hi anders, just did your test (even though I knew the results before that). my old zd shows far more noise than your example. the only difference: yours shows wurmy shapes - mine shows more normale noise in smaler spots.
purely based on the sample you mailed me, i think the problem your are experience is that you dont give the sensor the amount of light it needs. the sensor (like most mf-sensors) screems out for lots of light. Never mind using iso 400: I always expose for iso 25! and then you get clean shadows. i have seem this before on cameras like the sigma sd9 and the kodaks. if you want a camera for iso 400 (or underexposed iso 100) save yourself the money and get a canon.
otherwise: set it to iso 50, overexpose till the highlight-warning on the camera histogram blinks wildly. dont bother lookig at the camerameter (till last week, I didn´t even realise, I could meter with the ZD and old lenses). use the histogramm and expose to the right.
and: experiment with different raw processors. for example: RD hadles noise different to the mamiya software or acr.
as for the wurmy-shapes: I had a sigma sd9 exchanged for the same thing. dont know, perhaps they droped some fluff on the sensor when the were puting it together. You might want to get it exchanged for a sensor that shows more fine graine noise. but than again my warning: if you start seen those things - you propably didnt get enough light onto the sensor.
perhaps I got it all wrong though.
stefan
p.s. about the purple wurms: are they always in the same position?
(if they are, that would point to a bad sensor)
[a href=\"index.php?act=findpost&pid=137827\"][{POST_SNAPBACK}][/a]
Stefan,
Your reply is very interesting. Much kind thanks for taking a look at those samples and for doing that test. It would be really interesting if someone with a problem back did same test.
Reviewing the histograms from my test photos, there are two spikes not far from white and black limits, so the average metering worked rather good for that. I reran the test this morning to get one spike more towards the right limit and even exceeding it. It is difficult to say precise but if anything I would say shooting to far right made the problem issue worse, not better. It does not seem noise related, noise ware do not touch it.
Last night I got home at around sunset, a cloudy one with openings between clouds. I tested shooting towards right then also, with city view in partial shade and the sky. Seems around +1.0 on camera made best and cleanest. Thanks for advise. I recall I read that Michael Ezra even shoot by +1.5 or more outside. Same time I recall having read a review somewhere that with the ZD (or is my memory failing) one could expose like film rather for the dark, or is that complete wrong, or for certain circumstances?
You mentioned you shoot at ISO 25? My ZD starts from ISO 25. Any trick?
I have tried SilkyPix, CS3 with camera raw, Lightroom, Noise Ninja (I am on PC, so no RD). Of them I like the results and what I can do from SilkyPix best, also with regards to its NR.
Again, very much thanks for running the test, also for the other hints in your post.
Regards
Anders
Henry,
Perhaps the ZD camera at moment is more stable. If it wasn't for the issue I found I would for sure not worry. I really like the design and layout of controls of the ZD camera. I think controls are better than my D200 was, much more for what I feel comofortable with as a camera. The ZD camera is logic, it is simple to pick up if I leave it without shooting for awhile also. I think both Nikon and Canon could learn from Mamiya on this
I do wish the any issues were nonexistent, and instead that Mamiya was improving the noise performance. If both those happen, it is an amazing camera.
Regards
Anders