D90 back/front focus adjustment

[simonr23]

i bought a 50/1.8 last weekend and i noticed that at the wider aperatures it was front focusing quite severely.

to make things interesting, after downloading some focus aid thingos, i also tested my other 2 lenses (18-105VR and 70-300VR) and noticed a slight amount of back focus!

i found a guide on how to adjust the main autofocus/manula focus systems, using an allen key and gave it a shot.

wierdly, when i adjusted the camera for the 50mm lens, it also brought the other 2 lenses into 'spec' too.

it really wasnt hard, or risky. anyone who doesnt have the shakes could do it themselves. it was just a bit of trial and error to find which way to turn the allen key and then go in small increments until i went too far, and then i went back a smidgen (technical measurement)

i was pleased with my efforts

[old dog]

Hi Simon. Any links for doing this. I have a D80 and I wonder if it can be adjusted as well.

[arthurking83]

Hi Simon. Any links for doing this. I have a D80 and I wonder if it can be adjusted as well.

there are two allen key screws on one of the sides inside the miror box Graeme. The allen screw head is(from memory) 1mm, where the 2mm tool described in my link is to big for the D300(dunno about the D80?).

(AFAIK) all Nikon bodies work the same way, so here's the D70 adjustment link I have. The screws work the same for the D300 too.

Note: As I understood the D70 adjustment procedure described in that link, adjusting stopper #2 affects manual focusing only, and the AF system relies purely on stopper #1.

I adjusted my D300 in that manner, where (with manual only lenses) the manual focus light in the viewfinder would light up way before the image was in best focus, so i adjusted stop#2 to get a better manual focus confirmation. This worked a treat with the 50/1.2(wide open) and the 500 mirror(at close range) where DOF was very shallow.

The only thing i really don't like tho(for perfect accuracy) is using those focus charts at a 45° angle.
I used the 'battery method' much better(for me).

Line up a series of AA batteries in a line, but where the central one is forward of the next two, either side of the front central one, and then the next two on either side of them wil be further back from those two again.. so you end up with a series of batteries in flight formation. Focus on the central battery, and if the focusing is off and clearly backward of the correctly focus subject then then you're ready to test. If all batteries are OOF then you probably have front focusing issues so revrese the order of the batteries to suit.

keep them at a sensible distance back from the camera too. say 4 meters or whatever.. ie. not too close(unless you use those lenses only at close range. Remember the very narrow DOF is going to have a more pronounced effect on the focusing system(both AF and MF).

camera on tripod mirror lockup(exposure delay) for clearer test images to work with, and download each image as you capture it and you need to pixel peep them to get the best results. The tolerances are very slight, so you may almost invariably overdo it the first time you rotate the stopper.

hope that helps, and with perseverance you'll either get it right, or as a minumum you won't stuff it up.

ps. I didn't tape up the mirror, it's safe to hold it up with a finger from the edges.

pps. the first time you flip the mirror up you will see/hear the (hidden)AF mirror flip up into the back of the main VF mirror as well.. perfectly normal.

[old dog]

thanks Arthur for all that info. I`ll test it out this week.

[arthurking83]

This NEW link to that D70 article may be more accurate too.

[simonr23]

this is the link i used. it's just a different section of the same site as above. http://leongoodman.tripod.com/d70focuspart3.html

2mm allen key for d70-d90.

i have found that the best test is to use a black tv remote, that uses white characters. line it up and aim at particular buttons and look at the results. having some dust on it helps too
this and just aiming at a few certain things in my lounge had me set to go after about 30-40mins of tinkering. i used the test charts a bit, but found better, more consistent results the other ways.

the only annoying thing is the amount of 'throw' is limited in the camera box. you only JUST get enough room to kinda wiggle the allen key back into position for the next amount of adjustment- if needed.

[N*A*M]

interesting stuff

i wonder if my dads d80 could do with a bit of calibration

[etherial]

Ok, I have a stupid question that I'm sure there is a simple explanation for. Why do we even need to adjust for front or back focusing? Doesn't the camera focus through the lens and if it detects it is not in focus it makes an adjustment?

I understand the whole adjustment process both here manually and the Canon microadjust in the software but I'm struggling to understand how the camera gets it wrong to start with?

Please enlighten me oh wise ones!

[arthurking83]

The rotating cams will make very slight adjustments to the distance from the mirror, to the actual sensor.
(apparently???) there is scope to adjust the actual position of the sensor, but for all intents and puposes the distance to the sensor is a given value(ie. not adjustable!.. forget what I just said.. belive me that it is(even though some people have made adjustments to the position of the sensor.. it;s not for the faint hearted).
So if the postion of the mirror is adjustable, then the distance to the AF sensor must be different than it is to the sensor.. even if it's supposed to be the same distance, the fact that there is allowance to adjust the mirror(s) means that the distances to the image sensor and AF sensor don't necessarily have to be the same. Cameras get knocks, and can be skewed out of alignment, or misalighned at the factory(manufacturing tolerances.. etc).

If the AF sensor was on the same plane as the image sensor, or if the image sensor did the actual focusing, then there'd be no need for such technical endeavours(Live view rocks, but is too slow).
One day when live view focusing is as fast as the way the current technology AF sensors are, then you'd have to say that the manufacturers would surely dispense with such complicated mechanical systems.
So, where there's mirrors, there'd have to be an element of error. And the amount of adjustment you make with those adjustment cams is so infintessimally small, probably measured in 1/100ths of a millimetre.

I have no idea on how the manual focus system works, ie. the electronic rangefinder confirmation dot in the viewfinder, but after adjusting my D300, the two lenses most affected by the slight error are better.. but it's not an accurate way with which to focus anyhow, for example if the DOF is really thin.
The adjustment I made was for handheld work with my manual lenses, and for tripod use with the 500 lens, just to get a decent starting point. I don't know/understand where the sensor is for the electronic rangefinder, and only assume it's in the viewfinder, as it can't be coupled to the AF sensor!

[etherial]

Thanks Arthur, that kind of explains a reason for it, and adjustments in the camera like that make sense. What still confuses me is why different lenses would then be different?

[arthurking83]

different lenses.. not 100% sure but there has to be a few at least, that I can think of.

Autofocusing lenses usually have a CPU, camera to CPU exchange of information may be a cause.

Tolerances. every piece of mechanical equipment will have a degree of mechanical tolerance allowance. Slack in screw driven interfaces, etc.

Different lenses can mean different aperture ranges/values. Thinner DOF, more contrasty lenses with more contrasty coatings.. etc.

Andrew(I@M) had two of his lenses rechipped to get them to communicate more accurately with the camera body(then D200), and from memory, that being the only change to the system, they now work much more effectively(accurately).

[ScottM]

after reading an earlier thread about microfocussing, i looked it up. it seems the d90 isn't capable of it - yet. from my shots, i feel my mostly-'cheap' lenses cause a couple issues. there doesn't yet seem to be a firmware update of the d90 from v1.0.0 Given it's been in the market for a couple years, i'd say users and nikon haven't quite found enough serious issues or enhancements to make this worthwhile.
i'd be interested in trying such adjustments out, but am happy ATM with the focusing done by my camera. I use it 95% of the time on autofocus.

[simonr23]

something else that is explained for the reason as to why it may need adjusting in some cameras, is that, due to tolerances, etc, the actual autofocus sensor thingys arent always aligned with the markings in the viewfinder.

in the link i posted a few posts up, it goes on to explain it a bit better and with diagrams.

[mongo]

does anyone know if the D200 has such an adjustment ??

[simonr23]

i havent seen whether the d300 can be adjusted in this way, so i'm guessing the d200 might be the same.

is the d200/d80 a similar system to the d300/d90?

[mongo]

not sure about the system but MOngo understands it may use a cam2000 AF system if that means anything

[kiwi]

Interesting, first time I have ever heard of adjusting in this manner


Sent from my iPhone using Tapatalk

[arthurking83]

Mongo: yes the D200 can be adjusted in the same manner.

I found a link to a chap that adjusted his D200 and used that as my guide to adjust the manual focus stop on the D300.
Seems that Nikon must be changing the actual placement of the screws as they're different in my D70s compared to the D300. Same side same orientation(MF/Main mirror stop at the front, AF mirror stop at the rear). D300 uses a 1mm allen key, whereas it seems that the D70/80/90 use 2mm tools. It's almost a dead certainty that the D200 uses the same 2mm keys.
D200 AF system is exactly the same as the D90 Cam2000, D80 uses the CAM1000(I think inherited from the D100?) The focusing module makes no difference to these adustment stops.

The stop that adjusts the AF mirror distance is very deeply recessed and doesn't allow a large degree of turn in one hit.
Turn as far as the mirror box allows you to go with the allen key and take test shots to see if you;ve moved it the correct way.

Easy done, and just remember to count how many turns of the stop screw you're making, so that if you go the wrong way, you can easily undo that adjustment.

Note that adjusting the MF/main mirror stop which is located closer to the front of the camera won;t affect AF.. shouldn't affect AF.. and hasn't affected AF on my D300.

This adjustment is so easy and non destructive to do. Much less risk of doing any damage to the camera, than say changing focusing screens, or even sensor cleaning.
Without the tools that Nikon have at their disposal, we only have trial and error to help us get it right. where they probably have some kind of alignment doodad that gets it right in a jiffy.

Are you getting consistent focusing innaccuracies with all your lenses?

[N*A*M]

the f100 has the same stops and appear to be the 2mm heads

[rwg717]

Thanks Arthur, that kind of explains a reason for it, and adjustments in the camera like that make sense. What still confuses me is why different lenses would then be different?

Well if that's the case, do any Canon users have a link to a similar process for Canon bodies
Richard