[boren]

[quote name='Brightcolours' timestamp='1300127105' post='6769']

We were talking PD-AF exclusively here[/quote]



That's right, you were talking PD-AF exclusively, but my post wasn't wasn't directed to you. I specifically addressed Klaus.

A Grman magazine tested the AF tracking performance a few years or so ago, it was an interesting read. Tested were the Canon 40D, 1D mk III, Nikon D80, D300, D3, Sony A700, Pentax K20 (I think) and Olympus E3.



I am not aware of if they have repeated the test again with more recent cameras. It was the only systematic and quite scientific AF tracking/FPS test I have seen/read.



I think it was ColorFoto, but it might have been Foto Magazin.

You didn't click the first link I posted, did you?

[Brightcolours]

[quote name='boren' timestamp='1300129188' post='6771']

That's right, you were talking PD-AF exclusively, but my post wasn't wasn't directed to you. I specifically addressed Klaus.







You didn't click the first link I posted, did you?

[/quote]

No, I did not click it. Sorry for the redundant information <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/wink.gif' class='bbc_emoticon' alt='' />

[Guest]

[quote name='Sebastian' timestamp='1300095607' post='6728']

Sorry - I had no mirror ghosting issues - even during aurora long time exposures:



[url="http://photos.sibbi.de/Europe/Finnland/15099506_PfrGw#1128622381_NkMFj-A-LB"]http://photos.sibbi.de/Europe/Finnland/15099506_PfrGw#1128622381_NkMFj-A-LB[/url]

[/quote]



I wouldn't expect it to occur in that shot anyway. There are no bright point light sources. But let's leave it at that - no need to fuel the flames any further. <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/smile.gif' class='bbc_emoticon' alt='' />

[Guest]

[quote name='Brightcolours' timestamp='1300127105' post='6769']

We were talking PD-AF exclusively here, and its predictive tracking capabilities (which relies on the AF cumputer, not the PD sensor itself). According to dpreview, the tracking capabilities do not match the sports capabilities of the high frame rate possible.

[/quote]



And according to IR and all other reviews, it does match. Where are your crops from DPRreview anyway, like you have been doing for IR?

[Guest]

[quote name='Brightcolours' timestamp='1300119058' post='6765']

And what are these images supposed to show?

[/quote]



They show AF tracking in 10 fps mode. Works just fine.

[Guest]

[quote name='BG_Home' timestamp='1300144064' post='6778']

I wouldn't expect it to occur in that shot anyway. There are no bright point light sources. But let's leave it at that - no need to fuel the flames any further. <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/smile.gif' class='bbc_emoticon' alt='' />

[/quote]



unlike what someone claimed, it's not bright light. It's small bright light with dark background, and even then lowering the exposure might solve it. It's not a "huge" issue, even is most night photos, as thousands of user night images on flickr show.

[Yakim]

[quote name='oneguy' timestamp='1300110458' post='6753']

You don't even know what SLTA33/A55 ghosting looks like, which is in exact pixel numbers in exact direction. That says a lot about your credibility on the topic.

[/quote]



I am shooting for 20 years. I think I know something about photography. Oh, and my experience with AF tracking with the A33 was also not good (where my 7D excels). I guess my "credibility on the topic" further degrades, right?



It could be much simpler though. I had AF problems but Klaus and others didn't. Either we shoot in different conditions or we have different expectation or my cam was somehow faulty (though checked by a Sony expert) or any combination of the above.



See, no reason to get personal. We can agree that we disagree in an honorable manner.

[Guest]

[quote name='Yakim' timestamp='1300172336' post='6790']

I am shooting for 20 years. I think I know something about photography. Oh, and my experience with AF tracking with the A33 was also not good (where my 7D excels). I guess my "credibility on the topic" further degrades, right?

[/quote]



There was no SLT related ghosting in the photos which has a particular look. Someone here proclaimed everytime there is bright light, it would be there, which was factually erroneous claim.

[Yakim]

[quote name='oneguy' timestamp='1300183859' post='6792']

There was no SLT related ghosting in the photos which has a particular look. Someone here proclaimed everytime there is bright light, it would be there, which was factually erroneous claim.[/quote]



Now we have a basis for discussion.



Please note that I said "Some of these pics clearly show ghosting" and not "Some of these pics clearly show SLT related ghosting". I must admit that I can not tell the difference between regular ghosting to SLT related one in most shots and that is why I wrote what I wrote. Nevertheless I am confident that some people are more proficient than me in this area.

[popo]

I thought we already covered SLT ghosting to death when the cameras first came out?



To recap from memory, so the numbers probably are not exact:

The mirror ghost only happens in one direction relative to the original due to reflections between the mirror surfaces.

The offset between each ghost generation was some small number of pixels.

The level loss between each ghost generation was quite a bit. I can't remember, but I got a figure around 8 EV in my head.



Put the above together, you need a bright light source and immediately next to it a darker area low enough to show the ghost. This can most easily happen if you have clipped point sources. In the samples I saw, this was quite obvious on distant streetlamps in a night shot for example. Similarly stars and other highlights can also show it.