Tokina 11-16mm AF problems on 50D

Messages
223
Edit My Images
No
Okay, so my Tokina 11-16mm just arrived. First impressions are that the image quality looks good, but the AF is leaving me confused to say the least.

With all 9 AF points activated the lens focuses at between 0.7 and 1 meter despite the fact that the closest thing in the scene is about 10 meters away and most of it is hundreds of meters away (I'm shooting from my balcony).

Switching to a single AF point seems to result in correct focusing, although it still produces some suspicious figures at times.

Disappointing as both the build and image quality seem excellent. Has anybody else experienced similar problems with this lens on a 50D?
 
I've just been doing some tests and even focusing with specific AF points produces terrible results at times. Take this photo for example. Using AF a focus has been locked on the orange object towards the center of the shot. According to the lens the object is about 1.5 meters away but in actual fact it is at least 50 meters away. Shot at f2.8, hence nothing in the shot is in focus except for the vacant space directly in front of the camera.

focusproblem.jpg


Absolutely awful. To be honest if I decide to keep this lens I can't imagine using AF at all. Without a doubt one of the most disappointing purchases that I've ever made.
 
Last edited:
seems very strange. my lens meters perfectly?
 
seems very strange. my lens meters perfectly?

Yeah, it's bloody annoying. I've tested it plenty since the original post and its AF is substandard under just about all conditions. Take for example close focusing. With my Tamron 17-50mm if an object is 30cm away it always focuses consistently. With this lens it achieves focus lock with unacceptable variations each time.

After trying to focus once on infinity I noticed that the lens achieved a lock at round 0.75 meters! It's like a practical joke in very expensive packaging.

I've just sent an email to the Japanese supplier but my written Japanese isn't what it should be. Hope I get a replacement. Urrrrrgh. I could however just bite the bullet and use it on manual.
 
Could just be a duff copy. If your other lens' are OK then it's not the camera body so you could just be unlucky. I'd send it back.
 
I had a problem like this, but in reverse, with my 24-70. If I focused on anything more than about 20' away the focus ring whizzed all the way off to the infinity stop and focus was rubbish. If I wanted to focus on something further away then the lens was unable to achieve focus at all becuse it was hard up against the infinity stop and wanted/needed to go further.

The problem in my case was that the lens was damaged during a fall onto concrete and required repair with replacement parts. In your case I guess there are a number of possibilities....

- The lens to camera calibration is miles out and maybe you can try AF microadjustment to fix it;
- The lens calibration is miles out and needs warranty correction or replacement.
- The lens has been damaged and needs repair;
- The lens has never been right from the off and needs warranty repair or replacement.
 
Have you tried using the 50D's micro-adjust?

Edit - just seen above!
 
Last edited:
I got one after all the rave reviews and must say i was very disappointed, flare/purple fringing and poor focusing. Just could not get on with it and in the end changed over too canon 10-22 and couldn't be happier.
Sorry not much help but good luck with getting it sorted.
 
Rich, if you've only just received it can't you simply return it? If you want the lens for landscape work then it strikes me that the f/2.8 aperture of the Tokina would be of less use to you than the far wider 10-22mm focal length range offered by the Canon ultra wide.
 
I got one after all the rave reviews and must say i was very disappointed, flare/purple fringing and poor focusing. Just could not get on with it and in the end changed over too canon 10-22 and couldn't be happier.
Sorry not much help but good luck with getting it sorted.

Funny you should say that Neil as the second thing I noticed while checking the focusing issue was the high CAs, particularly on the edges of the shot.

I was pretty close to going for the Canon, but the fact that they don't supply a lens hood as part of the package put me off on principle alone. Add to that the fact that I'm a big fan of my Tamron 17-50mm, which would naturally be complimented by an 11-16mm.

Seriously, I'm so.......off. How the hell did this thing make it through QA.
 
Rich, if you've only just received it can't you simply return it? If you want the lens for landscape work then it strikes me that the f/2.8 aperture of the Tokina would be of less use to you than the far wider 10-22mm focal length range offered by the Canon ultra wide.

Fair point there. I'll reconsider. Recently I did some indoor work - taking photos of bars at night and I realised that f2.8 on a wide lens would be advantageous.
 
Rich,

1) Is this problem exclusive to the Tokina - does it happen on any other lens?

2) Have you got it in continuous AF at all?

I wouldn't write it off yet; I'd return it if there doesn't seem to be a logical reason for it. Like said previously, it could just be a duff copy. It happens unfortunately, although in this day and age, it shouldn't. :)
 
Rich,

1) Is this problem exclusive to the Tokina - does it happen on any other lens?

2) Have you got it in continuous AF at all?

I wouldn't write it off yet; I'd return it if there doesn't seem to be a logical reason for it. Like said previously, it could just be a duff copy. It happens unfortunately, although in this day and age, it shouldn't. :)

Thanks specialman

1) Yep - exclusive to the Tokina. I was swapping it with the Tamron today just to ensure that the camera wasn't at fault (I suppose it could be, but the problem only manifests itself with the Tokina).

2) Continuous AF? Are you talking about my AF mode? I switched to 'one shot' to prevent it from drifting once a lock had been gained, and to narrow down the cause.
 
Thanks specialman

1) Yep - exclusive to the Tokina. I was swapping it with the Tamron today just to ensure that the camera wasn't at fault (I suppose it could be, but the problem only manifests itself with the Tokina).

2) Continuous AF? Are you talking about my AF mode? I switched to 'one shot' to prevent it from drifting once a lock had been gained, and to narrow down the cause.

Sounds like a problem then. I'd get it returned ASAP, you're within your rights, and just get another copy that will, I'm sure, be 100% fine. Is it new or used? :)
 
Okay, I've identified the cause (in a sense at least). I bought the lens directly from the Tokina / Kenkou online store. The price paid was the same as Amazon's price, but Amazon are out of stock so I went to the kenko store.

However, on checking the returns policy I came across this text in small print right at the bottom of the page.

■商品は充分に用意しておりますが、販売してる商品の殆どがB級品ですので、万一品切れや不良発生のため、販売継続不可能となる場合もございます。この場合はご容赦下さい。

It basically says 'Most of the products sold on this site are factory seconds hence supply / quality issues may apply'.

I got my wife to check the site (she's Japanese) and she confirmed that this is not written on either the home page or the main body of the product pages. Very cheeky.

So yes. It's a factory second, and no doubt the reason for it being a second is because the bleedin' auto focus doesn't work.

Where this places me with respect to a refund I don't know, but I'll find out tomorrow.

I still don't understand why Kenko / Tokina would want to sell such a lens, even as a factory second. It can't do any good for their rep.
 
I would have thought factory seconds would only involve minor blemishes and not the technical aspects of what is after all a very technical product. Very shabby to let a lens such as this onto the market no matter how it is described.

Get it replaced. Not fit for purpose etc.
 
I just fired off a few test shots with my one, and it's spot on, even with the distance scale. it's spot on. Although it sounds to me you have found the problem, already. I have had nothing but excellent shots from my copy. I hope you get it sorted, mate!
 
Last edited:
I too have some issues with my 11-16mm, not quite as you have experienced with it reporting incorrect focus distances, although shots should if in excess of 2 metres be showing focus being at infinity.

What I have found is that with the camera wide open - 11mm at F2.8, that although the camera reports an accurate focus lock (AF - using One Centre point AF on a tripod), shots can be blurred, almost consistently in poor light conditions - but shut it down to ~F4 or zoom to 12mm or greater and the situation is improved.

It seems from what you have said the focussing issue is being experienced in good lighting conditions.

But manual focus via Live view always shows very sharp results.................. I had before purchasing the lens, borrowed a copy and that exhibited the same characteristics of poor focusing at 11mm / F2.8 in low light conditions.

I might be writing to Kenko here in the UK to ask if it is a general characteristic of the lens, although as you say you can overcome the situation by using Live view / manual focus, but that is not easy or convenient in all conditions - most people would be taking shots in very good light conditions so may not be aware of any issues, but I do want to get out to do some night city photography.
 
Never ever use ALL AF points on Tokina wide lenses. In fact, that is hardly a good way to focus for any landscape shot.

Stick to the centre point which works extremely well. Manual focus is also a very easy way to get excellent results with a lens that wide.
 
Last edited:
Never ever use ALL AF points on Tokina wide lenses. In fact, that is hardly a good way to focus for any landscape shot.

Stick to the centre point which works extremely well. Manual focus is also a very easy way to get excellent results with a lens that wide.

Well, that's not really good enough. If I point a lens at a scene which contains no close objects at all then I wouldn't expect it to focus at between 1 and 2 meters. Also, in the shot I provided above I used the lowest AF point to focus on the orange object in the center of the shot.

Manual focusing is all well and good, but with a wide lens such as this it's quite difficult to focus manually on objects which are some distance away.

If a lens offers AF then it should work to a reasonable degree of accuracy. I'd expect far more than a simple random number generator. Yes, I do use manual focusing for landscape shots most of the time, but I still felt the need to test the AF function, and I'm glad that I did.

Oh, and on another point. I'm pretty sure that if you choose a single AF point on most Canon cameras the exposure is weighted around that point irrespective of the metering options.

What's with the bold type and underlining? People are perfectly able to digest your posts without the need for you to highlight certain parts.
 
Last edited:
I'm pretty sure that if you choose a single AF point on most Canon cameras the exposure is weighted around that point irrespective of the metering options.

On the 50D that is true only for the Evaluative metering pattern. Demonstration in this post - http://www.talkphotography.co.uk/forums/showpost.php?p=2618734&postcount=21. Spot, Partial and Centre Weighted Average are not influenced in any way by the focus point used.

On 1 series bodies you can have the Spot meter follow the active focus point, but Partial and CWA metering will still remain centred around the centre of the frame. All the other bodies operate as per the 50D.
 
Last edited:
On the 50D that is true only for the Evaluative metering pattern. Demonstration in this post - http://www.talkphotography.co.uk/forums/showpost.php?p=2618734&postcount=21. Spot, Partial and Centre Weighted Average are not influenced in any way by the focus point used.

On 1 series bodies you can have the Spot meter follow the active focus point, but Partial and CWA metering will still remain centred around the centre of the frame. All the other bodies operate as per the 50D.

Yep, I thought that that was the case after some early experiementation with the 50D. It's a feature which I don't really like to be honest as it can result in wild exposures at times. Clearing when taking a photo like the one at the top of this thread, where all objects are a fair distance away, I wouldn't want the camera to modify the exposure significantly based on the object which happens to be in front of the AF point.

Still waiting for a response from the Kenko shop.
 
I've been out with the lens this afternoon - this time using only manual focus. Mixed bag to be honest. I terms of pure resolution it seems to be very good, but CAs are high (up to almost 5/6 pixels on my 50D on some shots). For some reason all of the shots required positive exposure compensation to prevent the clipping of shadows...strange as if anything I've found the 50D to overexpose up until now.

The whole focusing story gets even stranger. If I focus manually on distant objects using the viewfinder and then check the focal distance it always seems to be similar to what the AF system selects - a distance of between 1 and 2 meters.

Given this info you may initially think that it would be safe to conclude that the printed scale on the lens is incorrect, however, when the shot is actually taken at a test aperature of f2.8 the distant objects are completely out of focus every time.

Take for example the orange object in the middle of the picture at the top of this post. Using manual focusing, if I view the object through the viewfinder it appears to be in focus at a focal distance of about 1.5 meters, however, take the shot and the picture is out of focus as any depth of field occupies a region not far from the front of the lens (as you'd expect for such a short focal distance).

It's almost as if the information that my eyes and the AF system are getting differs to that which the sensor is receiving. However...when focusing on relatively close objects, such as one which 2 meters away, the lens will usually correctly select a focal distance of 2 meters. In such cases both the resultant along along with that viewed via the viewfinder are sharp.

So lets just recap. For a distant object the lens selects a focal distance of 1.5 meters, and yet for a object 2 meters away it selects 2 meters. In both cases the objects looks sharp via the viewfinder, but only in the second is the object in focus.

Essentially there seems to be a variety of seemingly contradictory things going on here, none of which I can really piece together. The whole focusing thing is just......nuts, but all suggestions / explanations welcome.

I'll try with live view next.

By the way, I'm not new to manual focusing. I have at least 2 years of experience under my belt so this isn't some schoolboy error. I wouldn't even know how to produce such results accidentally.

Kenko have already agreed to refund it, including postage.
 
This won't really help much, but it might go a little way towards explaining some of the potential problems with focus accuracy in general. However it won't explain the bizarre behaviour you are seeing. I think only Tokina can do that.

Anyway, once the light passes through the lens and into the camera there are three independent optical paths that the light will follow. The main mirror is half silvered, relecting some light up onto the focusing screen and allowing some light to pass straight through, where it hits another mirror behind the main one and directs the light into the bottom of the camera where the AF sensors lie. When you actually take the shot the main mirror lifts and the light passes straight through to the sensor. The issues with focusing performance arise when, for whatever reason, the distances involved in these three separate paths end up out of calibration. What looks sharply focused to you in the eyepiece might not look to be in sharp focus to the AF system. Whether or not those two agree, it doesn't automatically follow that they will match up with the position of the sensor in terms of distance from the lens. The fact that they usually do is something of an engineering marvel, especially when you consider the moving parts involved and the thunk of the mirror belting up and down at up to 6.3 FPS. If the camera takes a nasty bump or the focus screen somehow gets jogged out of position then there is a good chance that something will get thrown out of whack somewhere.
 
Thanks very much. I was planning on researching how AF systems work over the next few days. Yes, I guess that these three distances could be out of sync.

You were right about the Canon by the way - a couple of times today I wished that I could shoot at 10mm.

By the way, in my DPP photo viewer all the image thumbnails which have been generated by the Tokina have a small picture of a lens in their top right hand corner. I'm sure that I've never seen this before. Does it mean that a calibration is being applied, or is required? Searching online for info...
 
By the way, in my DPP photo viewer all the image thumbnails which have been generated by the Tokina have a small picture of a lens in their top right hand corner. I'm sure that I've never seen this before. Does it mean that a calibration is being applied, or is required? Searching online for info...

I suspect it'll be something on the third tab of the editing panels - to do with lens aberration correction.
 
A shot which I did today with crops. Test shot is of the Toujin Grave on the other side of the island. ISO 100. f9.0

5277041093_43fdb18ffb_b.jpg


Japanese text (lower middle) seems fairly sharp:

5277040823_119b23c2f9.jpg


Crop of the bird (upper center). Again pretty good.

5277040923_977d65a416.jpg


Left of the image seems to be weaker. High CAs and soft image here (note the right side of the rider, under his arm).

5277040951_00e3b75c0a.jpg


So overally not so good. Lens with sub-functional AF which underexposes, exhibits high CAs and offers only slightly over average sharpness (and I may be being kind on this point).

http://www.flickr.com/
 
Last edited:
Back
Top