Hello, Always a good way to start out on a forum with some technical issues! Apologies for the lengthy post, but I always find it's better to provide too much information rather than too little. I recently purchase a used EF 70-200mm f/2.8L IS III USM as an upgrade to my Canon EF 70-200mm f/4L USM, which I have been using for the past 14-15 years. The f/4 is a great lens for the price, and I've been extremely happy with the image quality and focus performance throughout the time I've been using it - it's just now time for a faster lens with IS, and one that I can use with an extender. The reason I mention the old lens is because with it, I've never had the issue I'm now having with the f/2.8L. Now from what I've read when deciding which lens to upgrade to, the f/2.8L is pretty much universally praised, and I haven't seen much in the line of issues, unless someone has had a defective unit. What I've found with mine is that no matter what combination of camera settings I use, it can be very inconsistent when it comes to (I believe) focusing - that is, sometimes it's pin sharp, and sometimes appears to have focused, but actually upon review, seems to be ever so slightly out. Not enough for the camera to actually recognise it has slipped out of focus and make an adjustment, as it might in AI Focus and AI Servo mode. Bearing in mind I've been experimenting in all three focus modes, including One Shot. I mentioned above that I bought the lens with the idea of also using it, at times, with an extender, in my case the EF 2x III. I know reviews about extenders are a mixed bag, especially the 2x, so I did consider that maybe this was affecting focus performance and even that it was degrading IQ to the point that it looked totally out of focus, but again, it's inconsistent, where sometimes it'll be sharp (or as sharp as you can get with an extender on) and sometimes blurry to the point of being unrecoverable. I have also tested with the extender both on and off, and am still getting similar results. I've specified as such with the images posted below. I have also considered motion blur, but believe me, I've had my fair share of images that I've had to throw out as a result of that! I don't believe it is down to this, as I have tested this on static objects, as well as the fast moving aircraft I've also showed as examples. As it's used, you may ask about the condition of the lens - a few cosmetic scratches but nothing jumps out at me as being damaged, with the glass, mount and contacts all appearing to be in good condition. And a lot of the time it functions perfectly, other than the issue I describe. All images below are straight from the camera with no post-processing, showing a full frame and 100% crop of each one. I've included comparison images from the same sequence so that the lighting, subject and settings are similar, if not identical, and so you can see the difference between a "sharp" image (I understand it won't be razor sharp with the extender on!) and the out of focus image. I have also checked the focus points on the EXIF data and all seem to be suggesting they're focused on the subject and indeed that the camera has recognised it as being focused, although I understand that if the subject has shifted in the frame, then this may be a bit misleading. However, in all of these images, I can't see anything to suggest that is has focused elsewhere, or why it would even try to (a blank, featureless sky, for instance). General settings / techniques I'm unsure as to whether these have been recorded by the EXIF data, so I'm unable to confirm 100%, but I believe these settings were true for the example images: - AI Servo focusing. - Continuous shooting, however generally taken as focus > shoot > refocus > shoot. - IS on, setting 1 for static objects, setting 2 for moving objects. - Unsure as to whether these particular images were in Manual (1pt), Zone AF or Auto (19pt), but I've tried with all 3 settings in various scenarios and none seem to make any difference. Image 1 (out of focus) Settings: TV 1/1000 AV 6.3 ISO 200 200mm + 2x extender (400mm) Image 2 (in focus) Settings: TV 1/1000 AV 6.3 ISO 200 200mm + 2x extender (400mm) Image 3 (out of focus) Settings: TV 1/500 AV 5.6 ISO 500 200mm (no extender) Image 4 (in focus) Settings: TV 1/500 AV 5.6 ISO 400 200mm (no extender) Any thoughts would be appreciated, ranging from "your lens is broken" to "you're an idiot who can't use a camera"! Happy to provide any more info required. Thanks.
Well, for the second situation your focus was primarily picking up the tree, were you on Servo during this? Generally I shoot 1 shot, with a specific focus point. I use continuous if i plan to rapid shot and try to make sure I get a a specific tight focus on a moving subject, and I use servo to track that moving subject. Servo i think may have not helped on the second situation because it was waiting for movement, maybe why the head up was good, hard to determine if all this checks out as i suspect. When you say focus shoot, refocus shoot, are you holding halfway down the focus then completing the press to shoot? You should be for it to hold focus and track properly. That said focus will not be perfect on a long lens like this when tracking moving objects so That's where servo with rapid firing a few shots of the planes is a good approach. The bird is a more questionable situation as you can remove the servo tracking and shoot a high shutter and catch him as long as focus was targeting him and not front focusing slightly favoring on the tree, sometimes all the extra focus points don't help
Thank you for your response! Again, I understand the focus points recorded may be misleading if the subject has shifted in the frame, but I did take a fairly consistently framed set of photos of the bird, as we were both stationary, so would not have expected the focus to shift away as much as it might do with a fast moving object that you're having a hard time keeping where you want it in the frame. The below images show where the focus points were. It obviously tracked the bird slightly better for the second photo, but even the part of the log that it has supposedly focused on is also nowhere near in focus on the first photo. Even on these small crops you can see how much sharper the whole log is in the second photo. If it were only busier images such as this one where I was having the problem, then I'd probably also be leaning towards the same thought, that it was simply focusing elsewhere, but it's happening even with the planes, providing a high contrast subject against a plain blue sky. I'm not sure what else the camera/lens has for it to try to focus on! Another thing to point out, and please correct me if these calculations are incorrect, but at 200mm, f/5.6 and a distance of 22m, there should be a DoF of 2.59m - far wider than the bird, and even the log, so even if it focused a little short, I wouldn't have expected the bird to be as out of focus as it is. I believe it was on AI Servo, yes. Some interesting thoughts regarding this mode, but I have tested with both One Shot and AI Focus which yielded similar, inconsistent results. I'm with you on the point that One Shot may be more suited to what I'm doing, though. In terms of the technique - it'll be half-press to focus, listen for the beep to confirm the camera thinks it's focused, and watch for the focus points used (if in Zone AF or Auto, single point speaks for itself). I'll never take a photo if it hasn't focused where I intended. Then a full press of the shutter, full release and repeat.
Welcome to the forum, Adam, with the out of focus image of the pheasant, I have a similar problem with my 60D with my 75 - 300 mm lens, I regularly resort to putting it in live view and magnifying the image to check that it is in fine focus. I have a newish lens and gave the old one away.
Use single point focus so that you choose what you want the camera to focus on, if you let the camera choose then you will get frequent errors.
Again, thank you for the responses. I've tested it once more with a passing aircraft, so the subject is bold against a featureless background, using single point focus as suggested. Apologies for the poor lighting, but this is the UK after all! These have been shot wide open so I could get the highest shutter speed possible, to eliminate the possibility of motion blur. The following four images (100% crop) were taken within the span of a few seconds, each time the lens 'focused' instantly, with no sign of hunting or difficulty in focusing in the lighting conditions. In the fourth image, the aircraft has shifted away slightly from the focus point, so it's not visible in the cropped image, but still remains in focus. All shot at 1/4000, f/2.8. Image 1 Image 2 Image 3 Image 4 Images 1 and 4 aren't exactly razor sharp, owing to the poor lighting, but I would certainly suggest they have focused, and it had no trouble doing so. Images 2 and 3 however have lost any semblance of detail, and you can see the halo effect around the entire plane, which just simply isn't there in 1 and 4. Compare the markings on the tail between images 1 and 2 and you can see what has been lost. I also tested the focus if I aimed away from the plane, to see whether the lens would pick up on something else to try and focus on. All it did was hunt and fail to find anything, so again, there should be nothing in these images to confuse it - there's a plane, and that's it. It's this inconsistency that's confusing me. I have shot using the aforementioned 70-200 f/4 and a 400mm f/5.6L, across the 750D and a 400D, and have never ended up with results like this. I apologise if it sounds at all like I'm suggesting "it can't be me/my skills, it must be the equipment", but it's only really the equipment that's changed, so I'm just trying to see if anything is fundamentally different from what I've been used to in the past.
I think you need to show that this problem exits on a non moving subject before you can diagnose any further, find something a fair distance away take one shot then take a shot of something close up before taking another of the distant scene, that way it will have to refocus for each shot, repeat this a few times. It could be that your lens/body combination is focusing just on the edge of the DOF lomit and needs an autofocus micro adjustment which I don't believe you can do with either the 400D or the 750D
You're right, the 750D doesn't have that feature, there were a few suggestions about that when previously Googling the problem so I did look into it. I'm away today but as soon as I can I'll get a few more test shots on a static subject. Anecdotally, I don't recall having much of an issue with closer subjects, so it does seem like this is far more of a problem with distant objects (if you can call the pheasant at 22m 'distant'). I'll maybe try it at a few different ranges.
I guess the first thing I would do is conform focus at both end of the lens, you can use charts/ruler method to see if its back or front focusing. It could be very likely that the issue is cropping up at the long end of the lens. You can also look at some of the settings like IS and see if they are impacting performance, if your shooting fast enough shutter, you don't need IS, and in fact if your using a tripod it will cause blur instead of helping. All this said, maybe if this is a new lens you did get a bad copy and you should have canon look at it. If the 750 is doing well with other lenses i doubt its the camera, but even those can start to have issues over the years.
Right, as suggested by Johnsey, I got myself some charts and it's produced some curious (but not entirely unexpected) results... All images taken on a tripod with a 2 second delay, f2.8, 1/500, ISO 100. 100% crop. All focused on the centre part of the card, folded vertically. Image 1 - 70mm, IS off. Image 2 - 70mm, IS off (same as image 1) Observations - image 1 doesn't look too far off, but image 2 appears to be front focused. Image 3 - 200mm, IS off. Image 4 - 200mm, IS on. Image 5 - 200mm, IS on (same as image 5) Observations - image 3 starts as being front focused with IS off, as at 70mm, image 4 looks about right and then strangely, image 5 seems to switch to being back focused with IS on! Obviously I've had to cherry pick after taking around 70 test images, but it does seem like the front focusing is the general theme. I find it interesting that the IS could be having an effect and may be where the inconsistency is coming from?
Few things with the test, Im not clear about the vertical fold. the idea is to hit the ruler at an angle, which i think it is. So the folding comment confuses me, and all this is so close i cant really what all is happening and what the thing on the right hand is. It does seem like your front focusing, this however is a very narrow window so any shake with be very off. Also IS shouldn't be used in general when shooting on tripod as it will introduce shake to the image, IS vibrates the lens internals to counteract handheld movement, it destabilizes the lens when on a tripod and will introduce inconsistency in your shots.
A few more tests without IS should confirm its front focusing and I think you should contact canon to get it repaired under warranty if you cant return it to the store.
Sorry I wasn't entirely clear! This one came with a section that folds vertically, which pivots around the 0cm mark, so in theory if you focus on the flat surface in line with 0cm, you should be focusing at exactly the right distance. If you try to focus on the slanted part then it can introduce the possibility of the camera being aimed low/high, so I understand why this has been done. See below! I'll try what you said and get some more tests, only IS off. I may also try it with the moving subjects I was having the most trouble with, and see whether the IS was potentially introducing an error on certain images. All of the images I used as examples on the aircraft/birds had IS on, so it's worth exploring as a possibility.
Yeah I'm at a loss with this one. It's all over the place! Settings: 1/4000 f/2.8 ISO 100, IS off 200mm focal length On a tripod, 2 second delay in the absence of a remote shutter. All focused on the QR code with a single focus point, one shot AF. I've uploaded 11 full size images (6000x4000) which will go over the forum size limits, so I've used a Dropbox link. Let me know if you have any trouble with permissions/access. https://www.dropbox.com/sh/fgx6qn7fjwg59re/AAAK0hVpz12jY1kbIRiTua0fa?dl=0 Some seem to be ok, some front focused, some back focused. Look at the difference between IMG_1901 (in focus) and IMG_1905 (back focused) around the DSLRKIT text and QR code. This is exactly what I'm seeing in the difference in image quality/blurriness in normal shooting. Again, if it was consistently bad I'd be able to draw my own conclusions, but the inconsistency is what's getting me. What next!?
do you have a return policy on the purchase? Otherwise you need to have canon look at the lens, seems there must be some drift in its focus motor.
Unfortunately not as this is a used lens, but I'm happy to pay for repair if required. May have to cough up as I can't think of any other reason why it would be so on and off without a physical fault.
i wasnt sure what kind of return policy it had even being used. But yes, call canon, they may have it shipped in and looked at i would just be clear at the level of testing and that it seems to both front and back focus so it seems very odd. You can refer with the tech on the phone you tried working through as far as testing. They wont estimate till they get it in house and confirm they can reproduce, but I would expect it to be mildly spendy as it probably is more than an hour of labor plus parts. Still if it is able to be fixed it will be cheaper than buying new, your in a tough spot now.
I'm away today but as soon as I can I'll get a few more test shots on a static subject. Anecdotally, I don't recall having much of an issue with closer subjects, so it does seem like this is far more of a problem with distant objects (if you can call the pheasant at 22m 'distant'). I'll maybe try it at a few different ranges.
I would say that pheasant is pretty telling, it and the log are large enough to have the focus nail it and be sharp, so unless the blur is from hand holding it at 1/60 with no IS I think you can safely say there is at least an issue at the long end, and I would expect tack sharp performance from this lens, the 70-200s are known for being spot on and very sharp.
