physically impossible - like fitting a quart into a pint > pot! Bad wording. I meant a 1.33:1 ratio sensor the size (increase the height) of an APS sensor. Just to get away from the 3:2 ratio. > > It make FAR more sense > >than continuing to support the ancient 3:2 format) to progress against > >increasingly sophisticated and strong competition. So, implementing > >an APS sized sensor with a 4/3rd configuration, at 14 megapixels > > 4/3 format is NOT the aspect ratio of the sensor! The images they produce at the moment are 1.33:1 > The 4/3 consortium DEFINES a 22mm image diagonal for the image. The > consortium name is derived from that fundamental definition. It comes [quoted text clipped - 6 lines] > they must produce an image which can be inscribed by a 22mm diagonal > circle. 22mm diagonal "circle?" Is that like, the other "side" of a pillar? I guess you meant diameter? >> In article <1190602810.351847.299...@19g2000hsx.googlegroups.com>, RichA >> <rander3...@gmail.com> writes [quoted text clipped - 16 lines] > >The images they produce at the moment are 1.33:1 4:3 is the aspect ratio. 4/3 is the format. >> The 4/3 consortium DEFINES a 22mm image diagonal for the image. The >> consortium name is derived from that fundamental definition. It comes [quoted text clipped - 9 lines] >22mm diagonal "circle?" Is that like, the other "side" of a pillar? >I guess you meant diameter? No actually, I meant a 22mm diagonal FRAME. SignatureKennedy Yes, Socrates himself is particularly missed; A lovely little thinker, but a bugger when he's pissed. Python Philosophers (replace 'nospam' with 'kennedym' when replying) |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2008 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.