 |
 | 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 2005Is it freaking Canon Day or something?
A lot of product roll outs, for sure. But, 5D samples! http://web.canon.jp/Imaging/eos5d/eos5d_sample-e.html  Signaturehttp://www.pbase.com/bcbaird Well, it seems Canon has fully punched out their camera line now. But I wonder how people will choose between the 5D and the big N, seeing as how the cost almost the same? > Well, it seems Canon has fully punched out their > camera line now. But I wonder how people will > choose between the 5D and the big N, seeing as > how the cost almost the same? Depends on their needs. The 1D Mark IIn is going to replace the 1D Mark II, so if people need a pro-level AF, weather sealing and a fast frame rate they'll go for that. The 5D brings full frame for under $4,000 - that will appeal to a lot of people who need true wide angle. Signaturehttp://www.pbase.com/bcbaird Hi Brian, > The 5D brings full frame for under $4,000 - > that will appeal to a lot of people who need true wide angle. Download the Landscape example from http://web.canon.jp/Imaging/eos5d/eos5d_sample-e.html and look at the edges. They are definitively unsharp to a degree which makes them unusable. My D100 and the Nikkor 12-24 beats this easily, and I guess also a D20 and the 10-22 is better. Therefore: If you need true wide angle (and quality), then you need new lenses (and a 1.5 or 1.6 crop camera). If you just need wide angle and do not mind mediocre quality, you may take the 5D or experiment with a WA-converter and a compact camera. Regards Benedikt > > The 5D brings full frame for under $4,000 - > > that will appeal to a lot of people who need true wide angle. [quoted text clipped - 3 lines] > and look at the edges. They are definitively unsharp to a degree which > makes them unusable. I agree that's a crappy picture. But my issue isn't really with the corners, it's more with the overall quality of the picture. Not great. But there are other lenses that could cover this focal length better. > My D100 and the Nikkor 12-24 beats this easily, and I guess also a D20 > and the 10-22 is better. The results from either those lenses never really dazzled me. You still have vignetting, you still have soft corners and distortion. I just think this sample photo was done poorly. > Therefore: If you need true wide angle (and quality), then you need new > lenses (and a 1.5 or 1.6 crop camera). If you just need wide angle and [quoted text clipped - 4 lines] > > Benedikt Let's not swear off an entire camera/lens combination because of ONE bad photo. I've seen other FF samples with that lens that did not exhibit nearly the same level of crumminess. Signaturehttp://www.pbase.com/bcbaird I commented on that picture's issues on another group. There is more wrong than just edge definition. Note the two dark vertical "tracks" left edge in the grassy area. Seems kinds of counterproductive, bringing out another full frame, whose major attribute is the preservation of the wide angle ability of wide angle lenses only to find the len(s) themselves produce poor results. >I commented on that picture's issues on another group. There is more >wrong than [quoted text clipped - 3 lines] >wide angle ability of wide angle lenses only to find the len(s) >themselves produce poor results. I think what you are seeing is a track in the grass itself, caused perhaps by someone walking through it; it does not seem to be continuous, or exactly parallel to the edge. Granted, the corners, and to a lesser extent the edges, go a little mushy. which would be expected at f/4 but is disappointing at f/8. David SignatureDavid Littlewood > The 5D brings full frame for under $4,000 - that will appeal to a lot > of people who need true wide angle. Generally speaking, Canon wide-angle lenses suck. Bad. There are folk who are using Contax/Zeiss/etc optics on the 1Ds2 with adaptors because of this issue. People considering a 5D should keep this in mind if the _only_ goal is in fact "wide angle"... > Generally speaking, Canon wide-angle lenses suck. Bad. There are folk I wouldn't go that far. They have some poor performers, yes. But to say they suck outright... > who are using Contax/Zeiss/etc optics on the 1Ds2 with adaptors because > of this issue. People considering a 5D should keep this in mind if the > _only_ goal is in fact "wide angle"... Well, often the case isn't just getting superwides to give you their intended FOV, sometimes it's just getting your 24-70 f/2.8L to behave like it should. Signaturehttp://www.pbase.com/bcbaird > The 5D brings full frame for under $4,000 - that will appeal to a > lot of people who need true wide angle. Canon has got very good telephoto lenses, but their wide angle offerings usually suck. Take a look at the "landscape" picture (sample image 3) that Canon posted on http://web.canon.jp/Imaging/eos5d/eos5d_sample-e.html . This is the EF 17-40mm f/4 L USM (from what I've read is regarded as one of Canon's better wide angle zooms) at what is probably its best aperture (f/8) - but the edges are so soft that this lens clearly obviously has no business on a full frame digital camera - it should really be designated EF-S. It is really so bad that I wondered why on earth Canon put it on the web. After seeing the leaked specs, I've seriously considered breaking the piggy bank and plunging for a EOS 5D with the EF 17-40 mm L - but that single image put me immedately off the idea. Nikon have some great wide angle lenses. This Canon sample just increased my cravings for a full frame Nikon. Signature- gisle hannemyr [ gisle{at}hannemyr.no - http://folk.uio.no/gisle/ ] ------------------------------------------------------------------------ Kodak DCS460, Canon Powershot G5, Olympus 2020Z ------------------------------------------------------------------------ > > The 5D brings full frame for under $4,000 - that will appeal to a > > lot of people who need true wide angle. > > Canon has got very good telephoto lenses, but their wide angle > offerings usually suck. They aren't the best, but the L stuff isn't the worst. From what I've seen and read the 14mm f/2.8L doesn't perform as well as it should for the amount of money it costs. > Take a look at the "landscape" picture (sample image 3) that Canon > posted on http://web.canon.jp/Imaging/eos5d/eos5d_sample-e.html . [quoted text clipped - 8 lines] > with the EF 17-40 mm L - but that single image put me immedately off > the idea. I've seen much better pictures from the 17-40 f/4 L on FF cameras so I'm going to say either the camera or the lens used in that picture was a dud. Or maybe something screwed up in post processing. In any case, if the final production model 5D produces output that crappy no one will buy it. What was Canon's marketing department thinking? "Hey, people will REALLY want to use this camera for wide angle work... let's show them this crappy photo... it's got field of view!" > Nikon have some great wide angle lenses. This Canon sample just > increased my cravings for a full frame Nikon. Nikon does indeed have some fine wide angle and super wide angle lenses... which has me shaking my head every time they go on and on about their new DX format. Signaturehttp://www.pbase.com/bcbaird > I've seen much better pictures from the 17-40 f/4 L on FF cameras On film.. > so I'm > going to say either the camera or the lens used in that picture was a > dud. So what's the chances YOU will end up with one equally as bad if this is the case? > Or maybe something screwed up in post processing. In any case, if > the final production model 5D produces output that crappy no one will > buy it. Have you actually looked at the full size images from a 1DsmkII using canon's wide lenses? They look just like this sample from the ones I've seen. > What was Canon's marketing department thinking? > > "Hey, people will REALLY want to use this camera for wide angle work... > let's show them this crappy photo... it's got field of view!" Ever consider this is the reality of Canon's wide lenses on a FF sensor? SignatureStacey > > I've seen much better pictures from the 17-40 f/4 L on FF cameras > > On film.. No, on digital. Stop trying to tell me what I've seen and haven't seen. Go check the samples from the 1Ds and 1Ds Mark II on dPreview.com. Plenty of samples with the 17-40 f/4L that aren't nearly as bad as the Canon sample. > > so I'm > > going to say either the camera or the lens used in that picture was a > > dud. > > So what's the chances YOU will end up with one equally as bad if this is the > case? I don't plan on buying either in the future, so I'll say the chances are zero. > > Or maybe something screwed up in post processing. In any case, if > > the final production model 5D produces output that crappy no one will [quoted text clipped - 3 lines] > canon's wide lenses? They look just like this sample from the ones I've > seen. I have, an no, they're not as bad as this shot. Sure, the corners are soft and there is considerable distortion present in the image - but the overall image quality is a hell of a lot better. There's a cathedral picture in the 1Ds Mark II preview gallery at dpreview.com shot wide open with the 17-40 f/4L. Look at that and try telling me it looks 'just like' the 5D sample shot. > > What was Canon's marketing department thinking? > > > > "Hey, people will REALLY want to use this camera for wide angle work... > > let's show them this crappy photo... it's got field of view!" > > Ever consider this is the reality of Canon's wide lenses on a FF sensor? It's the reality of superwide angle lenses and extreme FOVs, period. I'd like to see a comparison with some Nikon superwides mounted on the camera though - entirely possible with an adapter. Signaturehttp://www.pbase.com/bcbaird >> So what's the chances YOU will end up with one equally as bad if this is >> the case? > > I don't plan on buying either in the future, so I'll say the chances are > zero. So why would you even care? If the wide lens performance was really good, I'd be interested as this could replace my mamiya 645. I'm not seeing it from ANY of the samples I've seen from canon's wide lenses and their full frame sensors. WIth the tele lenses they look GREAT, the bitch is most of my work is done with wide lenses. >> > Or maybe something screwed up in post processing. In any case, if >> > the final production model 5D produces output that crappy no one will [quoted text clipped - 7 lines] > soft and there is considerable distortion present in the image - but the > overall image quality is a hell of a lot better. My point is and was, the corners are ALWAYS soft using these wide canon lenses on these full frame cameras. > There's a cathedral picture in the 1Ds Mark II preview gallery at > dpreview.com shot wide open with the 17-40 f/4L. Look at that and try > telling me it looks 'just like' the 5D sample shot. That's because they chose a subject with ZERO details in the corners! Downsampled to this size, you can't tell how sharp oo soft it really is. >> Ever consider this is the reality of Canon's wide lenses on a FF sensor? > > It's the reality of superwide angle lenses and extreme FOVs, period. Funny my fuji 6X9 doesn't do this sort of thing, neither does the 90mm SA on my 4X5. SignatureStacey > >> Ever consider this is the reality of Canon's wide lenses on a FF sensor? > > > > It's the reality of superwide angle lenses and extreme FOVs, period. > > Funny my fuji 6X9 doesn't do this sort of thing It wouldn't - it's a camera body. I wonder what lens you'd use on it? > neither does the 90mm SA on > my 4X5. Yeah, it delivers 20 degrees less FOV, too! (It would be closer to a 24mm lens on a 35mm camera, BTW) Signaturehttp://www.pbase.com/bcbaird >> >> Ever consider this is the reality of Canon's wide lenses on a FF >> >> sensor? [quoted text clipped - 4 lines] > > It wouldn't - it's a camera body. I wonder what lens you'd use on it? The fuji lens that's mounted to it? I've used a 47mm on 6X9 and it wasn't "Mush" in the corners either so it isn't WA perspective that's doing this. SignatureStacey >> >> Ever consider this is the reality of Canon's wide lenses on a FF sensor? >> > [quoted text clipped - 3 lines] > >It wouldn't - it's a camera body. I wonder what lens you'd use on it? I think it is a fixed lens rangefinder camera. >> neither does the 90mm SA on >> my 4X5. > >Yeah, it delivers 20 degrees less FOV, too! > >(It would be closer to a 24mm lens on a 35mm camera, BTW) SignatureDavid Littlewood > >It wouldn't - it's a camera body. I wonder what lens you'd use on it? > > I think it is a fixed lens rangefinder camera. Shows what I know! Signaturehttp://www.pbase.com/bcbaird >> >It wouldn't - it's a camera body. I wonder what lens you'd use on it? >> >> I think it is a fixed lens rangefinder camera. > > Shows what I know! Exactly, yet you continue to post your "Opinions".. SignatureStacey > > Shows what I know! > > Exactly, yet you continue to post your "Opinions".. Yeah, but amazingly I still come out looking better than you. Funny, that. I'll confess my ignorance fully when it comes to medium and large format gear. I don't touch it, use it or really want to know that much about it. Signaturehttp://www.pbase.com/bcbaird >> > Shows what I know! >> [quoted text clipped - 6 lines] > gear. I don't touch it, use it or really want to know that much about > it. Yes ignorance is bliss isn't it? SignatureStacey > > Yeah, but amazingly I still come out looking better than you. Funny, > > that. [quoted text clipped - 4 lines] > > Yes ignorance is bliss isn't it? Can't really say. I'd like to know more about medium and large format, but I'd rather stick to the stuff I'm working with now. Signaturehttp://www.pbase.com/bcbaird >> > Yeah, but amazingly I still come out looking better than you. Funny, >> > that. [quoted text clipped - 10 lines] > I'd like to know more about medium and large format, but I'd rather > stick to the stuff I'm working with now. So 2 hours ago you claim "I don't touch it, use it or really want to know that much about it." and now you say "I'd like to know more about medium and large format..."? LOL! SignatureStacey > > I'd like to know more about medium and large format, but I'd rather > > stick to the stuff I'm working with now. > > So 2 hours ago you claim "I don't touch it, use it or really want to know > that much about it." and now you say "I'd like to know more about medium > and large format..."? LOL! Oh Stacey, you'll find flaws in any remark that isn't your own... Signaturehttp://www.pbase.com/bcbaird >Ever consider this is the reality of Canon's wide lenses on a FF sensor? That's pretty much what it is, although the Sigma and Tamron lenses might be the same way. Everyone I've spoken to who uses 1.3x or FF Canons with super-wide lenses uses CA correction in PP or RAW conversion. Ideally, RAW converters should have, at least as an option, a way to handle RAW data as 3 meshes that are independently scalable; geometric corrections in a bitplane at the pixel resolution of the sensor are not ideal. One current option is to render at 200%, and adjust CA from there. Signature<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<> John P Sheehy <JPS@no.komm> ><<> <>>< <>>< ><<> <>>< ><<> ><<> <>>< SNIP > Ideally, RAW converters should have, at least as an option, > a way to handle RAW data as 3 meshes that are independently > scalable; Indeed, although there are post-processing alternatives available, and some Raw converters already (attempt to) correct lateral CA (or similar effects). Pixmantec's RSE 2005 does attempt to reduce it, as does Bibble and DxO, but some (optional) user control is preferable (to me anyway). I'm pleased that PTlens now also has a CA correction (http://www.epaperpress.com/ptlens/index.html), although the interface could be improved. > geometric corrections in a bitplane at the pixel resolution of the > sensor are not ideal. One current option is to render at 200%, > and adjust CA from there. Yes, although subsequent down-sampling could lose some of the improvement and what's worse introduce artifacts itself. A better option is a properly implemented resampling (Lanczos or Sinc) method with large support (and proper edge handling). Bart >> I've seen much better pictures from the 17-40 f/4 L on FF cameras > [quoted text clipped - 21 lines] > >Ever consider this is the reality of Canon's wide lenses on a FF sensor? It amply demonstrates the need for (near-)telecentric lens designs. <grins, ducks and runs> ;-) >>> "Hey, people will REALLY want to use this camera for wide angle work... >>> let's show them this crappy photo... it's got field of view!" [quoted text clipped - 4 lines] > > <grins, ducks and runs> Yet you'll never get these guys to admit this.. The people actually USING the 1Ds's have been bitching about this from when they were introduced. SignatureStacey >Yet you'll never get these guys to admit this.. The people actually USING >the 1Ds's have been bitching about this from when they were introduced. True. But it is only a problem with ultra wide angle lenses. I regularly rent a 1Ds Mk II for use with a 28mm Super-Angulon shift lens, and the CA is barely discernible. It is there, but it is certainly not severe. >>Yet you'll never get these guys to admit this.. The people actually USING >>the 1Ds's have been bitching about this from when they were introduced. [quoted text clipped - 4 lines] > lens, and the CA is barely discernible. It is there, but it is > certainly not severe. I don't really consider 28mm very wide. And isn't the main "point" of FF being able to use their old FF wides again? If 28mm FOV was all they need, this 17-40 on a 1.6 crop provides that and looking at this image, a crop from that part of this lens would be pretty good. SignatureStacey >I don't really consider 28mm very wide. Neither do I. That was actually my point. ;-) Tony Polson wrote: > >> I've seen much better pictures from the 17-40 f/4 L on FF cameras > > [quoted text clipped - 27 lines] > > ;-) Lack of telecentricity has nothing to do with the appearance of lateral chromatic aberration in wide angle lenses. What most people don't appreciate is that many normal lenses have exit pupil distances just as short as the least telecentric SLR wide angles. For example, the 55mm f/2.8 Nikkor and 20mm f/2.8 Nikkor have nearly identical exit pupil distances. That means they have the same lack of telecentricity. Despite this, the 55mm has no lateral color and the 20mm has a ton, thereby proving that telecentricity has no bearing on the appearance of lateral color. More evidence comes from LCD and DMD wide-angle projection lenses, which are all perfectly telecentric. Nevertheless, most of these lenses exhibit lateral color. Lateral color is by far the most difficult aberration to correct in reverse-telephoto lenses, and it requires skillful use of abnormal partial dispersion glasses. As far as I know the only manufacturer to fully address this problem is Zeiss, and even they only address it with a single lens: the 21mm/2.8 Distagon. Some of the recent wide angle zooms by Nikon and others are also very good with respect to lateral color, but typically not over the full zoom range. Brian >Tony Polson wrote: >> > [quoted text clipped - 30 lines] > >Brian I suspect we may be in danger of getting at cross-purposes here. Brian, you are describing the lens aberration. I believe the previous discussion (though this was not explicitly stated) referred to colour fringing effects caused by light interacting with digital sensors, microlenses, AA screens etc. Many people seem to have started calling this chromatic aberration, which shows they are not very technically aware, but since the phenomenon is similar in appearance this is not surprising. BTW, interesting point; my understanding is that LCA - that is, the real, lens-created aberration - gets worse with longer focal length lenses, which is of course likely to be the exact opposite of the position for sensor-angle artefacts. David SignatureDavid Littlewood > I suspect we may be in danger of getting at cross-purposes here. Brian, > you are describing the lens aberration. I believe the previous [quoted text clipped - 4 lines] > aware, but since the phenomenon is similar in appearance this is not > surprising. Did you not read the part where he discusses telecentricity and exit pupil distance Signaturehttp://www.pbase.com/bcbaird >> I suspect we may be in danger of getting at cross-purposes here. Brian, >> you are describing the lens aberration. I believe the previous [quoted text clipped - 7 lines] >Did you not read the part where he discusses telecentricity and exit >pupil distance Yes; what Brian (who is BTW a skilled lens designer) was referring to was an optical phenomenon quite different from the sensor fringing artefacts we had been discussing. Perhaps my post was not clear enough for you. David SignatureDavid Littlewood > Yes; what Brian (who is BTW a skilled lens designer) was referring to > was an optical phenomenon quite different from the sensor fringing > artefacts we had been discussing. Perhaps my post was not clear enough > for you. > > David See his reply to you. Signaturehttp://www.pbase.com/bcbaird Hi David: > I suspect we may be in danger of getting at cross-purposes here. Brian, > you are describing the lens aberration. I believe the previous [quoted text clipped - 4 lines] > aware, but since the phenomenon is similar in appearance this is not > surprising. There certainly are optical effects that are caused by the sensor, such as added light falloff and crosstalk between pixels. However, my point was that if you have two lenses with identical exit pupil positions and one shows lots of lateral color while the other shows none, then telecentricity or lack thereof can't be used to explain what's going on. In other words, many of the posts in this thread have blamed wide-angle color fringing on a lack of telecentricity, which I don't believe to be true. I have seen chromatic effects that probably are due to non-telecentricity, but these occur on a scale similar in size to the sensor as a whole and not on the scale of a small number of pixels. > BTW, interesting point; my understanding is that LCA - that is, the > real, lens-created aberration - gets worse with longer focal length > lenses, which is of course likely to be the exact opposite of the > position for sensor-angle artefacts. There are two distinct primary chromatic aberrations: longitudinal and lateral. Longitudinal chromatic aberration is the one you are referring to, and it does get worse with longer focal length lenses. It can be thought of as a change of focal length with color. It appears on-axis as well as in the outer parts of the image. Lateral color is always zero on-axis and increases with radial symmetry as you go off-axis. It can be thought of as the chromatic variation of magnification. Lateral color is automatically corrected by a symmetrical optical design, so it is most commonly seen in reversed telephoto and telephoto lenses. In the world of DSLR's, when people talk about "CA" they seem to be mostly referring to lateral color. If pixel crosstalk due to steep ray angles was the source of "CA", then I have a real problem understanding how it so closely mimics true lateral color when the pixel array is rectangular. If this were true I would expect the arrangement of the microscopic color fringes to change in a non-rotationally symmetric fashion. Another chromatic effect that people sometimes call "purple fringing" may be due to deep red or deep blue light leaking through the Bayer array and exaggerating the appearance of ordinary longitudinal or lateral color. Some people claim its due to "blooming", which may be true, although I don't fully understand what the term means or why it would result in a purple color. I've never seen "blooming" on my D1x or D1, so its difficult for me to do experiments to find out what it might be. On the other hand, I see examples of lateral color all the time. > David Brian > Some people claim its due to "blooming", which may be >true, although I don't fully understand what the term means or why it >would result in a purple color. Very simple; green-filtered pixels saturate first, since they are the most sensitive, and leak to the less sensitive red- and blue-filtered pixels, raising the red and blue above what it really is, while the green is limited to less than it really is. Signature<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<> John P Sheehy <JPS@no.komm> ><<> <>>< <>>< ><<> <>>< ><<> ><<> <>>< > Very simple; green-filtered pixels saturate first, since they are the > most sensitive, and leak to the less sensitive red- and blue-filtered > pixels, raising the red and blue above what it really is, while the > green is limited to less than it really is. But isn't "pixel leakage" only really an issue on CCDs due to the way the data is collected? Signaturehttp://www.pbase.com/bcbaird >> Very simple; green-filtered pixels saturate first, since they are the >> most sensitive, and leak to the less sensitive red- and blue-filtered >> pixels, raising the red and blue above what it really is, while the >> green is limited to less than it really is. >But isn't "pixel leakage" only really an issue on CCDs due to the way >the data is collected? I don't know. That doesn't sound familiar. Signature<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<> John P Sheehy <JPS@no.komm> ><<> <>>< <>>< ><<> <>>< ><<> ><<> <>>< > >But isn't "pixel leakage" only really an issue on CCDs due to the way > >the data is collected? > > I don't know. That doesn't sound familiar. I'm think specifically of blooming at high-contrast areas. This may or may not be the case, as my memory often does fail me. Signaturehttp://www.pbase.com/bcbaird >Hi David: > [quoted text clipped - 30 lines] >It can be thought of as a change of focal length with color. It >appears on-axis as well as in the outer parts of the image. No, I was referring to lateral chromatic aberration, and (unwisely) used an ambiguous acronym! I have no personal expertise in lens design, but the Focal Encyclopedia does say lateral CA becomes worse at longer focal lengths. This is the opposite of what you say; I must research it further. >Lateral color is always zero on-axis and increases with radial symmetry >as you go off-axis. It can be thought of as the chromatic variation of [quoted text clipped - 17 lines] >might be. On the other hand, I see examples of lateral color all the >time. Given that this seems to be a fairly important issue for digital imaging, I am surprised there is not more detailed technical material available. Well, perhaps there is, but it is not highly visible. Must look. Thinking back over the original purpose of the thread, the corner mush on some 5D images, they seemed to me to be lacking in resolution but not (so far as I could see) showing colour fringing, so I am not sure if it is the same issue at all. David SignatureDavid Littlewood >> The 5D brings full frame for under $4,000 - that will appeal to a >> lot of people who need true wide angle. [quoted text clipped - 17 lines] > Nikon have some great wide angle lenses. This Canon sample just > increased my cravings for a full frame Nikon. Kind of worries me about how the 16-35 f2.8 will do, it's soft on the edges on my 20D... SignatureSkip Middleton http://www.shadowcatcherimagery.com > Kind of worries me about how the 16-35 f2.8 will do, it's soft on the > edges on my 20D... I'd be worried as well. If Canon had some GREAT wide zooms made for use on a digital FF body, I'd seriously consider buying one of these. From the wide angle samples I've seen from the 1Ds using canon glass, this looks typical. Canon is going to HAVE to redo their wide lenses if they want to sell FF bodies to people shooting landscapes. People laughed at the "telecentric" stuff as being market hype but as the sensor densities increase, you're starting to see what they were talking about. I'm sure with normal to tele glass this camera will rock. It's the wide end that is going to suffer because of the glass they have to use on it. SignatureStacey Stacey babbles and babbles and babbles: > I'd be worried as well. If Canon had some GREAT wide zooms made for use on a > digital FF body, I'd seriously consider buying one of these. You clearly haven't the money to buy the medication you desperately need, so dream on, fruitcake. > Canon is going to HAVE to redo their wide lenses if they want to sell FF > bodies to people shooting landscapes. Nitwit: people are buying 1Ds2's and using third-party lenses on them. > People laughed at the "telecentric" > stuff as being market hype but as the sensor densities increase, you're > starting to see what they were talking about. Nitwit: sensor densities have nothing to do with it. (We know this because (a) the effects were visible on film, and (b) pixel pitches haven't changed signifigantly.) > I'm sure with normal to tele > glass this camera will rock. It's the wide end that is going to suffer > because of the glass they have to use on it. Baseless speculation (first sentence), false dichotomy (second sentence). Geez, why not think for a change? It's easy, it's productive, it's fast. Or is it against the tenets of the Nikon Funda-mentalcase Religion? >Stacey babbles and babbles and babbles: > [quoted text clipped - 8 lines] > >Nitwit: people are buying 1Ds2's and using third-party lenses on them. Yes, Zeiss and Leica lens for them not Sigmas. Lenses like the Ziess 21mm Distagon. ************************************************************** "There has always been war. War is raging throughout the world at the present moment. And there is little reason to believe that war will cease to exist in the future. As man has become increasingly civilized, his means of destroying his fellow man have become ever more efficient, cruel and devastating. Is it possible to put an end to a form of human behavior which has existed throughout history by means of photography? The proportions of that notion seem ridiculously out of balance. Yet, that very idea has motivated me. James Nachtwey War Photographer http://www.jamesnachtwey.com/ > >Nitwit: people are buying 1Ds2's and using third-party lenses on them. > > Yes, Zeiss and Leica lens for them not Sigmas. Lenses like the Ziess > 21mm Distagon. Or Contax or Nikon. There are a ton of lenses that can be adapted to the EF mount. Signaturehttp://www.pbase.com/bcbaird >> >Nitwit: people are buying 1Ds2's and using third-party lenses on them. >> [quoted text clipped - 4 lines] > > There are a ton of lenses that can be adapted to the EF mount. Yet none retain ANY automation. This again goes back to: the Legacy lenses people keep saying "I'm so glad I bought Full Frame glass" don't work good on a FF digital body anyway. The normal to tele lenses will work great, the wide film lenses are never going to work great unless the sensor design is changed or the lenses are and I'm not sure the mount size used by Canon is going to be large enough for telecentric wide lenses to be made for it. SignatureStacey > > There are a ton of lenses that can be adapted to the EF mount. > > Yet none retain ANY automation. This again goes back to: the Legacy lenses > people keep saying "I'm so glad I bought Full Frame glass" don't work good > on a FF digital body anyway. The normal to tele lenses will work great, the Some people don't care about automation. And if the image quality wasn't there, people wouldn't be gushing about it. > wide film lenses are never going to work great unless the sensor design is > changed or the lenses are and I'm not sure the mount size used by Canon is > going to be large enough for telecentric wide lenses to be made for it. That's just bunk. Signaturehttp://www.pbase.com/bcbaird > and I'm not sure the mount size used by Canon is going to be large enough > for telecentric wide lenses to be made for it. Well, Nikon is making them, with a smaller mount. SignatureJeremy | jeremy@exit109.com >> and I'm not sure the mount size used by Canon is going to be large enough >> for telecentric wide lenses to be made for it. > > Well, Nikon is making them, with a smaller mount. Then Canon -should- be able to as well. SignatureStacey >> and I'm not sure the mount size used by Canon is going to be large >> enough for telecentric wide lenses to be made for it. > Well, Nikon is making them, with a smaller mount. But so far, only for a smaller image circle. AFAIK, all Nikon's telecentric lenses are DX lenses. Same with Canon: The 10-22 EF-S is telecentric and from the samples I've seen, it produces better results on a 20D than this horrible 17-40 L example. The big question now is what gives the better result at the wide end: A 1.5x/1.6x crop and DX/EF-S designs, or FF and Canon's present EF line-up? There is IMHO no point in going for FF and L-glass if a APS-C sized sensor and telecentric DX/EF-S designs actually gives you better wide angle results. (Well, there is noise and there is DOF, but the high ISO-noise from the latest generation CMOS is so impressive at APS-C size that this no longer an issue for me, and if I want shallow DOF, I just up the focal lengh.) Canon has alreday announced that they plan to replace most of their present wide-angles with telecentric designs, and I've no doubt that this is doable - but these lenses are not avialable today. In other words, the question I am asking is this: Given that good wide angle performance is your top priority, is the EOS 5D /really/ such a bargain compared to, say, a D2x or a 1D MkIIn? Signature- gisle hannemyr [ gisle{at}hannemyr.no - http://folk.uio.no/gisle/ ] ------------------------------------------------------------------------ Kodak DCS460, Canon Powershot G5, Olympus 2020Z ------------------------------------------------------------------------ SNIP > The big question now is what gives the better result at the wide > end: A 1.5x/1.6x crop and DX/EF-S designs, or FF and Canon's > present EF line-up? One thing is clear, the '1.6x crop' needs to be magnified some 60% more to reach the same output size as a FF ... Bart > But so far, only for a smaller image circle. AFAIK, all Nikon's > telecentric lenses are DX lenses. The Nikon 17-35mm f/2.8 is a telecentric design with full-frame coverage. I don't know what other lenses may or may not be; unfortunately they don't mark them "telecentric" or anything like that. > Canon has alreday announced that they plan to replace most of their > present wide-angles with telecentric designs, and I've no doubt that > this is doable - but these lenses are not avialable today. It also means that you have to replace your lenses to get wide-angle on full-frame, which is exactly opposite to the reason people wanted full-frame in the first place (so their existing lenses would work like they did on film). I'd rather stay with cropped sensors so that my existing lenses keep working. > In other words, the question I am asking is this: Given that good > wide angle performance is your top priority, is the EOS 5D /really/ > such a bargain compared to, say, a D2x or a 1D MkIIn? Well, exactly. I can't answer the question, since super-wide isn't a priority for me, but at this point in the game I'd be inclined toward the APS-C side (or 1.3x, if that's small enough to avoid the issue). SignatureJeremy | jeremy@exit109.com "The Nikon 17-35mm f/2.8 is a telecentric design with full-frame coverage. " No, its not telecentric. > "The Nikon 17-35mm f/2.8 is a telecentric design with full-frame > coverage. " > > No, its not telecentric. No lens is telecentric, technically. But everyone says the 17-35 is a telecentric design, though no one seems to be able to say anything for sure. Do you have that information? SignatureJeremy | jeremy@exit109.com I wrote: >> "The Nikon 17-35mm f/2.8 is a telecentric design with full-frame >> coverage. " [quoted text clipped - 4 lines] > telecentric design, though no one seems to be able to say anything for > sure. Do you have that information? Actually, I've found that the exit pupil at the 17mm end seems to be 98mm, which sounds like a telecentric design to me, but I'm no expert. SignatureJeremy | jeremy@exit109.com > I wrote: > > [quoted text clipped - 9 lines] > Actually, I've found that the exit pupil at the 17mm end seems to be 98mm, > which sounds like a telecentric design to me, but I'm no expert. 98mm sounds about right, and its not telecentric. After all, that results in a chief ray angle of over 12 degrees (24x36mm format), which is alot bigger than zero. In a telecentic lens the exit pupil is at infinity. All that is required to achieve this is to place the aperture stop at the front focal plane. The front focal plane is found by tracing a ray parallel to the optical axis "backward" from the image plane back through the lens. True, you might misplace the stop location by a tiny amount during production and as a result place the exit pupil at a large finite distance, but I don't think that's the issue here. Many lenses really are designed to be truly telecentric, including projection lenses, microscope objectives, broadcast zoom lenses, certain types of eyepieces and magnifiers, and stepper lenses. As far as I know, none of the photographic lenses manufactured by Nikon, Canon, or Zeiss are telecentric. Brian > Many lenses really are designed to be truly telecentric, including > projection lenses, microscope objectives, broadcast zoom lenses, > certain types of eyepieces and magnifiers, and stepper lenses. As far > as I know, none of the photographic lenses manufactured by Nikon, > Canon, or Zeiss are telecentric. Okay. I'm a photographer, not a lens designer, so my understanding may well be flawed and I'll take your word for it. :) Having said that, my understanding was that a truly telecentric lens, having its exit pupil at an infinite distance, would be impossible in practice, and thus when we say "telecentric" we're really saying "more telecentric than usual". That's why I say "telecentric design" rather than "telecentric" in the absolute sense. Is that a flawed understanding? I've heard it said that an exit pupil at around 80mm is enough to call it "telecentric" (for 35mm). Also, Nikon's 17-55mm DX lens is claimed to be a telecentric design, at least (it doesn't cover full-frame, of course). In the case of the 17-35mm, one of its advantages was supposed to have been telecentric design. It's widely considered superior to most of the fixed-length wide angles in its range, and the telecentric part was supposed to have been an attempt to design it with digital sensors in mind. So, if what I've read about telecentric lenses is in fact incorrect in the lens-design context, is this a hint that the lens makers are abusing the term "telecentric" in ways that optical experts would disagree with? SignatureJeremy | jeremy@exit109.com > > Many lenses really are designed to be truly telecentric, including > > projection lenses, microscope objectives, broadcast zoom lenses, [quoted text clipped - 9 lines] > That's why I say "telecentric design" rather than "telecentric" in the > absolute sense. Is that a flawed understanding? If someone describes a lens as "telecentric", it really should be telecentric, IMO. Placing the exit pupil at infinity is possible with nearly *any* lens; all you have to do is position the stop correctly as I stated in an earlier post. I think its probably best to talk about exit pupil distances or chief ray angle when discussing non-telecentric lenses. > I've heard it said that an exit pupil at around 80mm is enough to call > it "telecentric" (for 35mm). Also, Nikon's 17-55mm DX lens is claimed > to be a telecentric design, at least (it doesn't cover full-frame, of > course). A large number of 35mm lenses designed well before the digital age already have exit pupil distances of 80mm or more. This is largely because of the large BFL requirement for SLR viewfinders. Some lenses in the 20-55mm EFL range do have exit pupil distances less than 60mm, but I'm not aware of any SLR lens with an exit pupil distance less than 50mm. The shortest exit pupil distance of any Nikkor that I'm aware of is the 25-50mm f/4, which has an axit pupil distance of a little over 50mm at the 25mm position. This would be a good lens to experiment with if you wanted to see what the effects of a short exit pupil distance are. By comparison, the 24/2, 50/1.8, and 55/2.8 are all around 57mm. The Zeiss 21/2.8 Distagon, which is widely regarded as being *superb* on full-frame digital, has an exit pupil distance of only 63.2mm. > In the case of the 17-35mm, one of its advantages was supposed to have > been telecentric design. It's widely considered superior to most of [quoted text clipped - 4 lines] > makers are abusing the term "telecentric" in ways that optical experts > would disagree with? Reading this thread is the first time I've heard the 17-35/2.8 described as telecentric, which it clearly is not. Its superior performance is due to superior aberration correction, particularly lateral color. Bear in mind that the first sensor-induced effect due to non-telecentricity is increased falloff. Sensor effects simply would not create anything that resembled lateral color, which is what most people complain about with regard to wide-angle lenses. If Nikon is actually calling the 17-35 a telecentric lens, then all I can do is groan. I suppose its no better than the marketeers at Schneider and Rodenstock using the term "APO" to describe non-apochromatic lenses. Brian >The Zeiss 21/2.8 Distagon, which is widely regarded as > being *superb* on full-frame digital, has an exit pupil distance of > only 63.2mm. *Superb* or just better than the WA options that canon provides? > Bear in mind that the first sensor-induced effect due > to non-telecentricity is increased falloff. Which they counter with micro lenses on the sensor. What we don't know is what effect on resolution these higher angles of the light rays striking these sensors has. When you go beyond 12 degrees, things start getting weird. With many of the canon WA optics on FF, you're talking about 19 degrees! > Sensor effects simply > would not create anything that resembled lateral color, which is what > most people complain about with regard to wide-angle lenses. http://nikondigital.org/articles/digital_lenses.htm "However, there is a downside to using our current lenses with digital. The fifth factor which influences how your lens will perform on a current generation digital camera is the most esoteric. Because digital cameras rely on a separate color filter array (CFA) placed in front of the CCD to create their notion of scene color, they are subject to a unique form of distortion. Light falling straight down on a pixel comes through the appropriate color cell and is correctly measured. But light coming from an angle may come through a different color cell in the filter, and the camera will be measuring the wrong color. " SignatureStacey >> Sensor effects simply >> would not create anything that resembled lateral color, which is what [quoted text clipped - 13 lines] > camera > will be measuring the wrong color. " Sometimes, Stacey, you are a wealth of information. SignatureSkip Middleton http://www.shadowcatcherimagery.com >> http://nikondigital.org/articles/digital_lenses.htm >> [quoted text clipped - 9 lines] >> camera >> will be measuring the wrong color. Sounds very close to what the Nikon Tech Rep told me earlier this summer. He recommended not using earlier film camera lenses below 46mm on our D1x and D2x bodies. He recommended only using the DX wideangle lenses below that focal range. > Sounds very close to what the Nikon Tech Rep told me earlier this > summer. He recommended not using earlier film camera lenses below > 46mm on our D1x and D2x bodies. He recommended only using the DX > wideangle lenses below that focal range. And again this isn't the end of the world for people with a bunch of "legacy" lenses. Only the WA lenses are going to be a problem and Canon just needs to get with the program and redesign their WA lenses for their digital FF models. If the 5D sells in the numbers I expect it to, they have no reason not to invest the R&D to do this. If they do redesign the WA end of their lenses, I'd seriously look at buying one of these. SignatureStacey " And again this isn't the end of the world for people with a bunch of "legacy" lenses. Only the WA lenses are going to be a problem and Canon just needs to get with the program and redesign their WA lenses for their digital FF models. If the 5D sells in the numbers I expect it to, they have no reason not to invest the R&D to do this. If they do redesign the WA end of their lenses, I'd seriously look at buying one of these. " Finally we agree on something: Canon needs to re-design their wide-angle lenses! Brian > " And again this isn't the end of the world for people with a bunch of > "legacy" lenses. Only the WA lenses are going to be a problem and Canon [quoted text clipped - 11 lines] > > Brian Are you daft? That is *EXACTLY* what I've been saying the whole time, then you started argueing that a Zeiss distagon on an manual adapter works sorta OK? Again have you actually LOOKED at the examples yet? We aren't talking about CA or lateral color shifts.. SignatureStacey >> " And again this isn't the end of the world for people with a bunch of >> "legacy" lenses. Only the WA lenses are going to be a problem and Canon [quoted text clipped - 16 lines] >OK? Again have you actually LOOKED at the examples yet? We aren't talking >about CA or lateral color shifts.. Spherochromatism. A combination of spherical aberration and chromatic aberration. Canon lenses (some of them) seem like teaching aids to study optical aberrations. http://www.vanwalree.com/optics/spherical.html -Rich "Bittorrents are REFUNDS for all the BAD movie products Hollywood never gave us refunds for in the past" "Spherochromatism. A combination of spherical aberration and chromatic aberration. Canon lenses (some of them) seem like teaching aids to study optical aberrations. http://www.vanwalree.com/optics/spherical.html " Its better to call spherochromatism a variation of spherical aberration with wavelength. Saying "a combination of spherical aberration and chromatic aberration" sounds like spherical plus axial chromatic, which is definitely not spherochromatism. In photographic lenses it tends to be unimportant, but is occasionally visible in defocused highlights and in extremely fast lenses used wide-open. It only becomes a really big deal with certain types of telescope objectives. For example, in Schmidt cameras and Schmidt-Cassegrain telescopes spherochromatism is often the dominant residual aberration. >> Sounds very close to what the Nikon Tech Rep told me earlier this >> summer. He recommended not using earlier film camera lenses below [quoted text clipped - 8 lines] > no reason not to invest the R&D to do this. If they do redesign the WA end > of their lenses, I'd seriously look at buying one of these. And I'd seriously look at buying a new WA lens, which plays directly into the hands of Canon... But really, Stacey, and I'm not taking a cheap shot, here, but what would you do with the E-300 and the glass you own for it? And what if Oly comes out with that "have to have" lens for the 4/3 system? SignatureSkip Middleton http://www.shadowcatcherimagery.com > And I'd seriously look at buying a new WA lens, which plays directly into > the hands of Canon... And lots of other people. Like I said, it's not a huge deal and why I thought the "You're buying lenses for a dead system" was kinda funny, as I was thinking these other people will probably be replacing their lenses by then as well! :-) > But really, Stacey, and I'm not taking a cheap shot, here, but what would > you do with the E-300 and the glass you own for it? Use it for tele and macro stuff, it does a great job there now. I only own 3 lenses for it so it's not the end of the world if I needed to add another camera. Plus it's smaller and lighter than a FF system would be etc. > And what if Oly comes > out with that "have to have" lens for the 4/3 system? The problem I have is: the E300 body doesn't resolve well enough to do large prints of landscapes shot at wide angles real well. The optics are fine, it just can't resolve fine details, not enough MP. At 8X10 it's OK but since it has to be upsampled to print 11X14, resolution is lost. It's about like I got from 35mm but that wasn't accpetable to me either. I'm still using my mamiya 645 for that stuff. I hate scanning film so if I could get a "landscape" body and lens setup to replace the 645 (or come pretty close), I'd be happy! By the time they come out with a good digital FF zoom (maybe something like a 20-50 F4), the 5D will have come down in price, had 4-5 firmware updates and the bugs squashed. Sounds like maybe next summer? Then again Nikon might be in the game by then as well and who knows, 4/3 may have some sensor development that solves my problem? SignatureStacey >> And I'd seriously look at buying a new WA lens, which plays directly into >> the hands of Canon... [quoted text clipped - 3 lines] > was thinking these other people will probably be replacing their lenses by > then as well! :-) I don't think 4/3 is a "dead system," just that it may go a different direction, one more indicated by the E-300 than the E-1. More of an uberPoint&Shoot, with options, as it were. >> But really, Stacey, and I'm not taking a cheap shot, here, but what would >> you do with the E-300 and the glass you own for it? [quoted text clipped - 3 lines] > lenses for it so it's not the end of the world if I needed to add another > camera. Plus it's smaller and lighter than a FF system would be etc. I thought you owned more than that... >> And what if Oly comes >> out with that "have to have" lens for the 4/3 system? [quoted text clipped - 7 lines] > like > I got from 35mm but that wasn't accpetable to me either. So the trade off is poor WA performance for FF because of light incidence or insufficient res or noise issues from smaller sensor. Pick yer poison? > I'm still using my mamiya 645 for that stuff. I hate scanning film so if I > could get a "landscape" body and lens setup to replace the 645 (or come [quoted text clipped - 3 lines] > next summer? Then again Nikon might be in the game by then as well and who > knows, 4/3 may have some sensor development that solves my problem? I may have to get that 20-35 f2.8L I saw at KEH's site. SignatureSkip Middleton http://www.shadowcatcherimagery.com "Sounds very close to what the Nikon Tech Rep told me earlier this summer. He recommended not using earlier film camera lenses below 46mm on our D1x and D2x bodies. He recommended only using the DX wideangle lenses below that focal range. " Only DX lenses? I wonder why he would rule out an excellent wide-angle like the 17-35/2.8, which is arguably better than any of the DX series? Or how about the far older 25-50/4, which I use all the time on a D1x? As I've pointed out above, this latter lens is the *least* telecentric of all Nikkors that I'm aware of, an yet it works great on digital cameras. At best, I would suggest that this Nikon Tech Rep has a conflict of interest when it comes to discussing new lenses vs. old lenses. Brian > "Sounds very close to what the Nikon Tech Rep told me earlier this > summer. He recommended not using earlier film camera lenses below [quoted text clipped - 7 lines] > of all Nikkors that I'm aware of, an yet it works great on digital > cameras. So you've tested the 17-35 vs these other DX lenses? And the 17-35 -is- a fairly telecentric lens and may have very well been designed with digital in mind. The other points are: you are using this on a 1.5X crop (and a 5.4MP) which has little to do with how the canon lenses they will work on a FF camera in the corners of a 10+MP camera. What looked OK at 5MP might not hold true at 10+MP, kinda like a really sharp lens can't resolve all it can using ISO 800 print film so will appear only as good as a much lower quality optic will. Put high resolution film in both and THEN you can see the problems with the lower quality optics. I thought I was getting "great" results until I got a really good designed for digital lens so ignorance is bliss. SignatureStacey > >The Zeiss 21/2.8 Distagon, which is widely regarded as > > being *superb* on full-frame digital, has an exit pupil distance of > > only 63.2mm. > > *Superb* or just better than the WA options that canon provides? Superb *and* better than anything Canon provides. > > Bear in mind that the first sensor-induced effect due > > to non-telecentricity is increased falloff. [quoted text clipped - 4 lines] > weird. With many of the canon WA optics on FF, you're talking about 19 > degrees! With or without microlenses, as you increase the incidence angle the first effect that occurs is light falloff. Actual crosstalk between pixels can only occur when light falls on the wrong pixel. Before this occurs the light will fall on the circuitry between pixels. This is what causes the light falloff. The Zeiss 21mm previously mentioned has an exit pupil distance of 63.2mm. That means the chief ray angle at the sensor is just under 19 degrees. Yet every test image I've seen from this lens is excellent in the corners. Exactly what is the problem you are seeing? Another excellent test would be to mount a Nikon 55/2.8 micro on a Canon FF camera. This lens has extremely good optical performance with essentially zero lateral color (less than 1 pixel at the extreme corner from red to violet). The 55/2.8 also has a fairly short exit pupil distance of 57mm, which would give a corner angle of incidence of almost 21 degrees on a FF camera. Any bizarre optical effects here would be due to the sensor and not the lens. Also, I believe that the Canon wide-angle zooms have more distant exit pupils than the Zeiss 21mm prime lens. If true, then you obviously need not worry about sensor effects creating the poor resolution with Canon wide-angle optics. > > Sensor effects simply > > would not create anything that resembled lateral color, which is what [quoted text clipped - 11 lines] > angle may come through a different color cell in the filter, and the camera > will be measuring the wrong color. " Stacey, even if you had some crosstalk going on it would't cause a radially-symmetric problem that looks just like normal lateral chromatic aberration. Why? - because the CFA covering the sensor has no rotational symmetry. Therefore, at a given radial distance from the center of the image you will get an odd mixture of crosstalk between green and red, green and blue, and red and blue. This clearly cannot produce the rotationally symmetric fringing created by lateral color. You can quote misinformation from the web all you like, and it will never substitute for careful observation and reasoning. > You can quote misinformation from the web all you like, and it will > never substitute for careful observation and reasoning. You can throw a bunch of big words out in hopes that no ones knows what they mean trying to make it appear you know what you're talking about. Doesn't change the FACT that these WA zooms suck on a FF body and their design -IS- the problem. Maybe you can blow off pixel cross talk as nothing to worry about, I can't imagine that is something you'd want going on or that it would be ideal. BTW I was reading an article by one of these guys who is using this "suberb" zeiss lens you keep going on about vs the 7-14 zuiko digial zoom, he said the ZD zoom is a better optic which goes against all logic that a "suberb" prime lens would get beat by a WA zoom! Couldn't be because the ZD was designed for digital use, cause you used some big important sounding words to "prove" it doesn't. SignatureStacey > > You can quote misinformation from the web all you like, and it will > > never substitute for careful observation and reasoning. > > You can throw a bunch of big words out in hopes that no ones knows what they > mean trying to make it appear you know what you're talking about. If you don't understand what an exit pupil distance is, or what lateral chromatic aberration is, then I'll agree that its going to be difficult for us to communicate very well. > Doesn't change the FACT that these WA zooms suck on a FF body and their > design -IS- the problem. Maybe you can blow off pixel cross talk as nothing > to worry about, I can't imagine that is something you'd want going on or > that it would be ideal. No, pixel crosstalk isn't good, but it won't explain the problems you are seeing. You are blaming pixel crosstalk and lack of telecentricity when you should be blaming poor aberration correction. > BTW I was reading an article by one of these guys who is using this "suberb" > zeiss lens you keep going on about vs the 7-14 zuiko digial zoom, he said > the ZD zoom is a better optic which goes against all logic that a "suberb" > prime lens would get beat by a WA zoom! Couldn't be because the ZD was > designed for digital use, cause you used some big important sounding words > to "prove" it doesn't. If its a good article with side-by-side test images I would like to read it. Brian > No, pixel crosstalk isn't good, but it won't explain the problems you > are seeing. You are blaming pixel crosstalk and lack of telecentricity > when you should be blaming poor aberration correction. But, are you seeing what we're seeing? It's not chromatic aberration or other color "fringing" in Canon super-wide shots that we're talking about; it's details blurring into nonexistence at the edges. Is that an artifact of the same thing? It's not "just" a little color fringe, it's picture- destroying blurring. SignatureJeremy | jeremy@exit109.com >> No, pixel crosstalk isn't good, but it won't explain the problems you >> are seeing. You are blaming pixel crosstalk and lack of telecentricity [quoted text clipped - 5 lines] > of the same thing? It's not "just" a little color fringe, it's picture- > destroying blurring. Exactly. They keep talking about lateral color shifts when it's details blured to oblivion and acting like software can fix this? SignatureStacey >> > You can quote misinformation from the web all you like, and it will >> > never substitute for careful observation and reasoning. [quoted text clipped - 5 lines] > chromatic aberration is, then I'll agree that its going to be difficult > for us to communicate very well. I do understand this, have you actually LOOKED at the samples we are talking about? Since when did lateral chromatic aberration turn ALL details into a mushy mess? We aren't talking about a touch of CA here, it's a unrecoverable mess! Do you understand what is going on when these micro lenses have light hitting them at a high angle? >> Doesn't change the FACT that these WA zooms suck on a FF body and their >> design -IS- the problem. Maybe you can blow off pixel cross talk as [quoted text clipped - 4 lines] > are seeing. You are blaming pixel crosstalk and lack of telecentricity > when you should be blaming poor aberration correction. So why are you defending these lenses (and the fact canon is ignoring this is an issue, just like they are ignoring the sensor dust issue..) that these aren't designed for this purpose as being good? Again it might be "acceptable" only because there isn't anything that is designed for this purpose so we must make do with what is avalible if you want to shoot full frame? And yes even the shots I've seen from this *superb* zeiss lens on a 1Ds were softer in the corners compared to the center, which DOESN'T happen on film. But according to you, this can't happen... >> BTW I was reading an article by one of these guys who is using this >> "suberb" zeiss lens you keep goin |
Enable EMail Alerts
Start New Thread
Reply to this Message