Key backfocus information/solution

Somasipper

Active member
Messages
89
Reaction score
0
Location
US
Before purchasing a D70 I researched the possibility of this issue thoroughly. I found the solution before I even had a problem. Hope it helps you too. This is my summation of what I found.

There are some important mechanics of focus you should remember before doing a backfocus test.

First off, and the most important thing to understand before claiming back focus, Anything closer to the camera then the focus distance will be blurred twice as much as something behind the point of focus. Let’s say you focus on a tree 20 feet away. A second tree that is 40 feet back will look just as focused as a third tree 10 feet away. In other words things behind the focus point are twice as clear as things in front of it. This is why we see a lot of people complaining about back focus but no front focus. It is more often then not a misinterpretation especially on the famous ruler test.

Secondly, The auto sensor (only the central one) covers a large amount of the frame. The camera will not necessarily focus on the middle of the sensor, if there is something further back but still within the auto focus zone, it may pick, and often does because its easier to find contrast in small thinner features. This is true even in closest subject mode, which chooses from which of the 5 auto focus zones are closest, but not what is closest in any one zone.

Finally, all lenses do not have infinitely perfect focus motors. They move in spurts, and will only reach an approximation, never the exact point. Combine this with the other two points and you will see back focus in any camera at random times. Higher quality lenses, and many macro lenses will have finer focus motors, but in the end the camera and lens can never really focus exactly where you want it to, nor will it know exactly where you want to focus at all. It’s a machine that tries to approximate a goal, and its your job to learn how to make the machine work properly.

This camera uses the same auto focus system as many other camera that have no problems other then occasional manufacturer defect.

With that said, one thing can be concluded. .When it is used with the same lenses used for years on other Nikons, there is not going to be a higher rate of focus issues then with any other camera with Cam900 auto focus(n55,n75,S2,D100,list goes on……).

Since the kit lens is not widely used on any camera other then this one I will not include it in this statement. It may have an unusually coarse auto focus motor, but I haven’t noticed it myself.
 
Yes good point! I read people complained backfocus... For me nothing
nothing wrong with it. Your message is make sense. I played and see
lot roses and I noticed, ah!..

Anyway but one thing, I wonder if they use CENTER with focus area? or Do they use select focus area (5 points area)? or do they complain backfocus with dymantec focus area such as scene landscape or AUTO scene? I had no idea what they try.
Before purchasing a D70 I researched the possibility of this issue
thoroughly. I found the solution before I even had a problem.
Hope it helps you too. This is my summation of what I found.

There are some important mechanics of focus you should remember
before doing a backfocus test.
First off, and the most important thing to understand before
claiming back focus, Anything closer to the camera then the focus
distance will be blurred twice as much as something behind the
point of focus. Let’s say you focus on a tree 20 feet away. A
second tree that is 40 feet back will look just as focused as a
third tree 10 feet away. In other words things behind the focus
point are twice as clear as things in front of it. This is why we
see a lot of people complaining about back focus but no front
focus. It is more often then not a misinterpretation especially on
the famous ruler test.

Secondly, The auto sensor (only the central one) covers a large
amount of the frame. The camera will not necessarily focus on the
middle of the sensor, if there is something further back but still
within the auto focus zone, it may pick, and often does because its
easier to find contrast in small thinner features. This is true
even in closest subject mode, which chooses from which of the 5
auto focus zones are closest, but not what is closest in any one
zone.

Finally, all lenses do not have infinitely perfect focus motors.
They move in spurts, and will only reach an approximation, never
the exact point. Combine this with the other two points and you
will see back focus in any camera at random times. Higher quality
lenses, and many macro lenses will have finer focus motors, but in
the end the camera and lens can never really focus exactly where
you want it to, nor will it know exactly where you want to focus at
all. It’s a machine that tries to approximate a goal, and its your
job to learn how to make the machine work properly.

This camera uses the same auto focus system as many other camera
that have no problems other then occasional manufacturer defect.

With that said, one thing can be concluded. .When it is used with
the same lenses used for years on other Nikons, there is not going
to be a higher rate of focus issues then with any other camera with
Cam900 auto focus(n55,n75,S2,D100,list goes on……).

Since the kit lens is not widely used on any camera other then this
one I will not include it in this statement. It may have an
unusually coarse auto focus motor, but I haven’t noticed it myself.
 
I do understand that these distance differences work like this, that was my first engineering instinct seeing all the "ruler tests".

But the problem has been reported as bad using the 50 mm AF primes, I do not understand how that fits the explanation?
 
When used wide open or nearly so, the 50/1.8,1.4 have a depth of field that is less deep than if they were closed down more. The kit lens for example does not have as large an aperture.

Over at the Canon 10d forum, many found that that camera's focus issues seemed worse with lens apertures larger than about F2.8/3.5 and in certain distance ranges.

AX
 
It fits it well, it suggests user interpratation error. The central auto zone picks somthing further back but still within the zone, then the 50mm focus motor missess slightly, perhaps in the same direction and voila.
I do understand that these distance differences work like this,
that was my first engineering instinct seeing all the "ruler tests".

But the problem has been reported as bad using the 50 mm AF primes,
I do not understand how that fits the explanation?
 
Before purchasing a D70 I researched the possibility of this issue
thoroughly. I found the solution before I even had a problem.
Hope it helps you too. This is my summation of what I found.

There are some important mechanics of focus you should remember
before doing a backfocus test.
First off, and the most important thing to understand before
claiming back focus, Anything closer to the camera then the focus
distance will be blurred twice as much as something behind the
point of focus. Let’s say you focus on a tree 20 feet away. A
second tree that is 40 feet back will look just as focused as a
third tree 10 feet away. In other words things behind the focus
point are twice as clear as things in front of it. This is why we
see a lot of people complaining about back focus but no front
focus. It is more often then not a misinterpretation especially on
the famous ruler test.

Secondly, The auto sensor (only the central one) covers a large
amount of the frame. The camera will not necessarily focus on the
middle of the sensor, if there is something further back but still
within the auto focus zone, it may pick, and often does because its
easier to find contrast in small thinner features. This is true
even in closest subject mode, which chooses from which of the 5
auto focus zones are closest, but not what is closest in any one
zone.

Finally, all lenses do not have infinitely perfect focus motors.
They move in spurts, and will only reach an approximation, never
the exact point. Combine this with the other two points and you
will see back focus in any camera at random times. Higher quality
lenses, and many macro lenses will have finer focus motors, but in
the end the camera and lens can never really focus exactly where
you want it to, nor will it know exactly where you want to focus at
all. It’s a machine that tries to approximate a goal, and its your
job to learn how to make the machine work properly.

This camera uses the same auto focus system as many other camera
that have no problems other then occasional manufacturer defect.

With that said, one thing can be concluded. .When it is used with
the same lenses used for years on other Nikons, there is not going
to be a higher rate of focus issues then with any other camera with
Cam900 auto focus(n55,n75,S2,D100,list goes on……).

Since the kit lens is not widely used on any camera other then this
one I will not include it in this statement. It may have an
unusually coarse auto focus motor, but I haven’t noticed it myself.
--



http://www.pbase.com/nowhereatoll
http://www.nowhereatoll.com
 
A larger aperture has a small DOF than if you stop down. Maybe you were thinking of the the F-stop number associated with an aperture...those go down as the aperture goes up.

AX
 
Today I returned the second kit lens tried, got the refund, and voila - no more backfocus here! Because the other 3 lenses I got (28-200, 70-300 and 105 Macro) do not suffer from it.

Your comment about closer subjects twice as blurred as remote ones is Ok for mid-range, mid-aperture, mid-focal distance shooting. At other extremes of range, aperture and focal distance this assumption is incorrect. For example, at close distance, longer zoom, wide aperture closer objects are equally blurred as the more remote ones: yesterday I posted a link here to a site with Java applet calculating DOF distribution in front and behind focal plane depending on those parameters, illustrating that. This link was kindly given by another poster earlier.

Regards,

Alex
 
And the reason this hasn't appeared before is that you have a much harder time pixel-peeping film negs and slides. It's probably been there all along.

I know that I've done such things manually with film cameras through lack of attention. If I take my time and fine-tune the focus, all is well. But apparently there is no time to do that when shooting these days, but plenty of time for testing for non-issues.

Just my $.02
 
Im glad you got your problem fixed. I is true that the depth of field changes and is not always 2 to 1 ratio, but all these ruler tests are midzoom.
 
Bingo! Pop Photo ran an expose about 8-10 years ago about autofocus. The other thread with the robot pictures shows a slight issue when printed to movie poster size. At 8X10 there is no problem, I'd say.

Whenever I've made a good HUGE enlargement I could not expect to get it right in ONE SHOT. I had to be very careful, play with settings and take various precautions. And then select the finalist images from a batch. Some seem to want to skip all that and get a billboardable / magazine cover image at 3 FPS ant any distance or lighting until the card is full. Well, we're not there yet!

AX
 
The D70 uses the CAM-900 AF module that has been used in several Nikon bodies, including the F100 and D100. It's perplexing that people think that it works just fine... until it's used in the D70!?!?!?
 
The D70 uses the CAM-900 AF module that has been used in several
Nikon bodies, including the F100 and D100. It's perplexing that
people think that it works just fine... until it's used in the
D70!?!?!?
Actually, the F100 (which I own and love) uses the CAM-1300 module, same as the F5, and substantially quicker than the CAM-900. But I'm using the D70 principally these days.

best...aa
 
That's where your assumption is wrong. some ruler tester might use mid zoom, many or most will use the largest aperture and distance around 1-3m. And hence the DOF will and should spread half to the front and half to the back.

I posted about this sometime ago but it look like there are still ppl that don't understand that most of the backfocus are problem when the DOF is small (and hence generally the DOF is divided into 1/2 in the front and 1/2 in the back).

Try this site for DOF calculation: http://dfleming.ameranet.com/dofjs.html

For example, I saw many pictures that are clearly backfocused. So I testested my D70 at various focal length and distance to the subject. For one instance, I use focal length of 70mm f/4.5 (kit lens) with distance from the subject of 3m. This should give me DOF about 32cm and almost divided half to the fron and half to the back (15cm in the front and 17cm to the back). But the real DOF I got was very much bias to the back. The theory about 1/3 to the front and 2/3 to the back ONLY works when the distance is far away which is NOT the case of the ruler tests (most will use distance less than 5-6m which is perfectly fine for the test itself.

Just to make thing clearer.

Mocca
Im glad you got your problem fixed. I is true that the depth of
field changes and is not always 2 to 1 ratio, but all these ruler
tests are midzoom.
 
It's also depend on the mfg torelence and quality control as well. AFAIK, this D70 is not produced at the same plant as the D100. My friend D100 is made in Japan (I'm not implying anything BTW) and the D70 is made in a new Nikon plant in Thailand.

Mocca
The D70 uses the CAM-900 AF module that has been used in several
Nikon bodies, including the F100 and D100. It's perplexing that
people think that it works just fine... until it's used in the
D70!?!?!?
 
I have no backfocus - even at 1.8!

I tested both my new 50mm (received yesterday) and the kit lens both at about 3 feet from a flat test object with an angled ruler next to the flat object. I tested the 50mm at 1.8, 2.8, 4, and 8 - and none of them showed backfocus at all! But when I tested the kit lens at 4.5 and 8 (both 50mm focal length and 70mm) I had backfocus on all!

I am going to see about swapping out my kit lens for another. I really don't think its the camera or user error in my case. These tests were pretty definitive.

Thanks,
Steve
A larger aperture has a small DOF than if you stop down. Maybe you
were thinking of the the F-stop number associated with an
aperture...those go down as the aperture goes up.

AX
 
Alex,
I hope my situation turns out like yours.

I tested both my 50mm prime and kit lens. I do not have backfocus on the 50mm lens - not even at f/1.8.

I DO have backfocus on the kit lens. I am going to swap it out - hope that corrects the situation!

Thanks for the hope,
Steve
Today I returned the second kit lens tried, got the refund, and
voila - no more backfocus here! Because the other 3 lenses I got
(28-200, 70-300 and 105 Macro) do not suffer from it.

Your comment about closer subjects twice as blurred as remote ones
is Ok for mid-range, mid-aperture, mid-focal distance shooting. At
other extremes of range, aperture and focal distance this
assumption is incorrect. For example, at close distance, longer
zoom, wide aperture closer objects are equally blurred as the more
remote ones: yesterday I posted a link here to a site with Java
applet calculating DOF distribution in front and behind focal plane
depending on those parameters, illustrating that. This link was
kindly given by another poster earlier.

Regards,

Alex
 

Keyboard shortcuts

Back
Top