Panasonic LUMIX G 14 mm f/2.5 review!

is this particular sample decentered? the left side looks horrible even stopped down, right side is ok though.
 
Summary - Pros:

-Very compact and well built
-Very high quality & sharpness in center
-Well controlled CA
-Good performance with into-the-light shots
-negligible astigmatism

Cons:

-Monstrous barrel distorition (Native, before correction in software - TN)
-Significant vignetting
-Quality and sharpness at the edges could be better
 
even slightly worse for corner sharpness.

Desining fast wide pancakes for m4/3 seems to be a real challenge, especially to get homogeneous results across the frame.

Hope the 12mm will do better that that.

--
Cheers,

Frederic
http://azurphoto.com/blog/
 
I could understand if the reviewer has a bias against software correction but if they are going to do a proper review why didn't they test with a camera with proper firmware? At the very least to see what the IQ would be with software correction? If the only cameras that have software correction is Panasonic cameras, then do some tests with a Panasonic camera. It's fine to test with a camera without software correction to see what an uncorrected image would be like but to give a conclusive result without testing with software correction is frankly a half baked review.

--
Hubert

My cameras: GF1, TZ3, recently fixed (I think) Minolta Hi-Matic 7s, broken Konica Auto S2(couldn't fix, who the heck GLUES screws in??), K1000 and my wife's old K110D



http://www.flickr.com/photos/peppermonkey/
 
Umm.. there is no indication that software correction wasn't accounted for in the conclusion, with the specific exception of the barrel distortion test. Which was done, I'm guessing, as a matter of curiosity since everyone who knows anything about mFT knows that software correction is an inherent part of the system.

There are lots of JPEG samples available that show the result after software correction for you to base your conclusions on.
I could understand if the reviewer has a bias against software correction but if they are going to do a proper review why didn't they test with a camera with proper firmware? At the very least to see what the IQ would be with software correction? If the only cameras that have software correction is Panasonic cameras, then do some tests with a Panasonic camera. It's fine to test with a camera without software correction to see what an uncorrected image would be like but to give a conclusive result without testing with software correction is frankly a half baked review.

--
Hubert

My cameras: GF1, TZ3, recently fixed (I think) Minolta Hi-Matic 7s, broken Konica Auto S2(couldn't fix, who the heck GLUES screws in??), K1000 and my wife's old K110D



http://www.flickr.com/photos/peppermonkey/
 
when is the 12mm schedule for? or is that just a wish? sorry im new to lumix m4/3 having just bought a GF1. I wonder if I should get the 7-14mm?
 
Seems I was wrong and that the cameras used did have software correction but it also seems they went out of their way to not use software correction (take a look at the thread that Jeanadriane listed below).

As for whether or not software correction wasn't accounted for in the conclusion, I think I lean towards them trying to dismiss it. Take a look at their conclusion, it seems more negative than positive. And they are complaining about; mainly:
  • monstrous distortion,
  • huge vignetting,
  • image quality on the edge of the frame could have been better.
other than the last point, the others are corrected via software. I.e. distortion corrected and in so doing, the area's with vignetting is mostly cropped out. This of course assumes that Andy Westlake is correct in his analysis of how Panny designs their lenses and how the reviewers at lenstip do their reviews. In that sense, their distortion test (and vignetting) weren't just a matter of curiosity but a major part of their evaluation. I don't particularly like software correction, since I much rather have optical correct lenses. But to complain about vignetting and distortion that is in general corrected via software, then make it the major reason for a mostly negative (albeit this could be a cause of lost in translation) conclusion is rather misrepresenting.
Umm.. there is no indication that software correction wasn't accounted for in the conclusion, with the specific exception of the barrel distortion test. Which was done, I'm guessing, as a matter of curiosity since everyone who knows anything about mFT knows that software correction is an inherent part of the system.

There are lots of JPEG samples available that show the result after software correction for you to base your conclusions on.
--
Hubert

My cameras: GF1, TZ3, recently fixed (I think) Minolta Hi-Matic 7s, broken Konica Auto S2(couldn't fix, who the heck GLUES screws in??), K1000 and my wife's old K110D



http://www.flickr.com/photos/peppermonkey/
 
If they gonna go out of their way to avoid software correction, they significantly undermine the credibility of their own test. One should simply dismiss their #1 complaint. On the other hand, vignetting is not corrected in software, so that's something one has to keep mind. Still, I'd wait for someone without an agenda doing the testing before jumping to any conclusion.

EDIT. On the other hand they justify their refusal to do distortion correction in the following way (google translate):
Of course we can take photos in the JPEG-ah, or correct the distortion in the computer. The problem is that if our field of vision is no longer 75.4 degree, but less. Rough estimate based on our test boards shows that even 5 degrees less. The lens is so then the equivalent of 28 mm full frame is not, but almost 31 mm.
In other words the field of view is not that of 28mm lens, but rather that of 31mm lens on FF. This is something many of us suspected all alone, but Andy Westlake claimed that this is not so. See the discussion here: http://forums.dpreview.com/forums/readflat.asp?forum=1041&message=36503587

If what this guys claim about angle of view is correct, it's a serious knock on Oly&Pany practices.
 
... On the other hand, vignetting is not corrected in software, so that's something one has to keep mind. Still, I'd wait for someone without an agenda doing the testing before jumping to any conclusion.
According to Andy Westlake the vignetting they claim to find is also mainly caused by their use of wrong software:
http://forums.dpreview.com/forums/read.asp?forum=1041&message=36943165
Correct me if I'm wrong, but aren't some of the things they're complaining about (vignetting and edge softness) likely to be better corrected on a Panasonic body than on the Olympus they used?
The key line for vignetting is this one:

'It should be mentioned here that if you save photos in JPEG format, the vignetting values are by several percent lower – it is due to aforementioned cropping of images after correcting the distortion.'

They forgot to add, though, that this comment applies equally to raw files which are converted using a program that supports the cameras correctly - i.e. the ones suppled with the cameras, or any of the major third-party converters (in fact you have to go out of your way to find one that doesn't apply corrections).

What they're complaining about seems to be that the extreme corners, which are designed to be cropped out of the final image after distortion correction, aren't fully illuminated. They're only seeing this by insisting on using dcraw and not applying the corrections that the vast majority of users will be using automatically, regardless of which brand body they use (Panasonic and Olympus necessarily apply distortion correction in the same way).

As for edge sharpness, to an extent this depends on where they are taking the measurement from. But the shots I've taken with the lens so far (including those in our GH2 14/2.5 preview samples gallery) suggest that it is genuinely a bit weaker towards the edges. Note software correction doesn't ever improve this.

--
Andy Westlake
dpreview.com
 
Whats next? Are they going to do image tests without demosaicing, white balance, hot pixel removal, etc.. because that's just software covering up for bad sensor design?

Software correction is an integral part of the u4/3 lens design. Any test without it is worthless.
 
yeah yeah .......ignore the software correction process and their is distortion

ever try removing the last element of an all optical lens?

you get the same problem- just there isn't a trick to do it without physically dismantling the lens.

wow....so it we don't use the whole lens system it doesn't work correctly......duh!

at least dpreview has sufficient technical understanding to not get such basic things wrong. when a reviewer cannot understand using software as a lens element, why would you trust any of their other comments
 
Seems good, thanks for the link,

Alan.
Merry crimbo.
 

Keyboard shortcuts

Back
Top