16-35mm f/2.8 L - Image quality issues

Craig_Hartley · #8 05-22-2005, 10:12 AM

That's strange. My 16-35 doesn't show any fringing problems like this; your lens almost looks fogged. Have you looked closely at the internal lens elements under a bright light? Maybe some moisture or adhesive has fogged an internal element. I'd send it to Canon, or at least do a lens test by shooting newsprint tacked on the wall.

Tom_Wilk · #9 05-22-2005, 01:07 PM

My 16-35 has a rather small amount of fringing at the edges & corners at the widest range, 16-20 mm. But the effect isn't nearly as pronounced as your example seems to indicate. I do find that contrast and resolution are greatly increased at the wide end when stopped down quite a bit. F/11 seems to be a good spot at 16 mm if you're planning on large prints.

Center sharpness is good all across the zoom range at all apertures on my copy, even f/2.8, but the corners at the wide end just don't do so well at wider apertures. That's acceptable if you're shooting subjects where the periphery blends into the background, but you'll need to stop down for landscapes.

At 24 mm, it comes much closer to the performance of the 24-70, though it can't match the f/2.8 performance of that fine mid-range zoom. I think that the 17-40 might be a little better at the wider end of the range, but I didn't have both lenses together long enough for a thorough comparison.

Tom_Wilk · #10 05-22-2005, 01:09 PM

I have to add - the 20 mm f/2.8 performs similarly to the 16-35 at 20 mm, except that the consumer prime loses quite a bit of sharpness at f/2.8, across the entire image. Otherwise, they're similar with a slight edge in color & contrast going to the L zoom, but with less barrel distortion on the USM prime. Its not a bad lens if 20 is wide enough for you, and you don't plan on using f/2.8 very often.

KirkThompson · #11 05-22-2005, 11:03 PM

The 20mm f2.8 is one of the least successful primes that were tested for the Photodo site.

Tom_Wilk · #12 05-22-2005, 11:12 PM

Photodo gives the 20/2.8 a 3.4 rating. Given sample variation, I'm sure that the typical copy could be a bit higher or a little lower. Nevertheless, 3.4 is not a terrible lens. It's the same rating that the very good 300/4 achieves in their tests.

I won't go out on a limb and say that the 20/2.8 is a top-notch lens either, but its image quality is competitive with the zoom from f/4 up. Not so at f/2.8 where it stumbles a bit, at least in my experience. It has reasonably low distortion (as a prime should), controls flare quite well, and is worth consideration for landscape photography, at least on the 1D Mk II. Stopped down, It's pretty nice. Not sure on full-frame since I haven't tried it there.

AntonyHands · #13 05-25-2005, 08:16 AM

Hi all,

Thanks for the feedback - the lens is going to Canon tomorrow for service.

Best regards,

Antony

kblundy · #14 05-27-2005, 09:22 AM

Hi All,
I have had the 16-35mm for about a year - using it on a 1DmkII and I get fringing also.
I don't think it is the lens alone - but believe it to be backscatter reflections from the rear element of the lense and the sensor surface. On both a 10D and D60 its not evident.

I quick correction in photoshop fixes the problem.

Regards
Keith