Donald Duck wrote:
alatchin wrote:
Donald Duck wrote:
alatchin wrote:
Donald Duck wrote:
Since the units there are lp/mm, you just proved that the m43 combo resolves about 1/2 of what the FF one does (twice as many lines on FF, roughly speaking, to to the larger sensor).
No, because their tests depend on MP based bodies, not sensor area. The D3x needs about 50% more MP to match the E-3 for sensor density. (around D800 mp).
So, lets have a laugh here, an 8000$ body with a prime lens marginally beats a $1700 body with a zoom in the image centre, is just beaten in the mid area, and is destroyed on the image edge... Tha is why the test images appear as they do.
You missed my point.
Lines pair per mm is exactly what it sounds like. A 2x smaller system
must resolve twice as much per mm, just to offer the same resolution in the end. In this case, if you believe lenstip, it does not. The OP is so proud to post those charts without realizing that they make him look ridiculous and disprove his point.
Have you read their explanation of how they test cameras? Comparisons are best achieved with comparable pixel density, not sensor height.
Comparisons of what? Lenses? Yes. But the iphone lens resolves even more than a Canon L lens, and MF and LF lenses resolve less. What is under question here is what system resolves more. Display the image the same size, etc. Then the 4/3 lens has to resolve 2x more just to stay even. And the MF lens needs to resolve less, etc.
you are really all over the map... Yes, comparison of lenses, because sensors change from body to body, MP keep marching forward it seems. The 43rds lens only has to resolve twice as much on a lenstip test if the FF sensor has a lower pixel pitch by half. If, as in the case of the E-3 vs the D3x the difference is more like 20% then it only has to resolve 20% as much... Remember lenstip is testing LENSES not system resolution which is what you are talking about.
So in the examples provided the 14-35 does indeed resolve more than 20% as much wide open, and more so at the frame edges.
The highest resolving power on the sensor of smaller format lenses is not an advantage - it is a necessity.
See, this is where you are jumping from proving one point to another. You are right for the same MP count it will have to resolve twice as much because it could be assumed the pixel pitch would be half on the smaller sensor. But for the purpose of these tests it doesnt.
You are wrong about the border performance as well. FF does not have worse (or better) corners compared to the center. At equivalent f-stops, of course.
Well, if you dont want to use a lens at its intended focal length then sure you can crop a FF sensor all you like...
I was not talking about that. The soft corners of FF lenses on FF bodies is a myth.
Huh, hmmmm, have you looked at many lens tests for FF bodies?
Of course you start losing your theoretical DR, Noise and detail advantage. Quite a silly thing to do.
There is so much advantage to lose, and so much to stay... No - I would not crop the corners, I like the 3:2 format.
So keep the soft corners, you were the one who wanted to crop them in your last comment. ANd there isnt so much of an advantage for a 12mp D700, D3, D3s, There is some room to crop for the D3x, D800, 5Dm11, 5Dm111, A99, A850 and a900, and to arying degrees you would end up worse off for the exercise.
At equivalent Fstops for the equivalen image the FF sensor loses its noise advantage and its DR advantage... not cool for the $3000-10000 camra.
EDIT: The 4:3 format actually helps to get better borders but one can always crop an FF image to 4:3 as well.
See above for why that doesnt work out so well as an argument.
The truth is simple, it is easier to make a lens with even corner to corner performance on a more "square" sensor as there is less pressure to move to the edge of the image circle.
Correct, the only problem is - do you really like square images?
Hey, that is up to the individual, the point remains, the wider the format, the bigger the sensor, the more pressure you put on the lens for a given MP count.
It is easier to make a near perfect lens for a smaller sensor.
Wrong. It is "perfect" on the sensor only. When you enlarge to the same size, and want to keep the same DOF and total light, it is much harder, maybe even impossible, to make even an equal lens. For example, which m43 f/1.4 lens wide open can get close to a quality lens at f/2.8 on FF? Which Canon or Nikon lens at f/1.4 on a crop body can be as good as a good lens at f/2-f/2.2 on FF?
Hey, if you want to buy a fast lens, say a 50mm f1.4 to use it at f2.8 be my guest, lose shutterspeed or light gathering and be left with an image comparable to a smaller sensor... great result for the FF body wasnt it?
It is harder for a smaller sensor to compete with a bigger sensor for noise performance, however it is easier to make a faster lens for a smaller sensor.
The "faster" lenses on a small sensors are not really faster.
See you miss the point. It is always a trade off. Faster lenses on smaller sensors make up for the smaller sensors. Stopping down FF lenses to make up for poor edge to edge performance doesnt do anything for the FF sensor except cost light gathering or shutterspeed for more even resolution.
FF cameras enjoy their advantage due to larger apertures at certain AoV meaning more light, this also applies to the DoF differences. To enjoy those advantages you pay the price of lens issues outside the centre, along with extreme lens pricing etc.
For 43rds users to claw back some FF advantages they need to buy expensive large lenses with even wider apertures (25mm f0.95s, 35-100 f2) and deal with their own optical issues, or pricing issues, size issues etc.
You want to have your cake and eat it... That goes against the laws of the universe
