Microlenses and FF

[guidomo]

Advertisement (gone after registration)

Hopefully some of the more tech-savvy forum members can comment on this.

 

As we all know, when making the step to FF the microlenses would have to "intervene" much more to get the light rays into the right angle/spot onto the sensor, particularly for the wide angle lenses.

 

I was now explained be a friend, that this would most likely mean that image quality with longer lenses would be somewhat degraded because the microlenses' action was overcompensating.

 

I can see that the difference in incidence angle between wide and long lenses will get bigger with the M9 than was the case with the M8. I would presume that Leica will have tried to find a balance between corrections at both ends of the spectrum - just how good can such a balance be??

[adan]

It is possible. The diagram below is a very rough and somewhat exagerrated schematic, but as the microlenses offset more and more to the corners (to the right) to minimize wideangle vignetting, they could end up offsetting relatively perpendicular telephoto light rays too much, such that they miss the sweet spot of the pixel, and begin vignetting a bit themselves.

 

It is analogous to the "H" screens Nikon sold for F cameras - microprisms across the whole field. They had to make 4 versions, for different apertures and focal lengths, because any one version would vignette with some lenses.

 

Worst case: all lenses at the extremes of the range would need coding for some software correction of vignetting. Given the fairly small range of M lenses (not even 9:1 - 16 to 135), I think firmware could easily handle it.

 

I should add that Leica's newer wides likely have some digital friendliness designed in. They may well put a cut-off on lenses the M9 supports, as they did to some degree with the M8. If someone wants to use a Super-Angulon on the M9, or even the C/V15mm, they may be on their own regarding vignetting correction. (I just hope they support my Mandler 21)

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[delander †]

You mention the 15mm CV. What about the WATE, I think it would be a disaster if that recent and expensive lens did not work correctly on the M9.

 

Jeff

[plasticman]

Guidomo - search this forum with the phrase 'Laws of Physics'. You'll immediately see that the M9 is actually directly contrary to these in every conceivable way, and therefore currently a figment of our collective imagination and will never happen (not before 2012 at least).

 

PS: solving the IR problem is also impossible without sacrificing image quality - so that's out too.

[Olsen]

You mention the 15mm CV. What about the WATE, I think it would be a disaster if that recent and expensive lens did not work correctly on the M9.

 

Jeff

 

I think both these two lenses will work very well on M9, the WATE slightly better than the Super Wide. It is first of all 'fast and wide' that is problematic to design for a digital camera - without making them huge. - And that is what Leica has done to their fast/wide lenses.

[adan]

Jeff: "Leica's newer wides" was intended to include the WATE, 28 'cron, etc. Anything currently in the lineup. I'd expect that if the WATE is covered, the 15 c/v is also probably OK, or at least no more vignetting than on film - but I doubt Leica bothered to test non-Leica lenses or lenses not already codable for the M8 - not their problem.

 

Realistically - this will all become clear in 5 days. Take deep breaths!

[haroldp]

Advertisement (gone after registration)

The WATE appears to be a retrofocus design, which is closer to optimized for digital sensors than conventional RF wide angle lenses.

 

Unlike the MATE, the WATE is actually a true zoom ( across it's range ) with stops.

 

It will probably be fine with FF sensors, it is the older designs that are more ideal optically that will be harder to correct.

 

All this is fun, but I would like to see what the clever folks in Solms, and their suppliers have come up with before spinning too many cycles on these issues.

 

Regards .. Harold

[lars_bergquist]

All damn M lenses shorter than 35mm (but including modern 35mm ASPH lenses) are of necessity retrofocus designs, because of the fixed flange-to-sensor distance of 27.8mm. Remember, a non-retrofocus lens has got to have its optical main plane in the vicinity of the physical/mechanical center of the lens, close to the aperture mechanism. Putting all or even most of the optics >inside< the camera body flange would necessitate an extremely small and thus slow lens. This is in fact how early wide angle lenses were designed, like the 2.8cm Hektor and Summaron lenses -- they were tiny! Not to speak of the effects on TTL metering.

 

The position of the optical main plane (to over-simplify a bit) determines directly the focal length. In a retrofocus lens, the plane is shifted to the rear, in extreme cases to a position outside the optical cell. The case is exactly the reverse one in the case of (true) telephoto designs.

 

The old man from the Age of the Zeiss Hologon

[ljclark]

I doubt Leica bothered to test non-Leica lenses or lenses not already codable for the M8 - not their problem.

 

 

My guess would be that they did test some non-Leica lenses -- especially the very wide lenses. It just doesn't make sense to ignore lenses that are already in production -- when it is so easy to learn something about the real world performance of your product.

 

But I wouldn't expect any of the performance data to make it outside of the lab. That would be asking a bit much.

[adan]

lj: Actually, you're probably right, especially if one includes "informal" testing by staff and/or the beta-testers (well, the official beta-testers, as opposed to first-adopters like me).