I haven't been around much lately, due to a disasterous series of photographic related events, which i won't bore you all with . These events have lost me almost an entire springs-worth of shooting , but i am hoping my run of bad luck will soon be over {but i'm not holding my breath }. If not, i am contemplating giving up photography.
A part of my problems has been a back-focus issue, and although i don't consider myself a pixel peeper, i don't think i am being unreasonable in expecting a 'fairly' accurately calibrated AF/MF.
After quickly realising this problem i set up a series of static 45' tests, at a common shooting distance of mine, and using my usual tele at widish apertures. It seems VERY OBVIOUS to me that true focus is actually falling about 20mm behind my intended focus point. This back-focus issue is clearly visible through the viewfinder {as well as on a downloaded 100% computer image/print}.
This 'may not' seem a huge amount to a landscape photographer , but i feel it is almost unusable, as it is enough to render a birds eye soft, and the back of the head sharp.
I have just returned my camera for the second time for this problem [first time made no difference}, and am dreading being told that this amount of back focus falls within their accepted tollerence levels.
Having gone through a Zenith B, K1000, two ME super's, LX, Mamiya 645, and D70 without ever suffering this problem, i'm beginning to think i am due some bad luck.
Just wondering if others feel i am expecting to much for these days, and whether a 20mm focusing error is about average through a long/fast tele.
Ive held off putting my head in the oven for now, but i can't hold out much longer
Fingers crossed
Alan
Posted by David Kjaer on April 26, 2007 at 2:31 PM
Hi Alan
You haven't said what camera and lens, please let us know. A friend recently sent me his Canon 1D MkII having returned it twice to Canon only to be told it was within their tollerence levels - I could not get a sharp image out of the camera. He kept on at Canon and eventually they replaced it with a new one. Although I really rate Canon gear, if you are one of the inlucky ones with faulty equipment, I don't have the same confidence in their repair service.
Hi Alan,
Just a quick message to wish you all the best with getting your issues sorted.
In answer to your question,is it acceptable? Most definately not in my opinion,an auto focus camera that doesn't auto focus (at least not accurately) is IMHO useless,and especially so with bird photography where we demand pin point accuracy,particularly with the maximum apertures we often find ourselves shooting at.A 5mm inaccuracy would make or break a shot,but 20mm is way,way inadequate.
Hope you get it resolved sooner rather than later.
Mmm.
with a 400mm at 2.8 at a distance of 4m you should be getting 20mm of acceptable sharpness. in this case a 20mm focus error (either back or front) would render the shot useless)
Same thing goes with a 600 at f4 at 6m
For close distance work you need as accurate a focus as you can get.
Canon will say that it falls in their accepted tollerance levels, but when you tell them you don't shoot sports at 10-20m they will revise their levels to take your shooting circumstances into consideration.
Hi it is 1/3rd DOF @ point of focus, so the camera can back focus or front focus 1/3 the depth of field so you would need at least 60 millimetres depth of field to get 20 mm back focus, if it is your d70 that need calibrating it is very easy to do it yourself with a 1.5 mm Alan Key.
I suffered the same with my 1DsII with my 500 F4.5 it was noticeably back focussing, I hadn't had the problem with film cameras and a 1D using the same lens, testing with my 100-400 proved it was the body at fault as the combination back focussed as well.
From the various American forums the recommended solution is to put both the body and lens in for matched calibration. Canon reckoned my lens needed servicing (doesn't exactly explain why the 100-400 also backfocussed....), Whatever the problem with the body was putting both in together resolved the problem.
Its a bit of a long story, but i shall make it as short as possible.
A couple of months ago i realised i needed a second body {fed up swopping between my two main lens's}.
I ordered another D70, and then discovered a S/H D2h in mint condition was available. I decided to bite the bullet, and swiftly cancelled the D70.
On arrival the D2h was as advertised {superb condition, with very few accutations}, and i had a wonderful couple of weeks with it. I was absolutely thrilled with it, and when teamed up with my 400mm/2.8 and tc-16a the combo worked like a charm. AF was pretty fast {thanks to the power/torque of the d2h motor}, and both manual and AF were EXTREMELY accurate.
After a couple of weeks the camera suffered a metering problem. This was a small disappointment, but i was still happy {as i knew some d2h's suffered from this, and it was a free fix by nikon}, and i sent it back to the dealer {and it was then forwarded on to nikon}
Five weeks later the camera arrived back to me in perfect working order {or so it seemed}.
I eagerly headed to my local coast, and had a fine days shooting. All my shots were manually focused, but as i headed home i decided i ought to take some AF flight shots, as the light levels were still good.
On my return home i loaded all my shots on to my computer, and noticed that although my MF shots seemed nicely focused, ALL my AF shots were back-focused .
I immediately set up some static targets, and did some controlled testing at 8m distance, using my 400mm/2.8+ 1.6 tc @ f4.5 {wide open} with a 45 degree target This confirmed to me that there was an AF problem.
Prior to the meter trouble, i had perfect AF/MF focusing, but i now appeared to have a very slight front focus when manually focusing {perfectly usable}, and a VERY noticable back-focus during AF. When focusing on a targe with the central focus area, and slightly depressing the shutter, the lens zips into 'apparent' focus and the focus confirmation light glows as it should. However, the true focus is clearly visible through the viewfinder {which is behind the intended focus point.
Yet again i sent the camera away {this time for back-focus}, and it arrived back with me today. However, it appears to be exactly the same.
I returned it today, after enclosing a print showing very clearly the back-focusing, and explaining that i use long lenses at shortish distances, with often fairly wide apertures.
I don't think it is lens related, as my sigma 300mm shows the same problem on the D2h. Both the Sigma 300mm and my nikon 400mm focus perfectly on the D70 {although i can only use MF with the 400mm on the D70, as the tc-16a does not AF with this body.
Well, thats it, in a nutshell.
Hopefully i shall get it back shortly, and all is well {but its a nerve-racking time for me at the moment}
Without wanting to hijack Alan's thread, what is it in the camera body that can be adjusted to resolve this issue?
I'm asking this because my 1Ds2 is exhibiting a greater depth of filed in front of the point of focus, than behind. I think this and another problem became apparent after the camera was serviced by Canon after I came back from the US in October last year.
I suffered the same with my 1DsII with my 500 F4.5 it was noticeably back focussing, I hadn't had the problem with film cameras and a 1D using the same lens, testing with my 100-400 proved it was the body at fault as the combination back focussed as well.
From the various American forums the recommended solution is to put both the body and lens in for matched calibration. Canon reckoned my lens needed servicing (doesn't exactly explain why the 100-400 also backfocussed....), Whatever the problem with the body was putting both in together resolved the problem.
Don't know about Canon Phil, but ive obviously been searching for nikon info on this subject over the last few weeks.
Not sure if its true, but apparently MF calibration is done via the shims, with the main mirror being set to 45 degrees using a form of template {requiring some disassembly of the body}.
The AF apparently is done via a computer , and does not require any disassembly.
I have also read that the main mirror has no effect on AF calibration, although i find that hard to believe, as the AF mirror is hinged to the main mirror {and surely any missalignment of the main mirror would have some effect on AF }.
Its all a mystery to me, and there is so much conflicting views on the internet.
Hi Alan both the main mirror and the A. F. mirror are adjustable independently of each other with the d2 series of cameras actually with most cameras that have AF the small shim you talk about is to make sure the distance between main mirror and focusing screen is correct, if you lift your mirror up carefully and look into the mirror box you'll see on the right handside a small pin sticking out, on the back of the AF mirror there is a small prong which fits over the pin this addjusts the angle of the AF mirror. Towards the front of the mirror box there is another Alan screw this is the stop for the main mirror which projects the image up to the focusing screen this is also adjustable, trouble is getting hold of the two special tools you need to do this.
I did look at the mirror box on the d2h, just in case there was the slightest chance that i could tweak the miss-adjustment myself {as a last resort, if nikon could not achieve it to my satisfaction}, but this seems impossible for me.
Had this been a problem with my D70, i would be quite confident of doing this, as both mirrors have easily accessable adjustment screws {although it would be quite time-consuming, with a lot of testing}.
I certainly hope nikon can get to the bottom of this problem this time, as i certainly expect a £2000 body to be capable of decent focus {through a high quality, and originally very expensive lens}, especially as i have no problems through my 'relatively' cheap body.
Not sure where this will leave me, if it is not fixed . My garauntee has expired from the dealer {even though the camera has not been in my hands for the majority of this time}, and i certainly can not afford to replace it with a similar body for a long time. I guess i could continue to use it using MF only, but this defeats the object of me buying the thing in the first place.
I am praying that i am worrying too much, and all will be resolved this time {although i will be as nervous as hell, as i try it out}
Tom, i had thought of selling it on, if it is still not up to scratch.
If Nikon consider this error to be within their tollerences, i guess i could sell it with a clear conscience {however i would feel uncomfortable doing so}.
I guess i had to have a run of bad luck eventually , but i shall have to wait and see.
If thats as good as Nikon can make it - what better can you say!
Slight errors on focusing must be present on some individual Canon cameras otherwise they would not have but that calibration feature in the new 1d3.
What I have never bottommed out is why it is only digital cameras, why did we not have this problem with an film camera?
Ive rarely heard of these problems during the film era {more likely to hear of film flatness issues}.
I would also have thought that tweaking AF precisely would be fairly simple today {given the present technology}. I would have thought that getting MF accurately calibrated to the viewfinder screen to be more difficult, as it is not a stepless procedure {varying shim thickness's etc}.
Blimey, i had better do some work, or i shall have to sack myself
Why no focus issues in film era - 2 reasons really
1 - Film cameras have a solid piece of machined metal - the film guide rails and a metal back plate to hold the film flat (although one of either Leica or Contax made a model with a vacuum system to hold the film even flattter - I can't remember which), it is much simpler to get computer milling tolerences finer than making a sandwich sensor plate - I should think there are thickness variations of sensors being mounted into the body chassis giving the micro variations in focus between bodies.
2 - Pixel peeping, a 8x loupe does not magnify to the same extent as 100% pixel peeping on a 22" screen - we view at least 6x final 300DPI print size -
Discussion
21 repliesI haven't been around much lately, due to a disasterous series of photographic related events, which i won't bore you all with
A part of my problems has been a back-focus issue, and although i don't consider myself a pixel peeper, i don't think i am being unreasonable in expecting a 'fairly' accurately calibrated AF/MF.
After quickly realising this problem i set up a series of static 45' tests, at a common shooting distance of mine, and using my usual tele at widish apertures. It seems VERY OBVIOUS to me that true focus is actually falling about 20mm behind my intended focus point. This back-focus issue is clearly visible through the viewfinder {as well as on a downloaded 100% computer image/print}.
This 'may not' seem a huge amount to a landscape photographer
I have just returned my camera for the second time for this problem [first time made no difference}, and am dreading being told that this amount of back focus falls within their accepted tollerence levels.
Having gone through a Zenith B, K1000, two ME super's, LX, Mamiya 645, and D70 without ever suffering this problem, i'm beginning to think i am due some bad luck.
Just wondering if others feel i am expecting to much for these days, and whether a 20mm focusing error is about average through a long/fast tele.
Ive held off putting my head in the oven for now, but i can't hold out much longer
Fingers crossed
Alan
You haven't said what camera and lens, please let us know. A friend recently sent me his Canon 1D MkII having returned it twice to Canon only to be told it was within their tollerence levels - I could not get a sharp image out of the camera. He kept on at Canon and eventually they replaced it with a new one. Although I really rate Canon gear, if you are one of the inlucky ones with faulty equipment, I don't have the same confidence in their repair service.
David
Just a quick message to wish you all the best with getting your issues sorted.
In answer to your question,is it acceptable? Most definately not in my opinion,an auto focus camera that doesn't auto focus (at least not accurately) is IMHO useless,and especially so with bird photography where we demand pin point accuracy,particularly with the maximum apertures we often find ourselves shooting at.A 5mm inaccuracy would make or break a shot,but 20mm is way,way inadequate.
Hope you get it resolved sooner rather than later.
Justin
Borrow another body and test it with it... go to a camera shop with the lens and square paper ruler.
T
with a 400mm at 2.8 at a distance of 4m you should be getting 20mm of acceptable sharpness. in this case a 20mm focus error (either back or front) would render the shot useless)
Same thing goes with a 600 at f4 at 6m
For close distance work you need as accurate a focus as you can get.
Canon will say that it falls in their accepted tollerance levels, but when you tell them you don't shoot sports at 10-20m they will revise their levels to take your shooting circumstances into consideration.
Hope you get it sorted out soon
Phillip.
From the various American forums the recommended solution is to put both the body and lens in for matched calibration. Canon reckoned my lens needed servicing (doesn't exactly explain why the 100-400 also backfocussed....), Whatever the problem with the body was putting both in together resolved the problem.
Its a bit of a long story, but i shall make it as short as possible.
A couple of months ago i realised i needed a second body {fed up swopping between my two main lens's}.
I ordered another D70, and then discovered a S/H D2h in mint condition was available. I decided to bite the bullet, and swiftly cancelled the D70.
On arrival the D2h was as advertised {superb condition, with very few accutations}, and i had a wonderful couple of weeks with it. I was absolutely thrilled with it, and when teamed up with my 400mm/2.8 and tc-16a the combo worked like a charm. AF was pretty fast {thanks to the power/torque of the d2h motor}, and both manual and AF were EXTREMELY accurate.
After a couple of weeks the camera suffered a metering problem. This was a small disappointment, but i was still happy {as i knew some d2h's suffered from this, and it was a free fix by nikon}, and i sent it back to the dealer {and it was then forwarded on to nikon}
Five weeks later
I eagerly headed to my local coast, and had a fine days shooting. All my shots were manually focused, but as i headed home i decided i ought to take some AF flight shots, as the light levels were still good.
On my return home i loaded all my shots on to my computer, and noticed that although my MF shots seemed nicely focused, ALL my AF shots were back-focused
I immediately set up some static targets, and did some controlled testing at 8m distance, using my 400mm/2.8+ 1.6 tc @ f4.5 {wide open} with a 45 degree target This confirmed to me that there was an AF problem.
Prior to the meter trouble, i had perfect AF/MF focusing, but i now appeared to have a very slight front focus when manually focusing {perfectly usable}, and a VERY noticable back-focus during AF. When focusing on a targe with the central focus area, and slightly depressing the shutter, the lens zips into 'apparent' focus and the focus confirmation light glows as it should. However, the true focus is clearly visible through the viewfinder {which is behind the intended focus point.
Yet again i sent the camera away {this time for back-focus}, and it arrived back with me today. However, it appears to be exactly the same.
I returned it today, after enclosing a print showing very clearly the back-focusing, and explaining that i use long lenses at shortish distances, with often fairly wide apertures.
I don't think it is lens related, as my sigma 300mm shows the same problem on the D2h. Both the Sigma 300mm and my nikon 400mm focus perfectly on the D70 {although i can only use MF with the 400mm on the D70, as the tc-16a does not AF with this body.
Well, thats it, in a nutshell.
Hopefully i shall get it back shortly, and all is well {but its a nerve-racking time for me at the moment}
Alan
I'm asking this because my 1Ds2 is exhibiting a greater depth of filed in front of the point of focus, than behind. I think this and another problem became apparent after the camera was serviced by Canon after I came back from the US in October last year.
From the various American forums the recommended solution is to put both the body and lens in for matched calibration. Canon reckoned my lens needed servicing (doesn't exactly explain why the 100-400 also backfocussed....), Whatever the problem with the body was putting both in together resolved the problem.
Not sure if its true, but apparently MF calibration is done via the shims, with the main mirror being set to 45 degrees using a form of template {requiring some disassembly of the body}.
The AF apparently is done via a computer
I have also read that the main mirror has no effect on AF calibration, although i find that hard to believe, as the AF mirror is hinged to the main mirror {and surely any missalignment of the main mirror would have some effect on AF
Its all a mystery to me, and there is so much conflicting views on the internet.
Alan
Phillip.
I did look at the mirror box on the d2h, just in case there was the slightest chance that i could tweak the miss-adjustment myself {as a last resort, if nikon could not achieve it to my satisfaction}, but this seems impossible for me.
Had this been a problem with my D70, i would be quite confident of doing this, as both mirrors have easily accessable adjustment screws {although it would be quite time-consuming, with a lot of testing}.
I certainly hope nikon can get to the bottom of this problem this time, as i certainly expect a £2000 body to be capable of decent focus {through a high quality, and originally very expensive lens}, especially as i have no problems through my 'relatively' cheap body.
Not sure where this will leave me, if it is not fixed
I am praying that i am worrying too much, and all will be resolved this time {although i will be as nervous as hell, as i try it out}
Alan.
Alan, if it doesn't work to your satisfaction simply sell it on, as it will work with no proplems for sport-shooters.
T
Twice recently you've mentioned them in -ve light]
T
Tom, i had thought of selling it on, if it is still not up to scratch.
If Nikon consider this error to be within their tollerences, i guess i could sell it with a clear conscience {however i would feel uncomfortable doing so}.
I guess i had to have a run of bad luck eventually
Alan
Slight errors on focusing must be present on some individual Canon cameras otherwise they would not have but that calibration feature in the new 1d3.
What I have never bottommed out is why it is only digital cameras, why did we not have this problem with an film camera?
Ive rarely heard of these problems during the film era {more likely to hear of film flatness issues}.
I would also have thought that tweaking AF precisely would be fairly simple today {given the present technology}. I would have thought that getting MF accurately calibrated to the viewfinder screen to be more difficult, as it is not a stepless procedure {varying shim thickness's etc}.
Blimey, i had better do some work, or i shall have to sack myself
Yes i have stumbled across that link.
In fact the severe back-focus image on that page seems remarkably similar to the error on my D2h.
Unfortunately the adjustment screws on my d2h are not as accessable as the D70 body.
1 - Film cameras have a solid piece of machined metal - the film guide rails and a metal back plate to hold the film flat (although one of either Leica or Contax made a model with a vacuum system to hold the film even flattter - I can't remember which), it is much simpler to get computer milling tolerences finer than making a sandwich sensor plate - I should think there are thickness variations of sensors being mounted into the body chassis giving the micro variations in focus between bodies.
2 - Pixel peeping, a 8x loupe does not magnify to the same extent as 100% pixel peeping on a 22" screen - we view at least 6x final 300DPI print size -