D7100 Inconsistent Focus Tracking Errors

Paganman2

Senior Member
DSC_0872_DxO.jpg

And a static again with AF-A

P.
 

Eyelight

Senior Member
View attachment 142016


View attachment 142017

These were both taken within a few seconds of each other with exactly the same settings, the only difference was with the first one i just shot the plane without following it with tracking, in the second shot i followed the plane with the center focus bracket on the plane, but as normal it was not locking focus as the focus confirm brackets were dancing like crazy.
Bot these shots are 100% crops direct from Nef files and both have had exactly the same editing done - noise control sharpening etc, so in theory they should look exactly the same.

Any ideas folks please...

Pag.
@Paganman2 Curious if you know the aircraft and what your guess (or knowledge) on height above ground would be? I'm just fiddling with some numbers.
 

Paganman2

Senior Member
@Paganman2 Curious if you know the aircraft and what your guess (or knowledge) on height above ground would be? I'm just fiddling with some numbers.


Hi there, yep Airbus A346 Virgin Atlantic flying at the time of exposure about 34.000ft, it had just been given its way point clearance to track to a reporting point before starting its oceanic crossing over the pond.

P.
 

Paganman2

Senior Member
What happened between the two shots was - i took the first and sharp one, then brought my camera away from my eye, i then decided while the plane was still within range to take another shot as it was getting more overhead at this point, i put the center focus bracket dead on the center of the plane belly and followed with the plane for a few seconds to let the Vr settle, then i breathed in and took the shot and held focus on the plane even after exposure.

P.
 

Paganman2

Senior Member
I know i have raised this so many times but i dont understand why my AF-C mode is acting up so much? it is being so super sensitive and just not wanting to confirm focus, its made me completely loose faith in its ability and resort to using the AF-A mode all of the time unless i am in doors shooting still life.
Just to expand further my focus does not hunt or go in or out of focus once achieving initial focus in AF-C mode, but what does happen is the focus indicators in the viewfinder either side of the focus confirm lamp just go crazy, dancing from side to side without the focus confirm lamp lighting up.
This tells me as does the manual that focus can not be made - try another medium or manual focus.

P.
 

Eyelight

Senior Member
Hi there, yep Airbus A346 Virgin Atlantic flying at the time of exposure about 34.000ft, it had just been given its way point clearance to track to a reporting point before starting its oceanic crossing over the pond.

P.

I should have guessed. I had it as a A340 and calculated the height at 34,248.78 feet based on the image size.

At that distance, you could set your focus manually and not use AF. Reason is using f/7.1 at 240mm focal length the hyperfocal distance is 1,326.53 feet; meaning everything from 1326.53/2 (half the hyperfocal distance) to infinity would be in an acceptably sharp focus.

You can do this on a lens with a distance markings and DOF scale by setting the infinity mark just inside the aperture value you are using. If the lens does not have a DOF scale, you can use a DOF calculator to find the hyperfocal distance, and then focus on an object that is at a greater distance.
 

Eyelight

Senior Member
I know i have raised this so many times but i dont understand why my AF-C mode is acting up so much? it is being so super sensitive and just not wanting to confirm focus, its made me completely loose faith in its ability and resort to using the AF-A mode all of the time unless i am in doors shooting still life.
Just to expand further my focus does not hunt or go in or out of focus once achieving initial focus in AF-C mode, but what does happen is the focus indicators in the viewfinder either side of the focus confirm lamp just go crazy, dancing from side to side without the focus confirm lamp lighting up.
This tells me as does the manual that focus can not be made - try another medium or manual focus.

P.

Cameras don't see as we see. It may not be "seeing" the craft and shooting at skies can be finicky. Try the manual focus technique I mentioned in the last post. You could practice this on some distance objects on the horizon.
 

Paganman2

Senior Member
Cameras don't see as we see. It may not be "seeing" the craft and shooting at skies can be finicky. Try the manual focus technique I mentioned in the last post. You could practice this on some distance objects on the horizon.

Thank you for your input and advice, i guess i just like the auto convenience with AF and feel that it should not have a problem, my d90 never did either with a 55-300Vr or my current 70-300Vr, i am currently working through my old d90 pictures cropping and editing them, and i cam tell you with identical shots to those i have shot with my d7100 from the same location weather etc, i have only come across about 5% that are out of focus or show any signs of shake, this compares to about 40-50% out of about 300 shots of aircraft taken with my d7100, that is quite shocking to me.

P.
 

Eyelight

Senior Member
Thank you for your input and advice, i guess i just like the auto convenience with AF and feel that it should not have a problem, my d90 never did either with a 55-300Vr or my current 70-300Vr, i am currently working through my old d90 pictures cropping and editing them, and i cam tell you with identical shots to those i have shot with my d7100 from the same location weather etc, i have only come across about 5% that are out of focus or show any signs of shake, this compares to about 40-50% out of about 300 shots of aircraft taken with my d7100, that is quite shocking to me.

P.

Though we have considered this and conjectured that, the problem will only be solved when it is eliminated. It remains a possibility it is focus related.

One thing I thought of is the D7100 focus points may actually be smaller and therefore harder to keep a single point on a small target. I looked but could not find any info on sizes.
 
Last edited:

Paganman2

Senior Member
Though we have considered this and conjectured that, the problem will only be solved when it is eliminated. It remains a possibility it is focus related.

One thing I thought of is the D7100 focus points may actually be smaller and therefore harder to keep a single point on a small target. I looked but could not find any info on sizes.

I too can understand that, the d90 had an option to increase the size of the center focus box making it almost twice the size, but i assume it had a single x type sensor but only active at f5.6 or larger(i know the d7100 is f8) i have tried dynamic 9 but i wonder if that again mite throw the focus off i have seen where it has chosen the trail over the plane and that has been in focus in stead, such a shame it isnt just a bigger f8 x sensor with additional sensors normal line sensors or switchable(say you could switch x or line sensor on or off or mix?

P.
 

Paganman2

Senior Member
It may not help but look at page 232 of the manual ref AF-C settings and try putting a short delay in the refocus time.


Hi there mike, thanks for the input, i have decide after some other replies to leave focus tracking off, i have also decided after doing a full reset and a power off reset like you can do with laptops etc, i have also changed the focus point allocation from the 51 point to the 11 as perhaps that may enlarge the center point like it does in other cameras as i think this may be one of the 7100 problems - the center focus sensor is too small yet the 9 point dynamic has spaces where the focus can drift off, it needs a joined together cluster or larger central one, - NIKON Are You Listening........


P.
 

Paganman2

Senior Member
I have been doing a bit of studying about these types of Focus modules - D4, D800, D750, D7000 and D7100 and it appears that the cluster of focus points around the center one are only line sensors above f5.6 eg f6,f7.1,f8, and become cross type at f5.6 and below, only the center one is cross type at f8, so this i assume will have an affect on the quality of focus capture if using Dynamic and the focus moves off the center point.
Any thoughts or knowledge folks...

P.
 

Paganman2

Senior Member
I always understood the center point was the most sensitive,the best one for slower lenses or poor light.

Yep f8 center is a good thing, but looks like to get a decent response with tricky subjects when using dynamic i will have to shoot at f5.6 not the best setting with my 70-300Vr, so i guess shortening the zoom a bit to say 250mm will be a necessity to keep the IQ reasonable?

P.
 

Eyelight

Senior Member
I should reiterate the idea of using manual focus. I'm thinking the lens in use has a distance scale, so all you need to do is set it at or just inside the infinity mark.

I also have some new, albeit old info, from my circa 1986 Kodak Professional Photoguide I located yesterday while rummaging through the gear box.

The guide has a formula for calculating the stop-action shutter speed for a given distance, focal length, subject speed etc. Took me a bit to get it set up in Excel so I could see how it works. It uses the same CoC (circle of confusion) used for DoF and other blurry calculations.

Cutting to the chase, I input the following:

Distance = 34,000 feet
Speed = 400 MPH <<<< should be a bit high
Focal Length = 240mm (DX)

and the shutter speed outputted is: 1/84 seconds

So, theoretically if you were to set the camera on a tripod and catch the plane as it traversed the view, 1/84 second shutter speed would be enough to "stop the action". I'd be real curious if this would work.

Assuming the calculation is correct, what it means is, as long as the shutter speed is enough to overcome camera shake, it would be plenty to stop-action the plane.

And I have an idea to check something. If the lens has a distance scale, point it at the clear blue sky and engage the AF while watching the scale. If the lens is trying to focus past infinity, that could explain how a missed focus at 34,000 feet is blurry. Someone else might confirm this, but I'm thinking the limit on the turn of an AF lens is actually past the focusing capability, so if the FP misses the airbus, you get the blur.
 

J-see

Senior Member
Yep f8 center is a good thing, but looks like to get a decent response with tricky subjects when using dynamic i will have to shoot at f5.6 not the best setting with my 70-300Vr, so i guess shortening the zoom a bit to say 250mm will be a necessity to keep the IQ reasonable?

P.

Your AF works at your lens wide open regardless what aperture you set. You can close down to f/40 and it'll still AF at 5.6 if that's the widest the lens can go. That's also why there's this button to check DoF. Only when you trigger it'll pick the f/ you set.

You don't need to worry about setting a certain aperture depending the focus method since it makes no difference.
 
Last edited:

Paganman2

Senior Member
I should reiterate the idea of using manual focus. I'm thinking the lens in use has a distance scale, so all you need to do is set it at or just inside the infinity mark.

I also have some new, albeit old info, from my circa 1986 Kodak Professional Photoguide I located yesterday while rummaging through the gear box.

The guide has a formula for calculating the stop-action shutter speed for a given distance, focal length, subject speed etc. Took me a bit to get it set up in Excel so I could see how it works. It uses the same CoC (circle of confusion) used for DoF and other blurry calculations.

Cutting to the chase, I input the following:

Distance = 34,000 feet
Speed = 400 MPH <<<< should be a bit high
Focal Length = 240mm (DX)

and the shutter speed outputted is: 1/84 seconds

So, theoretically if you were to set the camera on a tripod and catch the plane as it traversed the view, 1/84 second shutter speed would be enough to "stop the action". I'd be real curious if this would work.

Assuming the calculation is correct, what it means is, as long as the shutter speed is enough to overcome camera shake, it would be plenty to stop-action the plane.

And I have an idea to check something. If the lens has a distance scale, point it at the clear blue sky and engage the AF while watching the scale. If the lens is trying to focus past infinity, that could explain how a missed focus at 34,000 feet is blurry. Someone else might confirm this, but I'm thinking the limit on the turn of an AF lens is actually past the focusing capability, so if the FP misses the airbus, you get the blur.

Thank you for the hard work and time you have put into this - i do appreciate it, i will consider manual focus but i guess i am a bit "stick in the mud" and had to manage with manual focus for many years before AF came out, i am also of the option that if this camera can't do what i want it to then no camera will, as it is supposed to be one of the best and equal to the D4S.

P.
 

Paganman2

Senior Member
I was playing with the focus and those shiny bird things - and in overcast conditions with less sun, my AF did seem more accurate and locked on better, its as iff it doesn't like the sun or reflections?

P.
 

J-see

Senior Member
I was playing with the focus and those shiny bird things - and in overcast conditions with less sun, my AF did seem more accurate and locked on better, its as iff it doesn't like the sun or reflections?

P.

It needs something to focus upon and sun/reflection makes it very hard. Anything with a lot of contrast is easy for the cam while anything monotone or less defined makes it hard.

If you have the chance, point it at a brick wall to focus and then try that same on a solid white wall or whatever wall that is featureless.
 
Last edited:
Top