I said "Now whether you *need* or *want* that extra resolution...no argument there.", which was my attempt at alluding to your #3. And I absolutely am not trying to say that anyone should use this capability.
For #1, diffraction is the same whether you are shooting a 45 MP shot or you do pixel shift. Meaning the "amount" of diffraction is the same. Now how much that diffraction affects the pixel shift combined version, which inherently should give you more color resolution (at the minimum), I can't say. But there would have to be something wrong with the implementation if you shot let's say at f/11 and there was no increase in resolution between the native 45 MP RAW version and a 400 MP pixel shift version. I'm pretty sure I see a resolution difference even in the f/16 shot I posted above, between the 61 MP and 240 MP version. Now how "useful" that increase is.....again, each person really needs to make that call.
And for #2 it's impossible to maximize everything right? We are constantly making decisions on what to trade off. I tend to shoot higher resolving lenses, but that may have impacts on other qualities of the lens. But to say that something doesn't provide benefit unless it is maximized seems too black and white, which I assume you're not saying. I tend to stop down to maximize depth of field, at the expense of sharpness/resolution. I see in the forums people who say they shoot all their landscapes at wider apertures and then stack so they avoid diffraction.
I'm in no way lobbying that this is useful to anyone, that's personal preference. But it seems worthwhile to at least have an understanding of it, what the benefits and downsides might be, whether you use it or not.
In any case I'm not really trying to disagree with anyone. But I'm just trying to separate what benefits "should" be seen, from the decision as to whether those benefits are useful. To me the latter is something everyone decides based on their shooting style, preferences, etc.
Tim
gdanmitchell wrote: tsdevine wrote:
I don't know Canon's specific implementation, but I assume it will end up sampling all colors at each pixel, which should give you higher color resolution than a single shot. (Even if you were to stop down further than when diffraction starts to become limiting.). At least that's what I would tend to think.
If there was no resolution benefit shooting in pixel shift at apertures higher than wide aperture, it would be really limited in terms of application.
Heck the sample I posted was f/16 on a 61 MP camera. I think there is more usable detail in that 240 MP shot than in a straight 61 MP shot at f/16.
How did you judge resolution difference? Did you uprez the 45 MP shot?
TakesRandomPics wrote:
I tried some test shots of flowers. Either I'm hitting the limits of the lens or my monitor but I see no appreciable difference in detail between 400 and 45. How much would you stop down when using something like this?
A few thoughts...
1. It is true that no image gets worse from higher sensor resolution (where other capabilities are the same) and that generally there is at least some improvement in all cases, even those in which the sensor "out-resolves" the lens. (That's not a bad thing, by the way.)
2. To maximize any potential image detail advantages of the higher resolution system (whether from a higher resolution sensor or from the system used here), more attention needs to be given to everything that affects image sharpness — lens quality, aperture choice, camera stability, subject motion, etc.
3. Given that current high resolution sensors can produce really excellent results already, the detail-enhancing ability of things like 400MP images are perhaps more theoretical than visible, and it is a serious question whether they are worth the trouble in all but a few rare outlier cases.
Dan
Apr 10, 2023 at 11:50 AM
Previous versions of tsdevine's message #16215234 « R5 - 400 MEGAPIXELS Vs Sony A1 pixel shift »
