12-16-2010, 06:25 AM
[quote name='genotypewriter' timestamp='1292470691' post='5035']
But apparently the way you choose your "extreme borders" more extreme than mine accounts for the photozone results. But as I explained about... what is the meaning of the [url="http://www.opticallimits.com/lens-test-faq"]photozone borders[/url] if they account only for a very small area of the image frame?
http://www.flickr.com/photos/genotypewriter/5207301499
![[Image: 5207301499_ed6d40dd74_m.jpg]](http://farm6.static.flickr.com/5282/5207301499_ed6d40dd74_m.jpg)
[/quote]
The point made there is of course perfectly valid, but I think many people are concerned that soft corners will ruin the apparent sharpness of an image overall. While this might be true when shooting brick walls and test charts, it is of course not so much of a problem in real world shooting, especially at wide apertures.
However, even while being aware of this fact, I can't help but admit that still a lens with soft corners puts me off a bit.
[quote name='genotypewriter' timestamp='1292470691' post='5035']
Honestly I can't see how a single number ("optical quality") can describe distortions at different focal lengths, vignetting at different apertures, absolute sharpness and its distribution at different apertures and focal lengths, LoCA and CA at different apertures and focal lengths, bokeh, etc. in a way that's fully consistent and comparable across lenses. So why give it at all?
[/quote]
As I wrote in my other post above, I think that any reduction of the properties of something as complex as a camera lens to only a few numbers can only inevitably be a lossy abstraction from the details. And yes, I don't see too much value in that "optical quality" rating -- it suggest that you can compare a superzoom with a fast prime, which you really can't, at least not with a single number.
Perhaps the rating should be transformed into an "optical quality for the primary purpose of the lens" or "relative quality to the competition". So a superzoom could receive five stars in the category of "flexible, bright-sunshine-only outdoor lenses for people that hate swapping lenses and should have bought a P&S in the first place" <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/wink.gif' class='bbc_emoticon' alt='
' />.
Seriously, for the purpose available-light UWA/WA (FF/APS-C) photography, the Canon is probably much better than the Zeiss.
So this is pretty much the "field quality" rating that recently appeared in some reviews. However, for many lenses, several "purposes" come into question, and the rating is often less dependent on the actual optical quality than on the general specs of the lens.
Of course we could have dedicated ratings for sharpness, optical defects (distortion, CA) etc. for those people that only look at the bottom of the last page of a PZ review, but this still leaves the question on how to judge zooms that rock at one end and and suck really hard at the other end of their range?
But apparently the way you choose your "extreme borders" more extreme than mine accounts for the photozone results. But as I explained about... what is the meaning of the [url="http://www.opticallimits.com/lens-test-faq"]photozone borders[/url] if they account only for a very small area of the image frame?
http://www.flickr.com/photos/genotypewriter/5207301499
[/quote]
The point made there is of course perfectly valid, but I think many people are concerned that soft corners will ruin the apparent sharpness of an image overall. While this might be true when shooting brick walls and test charts, it is of course not so much of a problem in real world shooting, especially at wide apertures.
However, even while being aware of this fact, I can't help but admit that still a lens with soft corners puts me off a bit.
[quote name='genotypewriter' timestamp='1292470691' post='5035']
Honestly I can't see how a single number ("optical quality") can describe distortions at different focal lengths, vignetting at different apertures, absolute sharpness and its distribution at different apertures and focal lengths, LoCA and CA at different apertures and focal lengths, bokeh, etc. in a way that's fully consistent and comparable across lenses. So why give it at all?
[/quote]
As I wrote in my other post above, I think that any reduction of the properties of something as complex as a camera lens to only a few numbers can only inevitably be a lossy abstraction from the details. And yes, I don't see too much value in that "optical quality" rating -- it suggest that you can compare a superzoom with a fast prime, which you really can't, at least not with a single number.
Perhaps the rating should be transformed into an "optical quality for the primary purpose of the lens" or "relative quality to the competition". So a superzoom could receive five stars in the category of "flexible, bright-sunshine-only outdoor lenses for people that hate swapping lenses and should have bought a P&S in the first place" <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/wink.gif' class='bbc_emoticon' alt='
' />. Seriously, for the purpose available-light UWA/WA (FF/APS-C) photography, the Canon is probably much better than the Zeiss.
So this is pretty much the "field quality" rating that recently appeared in some reviews. However, for many lenses, several "purposes" come into question, and the rating is often less dependent on the actual optical quality than on the general specs of the lens.
Of course we could have dedicated ratings for sharpness, optical defects (distortion, CA) etc. for those people that only look at the bottom of the last page of a PZ review, but this still leaves the question on how to judge zooms that rock at one end and and suck really hard at the other end of their range?