PHOTOOG Photography writings by Olivier Giroux

14Feb/093

Bad news for the G1/Leica-M combo

Recently Sean Reid (pay-for review site) tested the innovative and popular Panasonic G1 with some of his Leica lenses. While I’m not going to copy his protected images or repeat his lengthy and insightful discussion, I can tell you that it put a damper on my excitement for the G1 as an entry-level M camera. It might explain why we don’t have a Leica G1.

Sean’s most revealing test pitted the G1 against the M8 in a micro-contrast face-off, using the same lens on the same subject under controlled conditions. In theory the G1 should have an advantage here if the lens resolves finely enough for its pixel pitch, as does Sean’s 28mm Summicron ASPH. This is popular wisdom because the G1 doesn’t record pixels as far from the center as the M8 does (2.0x versus 1.3x crop), and hence stays in a comfort zone.

The outcome? It wasn’t close, the M8 humiliated the G1 in the corners.  I mean the Summicron looked like a bargain-bin lens on the G1.

I think we have a good explanation for why that is, and it goes back to the M8’s launch material. Look at the diagram for offset microlenses:

 
Image credit: Leica, via Dpreview.com.

 The G1 obviously doesn’t have these micro-lenses because it is designed for lenses that project rays that are more parallel - at least it's clear it's what it works well with.  The Leica telephotos are likely to work well and the wide-angles not so well.  This is a shame because the 2x crop factor invites the use of wide-angle lenses.

It's possible that when we see a Leica camera come out in the Micro4/3 format, it will have offset microlenses.


Comments (3) Trackbacks (1)
  1. Sean’s conclusion is based on the 28mm Summicron. He has said that the results may not be the same for other M mount lenses. I have tried various M lenses on the G1 and I agree that the longer focal lengths look better although my impression is that the 35 cron is fine.

  2. we are talking about sensors with different pixel densities, with (Panasonic) or without (Leica) antialiasing filter (loss in sharpness and microcontrast).

    And I’m not sure the resolving power of the Summicron is high enough for the small pixels and antialiasing filter of the G1 sensor.

    Leica lenses are known to tend to privilege microcontrast at the (slight) expense of absolute resolution.

    I beg your pardon, but this seems more like comparing apples to oranges.

  3. Hi Alfredo,

    There’s truth in that argument.

    A few thoughts come to my mind.

    In Sean Reid’s images the performance of the Summicron in the center of the Panasonic image is just fine, it’s quite sharp. The Summicron and the Panasonic sensor work very well together in the center. The problem is when you leave the center – then the M8’s image remains sharp all the way into the corners and the Panasonic turns to pudding.

    Note two VERY important things here:
    (a) The distance to the Panasonic’s image border is insignificant for a FF lens like the Summicron so we do no expect corner softening there (we expect a more even performance in fact).
    (b) The Summicron’s MTF charts at f/5.6 show almost no weakening until the borders of the FF image is reached, which are twice as far away as the m4/3 image border.

    So even if we ignore the ‘sweet spot’ effect, the expectation is that if the center on the small sensor looks good then so should the border. They should both be good or bad, not only good in the center. This is really the key to the argument that the Panasonic sensor is missing some ‘special sauce’ which the M8 apparently has.

    My parting note is on a different level. Irrespective of technical concerns about the validity of the comparison, my take-away from inspecting only the G1’s images is that the results were unacceptable considering the equipment involved. The pairing itself is technically wrong-headed.


Leave a comment