[Class A]

However, one has to acknowledge that a larger format is incredibly easier to focus, both manually and by an AF module.

But isn't that true only if you are talking about output sizes that do not take advantage of the MF format?

I see your argument, but if you are printing MF images at the same dpi settings as APS-C images (i.e., not at the same output size) then the AF accuracy required should be the same, right?

I see an analogy to exposure. In converting shooting parameters between formats, AFAIC, one should only change ISO values if one assumes that the larger format should be allowed bigger prints without a noise disadvantage. This is how "exposure" is traditionally defined, but I see a point for talking about identical output sizes in which case there wouldn't be a need to increase ISO, since a push in post would do the same without an effective noise disadvantage for the larger format.

N.B., the above "exposure" discussion involves the difference between increasing film ISO (which should increase effective sensitivity) and digital ISO (which more often than not is practically the same as pushing in post) and we do not need to go there, but I thought that there is a point to be made about either making the assumption that you can exploit the advantages of a larger format (i.e., print large without showing more noise and/or showing more AF inaccuracy) or not making that assumption.

[falconeye]

But isn't that true only if you are talking about output sizes that do not take advantage of the MF format?

Yes.
In format discussions, I first assume that equivalent images are discussed.

When comparing crops shot at the same pixel pitch, you loose one of the two factors of advantage. But a difference remains: with the larger format, the difference between aperture used for focus and the one actually used for the photo, is typically larger. E.g., in MF, it isn't uncommon to focus at f/2.8 and shoot at f/8 or f/11.

There are some sources for 80MP images from MF on the web, in full size. I never saw any focussing issues.

[Zygonyx]

Independently of the theory, the bigger viewfinder size(s) is in any case of crucial help in MF, when manual focussing too.

[deejjjaaaa]

There are some sources for 80MP images from MF on the web, in full size. I never saw any focussing issues.

I doubt that people will be posting 80mp samples w/ focusing issues... you rather need to read forums like getdpi, ll or similar or forums like phaseone's U2U

[falconeye]

I doubt that people will be posting 80mp samples w/ focusing issues... you rather need to read forums like getdpi, ll or similar or forums like phaseone's U2U

There are websites out there postings hundreds of such photos every day. Their entire work they actually shoot. 80 MP and no focussing issues. I leave it as an exercise to the reader to find the sources

[philbaum]

...
However, note that the K-7 and K-5 are advertised to have accurate AF independently of the light temperature. That's why they have a "+" in their "SAFOX" designations.

I looked at the only K5 press release i could find on the pentax imaging site and it was from 2010, and i couldn't find a similar phrase "independently of the light temperature"

Rather, in their press release:

Newly improved, highly responsive and accurate 11 point SAFOX IX+ autofocus system with dedicated AF assist lamp and light wavelength sensor.

Similar mention of the light wavelength sensor is made in the specifications. Maybe most of us assumed it was as good as the daylight AF, but they never said precisely that. On the other hand, there's probably a lot of advertising out there that i'm not aware of.

[Class A]

When comparing crops shot at the same pixel pitch, you loose one of the two factors of advantage. But a difference remains: with the larger format, the difference between aperture used for focus and the one actually used for the photo, is typically larger. E.g., in MF, it isn't uncommon to focus at f/2.8 and shoot at f/8 or f/11.

Thanks, Falk.

[Class A]

Similar mention of the light wavelength sensor is made in the specifications. Maybe most of us assumed it was as good as the daylight AF, but they never said precisely that.

Here's a quote from a review that quotes Pentax:

...and light wavelength sensor which, Pentax says, "factors the light source at the time of shooting into its calculations to improve the accuracy of autofocus operation."

It is clear what the light wavelength sensor is supposed to do, i.e., compensate for AF errors in other than daylight operation.

FWIW, In the context of his K-5 AF FF investigation, Falk figured out that the light wavelength sensor is most likely integrated with the metering sensor, i.e., one sensor playing a double role. It is the main reason why the K-5 suddenly cannot focus precisely anymore despite the fact that the AF module itself would be sensitive enough to work in worse conditions.

[deejjjaaaa]

There are websites out there postings hundreds of such photos every day. Their entire work they actually shoot. 80 MP and no focussing issues. I leave it as an exercise to the reader to find the sources

so somebody is posting openly full, not downsized, 80mp pictures (JPG proofs) and hundreds of them for anybody out there to download from their assignments ? please post URL...

PS: and even somebody does that - why he/she will post OOF shots ? are those studio shots w/ severely stopped down apertures so that OOF is rather not an issue ? are we talking about AF or MF and are you stating that 645D will not miss ever using SAFOX ?