Canon 50D Micro Adjustment, your methods?

Marcel

Marcel

Kim Jong Bod
Admin
Messages
29,411
Name
Marcel
Edit My Images
Yes
Rovers_Andy said:
A quick update.

Spent yesterday morning testing the kit with Phil (Purpleclouds) and we've got everything callibrated now using the Micro Adjustment Cfn.


The 400 is on +5 for both bodies, the 300 is on +5 and zero, and the 135 is at zero on both bodies

Results from Blackpool Vs Middlesbrough last night were much better than the PNE Vs Blackpool at the weekend :)
Click to expand...

Excellent :thumbs:

Got any tips on how to set the Micro Adjustment?
 
Marcel said:
Excellent :thumbs:

Got any tips on how to set the Micro Adjustment?
Click to expand...

Hi Marcel some handy info here from Tobers.

We tested on a static subject but i found on saturday that its not quite right yet. I'll be testing using Tobers method tomorrow morning, then fine tune it tomorrow night
 
Cheers Andy.

I'm using the infamous 45degree Focus Test Chart from here
And using a guide from dgrin here
I\ve just done the 24-70 and cycling through it seems some steps, the lens/camera has focused out of turn...ie as Im cycling through them, the point of focus moves smoothly along bit by bit with each image....then for the odd image jumps out. I can only assume that time the camera is focusing on the wrong thing (although Im using the centre focus point only...on the correct part of the page...ah well)

Done the 24-70...Will do some test shots of my lad playing on the xbox in a bit.
Just doing the 50mm now :)
 
Well I did them via that method....First I took photos then compared the photos and wondered why the moire pattern was the same on each one :doh: ! lol
Then I did it live via live view using EOS Utility, pointing the cam at the image on a second monitor.....

Basically set the image up, AF'd on the centre, then reached to the camera and nudged the focus...if it went slightly better then it would be out...if not, then try it again but turning the focus ring the opposite way.....

Did that with all four lenses (24-70, 50mm, 70-200, and 10-20)...and the AF gave the most distinctive moire pattern....turning the focus ring even a little either way on all of them meant the pattern became less distinctive.

That tells me that none of them need calibrating so I need to test further because Im not convinced :)
 
Marcel said:
Well I did them via that method....First I took photos then compared the photos and wondered why the moire pattern was the same on each one :doh: ! lol
Then I did it live via live view using EOS Utility, pointing the cam at the image on a second monitor.....

Basically set the image up, AF'd on the centre, then reached to the camera and nudged the focus...if it went slightly better then it would be out...if not, then try it again but turning the focus ring the opposite way.....

Did that with all four lenses (24-70, 50mm, 70-200, and 10-20)...and the AF gave the most distinctive moire pattern....turning the focus ring even a little either way on all of them meant the pattern became less distinctive.

That tells me that none of them need calibrating so I need to test further because Im not convinced :)
Click to expand...

Does the 40D have micro adjustment :thinking:
 
PS. I've just split this out from the other thread, save taking it horribly off topic...and so that this can go in a more appropriate forum.
 
I had to micro adjust for my 100 - 400 on my 50D. The lens was focussing slightly in front of any birds that I tried to photograph. I looked at all the various methods and they all seemed a lot of mucking about. I was lucky enough to be working in a house with plenty of birds in the garden so I took my camera with me and tried photographing birds each lunchtime. All I did was adjust the lens one point each day and see which got me the best real life results when I looked at them on the computer. I've ended up with it set to +2.

Possibly not the most scientific approach but I'm happy with the results.
 
Did anyone else get a virus warning from any of the links above? I opened them all in seperate tabs, and AVG decided that there was something dodgy with some javascript somewhere?
 
None here, but Im on a mac with no live virus checker. Firefox didn't say anything though.
 
Heres a linky to a previous discussion about microadjustment on a 7D, I am sure the same principle will apply to the 50D.
 
Rovers_Andy said:
Hi Marcel some handy info here from Tobers.

We tested on a static subject but i found on saturday that its not quite right yet. I'll be testing using Tobers method tomorrow morning, then fine tune it tomorrow night
Click to expand...

Andy using a static subject is the most accurate way to set it up, but you may not really want perfect focusing for sport.

Ah rubbish I hear everyone shout, but hang-on let me expand.

Most of what Andy shoots is moving action, football/ rugby and so on. So what you really need it the camera to front focus a little, so that the player runs in to the focus point.

Now I'm not going to do the distance maths, but a 50D has 59ms shutter lag so you need the distance an athlete will run in 59ms front focusing.
 
chris321 said:
Did anyone else get a virus warning from any of the links above? I opened them all in seperate tabs, and AVG decided that there was something dodgy with some javascript somewhere?
Click to expand...

Yeah, NOD32 went into overdrive with the first link I clicked on... kept telling me there was a trojan lurking behind the page.

Thank heavens for decent anti-virus! :D

Si
 
whiteflyer said:
Andy using a static subject is the most accurate way to set it up, but you may not really want perfect focusing for sport.

Ah rubbish I hear everyone shout, but hang-on let me expand.

Most of what Andy shoots is moving action, football/ rugby and so on. So what you really need it the camera to front focus a little, so that the player runs in to the focus point.

Now I'm not going to do the distance maths, but a 50D has 59ms shutter lag so you need the distance an athlete will run in 59ms front focusing.
Click to expand...

Thats a good point, but a runner would only move a few cm if that during the 59ms.
 
I've just got a 5D2 with micro-adjustment for the first time. Basically, I think I'm going to leave it alone! It's difficult...

The main problem I've had is that with all my lenses being f/4 max, even on full frame there's quite a lot of depth of field if you test at sensible distances (Canon recommends 50x focal length) and tiny focus shifts are hard to estimate. You could also say that therefore any discrepencies are less significant (which is true) but that's not the point of playing with it! (On a 50D, DoF at f/2.8 will be roughly equivalent to f/4 on full frame.)

Anyway, after an hour or so of fiddling, I set my 24-105L to +5 and it was spot on every time at 60mm. Each increment of 5 was shifting the focus point less than one inch from 10ft or so, which I think is incredibly fine for the focal length and likely real world subjects. I could detect shifts in increments of about 2-3 reasonably well, but you have to look damn hard.

Then I tried my 70-200L 4 IS and settled on +4 at 70mm, again at 10ft distance. I was nearly blind by this stage. Then without changing distance (kind of typical long-ish portrait range I thought, so therefore valid) I zoomed it to 150mm and then it needed to go back to zero, maybe a touch of minus!

I had a go with the 17-40L but it was impossible to see any shifts at all with that except at silly-close distance.

I then gave up and reset everything to zero, in the fairly confident knowledge that the factory calibration is probably as close to the best average setting, at different focal lengths and focusing distances, and that whatever discrepencies there may be are insignificant compared to all the other variables.

I need to do some more tests, mainly out of interest, but the point I want to make is that a) it's difficult, and b) there are real changes when you alter focal length and shooting distance. The obvious corollary is that if you yield to temptation and optimise it for an unrealistically close shooting distance (so you get shallow enough DoF to see the changes), or don't check the full zoom range, then you're quite likely to make things worse for a lot of real world shooting.

Canon's recommended test procedure is here, from Chuck Westfall, second question down http://www.digitaljournalist.org/issue0812/tech-tips.html

Edit: If this comes across as saying micro adjust isn't necessary and best left alone, that may be true for f/4 lenses of common focal lengths which I guess most people like me use. Providing they are not out of spec in the first place.

However, if you have f/2.8 lenses, or very long zooms, or certainly fast primes - all of which are more critical of focus - then you may benefit more from giving them a tweak. I would just say keep the distances real world, and check at different focal lengths.
 
Hi,

I followed this method that I found for my 50D and 400mm F5.6 L, 100mm F2.8 Macro and 70-200 F4 L lens' and it seemed to work well.........


This methods uses JPEG file size (bigger is sharper) to determine optimal micro focus adjustment. A sharply focused image has more high frequency content than a poorly focused image. JPEG compression attempts to preserve high frequency image information and for any given quality setting will produce larger files for images with greater detail. In order to determine optimal micro focus adjustment you simply step through the adjustments and capture a JPEG. You then review the images in-camera and select the one with the largest file size and set the MFA ( Micro Focus Adjustment ) to the value used for that shot. This is easily done using image review with the info screen selected that displays both file size and MFA adjustment value.

To produce accurate and reproducible results, however, a few critical setup details must be observed. The basic goal is to take images in which the only variable is the MFA setting.

Camera must be on a tripod. Use MLU and 2 second shutter delay.
Target illumination should be constant.
Use manual exposure with widest aperture for the lens.
Set ISO to 100.
Use any white balance other than automatic.
Use JPEG large and fine.
Use One Shot Autofocus

The MOST important part of the setup is the target and its orientation to the camera. Failure to observe the three first items below will produce inaccurate results.

The target must be large enough to fill the entire field of view.
The target must be flat.
The target must be perpendicular to the axis of the lens.
The target image itself is not that critical. It should be black and white with a pattern at the center suitable for autofocus to acquire lock and a bunch of detail elsewhere. A white wall "decorated" with black tape would be fine.

I use a three passes to arrive at my final MFA settings.

First pass: -15,-10,-5,0,5,10,15 (increments of five)

Second pass: Increments of two centered on the previous high file size
(example if high file size was on 5, then retake at -1,1,3,5,7,9,11)

Third pass: Increments of one centered on the previous high file size
(example if high file size was on 7, then retake at 4,5,6,7,8,9,10)

Set MFA to the setting used for the highest file size on the last pass. Done.

I manually defocus the lens between each test shot to ensure that the autofocus has to actually move the focus elements each time. I also put the MFA adjustment menu into My Menu for faster adjusting and I also delete all images between each pass so as not to get confused. I want my lenses accurate at the distances I most often use them so that is how I determine how far away to set the target. If you prefer to use a multiple of focal length that will work as well.


You will notice that it doesn't mention anything about distance only that you need to fill the frame.

Hope this helps anyone wanting to get the best out fo their lens' although if I had any lens that was worse than say +/- 10 then it should surely be calibrated by Canon, my main worry with the AF adjustment, handy though it is, will be the amount of dodgy lens' going about on the used market that are miles out but were 'cured' with AF adjustment. It might also open up a can of worms as far as used buyers and sellers are concerned when trying to buy or sell a lens that had to be adjusted 'in camera', would you buy a lens that required say +/- 20 adjustment?

Although, I assume the body might play a part in this figure too, or does it?

Mike.
 
mikeyb said:
<snip>
Click to expand...

Good post :thumbs:

I think that's probably a pretty good way to go about it in practise. Although it doesn't specify a shooting distance, it does mention shooting a wall which to me implies a large target that will naturally be a realistic distance away. It should also be easier to change shooting distance and focal length settings with a target like that, purely from a practical point of view. Quite easy to ensure that the camera is set up absolutely square too.

I don't think you need a target that fills the whole frame though. So long as it covers the focusing point compeletely, which is actually quite a bit larger than the little square shown in the viewfinder. And it needs a fair amount of detail to be able to read properly, more than just a cross of black tape - something like a sheet of newspaper stuck to the side of a perfectly flat carboard box, set dead square to the camera. If that is shot at a distance and focal length so that the target at least fills the centre circle in the viewinder, that should be bullet proof. I'm suggesting a carboard box as that is maybe easier to position than a wall, in such a way that you can change lenses and distance easily.

The set up I used was a bit more complicated, involving a flat central target square to the camera to focus on, and then an angled element next to it to read the focus shifts from. The advantage there is that you can see exactly where the focus point is moving back and forth as you adjust it. However, at the end of the day while that might be interesting to see, it's not relevant to the finished result.
 
For me I had a flat surface to focus onto and put a steel rule at 45degs on its edge.
Camera on a tripod (Uni-Loc) and to make life easer I shot tethered to my computer so all shot remotely, this keeping any movement to min along with mirror lock up
I found this set up to work very well and easy
 
To do it correctly time and time again , a little money has to be spend on something like the LensAlign the pro version being $180 .

That may sound a lot but a percentage of a body and a couple of L lenses it's not much at all really, I sure there would be enough togs on here who live in the same regions that could group together.
 
whiteflyer said:
To do it correctly time and time again , a little money has to be spend on something like the LensAlign the pro version being $180 .

That may sound a lot but a percentage of a body and a couple of L lenses it's not much at all really, I sure there would be enough togs on here who live in the same regions that could group together.
Click to expand...

$180!!! :eek: To do a focus check?! That's ludicrous. And it's completely unnecessary. Basically the same as my cerial packet and ruler set-up in the kitchen.

You just don't need all that pseudo science. Definitely one for pixel peepers.
 
HoppyUK said:
$180!!! :eek: To do a focus check?! That's ludicrous. And it's completely unnecessary. Basically the same as my cerial packet and ruler set-up in the kitchen.

You just don't need all that pseudo science. Definitely one for pixel peepers.
Click to expand...

I can see the need for the camera to be in perfect alignment to the test chart, which the LensAlign should I believe ensure.

When setting up longer lens it's going to be very hard to set-up with a bit of paper and a rule as the recommend distance from chart to lens is 50 x focal length so a 300mm lens should be 15 meters away (over 45 feet)

As I said, if you spent £4000 on a 1D IV, £3500 on a 300mm f2.8 , £6400 on a 400 f2.8 then another £150 is a drop in the ocean to get the whole lot in perfect focus. Then again after spending that much you'd probably get it all off to Canon for them to do it for you anyway.


As always what every works for you and your happy with is fine.
 
whiteflyer said:
I can see the need for the camera to be in perfect alignment to the test chart, which the LensAlign should I believe ensure.

When setting up longer lens it's going to be very hard to set-up with a bit of paper and a rule as the recommend distance from chart to lens is 50 x focal length so a 300mm lens should be 15 meters away (over 45 feet)

As I said, if you spent £4000 on a 1D IV, £3500 on a 300mm f2.8 , £6400 on a 400 f2.8 then another £150 is a drop in the ocean to get the whole lot in perfect focus. Then again after spending that much you'd probably get it all off to Canon for them to do it for you anyway.


As always what every works for you and your happy with is fine.
Click to expand...

I agree that 50x focal length is a tricky one with longer lenses. You've really got to apply some common sense and use a distance that is not artificially close just so that you can see the focus shifts more easily (because that may be asking for trouble at longer range) and shooting at a distance that is representative of the kind of stuff you do. Difficult - there are birders who will be using a 300mm lens at less than 3m for small species, then motor sports photographers who might be at 50m and more.

So to claim that using a focus target like the LensAlign that is hardly larger than the size of your hand just can't be optimum for everyone. LensAlign talk about 25x focal length which is probably okay for longer lenses, but too close for regular focal lengths.

I would also question the need for this obsession with absolute laser-accurate alignment. If you are at a sensible distance, say 2-3m at least with normal focal lengths, then an inch or so either way of camera position (but no more than that) is neither here nor there - it will make an infinitesimally small difference. That is very easy to achieve with a tape measure and a bit of common sense and in practical terms, no lens or AF system is anywhere near accurate to a fraction of one mm at that distance and it's not even important that it should be.

But that's really beside the point. Sorry, as you say, whatever works for you. The thing I really want to emphasise is that whatever method you use, it is important to check every lens at different distances and focal lengths, and set an average because there is not a single optimum setting for all situations. There are mechanical tolerances throught the system as Canon states clearly and my experience with top grade equipment - a 5D2 and three L grade lenses - bears that out.
 
whiteflyer said:
Most of what Andy shoots is moving action, football/ rugby and so on. So what you really need it the camera to front focus a little, so that the player runs in to the focus point.
Click to expand...

Im not convinced by this to be honest. I expect the camera to AF fast enough that i just need the focus to be set correctly on a stationary subject. If i setup my cameras how you mention then if i ever took a shot of something moving away from me it would be even further out of focus. And my stationary shots would be slightly front focused.

I see where your coming from, but from a pro series body i would just expect it to work.
 
dunganick said:
Im not convinced by this to be honest. I expect the camera to AF fast enough that i just need the focus to be set correctly on a stationary subject. If i setup my cameras how you mention then if i ever took a shot of something moving away from me it would be even further out of focus. And my stationary shots would be slightly front focused.

I see where your coming from, but from a pro series body i would just expect it to work.
Click to expand...

I'm not convinced either, but thinking about it, it maybe depends how you work. Basically what you are doing is pre-focusing on a certain point and manually anticipating the delay and movement to get a sharp result. That isn't what micro-AF is about at all, though I can see that you might be able to adapt it to that way of working, in a quirky kind of way. And then having to accept that stationary subjects will therefore be a bit out, and subjects moving the other way will be even further out... :thinking:

On the other hand, if you're in servo AF mode and the camera is tracking a moving subject, then it is constantly monitoring all the AF points passively and using that data to measure speed and direction in an attempt to calculate the amount of focus shift required at the moment of exposure. But it needs to start from a point of perfect accuracy as that is where it aims to finally place the focus position.
 
dunganick said:
Im not convinced by this to be honest. I expect the camera to AF fast enough that i just need the focus to be set correctly on a stationary subject.
I see where your coming from, but from a pro series body i would just expect it to work.
Click to expand...

HoppyUK said:
I'm not convinced either, but thinking about it, it maybe depends how you work.

.
Click to expand...

I doubt it would make any difference really in running to be honest, and I'm not sure if the AF Servo keeps moving while the shutter is moving in that 59ms lag time, but just think an F1 car at 170mph will travel 13 feet in 59ms. It's now becoming something to think about.
 
I think part of the AI Servo algorithm predicts where a moving target will be in those milliseconds so should cover the F1 car, if not, how do they get all those sharp motor sport shots :D
 
artyman said:
I think part of the AI Servo algorithm predicts where a moving target will be in those milliseconds so should cover the F1 car, if not, how do they get all those sharp motor sport shots :D
Click to expand...

I guess your right. :)
 
Here's what I got with my 5D2. Target was a sheet of newspaper taped to the patio door in the kitchen which is quite long, so I could try different distances and focal lengths. It filled roughly one quarter of the image area. Floor is tiled so I used that and a tape measure to postion everything square. Lit with flash, tripod, IS off, defocusing between each shot etc.

17-40L at 35mm and 1.1m, +5.

24-105L at 35mm and 1.1m, +5.
At 80mm and 2.6m, +10.

70-200L 4 IS at 80mm and 2.6m, 0.
At 140mm and 4.9m, +5.

With Kenko Pro 1.4x TC at 100mm = 140mm and 4.9m, -5.

I found this by far the easiest test method so far, and reasonably easy to see the changes by chimping straight off the LCD. Doing it this way without an angled aspect to the target, you can't see where the focus actually is until it is dead sharp, but that's only of academic interest anyway, and that's the thing which makes other methods much harder to set up. I could have made finer adjustments, but given that things change a bit depending on focal length and shooting distance, it wasn't worth bothering with.

It's worth noting that depth of field for all these shots at f/4 is quoted at a whopping 23cm. Given that, I think this is a pretty good performance and I'd be very happy if my own technique allowed me to get anywhere near the camera's potential level of accuracy in practice. The camera's AF was usually better than me on manual even using live view at max magnification.

I was also impressed at how consistent the repeat focusing was - bang on every time without fail :)
 
You must log in or register to reply here.
Back
Top