Images not sharp

N

Nick Radley

Messages
7
Edit My Images
No
Hi all

I'm just new here. I have been taking photos for years now but am new to the world of calibrating lenses as I've always been quite happy with my focus results. I recently calibrated my lenses for both my D700s and got things as close as I figured they could be.
Now, it may be that I am now being over-fussy, or that I am doing something wrong when calibrating, or that there is something wrong in the set-up of BOTH my D700s, but I have begun to wonder whether my images are really as sharp as they should be.
Is it too much to ask that a 200mm lens (at furthest distance away, but still filling the whole frame), at F2.8 and shooting at 1/2000s in good light and static subject, is not pin-sharp? I have yet to go out and try it at F8 to see if that improves matters (which I figure it will).
The images look ok, but if I look at someones eye for instance, without zooming in in Lightroom beyond 1:1 it is soft and I wonder whether that is what is to be expected shooting so far away, zooming in and at F2.8?
I guess it could be the lens (which is a Nikon AF-S NIKKOR 70-200mm f/2.8G ED VR II) but could it even be the UV filter I have on?...perhaps that is of poor quality?
Any pointers would be greatly appreciated as I am fed up of sharpening my stuff (too much) in Lightroom.
Thanks
Nick
 
Post up some examples that concern you and we can discuss more effectively. :)
 
UV filter is a big no
I would expect if your subject is in focus the image be sharp without a UV filter with a little softening and shading in the corners. By F8 I'd execpt the sharpness (not just front to back) but edge to edge (often overlooked) to be more uniform with far less shading in the corners.
 
I have seen lens that don't focus properly to infinity, it might be worth checking with a different 200mm if you can, that'll give you a comparrison.
 
The settings you mention mean the lens is operating at the extremities... wide open and max focal length so it's unlikely to be producing it's sharpest shots, see how it compares at mid settings

Simon
 
What zoom length did you calibrate at? Reason I ask is that I have a different value, albeit not massive, at the 2 x extreme zoom ranges on the 70-200mm f2.8. As I tend to use the longer end mostly I've used a value at 180mm although the marginal difference through the zoom range on the type of stuff I primarily shoot is probably negligible. YMMV on portraits etc.

Agree with Gramps in # 2, lets see some examples.

GC
 
IME, poor/improper calibration is more likely to cause issues rather than provide any significant improvements. And even if done well/correctly it can provide improvements at one distance/for one situation while simultaneously degrading performance elsewhere.
 
Here is my RAW image...not edited in any way. If you look at the models eyes at 100% they are just not sharp enough for my liking..but am I being fussy? As you say @simonkit it may be that I am at the lenses limits so maybe coming back from the maximum may be better, but is that not good for a £1.5k lens. Perhaps the filter needs to go, it doesn't look dirty, but maybe it isn't helping. Thanks all for your thoughts...see what you think on viewing this image if you would be so kind. http://img.gg/ENvXyXG
 
Last edited by a moderator:
Welcome to TP Nick!

You're not fussy - that would be too blurred for me but it might be the result of the extremes of 200mm @ f2.8. I don't know the lens though. You're at 1/400sec which - again without knowing the capabilities of the VR would be too slow for *me* to get a steady shot hand held.

There's also a terrible haze to the image - was the weather poor? Reminds me of Los Angeles smog almost!

The only other thing I noticed was that the image does seem to be sharp in front of the model. Again, this could be my eyes. I've cropped a section for display here... Will be interesting to see what others weigh in with.

fFthGq1.jpg
 
Last edited:
What I see appears to be some form of veiling glare reducing contrast, that could be the filter. It is also apparent that the focus point is not on her face, it is more inline with her shoulder/hip (the focus point is showing as being on her left arm/elbow).

Also keep in mind that the AF system cannot focus on really small details w/in the frame (i.e. her eye/eyelashes) because it cannot discern them. They will be in focus as a consequence of some larger nearby detail being in focus (i.e. the outline of her nose or her hair).
 
Thanks so much for your thoughts @Harlequin565 . All that you have written seems fair enough. The haze is from a whole load of sunlight, so that's my fault to be fair. On this shot I was indeed handholding at 1/400s but I did this as the model was sat quite still...Whenever I shoot movement I am anywhere around 1/3000s+. So yes, perhaps some camera shake even at that speed. What I am really interested in is that you have found a sharp spot, so that immeditaely makes me ponder over whether I have calibrated right. Ill keep this thread open in the hope that others may give me some more things to think about. But thanks for taking the time out (and warm welcome) to have a look at this for me :)
 
Thanks lots @sk66 Yes, I wonder about that flare. What I do know is that the issue is not with my camera (unless both my D700s are wonky), so that puts it down to the lens or filter...I'm seriously beginning to think about the filter now...that would make sense. I'm inclined to agree with you about where the focus is locking in...my easiest point for reference is the seam of her shorts which then goes on to show that the seam of her top is slightly out...and both her top and her face are behind the focus point of her shorts. I'm not sure why this is, but maybe another dose of calibration is in order!
 
Also following on from @Harlequin565 If you follow the sharp point on the structure, downwards, there is a line on the wall that looks to be almost in line with her shorts. I need to calibrate, shoot at a faster handheld speed and consider getting rid of my filter I think!
 
Nick Radley said:
Also following on from @Harlequin565 If you follow the sharp point on the structure, downwards, there is a line on the wall that looks to be almost in line with her shorts. I need to calibrate, shoot at a faster handheld speed and consider getting rid of my filter I think!
Click to expand...
If you are going to shoot with seriously shallow DOF, then you need to be very specific/careful about where the selected focus point is placed. The image shows that the center focus point was used, either due to laziness or using focus/recompose... neither is optimal.
I would suggest you remove the filter, disable calibration, and see how it goes... AF/calibration is a finicky thing and it is never 100% consistent/accurate.
 
I quite agree @sk66 I think it sounds like a combination of factors that have been highlighted to me here...I use focus/recompose too so I think I may need to use back button in future! Thanks so much for your help. An indeed a big thank you to everyone else who replied :)
 
Nick Radley said:
I quite agree @sk66 I think it sounds like a combination of factors that have been highlighted to me here...I use focus/recompose too so I think I may need to use back button in future! Thanks so much for your help. An indeed a big thank you to everyone else who replied :)
Click to expand...

Using back button focusing won't stop you from using the "focus & recompose" method of focusing. You need to use the individual focus points and highlight the one over the subjects eye, thus negating the need to recompose the shot.
 
sk66 said:
IME, poor/improper calibration is more likely to cause issues rather than provide any significant improvements. And even if done well/correctly it can provide improvements at one distance/for one situation while simultaneously degrading performance elsewhere.
Click to expand...
This is why I’ve stopped using Nikkor lenses.

I’ve gotten so fed up of not being able to properly calibrate Nikkor lenses and the ones that I’ve used suffer from poor qc/poor copy variance and focus shift. Until Nikon can provide a dock similar to Sigma/Tamron their lenses simply cannot provide critical sharpness at ALL focus distances. Sigma/Tamron can.
 
ryanyboy said:
Using back button focusing won't stop you from using the "focus & recompose" method of focusing. You need to use the individual focus points and highlight the one over the subjects eye, thus negating the need to recompose the shot.
Click to expand...
My only issue with the outer focus points is that they are not always as accurate as the centre one. I know it’s body dependent. My D750 was not perfect on all the outer points so they became unreliable. My D600 is worse again and I can only really use the centre point if using wide open primes at closer distances.

I’d love to try a D810 to see how that compares.
 
If you got your hands on the d810 you would not want to let go best Nikon I have owned
don't know if this helps or not but here goes exif date left in. (Garden fence)

Fg88xD1.jpg

Hand held Nikon d810 with 70-200 mm Nikon lens
unedited in Jpeg fine
 
Last edited:
From personal experiences ,I would first off re-set one camera to zero and test it one for one with the second body to establish what’s going on ,use your focus points as needed as well .
Next even if you have bought your filter from a top notch store,take it off put it in a drawer and forget it .there are so many clones ,copies etc on the market it’s not worth the bother .and as far as I’m aware usually the front element of most lenses is simply plain glass anyway so basically a filter in its own right ..
Sharpness is usually defined a few stops down so f5 to f8 should be optimum
 
i took the filter off all my lens's and i think it makes a difference
 
Nick Radley said:
Thanks @ryanyboy After all these years of shooting at the centre focus point, this is likely to be quite a learning experience!
Click to expand...

Using the centre AF point with focus-recompose technique is usually a very reliable and accurate method. Focus-recompose only becomes significantly inaccurate when a) you're very close, b) using a wide lens, c) the main subject is well off to one side of the frame, and d) shooting at a low f/number. At other times, there are numerous other more likely causes of poor sharpness. In this example, there is zero possibility of focus-recompose being a problem per se.

On the other hand, shooting into the light, using a filter (particularly on a longer lens), and with a lens that may not be properly AF calibrated - they're all possible causes. A common problem with AF calibration is using a target that's too close, because it's easier to see errors that way. That's quite likely to put things out at longer/normal shooting distances.
 
HoppyUK said:
Using the centre AF point with focus-recompose technique is usually a very reliable and accurate method. Focus-recompose only becomes significantly inaccurate when a) you're very close, b) using a wide lens, c) the main subject is well off to one side of the frame, and d) shooting at a low f/number. At other times, there are numerous other more likely causes of poor sharpness. In this example, there is zero possibility of focus-recompose being a problem per se.

On the other hand, shooting into the light, using a filter (particularly on a longer lens), and with a lens that may not be properly AF calibrated - they're all possible causes. A common problem with AF calibration is using a target that's too close, because it's easier to see errors that way. That's quite likely to put things out at longer/normal shooting distances.
Click to expand...

It’s D the vast majority of the time. Combined with being close to the subject it’s incredibly easy to shift the plane of focus when you recompose. I don’t think anyone can say for sure if the focus/recompose method has caused the problem in this particular case but to say there’s “zero possibility” isn’t particularly helpful. I would strongly advise anyone shooting portraits at large apertures to avoid focus/recompose. To the OP, seriously - get used to using the full range of focus points, your ratio of in focus images will improve.
 
HoppyUK said:
A common problem with AF calibration is using a target that's too close, because it's easier to see errors that way. That's quite likely to put things out at longer/normal shooting distances.
Click to expand...

That is the problem with Nikon lenses/bodies you can only af fine tune at one distance. So the lens becomes soft at other distances. Your right about the close up showing more errors. My Tamron 85 1.8 needed -16 at mfd, -3 at regular portrait distances and -3 at infinity. It got all of that via the usb dock and I simply don’t have to accept the mediocrity of the Nikon 85 1.8 when using af with wide apertures.
 
ryanyboy said:
It’s D the vast majority of the time. Combined with being close to the subject it’s incredibly easy to shift the plane of focus when you recompose. I don’t think anyone can say for sure if the focus/recompose method has caused the problem in this particular case but to say there’s “zero possibility” isn’t particularly helpful. I would strongly advise anyone shooting portraits at large apertures to avoid focus/recompose. To the OP, seriously - get used to using the full range of focus points, your ratio of in focus images will improve.
Click to expand...
Focus and recompose generally causes a back focus error, not a front focus error as is evident in this example...
maybe I shouldn't have mentioned it...
 
ryanyboy said:
It’s D the vast majority of the time. Combined with being close to the subject it’s incredibly easy to shift the plane of focus when you recompose. I don’t think anyone can say for sure if the focus/recompose method has caused the problem in this particular case but to say there’s “zero possibility” isn’t particularly helpful. I would strongly advise anyone shooting portraits at large apertures to avoid focus/recompose. To the OP, seriously - get used to using the full range of focus points, your ratio of in focus images will improve.
Click to expand...

It would only be unhelpful if it wasn't true, but in this case it is.

The fundamental issue with focus-recompose is there's an angular shift when recomposing for final framing, which changes the effective focusing distance. Basic geometry* but it's only significant when the angular shift is large, ie when you're very close, with a wide lens, and the subject is well to one side of the frame. Then if depth-of-field is very shallow, it will show an error (focus moves behind the subject) but none of those conditions applies to this particular shot even at f/2.8 TBH.

Focus-recompose gets blamed for a lot of focusing errors when there are often other more likely causes. Most commonly, the subject and/or camera simply moved between focusing and shooting. A couple of inches movement is often imperceptible at the time but nobody stays perfectly still for more than the briefest moment. And in practise, that is the main problem I find with focus-recompose - you've got to keep focusing and refocusing constantly, it interrupts the flow and is easy to forget.

*Edit: for the benefit of others, the error with focus-recompose is the difference between the focused plane, that runs straight and parallel to the sensor, and the shooting distance that runs in an arc, pivoting around the camera. When the subject is in the centre of the frame they're both the same, but when the camera is pivoted and the subject recomposed off-centre, the focused plane moves progressively behind the subject as the angle increases.
 
Last edited:
You must log in or register to reply here.
Back
Top