Posts: 442
Threads: 25
Joined: Apr 2010
Reputation:
1
Thanks.
Maybe you should also measure sharpness at the mid-center (basically at the rule of thirds) when there's a gap between center & border sharpness.
Posts: 6,715
Threads: 236
Joined: Apr 2010
Reputation:
22
[quote name='youpii' timestamp='1304612990' post='8083']
Thanks.
Maybe you should also measure sharpness at the mid-center (basically at the rule of thirds) when there's a gap between center & border sharpness.
[/quote]
Then it makes more sense to switch to MTF curves, instead of bar charts.
However, Markus usually posts sample images (Klaus does that a lot less), and sometimes they do offer information regarding the sharpness situation across the frame.
Look at the barn/shed image, shot with f9. It shows good sharpness also towards the edges.
The metal rusty profile bar shot at f2.8, shows good sharpness in the center, ok sharpness kind of still at your rule of 3rds distance from the center, and really quite soft results towards the borders. So we can see that towards the 3rds areas already resolution is lower.
Posts: 442
Threads: 25
Joined: Apr 2010
Reputation:
1
[quote name='Brightcolours' timestamp='1304635239' post='8088']
Then it makes more sense to switch to MTF curves, instead of bar charts.
However, Markus usually posts sample images (Klaus does that a lot less), and sometimes they do offer information regarding the sharpness situation across the frame.
Look at the barn/shed image, shot with f9. It shows good sharpness also towards the edges.
The metal rusty profile bar shot at f2.8, shows good sharpness in the center, ok sharpness kind of still at your rule of 3rds distance from the center, and really quite soft results towards the borders. So we can see that towards the 3rds areas already resolution is lower.
[/quote]
Thanks for the tip about looking at the samples.
My suggestion is only about the handful of review where the sharpness collapses in the borders, not for all the reviews.
Posts: 8,010
Threads: 1,853
Joined: Apr 2010
Reputation:
49
[quote name='youpii' timestamp='1304659057' post='8089']
Thanks for the tip about looking at the samples.
My suggestion is only about the handful of review where the sharpness collapses in the borders, not for all the reviews.
[/quote]
Well, so we've got a mediocre lens and then try to make it look better again ?
Posts: 442
Threads: 25
Joined: Apr 2010
Reputation:
1
[quote name='Klaus' timestamp='1304663233' post='8091']
Well, so we've got a mediocre lens and then try to make it look better again ?
[/quote]
From a Physics class that I had years ago: when two samples are very far apart, get one more in the middle to get a better idea of how the curve goes.
Mediocre depends on how you use your lens. Many photographers compose their pictures to put the subject according to the rules of thirds and borders/corners only have sky or out-of-focus stuff. So dismal sharpness in the mid-areas is a no-go but can be less important in the borders if the subject (like portraits) allows it.
When the test show good border sharpness, we know it is OK too in the mid-areas. But in the case of the Sigma, we're not sure of how fast the performance drops from your test. Maybe reading MTF curves would be better then.
Posts: 2,441
Threads: 320
Joined: Apr 2010
Reputation:
19
You can always draw some conclusions if the lens is tested on both formats (DX review will follow soon).
Rule of thirds resolution figures will not happen. First of all because of time constraints. And then there's also a quite pragmatic reason: our resolutions charts are customized and optimized for our measurements. At the rule of thirds grid points, there is nothing to measure, it's just plain white.
-- Markus
Editor
opticallimits.com
Posts: 6,715
Threads: 236
Joined: Apr 2010
Reputation:
22
[quote name='mst' timestamp='1304672003' post='8094']
You can always draw some conclusions if the lens is tested on both formats (DX review will follow soon).
Rule of thirds resolution figures will not happen. First of all because of time constraints. And then there's also a quite pragmatic reason: our resolutions charts are customized and optimized for our measurements. At the rule of thirds grid points, there is nothing to measure, it's just plain white.
-- Markus
[/quote]
And that is why the samples you post at times are valuable (like in this case, of the 50mm Sigma).
Posts: 385
Threads: 14
Joined: Apr 2010
Reputation:
1
[quote name='youpii' timestamp='1304612990' post='8083']
Thanks.
Maybe you should also measure sharpness at the mid-center (basically at the rule of thirds) when there's a gap between center & border sharpness.
[/quote]
Second that motion. There is a blind-spot in the objective test procedure.
Posts: 8,010
Threads: 1,853
Joined: Apr 2010
Reputation:
49
[quote name='dave9t5' timestamp='1304711574' post='8103']
Second that motion. There is a blind-spot in the objective test procedure.
[/quote]
More test spots ... less reviews ...