philber wrote:
I agree with both of you,except I have a quetion for Ron. How does resizing affect corner smear. My hunch is that, once smeared, nothing can resurrect the sharpness. So resizing is not a magic wand to get out of trouble, and thus, the resizing does not affect the validity of the shots as proof of minimal smear. Right or wrong?
Resizing to 24mp will do diddly squat. Anyway, I've just uploaded a 36mp sample with the same combo.
philber wrote:
I have a quetion for Ron. How does resizing affect corner smear. My hunch is that, once smeared, nothing can resurrect the sharpness. So resizing is not a magic wand to get out of trouble, and thus, the resizing does not affect the validity of the shots as proof of minimal smear. Right or wrong?
The smear will have a certain scale, and when you resize, you affect that scale, so it should become smaller, or less noticeable (provided you are looking at 100%, and not at a print ). From 36MP to 24MP doesn't sound like it would change much though.
carstenw wrote:
Like I said, the scale is smaller, as long as you are looking at 100%. For a fixed size print you would lose sharpness as you downscale.
Strictly speaking yes.
But by downsampling like that, linearly you downsample 5 to 4 pixels, which is negligible in smearing. For smearing it is of larger scale than per pixel sharpness. So if you don't see smearing at 24MP, you won't in 36MP (at least you won't discern it between smearing with other sources of unsharpness)
wfrank wrote:
That would imply that smearing is the same for A7 and A7r. And that might so be, but I dont think we know that yet.
Are you speaking to me?
Well, I can't see where the "imply" comes from. My argument is from A7R 36MP downsampled to 24MP. And A7 is a completely different sensor which is not a simple downsampling from 36MP.
p.136 #10 · A7/A7r - performance with WA RF lenses
kolen wrote:
Strictly speaking yes.
But by downsampling like that, linearly you downsample 5 to 4 pixels, which is negligible in smearing. For smearing it is of larger scale than per pixel sharpness. So if you don't see smearing at 24MP, you won't in 36MP (at least you won't discern it between smearing with other sources of unsharpness)
Your logic doesn't follow for all sizes of smearing. If the smearing is slightly larger than one pixel, you will see it. Downscale it to 24MP and you may no longer see it. For all possible values of smearing, and all manner of detail in general, there are some which you will see at 36MP but not at 24MP. Otherwise, why even go to 36MP. 22% is not a trivial number, even if it isn't large.
p.136 #12 · A7/A7r - performance with WA RF lenses
carstenw wrote:
Your logic doesn't follow for all sizes of smearing. If the smearing is slightly larger than one pixel, you will see it. Downscale it to 24MP and you may no longer see it. For all possible values of smearing, and all manner of detail in general, there are some which you will see at 36MP but not at 24MP. Otherwise, why even go to 36MP. 22% is not a trivial number, even if it isn't large.
Please read the word in bracket. And also read the first sentence that I admit "strictly speaking yes".
See? I have tried to be as accurate as possible and be not overgeneralize things. So my logic still follows, if you follows the exact description I made.
I knew I shouldn't be joining this kind of discussion. Whenever I did it, people are just not paying attention to the fine details. People are usually in the "argue mode" thinking it has to one way or the other. But if you follow my words exactly you see I am agreeing with both sides.
Anyway, I will just shut up.
p.136 #13 · A7/A7r - performance with WA RF lenses
What is the root cause of smearing?
Do the A7 & A7R share the identical attributes that are responsible for smearing? If not, then I'm remiss to equate how downsampling the smearing of one camera would equate to the performance of the other, based solely on pixel size/quantity without also noting the differences in the root attribute.
p.136 #14 · A7/A7r - performance with WA RF lenses
A couple of weeks ago, I downloaded a few comparisons on flickr between 7 and 7R with RF lenses and downsized the 7R files to 4000x6000 pixels. I could still see that the 7R smears visibly more than the 7 with same lens and aperture, and that the smear area is larger as well. Of course this doesn't make much sense as theoretically they're supposed to be exactly the same but that's why there are things we don't exactly know and it makes more sense to look at photos and draw our own conclusions rather than trust in theories that are based on incomplete facts.
p.136 #15 · A7/A7r - performance with WA RF lenses
edwardkaraa wrote:
A couple of weeks ago, I downloaded a few comparisons on flickr between 7 and 7R with RF lenses and downsized the 7R files to 4000x6000 pixels. I could still see that the 7R smears visibly more than the 7 with same lens and aperture, and that the smear area is larger as well. Of course this doesn't make much sense as theoretically they're supposed to be exactly the same but that's why there are things we don't exactly know and it makes more sense to look at photos and draw our own conclusions rather than trust in theories that are based on incomplete facts. ...Show more →
Actually theoretically they are supposed to be not the same. Agree with the rest
p.136 #16 · A7/A7r - performance with WA RF lenses
kolen wrote:
Actually theoretically they are supposed to be not the same. Agree with the rest
I might be wrong on this, but in theory, the smearing should be the same because native lenses have been designed to work equally well on both sensors, taking into account any field curvature and astigmatism caused by the sensor cover. Therefore, even though one has an AA filter and the other doesn't, the thickness is normally supposed to be the same on both cameras, and the refractive index of the glass used is also supposed to be similar. Example is Nikon D800 vs D800E. Also in theory, if there are differences, they should be in favor of the AA less 7R. That the 7 shows less smearing goes against all logic, but it's a fact nonetheless, and we don't know yet why, but I'm sure when someone takes the sensor apart, we will know
p.136 #17 · A7/A7r - performance with WA RF lenses
edwardkaraa wrote:
I might be wrong on this, but in theory, the smearing should be the same because native lenses have been designed to work equally well on both sensors, taking into account any field curvature and astigmatism caused by the sensor cover.
no, native lenses just have to be designed to reach a certain level of performance for the worst sensor. i sincerely doubt they are designing the lenses so that they would perform worse with thinner cover glass. PDAF issues don't come into play in the design of native lenses since they all use on sensor focusing methods.
I think we need to see it at infinity 30 meters + to really see what's happening. I have the lens, so I'll give it a try too. I'll be shocked if it doesn't smear well before the edge @f/4, but in a good way
Your test shows it WAY beyond my expectations @2.8: I'll be thrilled if it can do that in a proper landscape
Wow, just tested indoors WO at 3m and it's pretty good. So glad you made this comment. I will test on mountain tomorrow
Will be so cool if this turns out to be a lens the A7r really likes.
p.136 #19 · A7/A7r - performance with WA RF lenses
Steve Spencer wrote:
Please no, I want to be able to move the focus magnification around. I actually really like the system on the OM-D. They have a number of boxes that you can make bigger or smaller (bigger boxes means fewer boxes) and you move between the boxes with a four point arrow. I find it easy to move to the spot without even looking and once you are at the box you want you just turn the focus ring for magnification. It is quick and lets you move anywhere on the screen.
I'm suggesting it as an extra mode not to replace what already exists. i.e you could toggle between normal movement or the quick move.
p.136 #20 · A7/A7r - performance with WA RF lenses
sebboh wrote:
no, native lenses just have to be designed to reach a certain level of performance for the worst sensor. i sincerely doubt they are designing the lenses so that they would perform worse with thinner cover glass. PDAF issues don't come into play in the design of native lenses since they all use on sensor focusing methods.
I'm sorry but that sounds like you just want to be contrarian. The glass cover thickness is calculated into the lens design, and a thinner cover glass will result in image degradation.
). From 36MP to 24MP doesn't sound like it would change much though.
