HELP - Canon 24-105 Back Focus

MJ Stebbings

MJ Stebbings

Suspended / Banned
Messages
2,444
Name
Mike
Edit My Images
No
Hmmm, I have a Canon 24-105 F4L, bought off here from someone, and it had slight back focusing issues. So, Sent it off to Lehmans for repair. Its come back, and I've shot with it and its still back focusing.
They said they reset the focus calibration and tested, however I've tried it on both my 50D and 30D and its the same with both. At first I thought it was my 50D but the shots I take with the 70-200 F4L are lovely and sharp, and the ones from the Sigma 24-70 are nice too, so it surely cant be my 50D

Cant anyone give me any advice? Should I just contact Lehmans and tell them its still back focusing?

Mike
 
Your 50D supports AF microadjustment, use it. This is exactly what it was invented for

Otherwise, send both lens and body for calibration
 
Are you in One Shot or AI Servo focus? I ask because the 24-105 seems to have a little glitch when using AI Servo on static (or near static) subjects. The focus goes slightly backwards and may well be what you're seeing but this isn't the case in One Shot though.

Bob
 
Last edited:
Send it back and tell them you are still not happy with the adjustment. If all your other lenses are correct on two bodies, they haven't fixed the lens properly.
 
Canon Bob said:
Are you in One Shot or AI Servo focus? I ask because the 24-105 seems to have a little glitch when using AI Servo on static (or near static) subjects. The focus goes slightly backwards and may well be what you're seeing but this isn't the case in One Shot though.

Bob
Click to expand...

Hmmm, not sure actually. I usually use one shot for most of my stuff.

Does anyone have any JPEGS, converted straight from RAW with no sharpening added, taken wide at 24mm. It would seem that the problem is worse at the wide end.

Obviously if i'm not happy with the lens calibration its going straight back to Lehamans for them to recheck it because if i'm honest, it doesnt look any better than it did when I sent it off
 
anyone?
 
Big help mate

Sending mine back to Lehmans. Not happy at the moment and sent them a service note saying so.

hopefully they will come back to me and get the problem sorted. I dont care if its a case of something wrong with the lens and it gets repaired, I'm just annoyed that its cost me £76 to be the same as it was before
 
Glad it was useful :)
 
Big help mate, thanks
 
arad85 said:
This do you? f4, 24mm convert RAW->JPEG and then upload after import into lightroom. No other processing applied (apart from LR default sharpen of 25). This is on a 5D2, so full frame as opposed to crop....

http://www.arad85.co.uk/hosted/talkp/_9006.jpg
Click to expand...

Yikes !
I'm sure my 24-105 is better than that at 24mm f4 :suspect:

I searched all my images and it seems I've mostly only kept the shots taken at 24mm f4 in low light at ISO 3200 hand held complete with camera shake.
There's only one taken wide open in good(ish) light that I've kept.
This One - I couldn't work out how to get a direct link to the original; the link goes to a reduced size which will fit your screen; then click 'O' at the top to view the original size (warning, over 10Mb).
Focus was on the tree, not the slate in the distance, but you still get a feel for sharpness and chromatic aberration.
There seems to be a fair bit of difference compared to Arad's.
Thoughts please!

Note - there is a fair bit of batch variability with the 24-105s and mine was initially much worse than Arad's.
I sent mine in to Canon Elstree who did a cracking job and my 24-105 is now very satisfactory.

Note that to get sharp pixels at 100% on the 5DII takes good technique, not just good glass. I wrote at length about this in another thread.
 
I have to say it's not my best image... Mine has already been to Elstree and came back much happier with it. Have a look here:

http://www.arad85.co.uk/hosted/talkp/_9233.jpg (focus point on axle of cart)

or


http://www.arad85.co.uk/hosted/talkp/_8613.jpg (focus point on wall, although I'm not square onto it)

It may be that my lens is slightly out at infinity focus, but I tend to think atmospherics can get in the way of a shot that's covering many miles. Perhaps that's just me making excuses though.
 
I also don't normally shoot wide open at the lens extremes either.... unless I'm after a DoF phenomenon. No idea why I shot the canyon/building at f4... The shots either side of those are at F8 and f10 for both the building and the canyon. Must have knocked the aperture down to f4 - there is no need as the canyon shot (for example) is at 1/800th at ISO 100 and 24mm....
 
arad85 said:
I also don't normally shoot wide open at the lens extremes either.... unless I'm after a DoF phenomenon. No idea why I shot the canyon/building at f4... The shots either side of those are at F8 and f10 for both the building and the canyon. Must have knocked the aperture down to f4 - there is no need as the canyon shot (for example) is at 1/800th at ISO 100 and 24mm....
Click to expand...

I know why I took this one at f4 - driver error :bonk:
It was f4 because I had been photographing in some tunnels and forgot to change it.
I only kept that shot to show someone what the interior of that quarry looked like.
Normally I get rid of my mistakes :D
 
Are these Sharpened finals?
 
MJ Stebbings said:
Are these Sharpened finals?
Click to expand...
Import to LR3 -> Export JPEG full size at 70 quality for mine. That means default LR sharpening of 25. No other processing...
 
arad85 said:
Import to LR3 -> Export JPEG full size at 70 quality for mine. That means default LR sharpening of 25. No other processing...
Click to expand...

Very similar - my save full size JPEG Export is normally for transferring an image so emphasis is on quality.
Settings are JPEG quality 90%, sharpening says "Sharpening Standard, for Screen" which is the default.
 
DuncanDisorderly said:
sharpening says "Sharpening Standard, for Screen" which is the default.
Click to expand...
Yes, I forgot that. Standard sharpening for screen here too.

The impression I get is that our photos are significantly sharper than Mikes. Is this the case Mike?
 
Yes indeed, as well as the back focusing issue.

Here is a sample shot with the 24-105:

sample-0001.jpg


And here is a crop of the image, you can see the focus is on the chair and not where I put it on the couple

sample-0001a.jpg


These are RAW files, converted straight to JPEG from lightroom with the standard 25 sharpening etc.
These are taken on the following:
ISO 800 or 1600 (cant remember) 24mm @ F8 (Can check settings later, was meant to post this up last night but fell asleep on the sofa :lol: )
 
I don't think I would make any judgement based on that image.
The faces and sofa both look soft.
This level of softness at 100% would be typical of my 60D at high ISO, regardless of lens.
The reason I had to hunt for an image to share is that my 5DII images look soft at high ISO too!

Need an image taken in good light at ISO 100 with plenty of fine detail and a clearly identifiable focus point.
I linked earlier to another thread about lens testing; it has to be done methodically to be meaningful.
 
You're right, probably not the best image to judge on. Should have put up one of the group shots taken outside in natural light. The building behind them is more in focus than the group. Will try and rembember to post a sample up tonight..... should I manage to stay awake :lol:

I have just got off the phone to H Lehmanns, and have a list of things to do to test the lens, and then send them off the JPEGS. I'm doing tests on both the 50D and 30D. So, will see what happens then
 
Gunna have to do the tests at the weekend. Damn weather and short days :lol:
 
Right, here we go. ISO 400, F11, 1/200 @ 34mm

Full Original

and cropped:

Cropped Image

Now, is it just me or is the wall more in focus?
 
Last edited:
MJ Stebbings said:
Now, is it just me or is the wall more in focus?
Click to expand...
Umm... I have no idea what the camera is focusing on. It certainly doesn't appear to be the couple in any shape or form. The wall looks to be more in focus, but it still doesn't look right. I assume the crop is a 100% crop...

Have you tried microadjusting the lens? I ended up taking a set of shots on mine at various focal lengths, various apertures and various microadjusts before fixing on +3 I think. I got frustrated with the tripod method and ended up hand holding on near, middle and far distance objects in good light and selecting the best compromise. I had to do this a few times to really understand what the lens was doing...
 
going to do some serious testing this weekend, compare results against the Sigma 24-70 and the 70-200 at relevant focal lengths. and I will set everything up as instructed by the people at Lehmanns and see what the results come out as. I will also test it out on both the 50D and 30D to make sure its not the camera.
 
Just an update.

After doing some tests, it would appear that the lens works better on the 30d than the 50d. The test images have been submitted to the engineer who originally serviced the lens and agrees there is a problem. He also looked at an image that was taken with my 50mm to show thats its not the camera on its own.
He has asked me to send the 50d and the 24-105 in for caibration together. If its the lens, then its FOC, if its the camera on its own, its chargeable (which is fair enough).
I have a shoot this Sunday, so will be sending the 50d and the lens off Monday. They also agreed to turn it around as soon as possible too, which is nice :)
 
Good news Mike. Let us know what happens.
 
Will do mate! I might have to post samples up later to show the tests, if I can bbothered :lol:
 
DuncanDisorderly said:
Yes please!
Click to expand...

Sorry bud, never did, couldnt be bothered :lol:

Anyway, sent the camera and lens in yesterday, it arrived today, they have done tests, the lens is fine but the camera is back focusing with my lens on AND other lenses they have tried on it. They are going to sort that out for the sum of £72. It should be in the post back to me tomorrow to arrive Friday.

If this is true, H Lehmanns are legends. I've had nothing but great communications from them and been kept up to date with everything.

Fantastic service, and I really hope that my camera comes back working fine. I thought that everything looked ok, Until I did a shoot Sunday with the 50mm F1.8 and thought the focus wasnt quite sharp enough and thought it was either the camera, or more than likely user error :lol: at least its not the latter :D
 
No worries :thumbs:
Fingers crossed they are as sharp as you hope.

I've been very happy with my 5DII, but it went for a paddle in 2 inches of water and has been written off, so I'm now keeping my fingers crossed that the replacement is as good as my old one.
 
All this testing gets me worried.

How on earth is the average Joe meant to know if he's got a dud lens/body? He'll just think the camera/lens is poop and sell it, or he'll think it's his technique, or not notice at all.

That image of the canyon has some serious vignetting from what I can see. I wouldn't be happy with that.

Anyway, has anyone got a definitive guide to testing your body/lens?
 
tomah said:
All this testing gets me worried.

How on earth is the average Joe meant to know if he's got a dud lens/body? He'll just think the camera/lens is poop and sell it, or he'll think it's his technique, or not notice at all.
Click to expand...
The average Joe doesn't buy £2.2k's worth of camera/lens and then micro analyse it. The average Joe buys a compact and shoots with it and is very happy - including all the whizzy processing they apply to make the image pop and correct for the lens defects.

tomah said:
That image of the canyon has some serious vignetting from what I can see. I wouldn't be happy with that.
Click to expand...
It's a raw, straight out of camera. No processing, no nothing applied - a camera processed JPEG would be much brighter, sharper and "poppier" and - if you have lens correction switched on - much less vignetting.

ALL lenses vignette (fundamental law of physics) and it is worse when they are wide open and at their wide end for a zoom. It's one of the design tradeoffs that enable you to buy the lens so that it covers a full frame sensor and is still affordable and portable and has a wide usable range. Easily correctable in post though - just click a tick box ;)

tomah said:
Anyway, has anyone got a definitive guide to testing your body/lens?
Click to expand...
My method is to take photos of a number of different things at different distances and zoom settings and see how many photos have the focus nailed. Change microadjust and repeat. Boring, but I get to know the lens then.
 
PS. This is what the canyon image looks like with a couple of quick clicks in LR (lens profile correction, bit of fill light, apply landscape as the camera profile and sharpen it up a bit). If you look closely (i.e. at 100%) at the clouds on the top left, you'll see some chromatic distortion corrected too.

http://www.arad85.co.uk/hosted/talkp/IMG_9006.jpg
 
Last edited:
tomah said:
All this testing gets me worried.
Click to expand...

Cant say there has been a lot of testing :lol: I did two test shoots, sent the results to Lehmanns and they agreed something was wrong so asked for the camera and the lens to go in.
Now, it turns out the bodies af was wrong, so they have sorted it for me (touch wood) and hopefully it will be back with me tomorrow.
They could have been the result of taking a bump or bang. I'm not sure as I've been happy with the results up until recently, but I do lots of shoots and could be anything from a bump when shooting a wedding, a live music event, helping a model up or down from a strange location etc. Kit gets bumped and with the amount of bits and whirly things inside (yes, thats the technical term) things can just drop out of place etc. Its not a design fault (Like the early 1DmkIII focus issues we all know of), its just my camera :)

And with regards to the average Joe, I think Andy summed it up perfectly. The average Joe doesnt go and spend all this cash on getting gear unless they A) know what they are doing or B) have more money than sense. If its the latter, they will buy new, if they arent happy with the results they will take it back.
 
You must log in or register to reply here.
Back
Top