Welcome to ReidReviews
The comparison images put me in a bit of a quandary.
Admittedly Sean's sample images aren't the final word, because he didn't have the final firmware, nor an Adobe M profile to use in LR/ACR; so refinements may later reveal more difference. But viewing his samples, I was surprised that IQ is so very close for the different sensors, at least when viewing detailed crops on a monitor. The main difference is, as expected, higher ISO performance from the CMOS sensor – but little or nothing else that seems important to image quality. Just a tiny increase in resolution in the same-size images, and some color-rendering differences that are probably profile-dependent and/or easily balanced in post-processing.
Here's the difficulty for me: I've already acquired Summiluxes in the focal lengths I use (or Summicron for 28), and I was 'raised' on Tri-X; so an ISO limit of 400 or so has never bothered me much. Indoors, F1.4 at 1/30 or 1/60 is just a habit. In low light I use ISO 320 with M9, and only rarely 640 (because of shadow noise that can show up in medium-large prints).
But in Sean's samples, 1250 is a fully acceptable M240 ISO; maybe 1600, before I'd notice any noise in 14x21" prints.
So here's the hard choice for me:
--Keep M9 and use the fastest lenses (as is a habit); or
--Move into 'modern times' with the M240 sensor, and then be able to stick with the pre-aspherical Summicrons that I like best? (Most of my shooting is with 35 Cron v4 or 40 Summicron-C. I've never really bonded with my 35FLE, because it's heavier and IMO 'clinical' in rendering.)
The irony is that there's really no cost involved, because the M240 would be 'free' if I sold the Summiluxes – only the inconvenience of Internet buying/selling.
This is a toss-up for me. I don't know if anyone else is looking at sensors and lenses from the same perspective, but I'd like to hear about similar/different responses to Sean's review.
Edited by thompsonkirk, 24 February 2013 - 19:42.