7D Mark II - Focus Discussions

skid00skid00 wrote in post #17829002
I believe the AF algorithm will pick the area with the greatest contrast, too. IE, the wet dark rock/snow interface. They might have placed some weight on the nearest subject, though. I'll have to test this.

Previous tests of mine seem to suggest that the AF pixels are similar in resolution to the image sensor in my 5D3. IE, saturated black lines that were one sensor pixel in width could be focused by the AF module.

With fur having regular, repeating patterns, the AF can be confused, and be off. The AF matches light/dark patterns from two opposing sides of the lens. Think of holding two hair combs together, and shifting one of them left/right, so you can see thru the open areas of the teeth. Many positions will align.

Yes, it is not just contrast, but also orientation of the shapes, and repetition (like rows of windows on buildings or patterns in a fence) can fool AF.

digital paradise wrote in post #17829016
What are we talking about, single, expansion or zone?

skid00skid00 is referring to the geometry involved in phase detection AF. So, all of the above, but only one AF point at a time.

This AF works a bit like an old rangefinder. Image portions obtained from the top and bottom of the lens are converged by the AF system to achieve focus.

My replacement 7D2 showed up today. I was a bit surprised when I got a call from CPS after I emailed them this afternoon. The gentleman with whom I spoke apologized that a notification email didn't go out as expected.

There was very little information on the repair sheet. I'll quote it: "Due to other arrangements already made directly with you or due to certain necessary components not being available to complete your repair, our Factory Service Center has shipped a replacement product to you in exchange for your original unit at no additional charge."

The rest was just the new serial number... call if you aren't pleased with this product, etc.

This ends my saga. I have decided to keep my 1D IV. I will put the new 7D2 up for sale tomorrow together with a couple EF-s lenses and my 35 f/2 IS. That should help offset the cost of the 1D IV and 16-35 f/4 IS I picked up for landscapes.

I do regret I wasn't able to assist anyone by determined what was going on with my 7D2. I am relieved that Canon was able to confirm it wasn't working properly and decided to replace it.

Based on the very active main 7D2 thread here, I think what happened to my camera is an anomaly. The vast majority of 7D2 owners are happy and producing very nice photographs. If someone has one that isn't performing as expected, getting Canon involved is always an option. It worked out for me.

Archibald wrote in post #17829032
skid00skid00 is referring to the geometry involved in phase detection AF. So, all of the above, but only one AF point at a time.

This AF works a bit like an old rangefinder. Image portions obtained from the top and bottom of the lens are converged by the AF system to achieve focus.

I see. Thanks.

Glad things worked out for you Kim.

Lupo-Lobo wrote in post #17829155
Glad things worked out for you Kim.

Thank you, Lupo.

Phoenixkh wrote in post #17829081
This ends my saga. I have decided to keep my 1D IV. I will put the new 7D2 up for sale tomorrow together with a couple EF-s lenses and my 35 f/2 IS. That should help offset the cost of the 1D IV and 16-35 f/4 IS I picked up for landscapes.

I wish you all the best with your 'new' 1D IV. Hope to see what it can do.
And no, I won't be bidding on your 7D2 no matter how good you make it sound.

Can you post your shots to the 7D2 thread anyway so I don't have to follow another thread. You can tell people that after Canon removed all the fancy new parts that didn't work you were left with a 1D4

Rod

TeamSpeed wrote in post #17828427
Depth of focus (which is what Canon states MFA works from) vs Depth of Field (what the rest of us really care about) are related but very different. The depth of focus (at the sensor level) is much smaller in overall depth than the depth of field. Thus why these adjustments are very, very tiny, but do produce noticeable results in the final image.

Here is a diagram that might help. Slice just one side of the field part of the diagram into 8 points.... MFA adjusts the right side of this diagram, not the left, but has an impact on the left.

| Byte size: ZERO | Content warning: NOT AN IMAGE

Aren't we say same thing? I agree the MFA adjusts the depth of focus on the right side of the diagram. The the front part of depth of focus will always correlate to the front depth of field, and same for back dofocus and back dofield. Therefore adjusting the front depth of focus by <=8 mfa units will correspond to 8<= same number of units spanning the distance of front depth of field. Since they are proportionate, and we observe the depth of field of our images, and can look up expected dofield values (total, front, back) , we can derive the subject do field movement each mfa unit will create. . Agree?

Pondrader wrote in post #17828440
Sometimes they can blend into the back ground so well, always is not a word I would use when it comes to shooting animals it good cover. there are many struggling with that one all the time.

Hosted photo: posted by Pondrader in
./showthread.php?p=178​28440&i=i241113512
forum: Canon Digital Cameras

I don't know much about Pine Martins. Do they change color in winter from brown to white, or is that a different kind?

digital paradise wrote in post #17828458
The move at the sensor level has to be smaller. Extremely fine moves. I did figure that out very quickly. I would have preferred that Canon also provided a value of an MFA move at the sensor level. I imagine it is not variable based on distance to target. It can't be. That was what threw me off. Canon documentation is limited.

DP, I do not believe you are right.
MFA increments can remain constant in absolute value at sensor.
However, mfa observed changes at the target are directly related to, but changes based on distance and focal length for the of field calculation. See my immediate previous post.

TeamSpeed wrote in post #17828461
Sure, AI Servo is predictive, and tries to determine direction and keep AF. However that is a double-edged sword. When you use AI Servo on a static subject, it then tries to predict focus next and can move things around on you, sometimes imperceptibly, and that will cause focus issues. I rely heavily on AI Servo predictive analysis, and run any number of the AF use cases in the AF menu when shooting sports. That is considerably harder, in many cases, than moving wildlife. AI Servo on static subjects can produce results that are unexpected however, been there many times.

If you don't change your AF case settings, and try to shoot moving targets vs fairly slow or non-moving targets, you won't always get good results. That menu is a very important one on the 5D3, 1D4, 1DX and 7D2.

Also in regard to center point, the 7D2 is the best APS-C to date in using any of the AF points to track motion. I rarely run center point at games or events, I am almost always using a perimeter AF point. You cannot shoot sports with center point and be close to the action, your results will suffer.

I have shot still objects with servo for over 6 years. Never had any adverse effect, always reliable to shoot a coach, fan, score board, or performer with no motion.

RodS57 wrote in post #17829206
I wish you all the best with your 'new' 1D IV. Hope to see what it can do.
And no, I won't be bidding on your 7D2 no matter how good you make it sound.

Can you post your shots to the 7D2 thread anyway so I don't have to follow another thread. You can tell people that after Canon removed all the fancy new parts that didn't work you were left with a 1D4

Rod

LOL Rod.

That little guy . Is along tailed weasel, he's not very big. He's white in the winter and brown in the summer. Just so happens theres one down from the end of my driveway that I need to go shoot soon

Shoot with enough lenses, and you will see some strange behavior. Sigma lenses, at least the older ones, will shudder. Also some of the 7d copies had strange ai servo behavior that I believe Canon fixed in one of the firmware updates.

Phoenixkh wrote in post #17829081
My replacement 7D2 showed up today. I was a bit surprised when I got a call from CPS after I emailed them this afternoon. The gentleman with whom I spoke apologized that a notification email didn't go out as expected.

There was very little information on the repair sheet. I'll quote it: "Due to other arrangements already made directly with you or due to certain necessary components not being available to complete your repair, our Factory Service Center has shipped a replacement product to you in exchange for your original unit at no additional charge."

The rest was just the new serial number... call if you aren't pleased with this product, etc.

This ends my saga. I have decided to keep my 1D IV. I will put the new 7D2 up for sale tomorrow together with a couple EF-s lenses and my 35 f/2 IS. That should help offset the cost of the 1D IV and 16-35 f/4 IS I picked up for landscapes.

I do regret I wasn't able to assist anyone by determined what was going on with my 7D2. I am relieved that Canon was able to confirm it wasn't working properly and decided to replace it.

Based on the very active main 7D2 thread here, I think what happened to my camera is an anomaly. The vast majority of 7D2 owners are happy and producing very nice photographs. If someone has one that isn't performing as expected, getting Canon involved is always an option. It worked out for me.

Are you not even going to give the replacement a try Kim?