7D mk2 Auto Focus survey

[aceflibble]

I think the fact that some people have sent their 7D2s into Canon and Canon have actually bothered to replace the sensor and mirror boxto fix the AF says that yes, there is a mechanical fault with some 7D2s (maybe one specific production run?) and no, it can't just be written off as user error.

 

[Andrewccm]

Af system on the 1DX should have a 95-99% hit rate for this like birds in flight
5d3 about 95%
7d2 about 90-95%

I'll stop there as the less sophisticated AF systems drop off rapidly

Such generalisations are hokum to me (please don't take this personally).
As you cannot specify the situations in wich people shoot. Birds or bats or pigs in flight are not standardised procedures. Even as a maximum hit rate this percentages are not really reliable. Besides, the 7D2 has the autofocus system of the 1DX, so why should it be only equal or less good than the 5d3?

It should also be mentioned that the autofocus hit rate depends highly on the lens AF capabilities.
I tried a tamron 150-600 on a 5d3 and at 600mm f8 (max apt. f6.3) in continious AF it didn't even come close to 95% hit rate. But that was the lens (or my) "fault".

And with the 5D3 Tamron combination I had to do manual AF adjustment. Quite a lot of adujstment to be honest, it had enormous back focusing before. My 7D2 did not need any manual AF fine tuning with any lens I put on it so far (100mm f2.8; 100-400mm L II, Tamron 70-200 f2.8 )


There is a Tamron Firmware update that supposed to correct the 5D3 compatibility issue. Mine reacts the same way, but I have yet to send it in. The lens works great on my 7D2.

 

[East Wind Photography]

Af system on the 1DX should have a 95-99% hit rate for this like birds in flight
5d3 about 95%
7d2 about 90-95%

I'll stop there as the less sophisticated AF systems drop off rapidly

Such generalisations are hokum to me (please don't take this personally).
As you cannot specify the situations in wich people shoot. Birds or bats or pigs in flight are not standardised procedures. Even as a maximum hit rate this percentages are not really reliable. Besides, the 7D2 has the autofocus system of the 1DX, so why should it be only equal or less good than the 5d3?

It should also be mentioned that the autofocus hit rate depends highly on the lens AF capabilities.
I tried a tamron 150-600 on a 5d3 and at 600mm f8 (max apt. f6.3) in continious AF it didn't even come close to 95% hit rate. But that was the lens (or my) "fault".

And with the 5D3 Tamron combination I had to do manual AF adjustment. Quite a lot of adujstment to be honest, it had enormous back focusing before. My 7D2 did not need any manual AF fine tuning with any lens I put on it so far (100mm f2.8; 100-400mm L II, Tamron 70-200 f2.8 )


There is a Tamron Firmware update that supposed to correct the 5D3 compatibility issue. Mine reacts the same way, but I have yet to send it in. The lens works great on my 7D2.

I don't take anything personally. It is true if you use a lens with slow AF speed or has sloppy backlash your hit rate will drop considerably. The metrics above are estimates using canon equipment and lenses suitable for bird in flight photography and that proper AFMA has been at least verified. In addition, the amount of light available will have an impact on AF speed and AF accuracy.

The AF systems are similar but not the same between versions. 7d2 has extra AF features that the 5d3 does not have and requires more cpu head room to execute those functions. 1dx has its own dedicated processor for just AF no matter how you configure it.

One should use the referenced percentages as a guide as to what to expect between models...and as a matter of scale, a faulty 7d2 AF system will yield about 10% success rate. If you are a poor photographer you may always have only a 10% hit rate!

 

[Keith_Reeder]

Where's the "never had to send it in, and it's bloody fantastic" option?

I can do this all day long. Slow shutter panning too - about as tough a test for AF as you get.

 

[Keith_Reeder]

If the 7D2 had better ISO performance then I would be buying one too!

100% crop of 4000 ISO from mine. As-is from the converter, in noise terms - no additional PP NR applied.

If a 7D Mk II's end-results are noisy, the user doesn't know how to convert the files properly, or - most likely - is using a crappy converter (and yes, that includes Lr/ACR).

 

[Keith_Reeder]

So, if this is true, you'd get more good keepers when you shoot BIF with a 7D2 than with a 5D3, because the 7D2 shoots 10 fps, the 5D3 only 6 fps...

I routinely shoot shoulder-to-shoulder with a 5D Mk III user.

And I do get more keepers than him.

 

[CanonOregon]

My issue seems different than others but I did have a 'mushy' looking focus on some images at first but mine was returned for 'Lens Error 01' or it would simply freeze and nothing happening. I mentioned the 'mushy' image issue to them as well and that did seem improved afterward but the lens error issue continued. After the 3rd time in they decided to replace the lens mount, just got it back two days ago so it won't be tested for that issue until today. I guess there must be something that stores camera error info as they knew I had a problem although they couldn't get it to repeat, this time. Last time they were sure they 'fixed' it with the replacement of the diaphragm assembly in my 300mm f4 L, but it came back. Repaired that free, under CPS- worth every penny spent on that!

 

[bluemoon]

Where's the "never had to send it in, and it's bloody fantastic" option?

I can do this all day long. Slow shutter panning too - about as tough a test for AF as you get.

Huh? Option 3!!!

Pierre