30D...adjust focus calibration

[RicksAstro]

I recently bought a new 85 f1.8 and noticed it was backfocusing severely on my 30D. I had several other lenses, and they seemed to be fine, so I assumed it was a lens issue.

Being a geek, I decided to test my lenses (85 1.8, 17-55IS, 70-300IS, 300 f/4, Sigma 150 f2.8 macro) using a focus test target (like http://www.photo.net/learn/focustest/ ). The 70-300 and 17-55 were in great focus, but I found the 300 and 150 were also just a touch off and the 85 was waaay off.

I've been a big fan of Christian Buil's great camera test pages (being an astrophotographer as well), and noticed he told how to adjust the autofocus for the 350D, and after examining the 30D, it appeared the same method should work:

http://astrosurf.com/buil/autofocus/adjust.htm
(translated to english:

http://translate.google.com/translate?u=http%3A%2F%2Fastrosurf.com%2Fbuil%2Fautofocus%2Fadjust.htm&langpair=fr%7Cen&hl=en&ie=UTF-8&oe=UTF-8&prev=%2Flanguage_tools )

Being a brave/stupid soul, I decided to give it a go.

My main objective was to get the 85 dead on and figured I'd see if it made the others better or worse. I was hoping it wouldn't throw the good focusing ones off.

After some trial and error, I learned that moving the hex wrench up/CCW pushed focus back (increased backfocus) and moving it down/CW pushed focus forward (decreased backfocus)...it works (whew!). I was able to get the 85 just dead-on at f1.8. With trepidation, I tried the other lenses. To my great joy, all were dead on, even the ones that focused well before I started!

I learned that different lenses react quite differently to a miscalibrated camera. Some were far less sensitive to it than others. There was a general correlation to max aperture (wider=worse), but not 100%.

I used sensor cleaning mode to allow the adjustments and ran the camera using the AC adapter I have. I tell you, having a metal object that close to the sensor is nerveracking, but some patience and finesse rewarded me with a now perfectly calibrated 30D, with all my lenses being spot-on. In just doing some random non-geek testing, the sharpness increase is definitely noticeable...OOF images I would normally just blame on "difficult situations" are far fewer and farther between.

This isn't for the faint of heart, and I image there certainly are many ways to mess up the camera, but wanted to confirm that it is possible to calibrate the 30D. There is very little info besides Christian's (french) page out there.

Rick

 

[Gao Gao]

Glad you have your camera working "dead on". I have 2 quick questions -

1. What's the size of the hexagonal key you used?

2. How do the rest of the AF points work after this physical adjustment?

--

. 。o O o 。 . 。o O o 。 . 。o O o 。 .

 

[RicksAstro]

Glad you have your camera working "dead on". I have 2 quick
questions -

1. What's the size of the hexagonal key you used?

I used a 0.05 inch hex key, which is very close to 1.3mm

2. How do the rest of the AF points work after this physical
adjustment?

I tested them before calibration, and they acted similarly to the central, but I failed to check afterwards...shame on me! I'll do that tonight.

--
. 。o O o 。 . 。o O o 。 .
。o O o 。 .

 

[Paul De Bra]

You can gently lift the mirror and do this with the shutter still closed. That makes it a bit less scary than in cleaning mode, but it also means you need a hand to keep the mirror flipped up.

I tried the adjustment on my DRebel/300D but it has a strange square hole and the screw doesn't move at all, even with the correct wrench.

I recently bought a new 85 f1.8 and noticed it was backfocusing
severely on my 30D. I had several other lenses, and they seemed to
be fine, so I assumed it was a lens issue.

Being a geek, I decided to test my lenses (85 1.8, 17-55IS,
70-300IS, 300 f/4, Sigma 150 f2.8 macro) using a focus test target
(like http://www.photo.net/learn/focustest/ ). The 70-300 and 17-55
were in great focus, but I found the 300 and 150 were also just a
touch off and the 85 was waaay off.

I've been a big fan of Christian Buil's great camera test pages
(being an astrophotographer as well), and noticed he told how to
adjust the autofocus for the 350D, and after examining the 30D, it
appeared the same method should work:

http://astrosurf.com/buil/autofocus/adjust.htm
(translated to english:

http://translate.google.com/translate?u=http%3A%2F%2Fastrosurf.com%2Fbuil%2Fautofocus%2Fadjust.htm&langpair=fr%7Cen&hl=en&ie=UTF-8&oe=UTF-8&prev=%2Flanguage_tools )

Being a brave/stupid soul, I decided to give it a go.
My main objective was to get the 85 dead on and figured I'd see if
it made the others better or worse. I was hoping it wouldn't throw
the good focusing ones off.

After some trial and error, I learned that moving the hex wrench
up/CCW pushed focus back (increased backfocus) and moving it
down/CW pushed focus forward (decreased backfocus)...it works
(whew!). I was able to get the 85 just dead-on at f1.8. With
trepidation, I tried the other lenses. To my great joy, all were
dead on, even the ones that focused well before I started!

I learned that different lenses react quite differently to a
miscalibrated camera. Some were far less sensitive to it than
others. There was a general correlation to max aperture
(wider=worse), but not 100%.

I used sensor cleaning mode to allow the adjustments and ran the
camera using the AC adapter I have. I tell you, having a metal
object that close to the sensor is nerveracking, but some patience
and finesse rewarded me with a now perfectly calibrated 30D, with
all my lenses being spot-on. In just doing some random non-geek
testing, the sharpness increase is definitely noticeable...OOF
images I would normally just blame on "difficult situations" are
far fewer and farther between.

This isn't for the faint of heart, and I image there certainly are
many ways to mess up the camera, but wanted to confirm that it is
possible to calibrate the 30D. There is very little info besides
Christian's (french) page out there.

Rick

--

Slowly learning to use the DRebel (only around 26.000 shots) and now also the Fuji F11.
Public pictures at http://debra.zenfolio.com/ .

 

[RicksAstro]

I considered doing it this way, but had the thought that the shutter would be far, far easier to damage, and is closer to the screw than the CMOS. a slight tap of the CMOS with the key will likely do no damage. Slightly bending a very delicate shutter blade could really mess things up.

That, combined with less chance of torquing the mirror assembly, cleaning mose seems safer. On advantage of manually flipping the mirror is that, if you damage the shutter, it's fixable...scratches to the CMOS likely are not.

If you use cleaning mode, I would recommend using the AC adapter to avoid running the low battery issue. But the camera does start beeping when the battery is low in cleaning mode to warn you to get out of there.

But either would work, obviously.

Rick

 

[Gao Gao]

I recently replaced the broken secondary mirror hinge with a segment of paper clip. The square-holed screw is visible on both sides of the mirror box so I turned it a bit from the outside. I didn't feel it moving but it seems the focus was adjusted quite a bit.

I recently bought a new 85 f1.8 and noticed it was backfocusing
severely on my 30D. I had several other lenses, and they seemed to
be fine, so I assumed it was a lens issue.

Being a geek, I decided to test my lenses (85 1.8, 17-55IS,
70-300IS, 300 f/4, Sigma 150 f2.8 macro) using a focus test target
(like http://www.photo.net/learn/focustest/ ). The 70-300 and 17-55
were in great focus, but I found the 300 and 150 were also just a
touch off and the 85 was waaay off.

I've been a big fan of Christian Buil's great camera test pages
(being an astrophotographer as well), and noticed he told how to
adjust the autofocus for the 350D, and after examining the 30D, it
appeared the same method should work:

http://astrosurf.com/buil/autofocus/adjust.htm
(translated to english:

http://translate.google.com/translate?u=http%3A%2F%2Fastrosurf.com%2Fbuil%2Fautofocus%2Fadjust.htm&langpair=fr%7Cen&hl=en&ie=UTF-8&oe=UTF-8&prev=%2Flanguage_tools )

Being a brave/stupid soul, I decided to give it a go.
My main objective was to get the 85 dead on and figured I'd see if
it made the others better or worse. I was hoping it wouldn't throw
the good focusing ones off.

After some trial and error, I learned that moving the hex wrench
up/CCW pushed focus back (increased backfocus) and moving it
down/CW pushed focus forward (decreased backfocus)...it works
(whew!). I was able to get the 85 just dead-on at f1.8. With
trepidation, I tried the other lenses. To my great joy, all were
dead on, even the ones that focused well before I started!

I learned that different lenses react quite differently to a
miscalibrated camera. Some were far less sensitive to it than
others. There was a general correlation to max aperture
(wider=worse), but not 100%.

I used sensor cleaning mode to allow the adjustments and ran the
camera using the AC adapter I have. I tell you, having a metal
object that close to the sensor is nerveracking, but some patience
and finesse rewarded me with a now perfectly calibrated 30D, with
all my lenses being spot-on. In just doing some random non-geek
testing, the sharpness increase is definitely noticeable...OOF
images I would normally just blame on "difficult situations" are
far fewer and farther between.

This isn't for the faint of heart, and I image there certainly are
many ways to mess up the camera, but wanted to confirm that it is
possible to calibrate the 30D. There is very little info besides
Christian's (french) page out there.

Rick

--
Slowly learning to use the DRebel (only around 26.000 shots) and
now also the Fuji F11.
Public pictures at http://debra.zenfolio.com/ .

--

. 。o O o 。 . 。o O o 。 . 。o O o 。 .

 

[Searching]

Where did you buy the hex key, and did you have to bend it 90 degrees or the way it was already bent OK to use.

Also, can you tell us how much a turn CW/CCW correlates to the amount of movement adjusted.
i.e. did it take you 1,2 3 etc turns to get to the right spot.

 

[Steven]

nt
--
Steve

 

[MarcusR]

n/t

 

[RicksAstro]

I had a set (much like this one
http://www.cvfsupplycompany.com/alwr30pcalwr.html )

that had the .05" wrench. It's a little small (a bit of play in the movement) so maybe the 1.3mm (which I didn't have) would have worked better.

Very little movement produced fairly large results. I felt lucky that the set screw was set such that the wrench was sticking straight out of the camera, so I could adjust it slight up or down. As it ended up, I could just use it in the default 90 degree configuration.

There was a lot of static friction (which is probably a good thing), so it was a little difficult to move in small increments. If you're close to good focus, only the tiniest of movements is necessary, particularly if you're almost right on. I probably only moved it about 10-15 degrees (out of 360 for a full turn). With some of the lenses (70-300IS), it made almost no difference..they started out good, and ended up good. With others (85 1.8) it made all the difference in the world.

Rick

 

[Searching]

Some may want to try this on an older camera. Great information.

 

[RicksAstro]

I ran the camera through a full gamut of walk-around shots and test shots during daylight and at various focal lengths after my calibration last night.

I must say every lens focused perfectly, and all focus points worked great...it's like a whole new camera!

So back to enjoying the camera with full confidence.

Rick

 

[David Yatim]

Wow, brother... All I can say is, you ARE quite brave!!! Good for you!!!

Just let me ask you... Is this way to calibrate the focus of a camera usually part of the documentation of a camera? I will be deciding between a Canon 30D and a Nikon D80 in the next few weeks, and I wonder if I will run into something like this in the near future...

Thanks so very much for sharing this information!!! I'll make sure I keep the thread for reference!!!

Davey

 

[Cal Dawson]

--
Cal

Put a Canon to your head, You deserve it....

http://funshots.smugmug.com/

 

[Gao Gao]

The quality control on 300D/Rebel wasn't very good and from my experience quite a number of them back-focus consistently. However, such instances are rare among 30D/D80. You can use a 85mm prime to test the AF in store if you're buying from a traditional retailer. In case you're buying online and get a lemon, send it back for exchange, or send it to Canon/Nikon for servicing. As far as I know, Canon's AF can be adjusted using some "secret" software.

No, such adjustment isn't documented in the users' manual and isn't meant to be performed by yourself.

--

. 。o O o 。 . 。o O o 。 . 。o O o 。 .

 

[RicksAstro]

Bravery, or stupidity...it can be debated Of course, I'm used to the Astro crowd, who totally takes the camera apart to remove the IR filter from the CMOS to make it a better camera for space...so adjusting this is very safe compared to that surgery!

This is really completely undocumented other than the provided link (which was just for the 350D)...it was a guess/hope that it was the same for the 30D, and it was.

I think most sane people send their camera/lens in for calibration if it's under warranty. I figured if it didn't work, I could send it in. As it sits now, I'm not going to touch it, since it's dead on.

I don't think any camera (N or C) is immune to slight miscalibration.

Rick

Wow, brother... All I can say is, you ARE quite brave!!! Good for
you!!!

Just let me ask you... Is this way to calibrate the focus of a
camera usually part of the documentation of a camera? I will be
deciding between a Canon 30D and a Nikon D80 in the next few weeks,
and I wonder if I will run into something like this in the near
future...

Thanks so very much for sharing this information!!! I'll make sure
I keep the thread for reference!!!

Davey

 

[thw]

-------------------------------------------
See the colors of my world in:
thw.smugmug.com

 

[JimH]

The purpose of that adjustment is to properly set the angle at which the reflex mirror and the relay mirrors rest.

When you adjust that stop, it will alter how the camera focuses. But, it won't do what you think it does!

The actual focus calibration for these bodies is accomplished by adjusting the values of some stored calibration numbers.

The autofocus systems in these bodies use sets of split-prism optical systems which project the "split images" onto pairs of linear arrays of photodetectors.

The computer in the camera analyzes the patterns projected onto those linear arrays and attempts to match the patterns and determine how far off and in what direction the focus is. It then issues a movement command to the lens telling it how far and in what direction to move.

It's a fairly simple matter to adjust some calibration constants to be used in the calculations so no mechanical adjustment is required. Canon connects the camera to a computer and uses a "tool lens" that is known to be calibrated and then they adjust the AF constants until things work properly.

The same sort of thing is done for the lenses.

Ok, back to that cam that you're adjusting.

Yes, if you only want the center AF point to work, and you don't mind having the manual focus messed up, you can adjust the angle at which the mirrors rest and it might seem to make things work better.

But, by getting the resting angle of the mirrors out of whack, you end up making it so that proper AF happens on one AF point, but not the others.

And further, you mess up any chance of getting useful manual focus.

These mirrors must rest such that they project a proper and flat image from the lens up onto the focus screen (reflex mirror) and also down to the AF sensors (relay mirror). This can only happen at one proper point.

If you send your messed up camera back to Canon, they probably have some way to realign the mirror and then go in and do a proper AF calibration. But I suspect that you will want to tell them what you've done so that they'll know what needs to be done to correct it.

And of course, they probably will not do this under any warranty. So you have to admit what you did so that they can properly fix it but by so admitting, you're doomed to have to pay for this service.

In sort, DON'T DO THIS!!!

--
Jim H.

 

[John Seymour]

As they say, a little knowledge is a dangerous thing... but I guess if his camera works better for him, he's been lucky.

When I had my body and leneses calibrated at Canon UK, they mentioned it's all a compromise anyway, as it's not possible to get best possible auto-focus at all distances and at all focal lengths (of zoom lenses). So the technique is to find the optimum that can be acheived. I guess if you also know you only have a few lenses, in theory you could remove some of the variables and end up with a body optimised for those lenses.

Your lenses and bodies are not directly calibrated to each other. They're calibrated to a reference lens/body as appropriate. At the end of the day, I think Canon just bung in on a machine paring with a reference body or lenses and the machine does it's stuff automagically. This is great so long as all their reference lenses/bodies/equipment is calibrated ok ;-)

--

http://www.thebaldphotographer.com/

 

[RicksAstro]

Thanks for the head's-up...

I will say, for MY camera, BOTH the af sensors (tried all of them!) and the image in the viewfinder are noticeably better sharper. Perhaps I lucked out and it wasn't correct before.

I have absolutely no reason to send the camera in...it is great in all respects.

I do appreciatebyour knowledge and cautionary advice...others may not be so lucky!

I really wish Canon would make their service manual and software available, even for a extra cost! Even if using it voided your warranty, so they wouldn't be obligated to fix any user messes...

Rick