[Klaus]

Canon forgot something ...

http://www.opticallimits.com/canon_eos_f...anon2470f4

 

[popo]

That focus shift would be a major deal breaker for me... so if I ever get a high(er) quality standard zoom I guess the 24-105L just went up the list again.

[soLong]

<p style="font-size:12px;font-family:Helvetica;">...i live in this fantasy world where the Canon reps read this and immediately jump in to leave the very important message here that…. "why yes we now know about our forgetful designer, a very sad case of alzheimers you know(we don't like to our let people go) and we have a fix for the problem that has been let through…. we've rushed out a free repair worldwide to this impossible focus shift problem and no new lens sold will be like this"….

<p style="font-size:12px;font-family:Helvetica;"> 

<p style="font-size:12px;font-family:Helvetica;">but so far this is only my delusion …. maybe someday soon it will come true and people will regain confidence that the lens they buy will be of the quality that they had hoped for - i'll update you all on any future movement in this direction

[Guest]

Hello Klaus,

when you did the LoCA test, you state the AF was off. Now, for real world situations when AF is on, I presume that Canon has implemented some kind of correction algorithm (presumably like in the 50mm f1.2) in their AF software. You state there was an effect even at moderate distances, which I presume means it wasn't properly corrected. Have you done a LoCA series with AF, to see how far off it is?  It might be nice to add a few more words to the review in this regard.

[Klaus]

Hello Klaus,

when you did the LoCA test, you state the AF was off. Now, for real world situations when AF is on, I presume that Canon has implemented some kind of correction algorithm (presumably like in the 50mm f1.2) in their AF software. You state there was an effect even at moderate distances, which I presume means it wasn't properly corrected. Have you done a LoCA series with AF, to see how far off it is?  It might be nice to add a few more words to the review in this regard.

 

I do AF reference checks during the MTF analysis.

The results were, as mentioned, fine at f/4 but beyond it was a roller-coaster ride which is typical for not sufficiently corrected RSAs.

[Guest]

I'm really at a lost why the cameras can't compenstate for this sort of thing. I mean it has a nice dandy computer with lens communication et all. Oh well; also that seems to be a huge shift; rather unusual.

[Guest]

I do AF reference checks during the MTF analysis.

The results were, as mentioned, fine at f/4 but beyond it was a roller-coaster ride which is typical for not sufficiently corrected RSAs.

 

 

As you2 said, one would have expected a better correction through software, if it's that bad. Seems indeed strange. Firmware update on the horizon?

[mst]

I'm not sure if software correction is really possible.

It would be simple in Aperture priority mode, but in Shutter priority and P-mode, software correction would require constant refocusing.

-- Markus

[Guest]

I'm not sure i understand; Wouldn't the camera know the aperture   (either by the user or the camera) at time of focus so couldn't it then compute a focus offset?

 

I'm not sure if software correction is really possible.


It would be simple in Aperture priority mode, but in Shutter priority and P-mode, software correction would require constant refocusing.


-- Markus

[mst]

The time of focus is not necessarily the time when the shutter button is pressed. In fact, there might be some delay in between, and at least in case of P and Tv modes, the aperture might change inbetween. Imagine you focus and then recompose...

You could of course suggest instead that the camera refocuses when you finally press the shutter button. However, that would probably lead to some sort of shutter lag.

-- Markus