Jump to content

17-35mm L not that sharp?


benstanley

Recommended Posts

<p>I have a new Canon 60D, a 50mm 1.8, a 100mm 2.8L macro, and a 17-35mm 2.8L.</p>

<p>The pictures from the 100mm 2.8L macro are super sharp, blows everything else away.</p>

<p>However the 17-35mm pales in comparison, in all focal lengths, and particularly in the outer third of the frame even though its a cropped sensor. While the centre will be moderately sharp, the outer regions of the photo have a fair bit of aberration - like a stretched sort of blur. I'm wondering if its worth upgrading to something with better optical quality, and what that better something would be? I think I'd prefer not to get a EF-S lens cos part of me expects that maybe in 5 years or so everything will be back to full sensor. The lens would be used for general purpose type stuff, and landscapes. I'd prefer to keep the costs down a bit after just buying the 60d and the 2.8L macro...</p>

<p>Also, am I expecting to much for the camera to handle focusing well on a moving target? For example, trying to get photos of a toddler running towards me using the 100mm macro, and most shots are out of focus. I've tried AI servo and repeadedly refocusing using one-shot focus.</p>

<p>Thanks!</p>

Link to comment
Share on other sites

<p>Ben, it's not surprising that your 100/2.8 L macro is sharper than your zoom. Macro lenses are optimized for sharpness across the frame. Also, normal and telephoto lenses have much less inherent distortion than do wide angle lenses.</p>

<p>It might be helpful if you could post an image or two so your problem can be better diagnosed.</p>

Link to comment
Share on other sites

<p>[[For example, trying to get photos of a toddler running towards me using the 100mm macro, and most shots are out of focus. I've tried AI servo and repeadedly refocusing using one-shot focus.]]</p>

<p>Unless you're pre-focusing at a point where you know the action will be, I'd think that AI Servo would be the only mode you should be using in this situation, preferably with center-focus-point. I can't speak to the AF speed of the new 100mm L, but it strikes me (based only on my experience with the non-L 100mm macro) that it may not be the right tool for the job. But again, no direct experience with the new lens.</p>

<p> </p>

Link to comment
Share on other sites

<p>Keep in mind you're comparing apples and oranges. One is a telephoto prime and the other a wide angle zoom. Primes will generally be sharper than zooms and your zoom, while a nice lens, is a little oudated. Also, are you shooting wide open or stopping down with the zoom? f/2.8 is a pretty wide aperture for a wide lens, the edges will definitely suffer. Try stopping down to f/8 or f/11 and see if that helps. If you are still dissatisfied, I'd consider selling the 17-35mm and putting the money towards a 17-40mm f/4L. I shoot mostly landscapes with my 17-40mm so a 2.8 aperture is not useful for me, and certainly not worth an extra several hundred dollars. </p>
Link to comment
Share on other sites

<p>Particularly at the wide end, you'll need to stop your zoom down a little. Also, softwares like DxO's Optics Pro do a great job of correcting for zoom lenses' geometric errors and vignetting and other distortions. You can download it for free to try it on that lens. I was ready to send my 24-105mm f/4L IS back until I saw how great the corrected images looked.</p>

<p>Of course, focus is critical on a wide-angle lens, but it's not always so apparent throught he viewfinder as it is with a 100mm prime, or any longer lens. The OOF part are more subtle, until you look at it full-screen or in a 100% blow up. So, make sure you and your camera are accurately focusing, but taking some static shots and then compare the edges.</p>

Link to comment
Share on other sites

<p>Ultrawide and zoom are both tricky things to get right, and putting the two together makes it even harder. I've never used the 17-35 but if I recall correctly, it didn't have a reputation quite as good as its predecessor (the 20-35L) or its successor (the first 16-35). I don't mean it had a reputation for being bad, but rather not quite what you'd hope for in an L lens.</p>

 

<p>Based on my experience with the 17-40/4L USM on a 20D, and on knowing how pictures on my 18 MP 7D look compared to how pictures with the same lenses looked on my 20D, that sample image you posted doesn't look good to me. Even wide open, the 17-40 gave me much sharper corners than that (assuming that whatever was in the corner was within <abbr title="depth of field">DOF</abbr>, of course).</p>

 

<p>Is this 17-35 something you just bought (in which case it's obviously used, and probably of unknown history) or a lens you've had for a while which worked better on whatever body you previously had? I suspect it may have had a rough life and is showing the effects. As has been suggested above, the 17-40 is a fine choice if you don't need an f/2.8 lens and wish to stick with full-frame lenses.</p>

 

<p>Autofocus - the AF system in the 60D is pretty good, and with a lens with a good AF design should be able to keep up with moving subjects. Use AI Servo mode if you're shooting something you know is going to move; this way, the camera continuously tracks the subject, whereas in AI Focus it assumes the subject is stationary until it has enough evidence to suggest otherwise. I don't know how good the 100 macro lens is at AF; even with USM, some macro lenses can be slower to focus than some other lenses because their focus hardware is designed with a long focus throw to maximize precision. Some other thoughts on your moving-subject issues:</p>

 

<ul>

<li>How much light is there? AF performance drops when light levels get low.

<li>Are you sure it's a focus issue and not a too-slow shutter speed issue?

<li>Toddlers don't necessarily move smoothly; their motion can be kinda jerky. I'm just guessing here, but this might be part of the problem, too. Before you take the picture (assuming you're using it as an SLR, not in Live View), AF reacts to the subject's motion. Once you fully depress the shutter release, a series of events happens which takes a certain time (the shutter release lag time), and the AF system can't see the subject any more. But it knows what the lag time is (I don't know the specs on that, but I'm guessing 60-70 ms), and it predicts where the subject will be based on its motion prior to that time and instructs the lens to focus on that distance. This works very well if the subject's motion is smooth, but if the subject happens to speed up or slow down (e.g. stumbles), then the camera will have focused on the wrong spot. There's not much you can do about this short of stopping down the lens to increase DOF in the hopes that the subject will be somewhere within DOF.

</ul>

Link to comment
Share on other sites

<p>I presume this is a 100% crop?</p>

<p>It looks to me as though there is a double image. My first thought was that it could be wind blowing the leaves, but your shutter speed is 1/160s, which should be adequate. You're at the extreme of the zoom range where the lens won't be at its best, but f/5 should be OK -- certainly better than you're seeing.</p>

<p>This problem COULD be caused by a very cheap filter. (That could produce a double image in the margins at wider angles.) Are you using a filter, or are you perhaps shooting through a window? (Filters aren't bad, BTW, as long as they're good ones.)</p>

<p>If this is a 100% crop, if a filter isn't to blame, and if you can replicate this problem on a tripod (same focal length and aperture) in a still environment (e.g. interior), I would suspect your lens is somehow out of whack. Canon can fix it for you for probably less money than you might imagine. </p>

Link to comment
Share on other sites

<p>Maybe the 17-35 needs a trip to CAnon for calibration and alignment. I had a 70-200 4L that was whack on the left side and Canon made it sharp across the frame.</p>

Sometimes the light’s all shining on me. Other times I can barely see.

- Robert Hunter

Link to comment
Share on other sites

<p>I purchased a Sigma 15-30 lens a few years ago, because I'd read that Canon had lousy wide-angle optics. It's sharp, and I've been pleased with the results.</p>

<p>But when the 17-40 II f4 L was introduced, I spent about $700 so that I would have all Canon optics. After getting it home and comparing the two lenses, the Sigma's performance summarily wiped that of the new Canon optic in both distortion and sharpness. I was really surprised. </p>

<p>A brief word about this comparison. I used a 14 megapixel camera mounted on a tripod. I placed a line art target against a wide, folding louvered door that encloses our entry room closet. The louvers enabled me to evaluate performance in the corners of each lens. The target enabled me to critically evaluate focus in the center of the field. I made this comparison for several focal lengths and apertures. Each exposure locked up the mirror, waited 5 seconds, and then exposed the image.</p>

<p>As I indicated, the Sigma 15-30 wiped the Canon optic at all focal lengths and apertures. There was really no comparison. So, I returned the Canon optic.</p>

<p> </p>

Link to comment
Share on other sites

<p>Hmmm. I own the EF 17-35 f/2.8L, and although it does have faults (mostly with CA), on a crop camera it should probably be performing better. Here is a 100% crops from mine shot at ISO 500, f5.6, 1/30th handheld. Keep in mind this is on a full frame 5D, so this is the actual corner without a crop. On a 60D you should be in sharper territory. Quality control does seem to have been and issue with the 17-35mm, and there have been many complaints about it, but a good copy really isn't that bad a lens. I tried to turn off all post-processing features in Lightroom for this crop.</p><div>00XQhI-287675584.jpg.338eed8ef8d5e852e028be75acccf1a7.jpg</div>
Link to comment
Share on other sites

<p>For reference, here is the full shot. The area I cropped from is actually starting to leave the plane of focus at f/5.6. Stopping down would have improved this, but I was looking for a shot that closely matched your criteria.</p><div>00XQhR-287677684.jpg.fea1967e84724eb5847644afaba49982.jpg</div>
Link to comment
Share on other sites

<p>Thanks for all the replies. I don't have any filters on the lens. I just shot a whole set at f/2.8 and f/8, tripod mounted, with a range of lenses. In the centre, the 17-35 sat beneath the 50mm/1.8&100mm/2.8L but was definitely better than a couple of kit lenses. However in the corners, the 17-35 was the worst of the lot. On the 35mm zoom it was ok, but 28mm or wider had the bad coners. Devon was yours at 35mm or was it wider? Your samples are definitely much sharper than mine when around 17-28 mm.<br>

Is that something that could really be adjusted somehow, given that the centre was sharp? I think I will take it to canon if you guys think that will help.<br>

ps. I bought the lens new 10 years ago.</p>

Link to comment
Share on other sites

<p>That was right at 17mm.</p>

<p>Like I said, I think I must have a good copy because I have always liked mine. I do, however, need to do a chromatic aberation adjustment on most pictures from it as you can see from the greenish horizontal line in the first crop I posted. I also have to say that I'm strictly a hobby photog. The newer versions have better weather sealing and are generally better optically from what I've heard, but, if you have a decent copy of the 17-35mm (or can fix yours), it's cheaper, lighter, does f/ 2.8 and has L construction, which is good enough for my needs. I also bought mine new around 10 years ago.</p>

<p>In another post I explored the possibility of calibrating this lens and was informed that in Canon's DPP software (comes bundled with Canon cameras, I think) they include a profile for this lens (CA and perspective distortion) much like the latest ones in PS and LR3. Not sure if that helps you, but you might want to check it out. I have never had a lens serviced to that degree, so I don't know what they can do. Good luck!</p>

Link to comment
Share on other sites

<p>Wild brainstorm here. I didn't see a full frame shot with your lens, but are you sure that your auto/manual focus control switch isn't broken in the off position or something like that? Maybe run some test to rule that out?</p>
Link to comment
Share on other sites

<p>Yeh autofocus is working fine, the centre of the frame is quite sharp.</p>

<p>DPP does have a lens profile for this lens. It works really well for chromatic aberration but won't fix blurred corners etc. Otherwise you could get a cheap lens and then turn it into an L series with software :)</p>

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...