70D focus issue & blurred pictures

meaton wrote in post #17077199
I disagree, I know my camera is suffering from the problem because I have just tested it with many shots and they are consistently front focussing.

Except the problem is described as the camera inconsistently focusing. Sometimes front, sometimes back.

If it is consistently front focusing, that sounds more like it just needs MFA.
Everyone of my lenses has required MFA on the 70D.

@ meaton :
These are the standard troubleshooting steps to follow when you have an AF issue try them out because there can be multiple causes :
http://www.bythom.com/​autofocus.htm

In my case the camera showed an issue with the point N°2 : "Do you know where they are ?". I found that the focus sensitive areas aren't well aligned with the marks visible in the viewfinder and in Canon's software DPP.

The links below are what I learned on my camera, amongst those links there are the steps to measure the position of AF area :
https://photography-on-the.net …?p=16947213&pos​tcount=719
https://photography-on-the.net …?p=17011802&pos​tcount=890
https://photography-on-the.net …p=17068543&post​count=1128
https://photography-on-the.net …p=17049246&post​count=1097
https://photography-on-the.net …p=17055295&post​count=1105
https://photography-on-the.net …p=17046799&post​count=1088
https://photography-on-the.net …380558&highligh​t=70d+poll


If you don't want to do the complete set of measurements, can you at least try :
- printing this picture (http://i1262.photobuck​et.com …eaTest-01_zps4366f833.png) on photo paper. Careful, you'll need flat colors, if your camera can focus on the black or on the white of the paper (because of the grain of the paper, or imperfections in the print), you'll need to use something else, like a piece of electric tape on some white plastic.
- Set up your camera on a tripod pointing to the ground like shown below with my 60D (of course, use a lens who can focus at the distance you're using).
- Select one of the focus points N°2, 3, 4, 13, 14, 15 (in the picture below, the green one is N°3)
- Switch to LiveView and select the "Quick AF" mode. This way, the camera will display the ideal position for the AF point you selected before going in LiveView and will use the same focusing method as through the viewfinder. This will remove any possible misalignment you can have between : the ideal marks' position (visible in LIveView), and the focus screen marks' position (visible through the viewfinder).
- Still in LiveView, align the black/white edge with the center of the selected mark (like shown below), and press AF-on button to perform autofocus (don't forget to turn on the back button autofocus in the menus). You can also simply half press the shutter button to perform AF if you don't want to change anything in the menus.
- The mirror will flip and try to focus on the black/white edge.
- If your camera hunts back and forth and can't lock focus (red rectangle), you might have the same issue as mine.

If you encounter a similar problem as mine, send your camera to repair with a letter explaining the issue :
- show tests performed with focus chart http://www.dphotojourn​al.com/focus-test-chart.pdf
- No effect from MFA values
- problem to lock focus in the middle of focus point N°2, 3, 4, etc... see my post to spot possible focus points with issue https://photography-on-the.net …p=17046799&post​count=1088)

You might need up to three trips and 7 weeks of immobilization to get your camera fixed.

CAREFUL : the goal of this last test isn't to get good focus (at least for those with an issue) ! It's to see if the camera can lock focus or not !

By "lock focus" I'm saying that the camera thinks it has achieved focus (by emitting a "bip" if you didn't disable the sound).


Note : the focus points I cited (N°2, 3, 4, 13, 14, 15) are those where my camera can't focus, your camera might have an issue but with other focus points. In the test procedure above, I just supposed that each faulty camera will have an identical defect. Normally the focus point numbers I cited above should be able to lock focus anywhere in the rectangle visible in LiveView (if I suppose that the 70D has the same AF sensor as a 7D http://www.aerien.ch …Nature_et_Oisea​ux.php%29.). So see if you can lock focus anywhere inside the marks visible in LiveView except for focus points N°1, 8, 9, 11, 12, 19, because those points might not be fully covered.

lozw wrote in post #17077368
Except the problem is described as the camera inconsistently focusing. Sometimes front, sometimes back.

If it is consistently front focusing, that sounds more like it just needs MFA.
Everyone of my lenses has required MFA on the 70D.

Not necessarily, in my case I often had front focus when aiming at subjects far enough. I tried to MFA but couldn't get it spot on because the resulting focus depended on the kind of subject I was aiming at :
https://photography-on-the.net …p=17068543&post​count=1128

Molybdo42 wrote in post #17077451
@ meaton :
These are the standard troubleshooting steps to follow when you have an AF issue try them out because there can be multiple causes :
http://www.bythom.com/​autofocus.htm

In my case the camera showed an issue with the point N°2 : "Do you know where they are ?". I found that the focus sensitive areas aren't well aligned with the marks visible in the viewfinder and in Canon's software DPP.

The links below are what I learned on my camera, amongst those links there are the steps to measure the position of AF area :
https://photography-on-the.net …?p=16947213&pos​tcount=719
https://photography-on-the.net …?p=17011802&pos​tcount=890
https://photography-on-the.net …p=17068543&post​count=1128
https://photography-on-the.net …p=17049246&post​count=1097
https://photography-on-the.net …p=17055295&post​count=1105
https://photography-on-the.net …p=17046799&post​count=1088
https://photography-on-the.net …380558&highligh​t=70d+poll

If you don't want to do the complete set of measurements, can you at least try :
- printing this picture (http://i1262.photobuck​et.com …eaTest-01_zps4366f833.png) on photo paper. Careful, you'll need flat colors, if your camera can focus on the black or on the white of the paper (because of the grain of the paper, or imperfections in the print), you'll need to use something else, like a piece of electric tape on some white plastic.
- Set up your camera on a tripod pointing to the ground like shown below with my 60D (of course, use a lens who can focus at the distance you're using).
- Select one of the focus points N°2, 3, 4, 13, 14, 15
- Switch to LiveView and select the "Quick AF" mode. This way, the camera will display the ideal position for the AF point you selected before going in LiveView. This will remove any possible misalignment you can have between : the ideal marks' position (visible in LIveView), and the focus screen marks' position (visible through the viewfinder).
- Still in LiveView, align the black/white edge with the center of the selected mark, and press AF-on button to perform autofocus (don't forget to turn on the back button autofocus in the menus).
- The mirror will flip and try to focus on the black/white edge.
- If your camera hunts back and forth and can't lock focus (red rectangle), you might have the same issue as mine.

If you encounter a similar problem as mine, do the measurements I described in one of the links above and send your camera to repair with a letter explaining the issue. You might need up to three trips and 7 weeks of immobilization to get your camera fixed.

Note : the focus points I cited (N°2, 3, 4, 13, 14, 15) are those where my camera can't focus, your camera might have an issue but with other focus points. In the test procedure above, I just supposed that each camera will have an identical defect. Normally the focus point numbers I cited above should be able to lock focus anywhere in the rectangle visible in LiveView (if I suppose that the 70D has the same AF sensor as a 7D http://www.aerien.ch …Nature_et_Oisea​ux.php%29.). So see if you can lock focus anywhere inside the marks visible in LiveView except for focus points N°1, 8, 9, 11, 12, 19, because those points might not be fully covered.

Thanks for going to all that effort.

This is a hobby and that really all is above my head. Whilst I am sure that I have problems, it may not be the one this thread relates to but there is something definately wrong in that the rear plane of acceptable DoF is the in cam focus point using the method I described earlier whereas in my understanding it should be 75% behind that and 25% in front?. This is consistent across all my lenses which are absolutely fine on the 600D. Perhaps the sensor is just out of whack slightly in one plane.

I think I will use FoCal to try and correct the problem, if that doesn't work, I shall get a refund on the grounds that I am too stupid to use it and stick with the 600D for now.

Good luck to the rest of you though. I will post later and let you all know the results of a good FoCalling on it!

meaton wrote in post #17077914
Thanks for going to all that effort.

This is a hobby and that really all is above my head. Whilst I am sure that I have problems, it may not be the one this thread relates to but there is something definately wrong in that the rear plane of acceptable DoF is the in cam focus point using the method I described earlier whereas in my understanding it should be 75% behind that and 25% in front?. This is consistent across all my lenses which are absolutely fine on the 600D. Perhaps the sensor is just out of whack slightly in one plane.

I think I will use FoCal to try and correct the problem, if that doesn't work, I shall get a refund on the grounds that I am too stupid to use it and stick with the 600D for now.

Good luck to the rest of you though. I will post later and let you all know the results of a good FoCalling on it!

I don't fully understand what you are saying.

In the 600D to have the best plane of focus slightly behind or in front what you are aiming is OK, but it shouldn't be regarded as a general rule. Because of manufacturing variations, the average position of the best plane of focus will always depend on the lens/body combination. The phase focus isn't as precise as you might think, but in principle, the camera should get you somewhere within the plane of best focus.
Also, on a camera, the plane of best focus should be considered as a flat sheet of glass parallel to the imaging sensor, so focusing on one thing (in ONE SHOT mode), keeping the shutter half pressed, then recomposing is a mistake. That's because the object on which you are focusing might be completely out of focus if it ends up in one corner of your picture. This is why it's often advised, with fast lenses, to limit recomposing to the bare minimum and to use the closest focus point available.

In the case of the 70D, the focus issue is mainly with fast lenses and the camera being in some cases completely unable to focus (yet giving you a focus confirmation). We are not speaking of being slightly front or back focusing, to fix that you can use MFA. We are speaking of that sort of focus issue :
https://photography-on-the.net …p=17068543&post​count=1128. Where aiming at different things might throw the camera completely out of focus even when you get a focus confirmation from the body.

If you buy FoCal, you might want to get the Pro version, that way you can test AF Consistency and do Mutipoint focus test.

But this is a lot of money for something you could do by yourself (http://www.youtube.com​/watch?v=E9fqWcgtHWA). Just place yourself at 50x the focal length you are using (distance in millimeters) to do your MFA, this is the recommended distance by Canon. You'll quickly see that +-1 MFA gives hardly any difference. In my opinion, anything above +-10 means that you have some sort of issue with : the lens, or the body, or both. Troubleshooting for AF issues isn't simple !

I do recommend you follow the simplified test method I proposed (it's way quicker than MFA), because once passed the warranty period, It'll be too late for you to get the camera fixed.

Molybdo42 wrote in post #17078057
I don't fully understand what you are saying.

In the 600D to have the best plane of focus slightly behind or in front what you are aiming is OK, but it shouldn't be regarded as a general rule. Because of manufacturing variations, the average position of the best plane of focus will always depend on the lens/body combination. The phase focus isn't as precise as you might think, but in principle, the camera should get you somewhere within the plane of best focus.
Also, on a camera, the plane of best focus should be considered as a flat sheet of glass parallel to the imaging sensor, so focusing on one thing (in ONE SHOT mode), keeping the shutter half pressed, then recomposing is a mistake. That's because the object on which you are focusing might be completely out of focus if it ends up in one corner of your picture. This is why it's often advised, with fast lenses, to limit recomposing to the bare minimum and to use the closest focus point available.

In the case of the 70D, the focus issue is mainly with fast lenses and the camera being in some cases completely unable to focus (yet giving you a focus confirmation). We are not speaking of being slightly front or back focusing, to fix that you can use MFA. We are speaking of that sort of focus issue :
https://photography-on-the.net …p=17068543&post​count=1128. Where aiming at different things might throw the camera completely out of focus even when you get a focus confirmation from the body.

If you buy FoCal, you might want to get the Pro version, that way you can test AF Consistency and do Mutipoint focus test.

But this is a lot of money for something you could do by yourself (http://www.youtube.com​/watch?v=E9fqWcgtHWA). Just place yourself at 50x the focal length you are using (distance in millimeters) to do your MFA, this is the recommended distance by Canon. You'll quickly see that +-1 MFA gives hardly any difference. In my opinion, anything above +-10 means that you have some sort of issue with : the lens, or the body, or both. Troubleshooting for AF issues isn't simple !

I do recommend you follow the simplified test method I proposed (it's way quicker than MFA), because once passed the warranty period, It'll be too late for you to get the camera fixed.

Sorry, I'll try to be clearer.

I use mainly a nifty fifty F1.8 and a Tamron 60mm F2.0 Macro. Both give acceptable to good results on my 600D. This to me means an acceptable area of focus behind and in front of the target, usually a garden bird. The target is invariably half way between near and far DoF at larger apertures, which is what I would expect given my limited knowledge.

On the 70D which has a 3 yr warranty (so I'm not to worried about time here), the results are different but consistent across the two brighter lenses that I listed.

The 70D results are that the acceptable DoF begins at the subject and extends forwards only. On the test listed on Page 2 of this thread, that means the scale 1-9 going forward not backwards is in acceptable focus, but nothing behind the subject. I would expect the region of acceptable sharpness to be -4 or 5 to +4 or 5. In the field, this means that many of my shots are unacceptable. I am using it a lot for garden birds using the wifi and iPhone so it wasn't really apparent when I first got the 70D a month ago and wifi/iPhone is one of the key reasons that I upgraded.

I hope that explains it a little better. I got the Plus ver of FoCal and I'll spend some time with it tomorrow.

meaton wrote in post #17078098
Sorry, I'll try to be clearer.

I use mainly a nifty fifty F1.8 and a Tamron 60mm F2.0 Macro. Both give acceptable to good results on my 600D. This to me means an acceptable area of focus behind and in front of the target, usually a garden bird. The target is invariably half way between near and far DoF at larger apertures, which is what I would expect given my limited knowledge.

On the 70D which has a 3 yr warranty (so I'm not to worried about time here), the results are different but consistent across the two brighter lenses that I listed.

The 70D results are that the acceptable DoF begins at the subject and extends forwards only. On the test listed on Page 2 of this thread, that means the scale 1-9 going forward not backwards is in acceptable focus, but nothing behind the subject. I would expect the region of acceptable sharpness to be -4 or 5 to +4 or 5. In the field, this means that many of my shots are unacceptable. I am using it a lot for garden birds using the wifi and iPhone so it wasn't really apparent when I first got the 70D a month ago and wifi/iPhone is one of the key reasons that I upgraded.

I hope that explains it a little better. I got the Plus ver of FoCal and I'll spend some time with it tomorrow.

OK, I think I understand. Did you use that chart http://regex.info/blog​/photo-tech/focus-chart ?
If yes, than you must use something else, because the 70D might also focus on the gray details. Now be careful, the numbers you gave me mean nothing and will vary depending on the distance between the camera and the chart (because the depth of field will vary with distance).
You should better use something like that : http://www.dphotojourn​al.com/focus-test-chart.pdf or even better, something like that :http://www.squit.co.uk​/photo/files/FocusChar​t.pdf

Very important : no matter the target you use, keep in mind that the sensitive area where you can achieve focus extends quite a bit outside the markers you see in the viewfinder. So the yellow zone shown in the picture below, should be :
- inside the white rectangle with a black strip if you use that chart : http://www.dphotojourn​al.com/focus-test-chart.pdf
- entirely on the standing (don't know if it's clear enough) target if you use that chart http://www.squit.co.uk​/photo/files/FocusChar​t.pdf

Also, from what you told me, it looks like the camera is front focusing. Maybe you'll need to use a +5 or +4 MFA value to get your lens in focus (or negative values, I don't remember).

Thanks again, I will try as you suggest tomorrow.

I know the numbers I gave are irrelevant, I simply gave them so you would have a point of reference given the chart that I used.

And yes, I think it is front focussing and I should be able to correct that automatically with FoCal Plus

Molybdo42 wrote in post #17078301
Also, from what you told me, it looks like the camera is front focusing. Maybe you'll need to use a +5 or +4 MFA value to get your lens in focus (or negative values, I don't remember).

So I did some testing today. Set up FoCal and got loads of unhandled application errors some of which related to light and some to distance. I don't really have the room so I'm going to have to do some more testing when I have cleared some space. I tested it with 60mm at 2m so maybe that was not enough distance.

I did use the test sheet you recommended from Tim Jackson with the rule printed on it.

LiveView perfect -14mm to +14mm in perfect focus
MF and AF v poor - +20 to +70mm (in front) in focus even then not as good as the liveview.

Whatever the problem is, that is a huge difference between live and AF/MF.

Live view is not effected by MFA. Also live view will always be better than through the viewfinder, that's why it takes so much longer.

travisvwright wrote in post #17079586
Live view is not effected by MFA. Also live view will always be better than through the viewfinder, that's why it takes so much longer.

That much is clear but it also shows there is nothing wrong with the lens or the camera, just the focusing using AF or MF.

For a camera to be front focusing as much as I have noted, that cannot be either expected, correct or acceptable IMHO.

I'm confused. MF typically stands for Manual Focusing. How can you blame the camera for your manual focusing? What are you meaning when you say MF?

travisvwright wrote in post #17079602
I'm confused. MF typically stands for Manual Focusing. How can you blame the camera for your manual focusing? What are you meaning when you say MF?

I am using the cameras manual focus confirmation and get the same results as with AF but that is a red herring. Focus confirmation when in MF I guess uses the same algorithm and mechanics as AF. Just take the AF results and see the huge difference. That is why I am concerned.

Yeah I've seen youtube videos who show that setting up FoCal targets can be difficult if you don't have proper lightning. https://www.youtube.co​m/watch?v=ALDWPtYwZ6I
I would suggest you to do the calibration at night (to avoid lightning variations) and with several desk lamps (preferably incandescent bulbs).

For parallelism you should use a bare mirror (no surrounding plastic frame) laid flat against the target, adjust the tripod head in order to center the image of the lens barrel on the center focus point.

Also for the distance it's advised to use roughly 50 times the focal length, so for a 60mm, you'll need 3000 mm or 3 meters. But that depends on people, some prefer to do the calibration at their shooting distance.

From the values you gave me looks like you need to put something like 10 or 15 in the MFA values (don't remember which sign to use). For me that's abnormal and it could vary a lot depending on the distance you shoot at. If the lens worked flawlessly on your 600D, you should send the camera to Canon (the lens isn't necessary).
Unless you can prove to Canon that the camera has an issue (maybe by measuring the AF coverage, I'm still waiting for my camera so I don't know if it'll work) ; I would wait a few months for the repair centers to get maybe some sort of notice on how to fix the issue and on what to look for. Because up until now, on first try the repair center will only do some software adjust, and you had to send the camera a second time and even a third time to get it fixed. But it now looks like some people are seeing their camera either replaced or repaired (might depend on where you live) and apparently one repair center is getting at least twenty 70D per week (https://www.flickr.com …/72157639283256​254/page2/). So maybe you wont have to wait too long.

Last thing : having great LiveView focus doesn't rule out a lens issue, because a lens could also need some MFA to perform well. The only way to rule out a lens issue would be to test it on several bodies, like you did with your 600D, so I think that in your case we can rule out a lens issue.