 |
 | 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 / August 200520D Canon focus advice
Hi all, I have a nagging suspicion that the focus on my new 20D isn't all that it should be. Do any of you have any suggestions on the best way to test this? I was thinking of photographing a sheet of A4 "Math - graph" type paper. If the camera's focus is out, will it be out consistently when focusing on close and distant objects? Should I have the lens wide open or stopped down for the test. Is it possible for an image to be in focus through the viewfinder, and yet be out of focus at the CMOS sensor? Any other thoughts (I only have the one kit lens at this time, so I can't do any lens comparisons). Thanks for your input. Cheers, CC > Hi all, > > I have a nagging suspicion that the focus on my new 20D isn't all > that it should be. Do any of you have any suggestions on the best way > to test this? I was thinking of photographing a sheet of A4 "Math - > graph" type paper. Try printing this: http://www.photo.net/learn/focustest/scale45.jpg Place it at 45 degrees and focus on the line in the middle. > Should I have the lens wide > open or stopped down for the test. A large aperture will show the results more clearly. > Is it possible for an image to be in focus through the viewfinder, > and yet be out of focus at the CMOS sensor? Yes, but IMO that would mean a major physical problem which I would expect to show up as other symptoms as well. -Mike > Hi all, > > I have a nagging suspicion that the focus on my new 20D isn't all that it > should be. Hi Colin... Not sure what you mean about the focus...whether you are talking about through the lens or the actual results. Don't mean to be condescending here...but if you are talking about the focus through the lens, have you checked the diopter adjustment on the viewfinder. The only reason I ask is that I had a friend with the same complaint and the diopter just got knocked a fraction. Cheers Steve > Hi all, > [quoted text clipped - 16 lines] > > CC I'm probably teaching old dogs new tricks but always ensure you know which AF point the camera is focusing on, I always lock it to the center one, and remember the AF sensors maybe bigger than the markings in the viewfinder. take a look at http://www.knighttrain.freeserve.co.uk/400.htm where you can see 10D AF markings and sensor sizes. This is based on 10D info but probably holds true for the 20D. > I'm probably teaching old dogs new tricks but always ensure you know which > AF point the camera is focusing on, I always lock it to the center one, and [quoted text clipped - 4 lines] > > This is based on 10D info but probably holds true for the 20D. Also, try the test with different lenses and versus your own manual focusing ability. I thought my camera was poor on focus, then I tried beating it with my own eye. The AF wasn't perfect, but I wasn't anywhere close! If you still have problems, send the camera with your lenses to a Canon repair facility. People have reported success after the techs re- calibrated their body and lenses.  Signaturehttp://www.pbase.com/bcbaird >> I'm probably teaching old dogs new tricks but always ensure you know which >> AF point the camera is focusing on, I always lock it to the center one, and [quoted text clipped - 10 lines] >I thought my camera was poor on focus, then I tried beating it with my >own eye. The AF wasn't perfect, but I wasn't anywhere close! Unfortunately, current AF bodies (except top-of-range models like Canon's 1-series) are stuck with screen almost totally unhelpful for manual focussing. This is, BTW, a major problem with using DSLRs for photomicrography, one of my main interests. It (interchangeable screens) is also one of the more exciting (to me) features of the EOS 5D, if the leaks prove accurate (and if one of the screens is a useful one with clear spot and cross-hair). >If you still have problems, send the camera with your lenses to a Canon >repair facility. People have reported success after the techs re- >calibrated their body and lenses. David SignatureDavid Littlewood > Unfortunately, current AF bodies (except top-of-range models like > Canon's 1-series) are stuck with screen almost totally unhelpful for > manual focussing. This is, BTW, a major problem with using DSLRs for > photomicrography, one of my main interests. I agree and this was one of my "shopping criteria" when buying a dSLR, something the review sites NEVER even mention. Most dSLR's assume you will ALWAYS use AF, have no idea how to focus a camera yourself so they design the focus screen to be as bright as possible (to brighten the view with the slow kit lens?) at the expence of manual focusing ability. And not all camera makers screens are like this. Some models get "blasted" by the reviewers for having a dimmer focus screen when the fact is, they designed the screen so you can actually see the focus plane for manual focusing. Just another thing people should check out when camera shopping. If you can't focus the camera well, it really doesn't matter how great the other features are. SignatureStacey > Unfortunately, current AF bodies (except top-of-range models like > Canon's 1-series) are stuck with screen almost totally unhelpful for > manual focussing. This is, BTW, a major problem with using DSLRs for > photomicrography, one of my main interests. > > David That should've read, come stock with an AF screen >> http://www.keoptics.com/Canon20D.htm Although, it will add some $$ to the orig purchase price, at least there is an option available. -- Slack >> Unfortunately, current AF bodies (except top-of-range models like >>Canon's 1-series) are stuck with screen almost totally unhelpful for [quoted text clipped - 7 lines] >Although, it will add some $$ to the orig purchase price, at least >there is an option available. Thanks for this. Unfortunately, split prisms and micro-fresnel screens are not much improvement for what I want to do. The only satisfactory screen is the clear spot/cross hair type, which gives precise focussing in extreme macro and photomicrography. The "I" screen in the 1-series film cameras is the one I use in those. If it doesn't have a changeable screen, bugger. David SignatureDavid Littlewood > Thanks for this. Unfortunately, split prisms and micro-fresnel screens > are not much improvement for what I want to do. If this page explains how to remove the factory screen, send to to Maxwell precision optics and have him make you a new screen. I'm don't know if you've ever used one of his screens, they are amazing. Very easy to see the focus plane yet still is a very bright screen. I doubt you'd need any kind of focus aide with it. Give him a call and tell him what your needs are, I'm sure he can come up with a very good solution. SignatureStacey snip > Unfortunately, current AF bodies (except top-of-range models like Canon's > 1-series) are stuck with screen almost totally unhelpful for manual [quoted text clipped - 10 lines] > > David Boy do you have that right. The screen in my Digital Rebel sucks for photomicrography. I can still use it but its not a pleasant experience. I miss the screens I used in my F1's. But on the other hand the sensor crop means my bellows mount macro lens give me a higher magnification. SignatureJohn Passaneau State College Pa. w3jxp@adelphia.net You might check out a software called "Imatest" This allows you test various lens for a variety of quality issues. www.imatest.com The test procedure is long and detailed but not necessarily difficult. I recently did it with a couple of new lens, a Canon 17-85 EF-S IS and a Canon 70-300 EF IS DO. Both came out with good results. What was interesting, the shorter lens' quality measurement went down about 5% with a high quality (and clean) filter. With the "DO" lens, the image quality actually improved. I ran the test twice over a couple of days taking two measurements each time. The data were consistent. I think what it told me is that my lenses are high quality and if my images are bad, it's probably the photographer and not the equipment. Steve >> I'm probably teaching old dogs new tricks but always ensure you know >> which [quoted text clipped - 17 lines] > repair facility. People have reported success after the techs re- > calibrated their body and lenses. > Hi all, > [quoted text clipped - 14 lines] > > CC Using an angled scale is not the best way to test your focus. 20D's have a really tight tolerance to distance from the rear element to the film plane (sensor). They are much less tolerant to errors than a 35mm camera is due to the closeness of these 2 points. According to Canon: correct focus is anywhere within the depth of field of the lens. If you bought a camera with a "Kit lens", the focus point could be anywhere within a inch of the centre at about 2 feet and still be within acceptable factory tolerances. Pretty poor stuff when you decide to shoot some flowers or small animals! The most common focus error on 20Ds is called "back focus" which is independent of lenses. All three of my 20Ds had back focus errors which only became evident when I used a 50mm f1.4 lens at close range. Canon fix it by plugging in a computer and making a few adjustments, no lens attached. Try to get the software yourself and see how friendly and cooperative Canon are not! You can get an *IDEA* of the error by using the chart Mike Warren posted a link to but the process of using a chart at an angle becomes imprecise when you use a wide angle lens or a wide aperture lens designed for full frame cameras (like a 50, f1.4 or 1.8). Sort of self defeating unless the error is substantial. All canon (non macro)lenses exhibit barrel distortion or a pincushion effect at close quarters so your results might *indicate* a back focus error but they cannot quantify it. Canon will take very little time in fixing the problem. You might discover a relapse of the problem if you have to remove the camera's "stay alive" battery when it dies on a lens change or from connecting your Speedlight when it's switched on, but inside 12 month, they'll do it free as many times as you need it. Just make sure the error is consistent between lenses or you need to take all the lenses in as well! In all the years I used Nikon cameras, I never saw one instance of faults or reliability relating to quality control during manufacture at any of the newspapers I worked on. Since buying Canon gear I've seen many photographers with 20D problems and experienced some severe problems with my own gear which should not have happened. People seem to think quality is related to price. Canon have disproved that with their 20D. Those who don't get the faults, don't push their cameras. SignatureDouglas, You never really make it on the 'net until you get your own personal Troll. Mine's called Chrlz. Don't feed him, he bites! > Those who don't get the faults, > don't push their cameras. Translation: Measurebators will always find faults with their cameras. Signaturehttp://www.pbase.com/bcbaird >> Those who don't get the faults, >> don't push their cameras. > > Translation: > > Measurebators will always find faults with their cameras. Actually you're the one who said you can't even manually focus your camera so how would you know? SignatureStacey > > Measurebators will always find faults with their cameras. > Actually you're the one who said you can't even manually focus your camera > so how would you know? Can't you go two seconds without trying to stir sh.t up, Stacey? Signaturehttp://www.pbase.com/bcbaird >>> Those who don't get the faults, >>> don't push their cameras. [quoted text clipped - 5 lines] >Actually you're the one who said you can't even manually focus your camera >so how would you know? Poor Brian. He's not actually very bright. ;-) > >Actually you're the one who said you can't even manually focus your camera > >so how would you know? > > Poor Brian. He's not actually very bright. Look you guys, if you're going to troll you'll just have to do a better job. Signaturehttp://www.pbase.com/bcbaird > Since buying Canon gear I've seen many photographers with 20D problems > and experienced some severe problems with my own gear which should not > have happened. People seem to think quality is related to price. Canon > have disproved that with their 20D. Those who don't get the faults, > don't push their cameras. There are too many people with money who buy cameras like the 20D and have no knowledge of photography. >>Since buying Canon gear I've seen many photographers with 20D problems >>and experienced some severe problems with my own gear which should not [quoted text clipped - 4 lines] > There are too many people with money who buy cameras like the 20D and > have no knowledge of photography. There are many people who buy expensive cars who have no interest in lifting the engine hood. When the DOD was dumping aircraft into the civilian market, I know people who bought T-28 North American built trainers, who didn't know how to fly. But the price was right, so they learned how to fly. >> Since buying Canon gear I've seen many photographers with 20D problems >> and experienced some severe problems with my own gear which should not [quoted text clipped - 4 lines] >There are too many people with money who buy cameras like the 20D and >have no knowledge of photography. 'Twas ever thus. A great many people bought high end 35mm SLRs like the Canon EOS 1N/V and the Nikon F5, only to use them with lousy 28-200mm lenses for taking family snapshots. It's like people who buy a Porsche and use it only for shopping trips. It's called "freedom". ;-) > Snip > Since buying Canon gear I've seen many photographers with 20D problems > and experienced some severe problems with my own gear which should not > have happened. People seem to think quality is related to price. Canon > have disproved that with their 20D. Those who don't get the faults, > don't push their cameras. Problems with the 20D? I haven't experienced any problems with my 20D. I use my cameras in ways that I've always used my cameras. I had problems with the BG-E2 battery pack not being usable with AA batteries, but that problem took about 15 minutes to go away. I drove to the Canon repair center and told them about the problem. They took my camera to their repair lab and before I had the chance to finish reading a magazine article while waiting, they came back with a NEW BG-E2 battery pack. While there, they asked to see my equipment and updated firmware and accomplished a series of checks on my camera bodies and lenses. Thanks all, My suspicions were aroused when I took a photo of a stream and surrounding vegetation whilst on a walking track (hand held) - it just looked plain blury. I went back later on that day and set the camera up on a tripod. On my 20D with kit lens (zoomed all the way out) At 100 ISO and F29? (or there abouts - stopped down as far as it would go) I ended up with a exposure of 2.5 sec (not a breath of wind, so not worried about anything moving in the breeze). Back on the PC EVERYTHING just didn't appear to be as sharp as I would have expected (shot in RAW too). The whole thing just appeared hard on the eyes - it wasn't until I ran an unsharp mask at 200% over it that it started to look "normal". Not sure if this is the camera, or if I'm expecting too much. I did take some photos of some graph paper later on that night, and it "sort of" appeared OK (difficult to tell). I could send it in for checking, but it probably means doing without it for a couple of weeks :( Cheers, CC > Hi all, > [quoted text clipped - 14 lines] > > CC > Thanks all, > > My suspicions were aroused when I took a photo of a stream and surrounding > vegetation whilst on a walking track (hand held) - it just looked plain > blury. Depending on what active focus point was looking at, the camera might have misfocused for the scene. IMHO using all seven focus points can be a trap if one of the outer points happens to be the one chosen by the camera, and it's looking at an object that is not representative of the scene distance. For this reason I always have my camera set for the centre point only. > I went back later on that day and set the camera up on a tripod. On > my 20D with kit lens (zoomed all the way out) At 100 ISO and F29? (or there > abouts - stopped down as far as it would go) I ended up with a exposure of > 2.5 sec (not a breath of wind, so not worried about anything moving in the > breeze). The lens stops down to f/22 at 18mm, and f/38 at the 55mm end - where you had the lens set - and at that aperture the image will be badly affected by diffraction, so it will not be anywhere near as sharp as at f/11 or f/8. Basically, that sort of aperture is unusable for normal photography. > Back on the PC EVERYTHING just didn't appear to be as sharp as I would have > expected (shot in RAW too). The whole thing just appeared hard on the eyes - [quoted text clipped - 5 lines] > I could send it in for checking, but it probably means doing without it for > a couple of weeks :( Set it up on the tripod - is it a sturdy tripod, or a $30 model that vibrates like a tuning fork? {:-) - and take a shot of the houses across the street from your front porch, or similar, at f/8 and again at f/11 using the Av option, at 100 ISO, using the delayed action to fire the camera, and then have a look at the resultant images. A Canon rep somewhere advocated using an unsharp mask at 300% and 0.3 pixels, no threshold. I use this setting, and the images fair leap off the paper at A4, even with Large Fine jpeg images, though I use RAW for all my serious stuff. Regards, Colin D. >The lens stops down to f/22 at 18mm, and f/38 at the 55mm end - where >you had the lens set - and at that aperture the image will be badly >affected by diffraction, so it will not be anywhere near as sharp as at >f/11 or f/8. Basically, that sort of aperture is unusable for normal >photography. True-ish; f/38 will be very much below optimum for anything other than a pinhole or bottle end. However, it may well be obligatory to get the depth of field required in high ratio macro work (don't forget that that diffraction is determined by effective aperture (which is measured aperture x (1+m)). Also, in large format work, f/45 is quite routine. Yes, it is sub-optimum, but at least you have a lot of resolution to spare. Finally, a ^good^ lens should be best (in resolution terms) at something a lot wider than f/11. F/5.6 for a really well corrected lens, f/8 is not bad. David SignatureDavid Littlewood > >The lens stops down to f/22 at 18mm, and f/38 at the 55mm end - where > >you had the lens set - and at that aperture the image will be badly [quoted text clipped - 10 lines] > Also, in large format work, f/45 is quite routine. Yes, it is > sub-optimum, but at least you have a lot of resolution to spare. <snip> True also, but the lens the OP is talking about is a zoom 'macro', not capable of high-ratio work, and of course f/45 in a 150mm lens is entirely different from f/45 in a 50mm lens, for two reasons; one, that diffraction is a function of the actual diameter of the aperture, and a longer lens has a larger aperture for the same f-number than a short lens; and two, there is less magnification of the image required from the larger negative. regards, Colin D. >> >The lens stops down to f/22 at 18mm, and f/38 at the 55mm end - where >> >you had the lens set - and at that aperture the image will be badly [quoted text clipped - 15 lines] >True also, but the lens the OP is talking about is a zoom 'macro', not >capable of high-ratio work, True > and of course f/45 in a 150mm lens is >entirely different from f/45 in a 50mm lens, for two reasons; one, that >diffraction is a function of the actual diameter of the aperture, and a >longer lens has a larger aperture for the same f-number than a short >lens; Not true. I used to think this until I actually went and studied the physics behind it, but diffraction is purely a function of f-number, not absolute diameter. > and two, there is less magnification of the image required from >the larger negative. True, which is why I said "but at least [with LF] you have a lot of resolution to spare". This is probably why you think (and I used to think) resolution depends on diameter: the lower magnification factor for a large film means that a higher f-number will be satisfactory. Rules of thumb exist, along the lines of "anything above 4mm (i.e. f/38 on a 150mm lens) is fine". They don't entirely cover the position though. SignatureDavid Littlewood <snip> > > and of course f/45 in a 150mm lens is > >entirely different from f/45 in a 50mm lens, for two reasons; one, that [quoted text clipped - 5 lines] > physics behind it, but diffraction is purely a function of f-number, not > absolute diameter. Very interesting. I dug out my old Ilford Manual of Photography (that shows how long I've been involved in this caper!) and looked up Airy Disc calculations, and you're right. I am guilty of not reading the formula correctly umpteen years ago. Since d, the aperture diameter, is a component of the equation, I assumed that diffraction was simply proportional to d - but I overlooked that v (image distance) is also in the equation. Since v/d is focal length over diameter, and v/d = f-number, then diffraction is indeed purely a function of F-number. My old Dad used to say "the more you know, the more you know you don't know." After 50 years, I am still learning ... thanks. Colin D. > Back on the PC EVERYTHING just didn't appear to be as sharp as I would have > expected (shot in RAW too). The whole thing just appeared hard on the eyes - > it wasn't until I ran an unsharp mask at 200% over it that it started to > look "normal". Not sure if this is the camera, or if I'm expecting too much. > I did take some photos of some graph paper later on that night, and it "sort > of" appeared OK (difficult to tell). IF you're shooting at really small apertures things aren't going to be as sharp due to diffraction. Running an unsharp mast will help a bit, and in general all DSLR images need some final sharpening to achieve optimum results. Signaturehttp://www.pbase.com/bcbaird Or you can just turn up the in-camera sharpening... ...Did you ever bother to look at the custom functions, Doug? Signaturehttp://www.pbase.com/bcbaird > Thanks all, > [quoted text clipped - 44 lines] >> >>CC According to Canon (again) application of an unsharp mask of 300% with a radius of 0.3 is required to offset the effects of their anti-alias filter (on a 10D) The AA filter on a 20D is less aggressive but still needs about 200% at 0.2 to reach the out of camera sharpness of a Nikon or Olympus image. SignatureDouglas, You never really make it on the 'net until you get your own personal Troll. Mine's called Chrlz. Don't feed him, he bites! > Thanks all, > [quoted text clipped - 48 lines] >> >> CC At F29 or so almost any lens will be un-sharp. This is due to diffraction from the small aperture. It is not a lens defect, it is a law of physics. At F29 or so the depth of field really means "every thing is fuzzy". The best aperture to check for focus errors would be wide open or if the lens not of the highest quality, at 1 or 2 stops down from the maximum aperture. Also with a zoom lens use the longest focal length available. If your basing the possible focus problem on a shot at the smallest possible aperture of your lens, I would expect it to be fuzzy. SignatureJohn Passaneau State College Pa. w3jxp@adelphia.net > At F29 or so almost any lens will be un-sharp. This is due to diffraction > from the small aperture. It is not a lens defect, it is a law of physics. At [quoted text clipped - 4 lines] > possible focus problem on a shot at the smallest possible aperture of your > lens, I would expect it to be fuzzy. Thanks for that. Unfortunately the "nature scene" wasn't intended as a focus test - it was one nice scene that I wanted as a "keeper". I wanted the best possible DOF so I went for smallest aperture - lowest ISO - shot in RAW - put the camera on a tripod - and set the camera for a delayed shot. I knew that a large aperture means poor DOF - I didn't know that small aperture has issues as well (something else to add to my bucket of wisdom!). So far, using an unsharp mask @ 300% @ 0.3 as suggested by Colin D seems to be producing a good result (I didn't know you could select anything under 1 pixel). I'll have to do some more tests to see it F8 to F11 are any better for me. Cheers, >> At F29 or so almost any lens will be un-sharp. This is due to diffraction >> from the small aperture. It is not a lens defect, it is a law of physics. [quoted text clipped - 30 lines] > > Cheers, Try www.bobatkins.com Dr Atkins has lots of info on lenses, dslrs, etc.... or try www.photo.net which he now runs. > Hi all, > [quoted text clipped - 16 lines] > > CC Stop worrying. If you cannot tell from normal shots then it is fine. But 'geek and ye shall find' ;) However if the measurebator urge is too strong, then go on over to www.dpreview.com and search on bf and ff and you'll find your comrades. Indeed there is a lot of very useful information on dpreview also try www.photozone.de . There are plenty of charts on dpreview which you can get off on - shoot. ;) You should know that it if there is a problem it could be the lens, the camera or both. Some wide angles lenses do not focus too well at a distance. Things look sharper in the viewfinder than on a PC screen on which they tend to look soft and even unsaturated and using a wide-angle cheapo lens will make them look even softer and duller. Also, you MUST post process the images to reveal their potential - that is an art in itself. You need some original images from the *same* camera and lens to compare with yours. Try on www.pbase.com use their search function - look for full-size unprocessed images. Stop worrying. It is most likely you simply don't know much yet... |
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
|
©2009 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.