08-26-2011, 04:28 PM
[quote name='Brightcolours' timestamp='1314370798' post='11054']
Still the actual CA and the "measured values" do not match. This is not concerning this review only. When you look at the CA, it is at least 3 pixels wide and that is not "the extreme corner".
To gnatsum, which Canon "L primes" are so horrible and yet get excuses?
Just to put things into perspective, this is a 24mm f4 lens, if you want to compare it to "Canon L" lenses.
[/quote]
Here's a CA spot that I measured at 12mm f/4 - border region:
![[Image: ca.jpg]](http://www.opticallimits.com/images/8Reviews/lenses/olympus_12_2/ca.jpg)
I don't know how Imatests calculates the CAs but 0.6px doesn't seem to be far off in this sample. It's difficult to tell visually from the tilted quadrant because the CAs will actually pollute the aliasing thus emphasizing the effect somewhat
Still the actual CA and the "measured values" do not match. This is not concerning this review only. When you look at the CA, it is at least 3 pixels wide and that is not "the extreme corner".
To gnatsum, which Canon "L primes" are so horrible and yet get excuses?
Just to put things into perspective, this is a 24mm f4 lens, if you want to compare it to "Canon L" lenses.
[/quote]
Here's a CA spot that I measured at 12mm f/4 - border region:
I don't know how Imatests calculates the CAs but 0.6px doesn't seem to be far off in this sample. It's difficult to tell visually from the tilted quadrant because the CAs will actually pollute the aliasing thus emphasizing the effect somewhat