Weird autofocus/sharpness issues: Any ideas?

I may have missed it, but have you performed AF calibration with this body/lens combination?
 
No I haven't as it doesn't seem to be a constant issue and therefore I don't want to cause more harm than good.
 
If technique is sound that suggests to me something loose or broken within camera or lens, which in turn leads to inconsistent focusing performance. Plus, if the lens is softer on one side it sounds like it's decentred. Has the lens been knocked or dropped, our could it have been without your knowledge?
 
I have always looked after both this lens and body extremely well. The lens is very rarely off the body too so I can't see it being the case. Thinking about it I reckon about 15 out of 1500 shots were affected on Saturday so I reckon if there was something on the lens it would occur more than this (?)..... Where might be a good place to take the lens and body to have it tested? (Good meaning quick and not costing the earth).
 
I can't help you with that one, but if it is something loose (or not) and only causing failures in ~1% of shots it may not be at all easy to diagnose - and that might either be costly or inconclusive. I'm afraid I'm a Canon man, so can't help you with Nikon advice, but if I had intermittent woes I couldn't pin down I'd send both body and lens to Canon for a look see.
 
Thanks for your input. I am going to swap bodies for an assignment to see how the lens functions on the other body- hopefully I should be know more this time next week.
 
What could be that deficiency though? Just checking out one of the shots it was 1/8000s 36mm f4.5 AF-S with a single focus point aimed at the chest of the bride (hitting half skin half dress) and the point of focus looks to be about 2ft behind.
1/8000 is waaaaay too fast for wedding photography / group shots! I'd say (while not a full answer to this particular issue, but would certainly help) you might be able to drop the ISO a bit, or tighten the aperture to sharpen the corners?!
 
Last edited:
Yes, I'd agree they're not ideal- settings were configured by the body really on the fly, I was in apertue with auto iso and it was extremely bright.
 
I have done lots more testing (as per the nikon guide) and I couldn't fault it as it was spot on. Taking into consideration out of 2000 images about 5 were out I am more inclined to blame my method. I am going to double careful moving forward to see if I can eradicate the issue.
 
The newer cameras have some weird A/F behaviors w/ BBF. If you have the latest firmware you should be able to use AF-S and single point with the AF-S menu setting to "focus release." With those settings the camera should not fire unless the camera believes the image (point of focus) *is* actually in focus (any other settings w/ BBF and the camera will fire regardless).

You could also test the focus point display for accuracy (i.e. size/position), but I doubt that's the issue. If it is still inconsistent, enable AF on the shutter and try using that instead of BBF. AF-C and any mode other than single point the camera is actually using 9 AF points to focus (the selected one and the 8 points around it) and may not be focusing where you think it is.
 
I had read this re the firmware also and will check it again, however I accidentally left myself in AF-S on a moving target shot last saturday and I held the BBF in and it would not release the shutter for love nor money so I suspect I already have the right firmware in place...... your post has only just jogged my memory and made me put two and two together so thank you.
 
Sorry I can add anything, but it's an interesting read, as I have had this occur to me...it turned out to be my fault. ;)
 
Yes, I am leaning towards the same (queue the older hands tutting)- Saturdays results are much better but still one or two instances of it missing but then on the next composition it being bang on. 99% sure it is me, however narrowing down how and why I couldnt say.
 
Yes, I am leaning towards the same (queue the older hands tutting)- Saturdays results are much better but still one or two instances of it missing but then on the next composition it being bang on. 99% sure it is me, however narrowing down how and why I couldnt say.

Use focus-recompose, with back-button AF. I prefer using the centre-point, as you always know where it is! And it's also more accurate in difficult conditions. Don't believe those that may claim this method can be innacurate - that is very rare in practise, only occurring when you're very close with a very low f/number. Then you know what is focused on, and it can't change.

You also need to tighten up on your settings technique - they should be optimum for the shot, not just okay from an exposure point of view ;)
 
I think some of previous misses occurred when rushing as the need and moment dictated. Certainly more being careful with adjusting settings when doing the shots where the issues had occurred previously has helped.
 
Excuse the question but, do you do a lot of wedding photography? I only ask as from a technical point of view, your settings seem to be way off kilter in terms of Necessary shutter speed and ISO?
 
Last edited:
You also need to tighten up on your settings technique - they should be optimum for the shot, not just okay from an exposure point of view ;)

1/8000 is waaaaay too fast for wedding photography / group shots!
I don't think there is generally any such thing as "too fast." *Especially* with group shots.

*I* might not have used the same settings... actually I know I wouldn't have because the camera would have chosen different ones. But, I can't really say there is anything "wrong" with the settings used. With a wide FL (36mm) and 10-20ft distance there is still a ton of DOF at f/4.5. On a D8xx ISO 100/200/400 all look about the same (100/200 test the same technically, so I seldom ever choose 100). And even ISO 800 will seldom show any significant detriment even with very critical evaluation.

Assuming f/4.5 was chosen for sharpness/DOF w/ a fast lens (36/4.5 on a 24-70/2.8 should be excellent), *my* settings would have been something like f/4.5, 1/5000, 200 (aperture selected by me, ISO/SS set by the camera). But IQ wise the image would look identical to the settings he used; f/4.5, 1/8000, 320.

So, what settings do you think would have been notably better? Because I don't see the issue... nor with the original image at 24/5.6, 1/4000, 320. And I certainly don't see them contributing to a focus problem. If there is an issue, it's with the hard/bright midday sun IMO.

(let's ignore ETTL/ETTR/recovery considerations...)
 
I wouldn't class myself as a seasoned pro by any means, hence why I am always happy to ask for advice on here. I would also say that my methods and thought processes do evolve and certainly have following this issue.
A lot of the shots where I have seen the issue has been shooting pretty much on the fly in aperture with auto iso set between certain parameters so the camera has been doing the decision making re shutter speed and iso..... this has been because time has been a factor. Please note though that with the shots previously that have been less sharp there have been good ones- primarily I wanted to know 'why' and 'how' this was happening with one eye on whether the resolution of the 36mp was at play.
Any input you have setting this up more appropriately and as to the logic behind it would be welcomed.
 
Thanks SK66. Yes I have merely increased aperture to 4+ to allow more DOF.
Yes and the sun etc. In the original example is shocking- not a shot I was happy with but it was really only included to show the technical issue, it was bang on mid day but circumstances (church g-yard layout, congregation, horse and cart and time etc.) meant it was pretty much there or nowhere.
 
I don't think there is generally any such thing as "too fast." *Especially* with group shots.

*I* might not have used the same settings... actually I know I wouldn't have because the camera would have chosen different ones. But, I can't really say there is anything "wrong" with the settings used. With a wide FL (36mm) and 10-20ft distance there is still a ton of DOF at f/4.5. On a D8xx ISO 100/200/400 all look about the same (100/200 test the same technically, so I seldom ever choose 100). And even ISO 800 will seldom show any significant detriment even with very critical evaluation.

Assuming f/4.5 was chosen for sharpness/DOF w/ a fast lens (36/4.5 on a 24-70/2.8 should be excellent), *my* settings would have been something like f/4.5, 1/5000, 200 (aperture selected by me, ISO/SS set by the camera). But IQ wise the image would look identical to the settings he used; f/4.5, 1/8000, 320.

So, what settings do you think would have been notably better? Because I don't see the issue... nor with the original image at 24/5.6, 1/4000, 320. And I certainly don't see them contributing to a focus problem. If there is an issue, it's with the hard/bright midday sun IMO.

(let's ignore ETTL/ETTR/recovery considerations...)
I would vary the settings depending on the conditions, but a speed of 1/8000, I've only used for ballistic photography, for freezing the action as much as possible (here we're talking bullets / missiles hitting objects and fragmenting).

That's the only reason I think it's a little too fast for a group shot.
 
Last edited:
I had this issue with a 5dmk2 many years ago and narrowed it down to technique and a very sensitive shutter button on the camera. Basically in-between focus and recompose there was enough of a release of pressure on the shutter to allow it to refocus once composed thereby focussing on something completely different , and that would be fairly random ie if the different thing was in the right plane you would not notice but occasionally as in your case it could be a wall behind someone hence tricking you into weeks of arsing about with lens calibrations ( so well done on not going that route ) as that is almost always a few cm at most and normally not that far out. I learnt then to slow down slightly and adjusted my camera technique as well as using back button if needed and the problem went away. I had a couple of assistants over the years have had the exact same problem. I use spot focus almost permanently now apart from the odd action shot. Group shots probably in that situation more like iso 400 tops and F8 with fill flash ( HSS ). Each to their own with settings though and one thing does not suit everyone.
 
Thanks SK66. Yes I have merely increased aperture to 4+ to allow more DOF.
Yes and the sun etc. In the original example is shocking- not a shot I was happy with but it was really only included to show the technical issue, it was bang on mid day but circumstances (church g-yard layout, congregation, horse and cart and time etc.) meant it was pretty much there or nowhere.
My only "recommendation" would be to refine the camera settings and the way you use them slightly.

*I* usually set up the D8xx aperture priority to use auto ISO set to 200-1600 with a minimum SS of "Auto+1" (2x FL). With shorter FL's/VR I might change the settings to either 200-800 or just SS auto (1x FL).
The way the D8xx (and most Nikons) behaves when set like this is it will first use the lowest ISO setting (200 in this case) unless it can't achieve the minimum SS (1 or 2 x FL). When it hits the minimum SS setting it will then increase ISO. And then when it get's to the max ISO setting it will allow the SS to drop below the minimum set... This is exactly the same choices in the same sequence as I would make 95% of the time. The other 5% of the time you just have to add some EC to make it do what you want... and that's simple to do if you enable "easy EC" on the front wheel (but "easy EC" is also *very* easy to inadvertently change and not notice).
 
Last edited:
Interesting Andrew thank you. I have been using BBF but I defintitely have unfocused/refocused so could see the potential of this. But as you said I essentially slowed down a touch with everything and it has improved things- I will also go tighter with the fstop.
 
I would vary the settings depending on the conditions, but a speed of 1/8000, I've only used for ballistic photography, for freezing the action as much as possible (here we're talking bullets / missiles hitting objects and fragmenting).

That's why I think it's a little too fast for a group shot.
Well, it certainly wasn't a "requirement."
 
Well, it certainly wasn't a "requirement."

Disagree. When the light gives you options, as it did here, dropping to base ISO100 would deliver better image quality, even if not by much.

With groups, it's obviously important to get everyone sharp and to optimise that needs a higher f/number than you might think. F/8 would be better, and would also cover any slight focus discrepancies, f/11 for two rows of people. Depending on the lens, this would also ensure sharp edges/corners. For best sharpness, getting the subject in focus, or as far within the depth-of-field zone as possible, is better than worrying about any losses from diffraction. I would also have moved back if possible, shooting at 50mm rather than wider.

There's still more than enough action-stopping or shake-minimising potential in 1/500sec or 1/1000sec shutter speed.
 
Disagree.
Richard, I was replying that 1/8000 wasn't a requirement...

We're taking about 36mm at f/4.5 here used from around 15 ft (guess)...that's around 20ft DOF. And much greater for the 24/5.6 image originally posted (probably at/beyond HFD). I get the difference between DOF and actual focus, but we are talking about a group shot, not a portrait....
I will say that with that much DOF (shorter FL's) there is no real benefit to using wider apertures most of the time; it's not like you're going to blur the BG into soft focus (unless it is *very* close). And yes, there would have been plenty of SS remaining even at f/11 (although the D8xx is diffraction limited at f/11, I agree it's not a primary concern for this type of shot). And I agree with using a longer FL stopped down farther for BG perspective/sharpness *if possible.*

But IMHO, given the shot as taken (distance/FL), I highly doubt anyone would see any significant difference between a (properly focused) image taken as it was (24/5.6, 320, 1/40000) as opposed to 24/11, 100, 1/800, *other than the BG being more in focus.
I have the 24-70/2.8 and D810... maybe I will take a couple test shots if/when the sun comes out.
 
Richard, I was replying that 1/8000 wasn't a requirement...

We're taking about 36mm at f/4.5 here used from around 15 ft (guess)...that's around 20ft DOF. And much greater for the 24/5.6 image originally posted (probably at/beyond HFD). I get the difference between DOF and actual focus, but we are talking about a group shot, not a portrait....
I will say that with that much DOF (shorter FL's) there is no real benefit to using wider apertures most of the time; it's not like you're going to blur the BG into soft focus (unless it is *very* close). And yes, there would have been plenty of SS remaining even at f/11 (although the D8xx is diffraction limited at f/11, I agree it's not a primary concern for this type of shot). And I agree with using a longer FL stopped down farther for BG perspective/sharpness *if possible.*

But IMHO, given the shot as taken (distance/FL), I highly doubt anyone would see any significant difference between a (properly focused) image taken as it was (24/5.6, 320, 1/40000) as opposed to 24/11, 100, 1/800, *other than the BG being more in focus.
I have the 24-70/2.8 and D810... maybe I will take a couple test shots if/when the sun comes out.

These detailed technicalities are not really the point though. For me, the thing is that when you use the extreme ends of any exposure settings, whether it's very high or very long shutter speed, lowest f/number or highest etc, there is usually a particular reason. It's done consciously and deliberately, not by accident, even in haste.

I can't think that any seasoned photographer would look at those settings and think, yes, they're perfect, that's just what I would use. Same goes for AF-C that is simply inappropriate in this situation and prone to the kind of errors that it seems like the OP has fallen into. It suggests a general lack of good understanding.
 
These detailed technicalities are not really the point though. For me, the thing is that when you use the extreme ends of any exposure settings, whether it's very high or very long shutter speed, lowest f/number or highest etc, there is usually a particular reason. It's done consciously and deliberately, not by accident, even in haste.

I can't think that any seasoned photographer would look at those settings and think, yes, they're perfect, that's just what I would use. Same goes for AF-C that is simply inappropriate in this situation and prone to the kind of errors that it seems like the OP has fallen into. It suggests a general lack of good understanding.
Indeed, hense why I questioned the 1/8000 specifically. No matter how well a camera performs, if you ask so much from it in one area (1/8000 is a pretty extreme speed that I would only use in very specific areas) physics dictate another area will suffer. In this case, quite unnecessarily :)
 
Last edited:
Has anyone checked what he has custom menus a1 and a2 set to? Even with AF-S you can still set release priority over focus priority. This WOULD explain intermittent focus accuracy. @Pegasus_Thrust have you checked this?
 
Has anyone checked what he has custom menus a1 and a2 set to? Even with AF-S you can still set release priority over focus priority. This WOULD explain intermittent focus accuracy. @Pegasus_Thrust have you checked this?
I mentioned it...
But it is important to note that the AF-C setting will apply with BBF enabled *only* if the focus point selection is set to single point (this is true of most newer Nikons). My D810 doesn't even apply the "release+focus" setting in AF-C with anything other than the single AF point setting.
 
I mentioned it...
But it is important to note that the AF-C setting will apply with BBF enabled *only* if the focus point selection is set to single point (this is true of most newer Nikons). My D810 doesn't even apply the "release+focus" setting in AF-C with anything other than the single AF point setting.

Is that in the manual? Can't see it anywhere (Assuming it's the same with the D800).
 
These detailed technicalities are not really the point though.
Hmm, it seemed to me that you were arguing "for" the technicalities. IMHO, if the aperture was the right one for the image desired, then the technical differences between ISO/SS options don't matter *in this case.* From my perspective aperture determines what an image *IS.* As long as SS is high enough, and ISO is low enough, they don't much matter 95% of the time (other than slow SS images).

I can't think that any seasoned photographer would look at those settings and think, yes, they're perfect, that's just what I would use. Same goes for AF-C that is simply inappropriate in this situation
I'm fairly well "seasoned;" and if I had chosen a fixed ISO of 320 on the D8xx due to the prevailing conditions, then I wouldn't worry about it again unless it became a problem (too low) or there was a slow time to change it to something more ideal for new prevailing conditions. IMO, manually selecting a fixed ISO for individual images in a changing environment is not practical. On my cameras it's a two handed operation with me looking at an LCD screen (there is "easy ISO," but I screw that up more than it helps).
FWIW *I* wouldn't set a fixed ISO and those aren't the settings I would have used, that's not how I work... ISO is the last setting I lock down, and only in a constant/controlled light situation. But that's kind of irrelevant (maybe it's not, but I'm hesitant to say that my way of working is "better").

And AF-C with BBF (which is what was used) is perfectly suitable for this situation... if used correctly.
 
Is that in the manual? Can't see it anywhere (Assuming it's the same with the D800).
Nope, and it's the same with the D800 and most (all?) newer Nikons. The way BBF works with newer Nikons is, once AF has been achieved with BBF it then acts as an AF/MF switch. You push it to activate AF or you leave it in "MF" and the camera doesn't care if the image is in focus or not. (This behavior is overridden if AF is also enabled on the shutter at the same time, but why would you do that?). And I mis-spoke... this behavior is the same for AF-C and AF-S.

They broke "trap focus" with the newer cameras, but enough people were unhappy about it that they enabled it in AF-S w/ single point *only* with the newer firmware (I'm not certain about the latest cameras like the D750, but I imagine it's the same). For those that don't know, "trap focus" is a Nikon behavior where you can prefocus at a distance and use the AF system to trigger the shutter when a subject comes into focus.
 
Hmm, it seemed to me that you were arguing "for" the technicalities. IMHO, if the aperture was the right one for the image desired, then the technical differences between ISO/SS options don't matter *in this case.* From my perspective aperture determines what an image *IS.* As long as SS is high enough, and ISO is low enough, they don't much matter 95% of the time (other than slow SS images).


I'm fairly well "seasoned;" and if I had chosen a fixed ISO of 320 on the D8xx due to the prevailing conditions, then I wouldn't worry about it again unless it became a problem (too low) or there was a slow time to change it to something more ideal for new prevailing conditions. IMO, manually selecting a fixed ISO for individual images in a changing environment is not practical. On my cameras it's a two handed operation with me looking at an LCD screen (there is "easy ISO," but I screw that up more than it helps).
FWIW *I* wouldn't set a fixed ISO and those aren't the settings I would have used, that's not how I work... ISO is the last setting I lock down, and only in a constant/controlled light situation. But that's kind of irrelevant (maybe it's not, but I'm hesitant to say that my way of working is "better").

And AF-C with BBF (which is what was used) is perfectly suitable for this situation... if used correctly.

Technicalities are importrant. No, as I said, it's the detail of the technicalities I'm less interested in, and more the general principles.

If you're shooting at 1/8000sec, then that would strongly suggest rapid subject movement, and there isn't any here. At f/4.5, that suggests there isn't a requirement for more than minimal D0F, when in fact that is the main priority here, and indeed where the OP has tripped up and failed to get all of the group sharp.

And why ISO320? Not that it's very important here, but these things shouldn't happen by accident. Why AF-C. Again, not needed here and prone to inconsistencies that the OP has fallen foul of.

Sometimes you have to take calculated risks with camera settings, mostly in fact, but with this situation the conditions are benign. So minimise the risks, and go belt-and-braces on the priorities.
 
Nope, and it's the same with the D800 and most (all?) newer Nikons. The way BBF works with newer Nikons is, once AF has been achieved with BBF it then acts as an AF/MF switch. You push it to activate AF or you leave it in "MF" and the camera doesn't care if the image is in focus or not. (This behavior is overridden if AF is also enabled on the shutter at the same time, but why would you do that?). And I mis-spoke... this behavior is the same for AF-C and AF-S.

That's not what I meant.. you said release priority and BBF only works with a single AF point. Where is THAT in the manual? Can't find it.
 
If you're shooting at 1/8000sec, then that would strongly suggest rapid subject movement, and there isn't any here. At f/4.5, that suggests there isn't a requirement for more than minimal D0F, when in fact that is the main priority here, and indeed where the OP has tripped up and failed to get all of the group sharp.
Agreed, but I don't think DOF was the problem nor would more have helped really.
And why ISO320? Not that it's very important here, but these things shouldn't happen by accident.
It does seem like an odd choice... if I were setting an ISO it would be a full stop setting (unless pushing max tolerance). Not sure that it matters though.
Why AF-C. Again, not needed here and prone to inconsistencies that the OP has fallen foul of.
Ok, you've lost me here.
What benefit is there to BBF if you do not use AF-C? How is BBF/focus-recompose any different in AF-C vs AF-S?
IMO, the whole point of BBF is to have instant choice between AF-C (hold BB), AF-S (release BB), and Manual (don't use BB) behaviors all with one convenient button.

The biggest risk to AF error is in using AF-S on the shutter and focus-recompose... either aspect could cause a problem (double press/angular error). And neither is actually prevented by using BBF (although double press is less likely).
And BTW, with recent Nikons you can't use AF-S with single point and BBF for focus/recompose... it won't work (unless AF-s is set to release priority).
Sometimes you have to take calculated risks with camera settings, mostly in fact, but with this situation the conditions are benign. So minimise the risks, and go belt-and-braces on the priorities.
Ah, so full manual w/ MF then? ;)

Of course I agree. When time permits there is no excuse for mistakes... but I still make them sometimes.
 
Last edited:
Back
Top