How I fixed my front focus

And those with the Bigma that don't like the sharpness, it appears to be very easy to adjust the lens itself for focus issues.

http://forums.dpreview​.com …rum=1029&messag​e=12833550

Hi chinch, I've never seen anything from Canon regarding this screw and for some reason I doubt they would ever verify what this screw does. If you do a google search for "20D front focus adjustment" (I wanted to just put the links here but I'm not sure if they sites want me doing that) there are two links I used, one to a translated page from French and one from Russia (I think). Anyway, both mention that this screw moves the mirror which reflects on the focus sensors. This seems to make sense, because the screw turns both ways without tightning meaning to me it's for an adjustment, and I can vary focus both front and back consistantly by turning the screw.

Also, if you look at infrared photography, there is a need to adjust the focus point since infrared light focuses at a direct point that visible. If I rememeber, there are some sites that mention this process and make reference to this screw as the method. And for what it's worth, it seems Nikon uses the same adjusment method, but also adds another screw to move the mirror used for manual focus.

My thought is that there has to be an easy, non intrusive method to adjust focus for Canon. Can you imagine having to tear apart, even somewhat, every camera that comes in for focus adjustment?

Hi TeamSpeed, I wonder on that post regarding getting other points out of whack if that person's problems were body or lens related. All I know is that all my focus points work fine, I did test for that both on a target and real world shots and my focus issues were due to the body being out of adjustment. And you are right about brining things back to where you started, the first time I turned the screw I went the opposite way I needed to go. It was a simple matter of turning it back the other way and then tweaking it.

Front/back focus issues Canon service centers fix via software which is not publicly available. MGW172 has adjusted position of the secondary mirror which reflects light to AF sensors. My guess is that now other AF points (especially upper and lower) will show some FF/BF. If not, camera wasn't properly calibrated at the factory.

Question...
I am starting to get a hunch that my camera body could use some calibration.
This is why: when I manual focus I rely on the confirmation beep to know when to stop twisting the focus ring (I can't just by looking through the 400d viewfinder).
I am trying to focus on a person's eye, and keep getting the forhead in focus.
It could be that the eye doesn't have enough contrast.

What do you think?

Erez

tiha wrote in post #3506615
Front/back focus issues Canon service centers fix via software which is not publicly available.

Do you have documentation/links on this? The reason I ask is, to my mind, focus accuracy is a function of mechanical parts operating in proper alignment. I don't see how you can open up software program to adjust something physical, unless that software controls a robot that turns the screw or makes some other mechanical adjustment.

software could do it, it is after all doing the AF, so if they adjust the af setup to be +/- whatever its off then its all set

cdifoto wrote in post #3506634
Do you have documentation/links on this?

IIRC, a few years ago software for 300D find a way to go to public . Newer software is better protected. Rough adjustment is done mechanically (that is what MGW172 did) and fine via software. It's a probably simple adjusting of a few parameters in camera firmware (something which 1D Mk III already has in the menu).

tekkie wrote in post #3506678
software could do it, it is after all doing the AF, so if they adjust the af setup to be +/- whatever its off then its all set

Getting correct optical reading with 0 offset is always better than not so correct than wrong optical reading with some offset. the amount of offset can have different effects in different occasions (focal length, distance of object).

My opinion anyway.

Erez

Katzer1 wrote in post #3506730
Getting correct optical reading with 0 offset is always better than not so correct than wrong optical reading with some offset. the amount of offset can have different effects in different occasions (focal length, distance of object).

My opinion anyway.

Erez

Same thing? Just like what the Mark III has, if you can adjust a lens to a body by knowing optically how far off focus is off "0", then you can tell the camera to adjust the lens focus ring a millimeter or more either way after camera thinks it has "0". The Mark III allows fine tuning for up to something like 20 different lenses. That is how they fix it with software, just change where "0" occurs on the focus ring, which is optically fixing the problem with no hard mechanical changes.

Katzer1 wrote in post #3506730
Getting correct optical reading with 0 offset is always better than not so correct than wrong optical reading with some offset.

It is better of course, but achieving it probably would require more precise manufacturing standards, quality control and more skilled workers - all of which would increase prices of cameras and lenses.

tiha wrote in post #3506714
IIRC, a few years ago software for 300D find a way to go to public . Newer software is better protected. Rough adjustment is done mechanically (that is what MGW172 did) and fine via software. It's a probably simple adjusting of a few parameters in camera firmware (something which 1D Mk III already has in the menu).

This is true...I forgot all about that.

MGW172 wrote in post #3505430
No, I didn't notice that, but then again I didn't test for that either. My test was to take three shots, tripod mounted, with indoor overhead lighting for focusing and then a flash for primary illumination. I used this method to make the adjustments (although it only took two adjustments of the screw to get it done). I then shot real life test shots both inside and outside, flash and no flash to confirm beyond a test chart that I had gotten it right.

Regarding light levels and front-focusing - when you tested with a flash, did you leave the AF assist beam on? The reason I ask is that in my tests of front-focusing, I found that I was seeing the most front-focusing with just indoor lighting or with a flash with AF assist beam disabled. Front-focusing was less with the AF assist beam but still significant and even less using a large flashligt to illuminate the subject.

Right. I had been assuming that bright light levels improve the accuracy of auto-focus, however, if this were the case, I would expect to see more variation in the auto-focus point in low light clustered around the same focus point I'd see in bright light on average. This is not the case - lower light conditions consistantly cause more front-focusing with my camera. So, does low-light bias Canon's auto-focus system to select a focus point in front of the target OR is this just a peculiar failure of my camera's auto-focus system? Its a mystery.

Never try this at home.
A mirror adjustment (if that screw is really a focus mirror adjuster) won't move the focusing plane back and forth, but lateral (or up-down), wich is of no use, and the actual focus points would be shifted from their painted positions on the viewfinder.

I hate too smart womens. 8)

JackProton - I used the on board flash, which did not fire any bursts for focusing. Focusing was done only with light from a lamp located above the test target. Actually, the flash was only because I was annoyed at the poor white balance with incandesent lighting and I wanted to see a white page with black lines.

As for your second point - wow I dont know! Actually I wish I knew more about this and what exactly I was moving ect........

To Adas- my results show different than your statement and I did test for this. I wondered if I was shifting away something from the visible focus markers and I didn't. Not sure about the too smart woman comment - since I'm a guy! The avitar is my beautiful wife if that is where that is coming from.

Also all the focus points focus consistantly with no back or front focus, including the top points.

All I can say is your mileage may vary and maybe the problem with my camera was with this adjustment and nothing else.