5 replies to this topic

[genotypewriter]

Hope you find what's said in the following article useful when looking at MTF charts/resolution tests:

http://www.flickr.co...ter/5207301499/

Feel free to share your thoughts.


GTW

PS: FYI, this is not meant to be a stab just at photozone's tests.

[Klaus]

Hope you find what's said in the following article useful when looking at MTF charts/resolution tests:

http://www.flickr.co...ter/5207301499/

Feel free to share your thoughts.


GTW

PS: FYI, this is not meant to be a stab just at photozone's tests.

When using a wide to ultra-wide for a landscape scene I don't really care about the center. Frankly the center is always sort of Okayish at least.
It would be possible to provide another figure for the "border of the center region" but I think we would face a diminishing return on investment here.

[PuxaVida]

Interesting perspective... But I didn't understand the drastic drop after 18mm.

Anyway, what I think is: the pixels beyond 18mm are too pixels like their brothers luckily staying in the 8-15mm range. If the corners are not sharp in a landscape shot, I can't just ignore them (as they are stying in the low pixel populated area) and focus my eyes in the midrange... But again, very interesting quantitative approach I must say...

Regards,

Serkan

[popo]

Interesting perspective... But I didn't understand the drastic drop after 18mm.

Imagine a circle expanding from the middle. Up to 12mm radius it is complete. Between 12mm and 18mm the top and bottom of the circle are outside the frame so don't count. After 18mm, you also lose the sides of the circle and are left with only the corners.

[genotypewriter]

When using a wide to ultra-wide for a landscape scene I don't really care about the center. Frankly the center is always sort of Okayish at least.

I mentioned this exact reasoning in the write up. It is quite sensible but the point I'm trying to make is that the critical 8.5-15.4mm region shouldn't be ignored... which is what most people do when doing tests and reading them. And haven't we all heard how some people bash artificial lens tests saying they don't mean much in the real world? Maybe they're right after all... most people aren't doing/reading the tests properly.

Interesting perspective... But I didn't understand the drastic drop after 18mm.

Good point... I updated the original article with more visualisation:
http://www.flickr.co...iter/5207301499

Anyway, what I think is: the pixels beyond 18mm are too pixels like their brothers luckily staying in the 8-15mm range. If the corners are not sharp in a landscape shot, I can't just ignore them (as they are stying in the low pixel populated area) and focus my eyes in the midrange... But again, very interesting quantitative approach I must say...

Glad you found it interesting, Serkan... yes, all pixels are important. The article is just trying to show how that middle radius segment is important because 50% of the pixels are in there... but most of us don't think about that part when reading MTFs or lens tests. For example, we should watch out for lenses showing this kind of behaviour

http://www.slrgear.c.../ff/tloader.htm

GTW

[PuxaVida]

Yes I understand that... But if you compare the drop after 12mm and after 18mm the delta is huge. Actually that's the point GWT is trying to make I suppose. But anyhow, what I think is: corners are important on a wide angle (regardless of the amount of pixels in there).

Serkan