D7000 AF Tune and Back Focus

[gerry]

Ever since owning the D7000 and in particular using the 17-55mm f2.8 lens I have suspected that there may be a focus issue or misalignment, I did correct this on the lens fairly earlier on but did not do a proper focus check (ie tripod, chart etc) until recently, at which time I thought I would do most of my AF lens. Note you can also MF lens but would need to use the default option I guess - this would particulary useful if you relied heavily on the focus indicator when using MF lens and the lens/body was out by a fair bit.

So first thing first, focus charts, there are a few freebie ones out there which you can download with varying levels of contrast and details, however these are all simple flat paper ones where you simply place the chart on a angle and focus on the contrasty bit. I did this originally but found to be a bit hit and miss since you are focusing on something that is on a slope which was/is a bit inaccurate, a contrasty item perpendicular to the lens would be a much better idea. Now, with most good ideas, of course someone has already deisgned and manufactured one - Datacolor Sypder people have a SpyderCal (from memory), this is a trick device with levels and tripod mounts and retails for at least AUD80, yep 80 bucks...no i am not spending my dosh on that...

A far simpler option and cheaper is to fabricate your own, it ain't rocket science and unless you get paid solicitor rates you will come out ahead if you can do it in half an hour :)

All I did was bend up a bit of 0.032" thick aluminium and take a corner out to hold the ruler and/or chart. I did size it up so my typical angle would be about 25 degrees. I did up a contrasty square to focus on (in a trusty cad program, but paint would work just as well) and stuck that to the vertical surface. I stuck part of a chart on the flat bit of AL and stuck a ruler next to that, ideally you would get the vertical face to line up with a major graduation on teh ruler, however that is for MK2. Cost, 30 minutes of my time.
http://gerry.avernus.com.au/slide/files/2391_o0aej/GJB_6504.jpg
components
http://gerry.avernus.com.au/slide/files/2392_q6kfs/GJB_6505.jpg

The D7000 has the AF tune feature, basically it has a default value, this is applied if a unrecognised/unsaved CPU chip lens is attached, contrary to what I intuitively thought, which was that it was a global default value applied to all CPU lens (and hence the saved value on top of that, that would be nice since you could get 40 units of adjustment). It has a saved value for each lens attached (up to 12 lens) which does not get applied until you make some AF tune adjustment (you can apply zero to get to be listed). The saved value gets you plus/minus 20 units adjustment where minus brings the focal plane towards the camera. I do not know if these relate to a actual measurement, it would be nice if they did, would make it easier to adjust.

So onto the actually results and the interesting part.

Two of my lens, the 17-55mm and the 28mm f2.8 maxed out the available AF tune at minus 20, with the 28mm f2.8 needing still more adjustment, pain in the rear I say.

All lens typically back focused with the exception of the 300mm and 50 1.8. The 50 1.8 is interesting since I never felt I got sharp results with it on the D80, however the 300mm was always sharp on the D80. I guess the next thing is to check all the lens on the D80 and see what that presents. I reckon that will show up two extremes of nikon tolerances.

Here is my list of lens and their AF tune values, all this leads me to believe that my D7000 is on one end of the manufacturing tolerances with it typically back focusing.

17-55mm f.8 - minus 20

55mm f2.8 - minus 5

11-16mm f2.8 - minus 7

18-200mm f3.5-5.6 - minus 15

50mm 1.8 - zero

28mm f2.8 - minus 20 (and still needing more)

300mm f4 - zero

300mmf4 + TC1.7II - minus 12

So here are a couple of shots of the 17-55mm - Note the focal plane is on the 122mm mark

The first is no adjustment and the crop
http://gerry.avernus.com.au/slide/files/2384_pvmgk/GJB_6464-17-55mm.jpg
http://gerry.avernus.com.au/slide/files/2383_pwpsd/GJB_6464-17-55mm-crop.jpg

The next one is with minus 20 adjust and the respective crop
http://gerry.avernus.com.au/slide/files/2386_qrsjn/GJB_6468-17-55mm_minus20.jpg
http://gerry.avernus.com.au/slide/files/2385_smgl6/GJB_6468-17-55mm_minus20-crop.jpg

So here are a couple of shots of the 28mm - Note the focal plane is on the 122mm mark

The first is no adjustment and the crop

http://gerry.avernus.com.au/slide/files/2388_um5nv/GJB_6487-28mmf2.8.jpg
http://gerry.avernus.com.au/slide/files/2387_mmuvm/GJB_6487-28mmf2.8-crop.jpg

The next one is with minus 20 adjust and the respective crop

http://gerry.avernus.com.au/slide/files/2390_g1wzs/GJB_6490-28mmf2.8-minus20.jpg
http://gerry.avernus.com.au/slide/files/2389_6fnjx/GJB_6490-28mmf2.8-minus20-crop.jpg

so, anyone else got any useful data on the AF tune? does anyone else here AF tune on any nikon body? if so what kinda values do you have.

[Lance B]

Both my D700 and D7000 AF perfectly with all my lenses, as far as I can tell, as I have not had any focus issues with either except for the D7000 + 300 f2.8 VRII when used with the 2x TCIII where I have dialled in +2 AF fine tune. However, I think this is only due to the fact that there is such narrow DOF at 600mm that when I am focusing on a bird, I don't/can't always focus on the birds eye and may focus on the birds side and this may mean that the eye is not 100% in focus. On other occassions where I have been able to get proper focus on the eye/s, the focus has been ok without the AF fine tune. I am still in the grips of testing this.

[gerry]

Both my D700 and D7000 AF perfectly with all my lenses, as far as I can tell, as I have not had any focus issues with either except for the D7000 + 300 f2.8 VRII when used with the 2x TCIII where I have dialled in +2 AF fine tune. However, I think this is only due to the fact that there is such narrow DOF at 600mm that when I am focusing on a bird, I don't/can't always focus on the birds eye and may focus on the birds side and this may mean that the eye is not 100% in focus. On other occassions where I have been able to get proper focus on the eye/s, the focus has been ok without the AF fine tune. I am still in the grips of testing this.

+2 adjustment is pretty small, I would suggest doing a consistent test (tripod, chart etc), this was the only way I worked out what adjustment to put in - otherwise there is too many variables when in the field to deal with.

[Sar NOP]

+2 adjustment is pretty small, I would suggest doing a consistent test (tripod, chart etc), this was the only way I worked out what adjustment to put in - otherwise there is too many variables when in the field to deal with.
For me, AF fine tune is a useless feature because indeed "there is too many variables when in the field to deal with".

[gerry]

For me, AF fine tune is a useless feature because indeed "there is too many variables when in the field to deal with".

I think useless is a bit harsh, at teh end of teh day if you have a lens that when on single point focus aimed at the eye of a portait and misses focus and you can't do nothing about it then thats useless, at least with the limited AF tune function you can 'correct' some back focusing issues.

[ecopix]

No more than + or - 10 for my 4 Nikons and one Canon. My little Oly E-520 was hopelessly out, but not even Olympus could do anything about that...
I agree fine tune is very important when using long lenses, especially fast ones. The logic is that yes, there are lots of variables in the field environment, but they should be random, that is, have a normal distribution of error around the mean. So it's worth getting the mean right to start with, otherwise what hope do you have?
A friend was devestated when he discovered his new 500mm f4 Canon was out of focus, and very relieved to learn about the focus adjust! It saves a lot of hassle for us and for the manufacturers.
So much for my cameras. It's me who needs the focus adjust - plus or minus 100 might do.

[arthurking83]

LOL! I partly agree with Sar NOP and his summary! .. OK then I did say only partly, and if I have ever or do ever use the term useless, it'd be partly in jest too.

But as Sar noted, there are far too many variables in determining a perfectly acceptable balance of focus accuracy and consistency.

I've done all the tests too, of course using tripods and correct procedure and one thing that has struck me as the only guarantee in all of that, is that focusing accuracy is probably going to be inconsistent!

That is, and gerry should be commended on his work above, but the scale of the markings made in the test bed, is only useful at close range, and as far as I'm aware, most if not all!!) lenses perform a little bit differently at close range compared to longer focused distances.

In an ideal world, we'd all be shooting our favourite scenes focused at the exact same focus distance, and nothing else, and you'd be seeing massively proportioned MTF and LWpH figures.

I have two lenses that work in clear opposition at variying focused distances. Up close both my Sigma 50/1.4 is variable in it's accuracy, but out beyond the 2m(approx) range it's dead on
Same with the Nikon 105VR It seems to have a massive ability to produce sharp images, but again only within a weirdly defined focusing distance. Too close(macro) and it seems to lose some sharpness, at infinity it's also good - v.good, but not excellent. In this (say) 0.5m to 1m range it seems to produce eyeball piercing detail sharpness sometimes. It's all very weird, and has been happening too often for it to be simple gear related or simple and basic user incompetence. While I'm no expert, I'm definitely not expert enough to get consistently excellent results only within a focus distance range of 0.5m and 1m.
I've been meaning to check this hypothesis out for a while now, but either keep forgetting too, or have been too busy too.. and too lazy to devise a proper test procedure to do so.

As for using af fine tune on a zoom. never.. ever again. Did this with my 80-200/2.8 with bad results, and then same deal with the Tammy 28-75/2.8.

The Nikon lens only required it at the 200mm setting and at a focused distance of about 3-4m, and using AF fine tune(or about -15 to -20) only created more problems than it solved. Good for 200mm and 3-4m, but very bad for any other focal length, or 200mm and near to infinity!
Same(ish) deal with the Tammy. @ 75mm it backfocused and using AF fine tune, resulted in much better(and close to astonishing) sharpness at 75mm, but all other focal lengths caved in to the pressure of a now different focus distance and produce nice soft looking high detailed bits(great for smoothing out wrinkly skin tho! :p)

Nup! I'm like Sar. I've stopped using AF fine tune, and only use it to confirm any lens defects(will use it to confirm my 105VR kens thoughts). Once a lens is 'confirmed' to be suss to misfocusing, notes are taken, and a new technique is attempted from this point on.
What I should do is to see if I can take the offending lenses to a service centre to get them properly recalibrated.

This last point is why I don't label the AF fine tune feature as useless. Because it's use(for me) is now only to confirm my suspicions.

One last point about focus tuning(that I've never seen commented upon) is the question of defocusing the lens after ever focus accuracy test run. Not so much do you defocus the lens before ever test run, but more on the point of in which direction do you defocus.. towards infinity, or towards MFD? Do you take note of this variable.

As MF lenses were mentioned, I too had an 'inaccurate' reading focus confirmation light when using my 50/1.2 on my D300. BUT! it turned out that the inaccuracy was only registering when I focused from either infinity or from MFD(can't remember which one now as it was a long time back now).
That is, (and lets say confirmation light was inaccurate if I focused from infinity back to subject distance).. when I focused to subject distance from MFD(or near, or infornt of the subject) the light woudl immediately indicate at the perfect focus point :th3:.. but if I focused from behind and gradually to subject distance, I would then see confirmation when I was slightly behind the focus point! :confused:
This idea that perfect focus was a zone.. on a 50mm f/1.2 lens never sat too well with me, as this zone was so razor thin, it made it impossible to achieve decent results.. even in a hit and miss manner! :p
I went with a KatzEye focus screen instead and have never looked back.
FWIW, I also tried AF fine tune to no avail, and had to resort to manually adjusting the MF mirror stop to get focus confirmation more in the central part of the focus zone, as opposed to the near end.. but this wasn't ideal.. and once I fitted the KatzEye screen had to re set the MF mirror stop again .. or back for perfect viewing pleasure again.


My only recommendation or CC of the tool the OP has made here is to add the possibility of larger graduation markings(ie. not just so fine) so as to be able to meaningfully test the focus accuracy from further distances too.(or more accurately distances more likely to be used in real world situations).

my summary would be more along the lines of 'pretty much useless' :D

[Sar NOP]

Thanks Arthur for developping the meaning of the word "useless" ! :D:D:D

[Lance B]

+2 adjustment is pretty small, I would suggest doing a consistent test (tripod, chart etc), this was the only way I worked out what adjustment to put in - otherwise there is too many variables when in the field to deal with.

I have had a lot of experience with regards to AF fine tuning with my days with Pentax cameras and adjusted them using both test charts and in the field testing. What I did find was the test charts only told a small part of the story and that in the field testing was the real answer. In fact, in the field testing was where I achieved the best results. However, my Nikon gear is pretty much spot on but having had experience with the D7000 and the 300 f2.8 VRII + 2x TCIII, +2 adjustment is as much as I need, and yes it is a small amount, but that is all that it actually needs. I mean, if you are under the impression that +2 is small and therefore shouldn't be taken seriously, then why have an adjustment for +/-1 or +/-2 and not have +/-5 as the minumum increment? The fact of the matter is, I have adjusted it by +2 and it seems spot on.

The trouble with the AF fine adjustment is that if you are using a zoom, then at one end the AF fine adjustment may be perfect, but at the other end it may be out of whack. Also, some lenses also tend to focus differently with differing apertures and more importantly, at differing subject distances!

What I do know is that the best way to get accurate focus is to get to know your gear.

[old dog]

I have thought about fine tuning my lenses but to date have avoided it. When I first got the camera I had a bit of difficulty getting correct focus but what Lance said about getting to know your gear is right on. Now I seem to get most shots in focus or what I consider to be acceptable focus. Might still try out the tests though. I`ll make a test chart with the rule next week and have a go.

[gerry]

LOL! I partly agree with Sar NOP and his summary! .. OK then I did say only partly, and if I have ever or do ever use the term useless, it'd be partly in jest too.
snip


Good information Arthur, put me to sleep, but good oil nonethless :p.

I definitely hear and agree with the issues surround zoom lens and trying to AF tune them, obviously testing throughout their range is a minimum. I did this with the 17-55 since this was the lens (in my case) that was giving me the most grief. In terms of focus technique etc, that is all well and good but when you have had lens perform a particular way on another body then start noticing consistent back focus (as was my case) on a newer body then the AF tune is handy.



I have had a lot of experience with regards to AF fine tuning with my days with Pentax cameras and adjusted them using both test charts and in the field testing. What I did find was the test charts only told a small part of the story and that in the field testing was the real answer. In fact, in the field testing was where I achieved the best results. However, my Nikon gear is pretty much spot on but having had experience with the D7000 and the 300 f2.8 VRII + 2x TCIII, +2 adjustment is as much as I need, and yes it is a small amount, but that is all that it actually needs. I mean, if you are under the impression that +2 is small and therefore shouldn't be taken seriously, then why have an adjustment for +/-1 or +/-2 and not have +/-5 as the minumum increment? The fact of the matter is, I have adjusted it by +2 and it seems spot on.

The trouble with the AF fine adjustment is that if you are using a zoom, then at one end the AF fine adjustment may be perfect, but at the other end it may be out of whack. Also, some lenses also tend to focus differently with differing apertures and more importantly, at differing subject distances!

What I do know is that the best way to get accurate focus is to get to know your gear.

Lance, don't take my reply as having a dig at you, that was certainly not the case. I barely noticed a few point increments on teh wider lens, on the 300 mm f4 i noticed the AF tune more prominent and a +4 made significant visible difference.

[Lance B]

Lance, don't take my reply as having a dig at you, that was certainly not the case. I barely noticed a few point increments on teh wider lens, on the 300 mm f4 i noticed the AF tune more prominent and a +4 made significant visible difference.

No problems, Gerry. I just wanted to make sure you realise that I am not a novice at this! :) In fact, I noticed FF/BF issues way back when I had a Pentax *stD (2004) and on my K10D (late 2006) and I was hoping there was something that could be done about it and then Pentax introduced this feature on the K20D (early 2008). What a Godsend, as it allowed me to adjust for BF/FF on my 2 of my favourite lenses! :)

I think +2 works for me and the 300 + 2x TCIII because it is such a long focal length, ie 600mm, and therefore any minor difference shows up readily. When I look at the difference it makes, we are talking about probably only needing to focus 10-20mm further back at this focal length at probably 4-5mts! :)