Does the 5d2 have the Banding with AI Servo like the original 5d had? I know Canon claims it only happens with certain lenses, but it pretty much happens on all of them. Some worse than others. Especially when you try higher ISOs (800 and up).


johnny

dhphoto wrote:
For me it's really only the focusing system which stops the 5DII from being superb.




+1


Although i'd love a one-series body


ps: 1Ds3 is not an option for most of us

Orestis.Ch wrote:
dhphoto wrote:
For me it's really only the focusing system which stops the 5DII from being superb.




+1


Although i'd love a one-series body


ps: 1Ds3 is not an option for most of us

I'm finding there are lots of times I really don't want a large 1-series body to cart about, looking very professional and attracting attention.

The 5DII IQ is so good that is no longer an issue, I just wouldn't feel comfortable using it for something that required really good focusing.

I'm not sure if it's to do with the sensitivity or the non-cross nature?

As I explained above, it's the non-cross nature. If you turn the camera so that the rectangle is perpendicular to linear contrast in the subject, those peripheral points are every bit as accurate as the center point with lenses slower than f/2.8.

Quality control is pretty ordinary also.
Although I think they intentionally send out cameras with slight faults thinking most buyers of this level camera won't notice.

There is always a certain percentage of error permitted in nearly all quality control systems, and the less error permitted the higher the cost...exponentially.

RDKirk wrote:

There is always a certain percentage of error permitted in nearly all quality control systems, and the less error permitted the higher the cost...exponentially.

Except if it's a 1D3

David

The only flaw on the 5D2 for me is that it is not the top model, so there is a risk I will upgrade to 1Ds3 or 1Ds4

RDKirk wrote:
I'm not sure if it's to do with the sensitivity or the non-cross nature?

As I explained above, it's the non-cross nature. If you turn the camera so that the rectangle is perpendicular to linear contrast in the subject, those peripheral points are every bit as accurate as the center point with lenses slower than f/2.8.

True... but there is more to it... The AF point in the VF represents a much larger area than you can see in the VF.

squeaky CF compartment. Really bugs me and i have not found a solution yet

Karokan wrote:
squeaky CF compartment. Really bugs me and i have not found a solution yet

KY Jelly

Daan B wrote:
RDKirk wrote:
I'm not sure if it's to do with the sensitivity or the non-cross nature?

As I explained above, it's the non-cross nature. If you turn the camera so that the rectangle is perpendicular to linear contrast in the subject, those peripheral points are every bit as accurate as the center point with lenses slower than f/2.8.

True... but there is more to it... The AF point in the VF represents a much larger area than you can see in the VF.

That's true of all the points, and more true of the center point. The sensed area of the center point is about nine times the area of the viewscreen marking. The sensed areas of the rectangles are about three times the areas of the viewscreen markings.

So in that respect, the peripheral points are no different--no better and no worse--than the central point.

The pixel arrays that make up the peripheral focus points are not inferior arrays to those in the center. They are the same type of arrays, the same type of pixels, the same number of pixels in each array, the same technology, using the same system.

The only difference at less than f/2.8 is the physical orientation--being a single-orientation array, it's only half as likely to be perpendicular to subject linear contrasts.

RDKirk wrote:
Daan B wrote:
RDKirk wrote:
I'm not sure if it's to do with the sensitivity or the non-cross nature?

As I explained above, it's the non-cross nature. If you turn the camera so that the rectangle is perpendicular to linear contrast in the subject, those peripheral points are every bit as accurate as the center point with lenses slower than f/2.8.

True... but there is more to it... The AF point in the VF represents a much larger area than you can see in the VF.

That's true of all the points, and more true of the center point. The sensed area of the center point is about nine times the area of the viewscreen marking. The sensed areas of the rectangles are about three times the areas of the viewscreen markings.

So in that respect, the peripheral points are no different--no better and no worse--than the central point.

The pixel arrays that make up the peripheral focus points are not inferior arrays to those in the center. They are the same type of arrays, the same type of pixels, the same number of pixels in each array, the same technology, using the same system.

The only difference at less than f/2.8 is the physical orientation--being a single-orientation array, it's only half as likely to be perpendicular to subject linear contrasts.


Did you forget about the 6 invisible assist AF points surrounding the center one?

On top of that... less light will hit the outer AF points compared to the center AF point. The outer AF points are less sensitive (to light) in this regard. This is of course true for every camera. But it sure doesn't help that the 5D2 outer AF sensors are only single axis.

mgipe wrote:
Taoguy wrote:
Daan B.

...
The only downside? Tough keeping up in these times with both families, expensive, but what's a guy to do?




Bigamy!

...good idea for a smash HBO mini series "Digital Love" Not content with Canon and Nikon, hard working Joe starts dating a Sony...

Daan B wrote:
RDKirk wrote:
Daan B wrote:
RDKirk wrote:
I'm not sure if it's to do with the sensitivity or the non-cross nature?

As I explained above, it's the non-cross nature. If you turn the camera so that the rectangle is perpendicular to linear contrast in the subject, those peripheral points are every bit as accurate as the center point with lenses slower than f/2.8.

True... but there is more to it... The AF point in the VF represents a much larger area than you can see in the VF.

That's true of all the points, and more true of the center point. The sensed area of the center point is about nine times the area of the viewscreen marking. The sensed areas of the rectangles are about three times the areas of the viewscreen markings.

So in that respect, the peripheral points are no different--no better and no worse--than the central point.

The pixel arrays that make up the peripheral focus points are not inferior arrays to those in the center. They are the same type of arrays, the same type of pixels, the same number of pixels in each array, the same technology, using the same system.

The only difference at less than f/2.8 is the physical orientation--being a single-orientation array, it's only half as likely to be perpendicular to subject linear contrasts.


Did you forget about the 6 invisible assist AF points surrounding the center one?

That's only in effect when AI Servo is on, and it does not change the level of accuracy--it reduces the likelihood of the camera losing focus on a moving subject.

On top of that... less light will hit the outer AF points compared to the center AF point. The outer AF points are less sensitive (to light) in this regard. This is of course true for every camera. But it sure doesn't help that the 5D2 outer AF sensors are only single axis.

No, that's not an issue. Canon has limited the area that they place the sensors to within the cone of light that they are within suitable limits--which is why there aren't any focus points farther than 15mm from the center in any of their cameras.

The only difference is the orientation. You can test it yourself: Switch on one of the peripheral points and turn the camera so that it's perpendicular to a strong line, like a door frame. Now dim the lights until you reach a light level so dim the camera can no longer focus on that perpendicular line.

When you reach the level of dimness that finally confounds the outer sensor, try to focus on it with the center point.

How many years have we had this AF system now and people are still confused about it? The outer points are pants because they aren't cross-type. RDKirk has given the more technical explanation, but that's really all it boils down to. Basically expect the center point to be able to lock on more quickly (or at all) in more circumstances than the outer points.

Yohan Pamudji wrote:
How many years have we had this AF system now and people are still confused about it? The outer points are pants because they aren't cross-type. RDKirk has given the more technical explanation, but that's really all it boils down to. Basically expect the center point to be able to lock on more quickly (or at all) in more circumstances than the outer points.

doesn't the center point also do high-precision 1/3 DOF instead of 1 DOF focusing unlike the outer points too?

I don't own any lenses slower than f2.0, so I don't buy the line that all the points are the same sensitivity, because clearly they are not for lenses faster than f2.8

Re: squeaky card door

I sprayed a little WD-40 onto a gauze pad and carefully went around the edges of the card door. Along with my own finger oils, sunscreen residue and some use over the last few months shooting - the squeak is gone. The door still flexes at bit, but silence is golden.

Mike V wrote:
I don't own any lenses slower than f2.0, so I don't buy the line that all the points are the same sensitivity, because clearly they are not for lenses faster than f2.8



+1

Like... when I use my 85L II @ f/2.8 (or larger), amd aim for an eye, using an outer AF point... it misses focus most of the time. When I do the same thing, but use the center AF point instead, I get everything in focus most of the time.

I am obviously no technician... but this can't be (entirely) explained by the difference in single axis vs cross type. Because an eye contains no straight lines and has both horizontal and vertical contrast.

Even though some have said otherwise, my best guess is that the outer AF points are less sensitive (to light -?-) than the center one.

Like I said earlier... for me it is no problem. I will either focus-recompose using the center AF point. Or just use the center AF point and crop afterwards. Plenty MP's to crop with.

Of course, it would have been better if the 5D2 had a all cross type sensors (of the same sensitivity). But it hasn't

My only nit...still doesn't have a 100% viewfinder (and maybe no external button for mirror lock up, although you can set it in one of the three custom modes, along with other settings). But like anything else, if one looks hard enough to find a flaw... Personally, considering the price/features/performance aspect, I find it to be one heck of a camera. Maybe I've just been lucky, but I have to have any issues with any Canon body or lens (unless I'm just not looking hard enough).

Mike Ganz wrote:
My only nit...still doesn't have a 100% viewfinder (and maybe no external button for mirror lock up, although you can set it in one of the three custom modes, along with other settings). But like anything else, if one looks hard enough to find a flaw... Personally, considering the price/features/performance aspect, I find it to be one heck of a camera. Maybe I've just been lucky, but I have to have any issues with any Canon body or lens (unless I'm just not looking hard enough).

98% isn't bad and you do have 100% instant review on the LCD

As far as mirror lock up, I just use Live view and the 2 second self-timer, that way I can check focus accurately too

David

Having just got one to complement my 1D III, I feel that it is going to be very few circumstances that I use the 1D, other than sports now (I don't do BIF or non landscape nature or aviation). I like it that much.

That said, I do worry I will miss the moment. There is a clear difference in the focusing speed -- which doesn't surprise me or necessarily bother me per se. There's currently a $1000+ difference between the two bodies new, and it was much more than that on release.

I do miss some other things, which I guess are possibly just hidden in the menus that I haven't found yet, a few of which are:

1. The well stated fact that the focus points are too clustered.
2. The fact that it doesn't appear in custom functions you can make taking the shot the priority other than focus lock.
3. Have exposure linked to the focus point.
4. Having the zoom to review linked to the particular focus point you selected.

I haven't used a flash yet, but I'm sure the synch speed will also affect me.

outer AF points are untrustworthy in low light
noise floor has a large scale criss-cross pattern visible in shadows at low isos
you still have to press play before zooming
you can't get spare batteries
the mirror sticks out further than it needs to
body geometry calibration may not be perfect
the power switch is in the wrong place
there is still no way of showing histograms superimposed on the full size image in review
contrast detect focussing is too slow
you can't choose the contrast detect and liveview magnify points by tapping a finger on the monitor

Not a lot really.

At the moment at least, batteries are available again.

brainiac wrote:
outer AF points are untrustworthy in low light

Check

noise floor has a large scale criss-cross pattern visible in shadows at low isos

At what ISO value will the criss-cross pattern disappear? I will be using mine mainly for high ISO low light shooting.

you still have to press play before zooming

That is one thing I love about Nikon cameras... not having to press play before zooming

you can't get spare batteries

They are for sale over here...

the mirror sticks out further than it needs to

I don't intend to use any other glass than Canon glass on it.

body geometry calibration may not be perfect

You mean creaky CF doors?