All kinds of pixel counts have been 'sweet spots' over the years. The competition has moved up to 30-ish MP and I suspect Nikon will want to follow suit, in case it gets branded with 'only' 24MP. Or maybe for no other reason that when it comes to lower end cameras they go with whatever Sony SS provides as a catalogue product.
39 might be enough for UHD at 16x9. They need 45 for DCI . We can hope Nikon is going to provide DCI on all future bodies. At least I hope.
In addition, it provides at least some upgrade incentive regarding IQ for owners of 24 MP cameras (noise and DR are stagnating for a long time).
There is an argument that the 47MP on the Z9 will become Nikon's new normal. They already have the vanilla BSI version. The basic design goes back to the D850, so all R&D investment is paid off and the basic sensor is low cost for them (it doesn't cost any more to fabricate a 47MP sensor than it does a 24MP one). They have the stacked variant for the Z9, but they can introduce lower cost stacked variants, with different circuitry in the stack, maybe less buffer and single channel readout. One advantage of a stacked sensor is that it can remove the need for a specialist sensor interface on the graphics processor (or at least changes it to a more standard high-speed bus) , which allows more flexibility in ringing the changes in processing power. I'm told that Nikon has trialled 67 and 100+MP variants for the high resolution model, whatever it's called, so if they continue with their two resolution model, 47MP would be the 'low' resolution one.
Totally agree on that one. Photography really hasn't changed much over the years. But the "sweet spot" seems to change. Prices change, storage changes, processing power changes, technology changes, etc.
Well only in the crop-sensor bodies AND -C is the same as super-35 in CINE bodies and the sensors for these are plowing forward -- I of course refer to the new ARRI Alexa 35.
There isn't yet a FF sensor that will yield more than 14 stops of DR at the pixel spatial frequency, so those extra two bits will just be encoding noise.
What value did you get for DLA? Diffraction limited region actually starts around f/11 for 3 um pixels.
Ah, that's what was meant by 'DLA'. Has that fallacy become a thing now?
YET !!
Indeed, but they've been stuck at the same point for five or six years. And they don't need 16 bits until they get there.
I can't see that makes much sense at all. Most lenses have an aberration limited region (at larger apertures) and a diffraction limited region (at small apertures). Aberrations decrease as the aperture is closed and diffraction increases, so in the middle there is a region where the one merges into the other, and that is generally where the peak resolution is. None of these is dependent on the pixel size. More pixels just allows the camera to reproduce more of the detail that the lens projects onto the sensor.
There's no such lens. And even if there were, a low f-number 'DLA' is a good thing, not a bad one. So ajm057's 'horrendously low DLA' makes no sense at all.
Thank you very much for letting me know. I would have never thought of that.
Cambridge in Colour and ScanTips.com - both of whom have Diffraction Calculators, both ignore optical performance of actual lenses, which I why I stated one has to wait for actual measurements.
The sweet spot has remained constant, it's always been whatever camera the poster in question is happy with.
If you look at this post of mine: https://www.dpreview.com/forums/post/66492494, F/5.3 for X-H2 might be a border between regions 1 and 2 while diffraction-limited aperture should be a border between regions 2 and 3 instead.Cambridge in Colour and ScanTips.com - both of whom have Diffraction Calculators, both ignore optical performance of actual lenses, which I why I stated one has to wait for actual measurements.
Fuji Film
Others
