76 replies to this topic

[echeng]

There's a review of the new Canon EF 16-35mm f/2.8 L II review online:
http://www.the-digit...ens-Review.aspx

Many of us use Canon 16-35, 17-35, or 17-40 wide angle zoom lenses, and none of us are really happy with it. The cropped sensor folks can't get wide enough, and the full-frame guys (myself included) are unhappy with edge performance.

"Canon's engineers have completely redesigned the optics on the EF 16-35mm f/2.8L II USM to deliver higher contrast levels and improved resolving power." [Canon's Press Release] Of course, they are referring to a redesigning of the Canon EF 16-35mm f/2.8 L USM Lens - a very good and very popular lens itself. The biggest physical appearance change is in the II's wider lens barrel at the objective end - going from 77mm to 82mm in filter size. The 16-35 L II is .2" (5mm) wider (for the just-noted reason), .3" (8.6mm) longer and weighs the same as the original 16-35 L. They feel identical out to and including the focus ring. The II's zoom ring is more firm (I like this better). Minimum focus distance remains the same - Maximum magnification remains the same as well.

Barrel distortion is noticeably reduced at the wide end while pincushion distortion is slightly increased at 35mm. Flare is reduced over the entire focal length range - improving contrast in some comparisons. Overall, I consider the 16-35 L II sharper than the 16-35 L I, but this amount of difference varies throughout the focal length range and distance/direction from the center of the image. The II is at least as sharp or sharper in the center of the image at all focal lengths. The II is especially improved in f/2.8 non-center sharpness on the wide end. For the most part, I would consider the II an improvement in non-center sharpness overall though the 16-35 L I holds its own or even surpasses the II at certain focal lengths/apertures/points within the frame.

The II has slightly less vignetting than the I. CA is very slightly reduced - but looks different as it is primarily present in the corners which are now sharper at many focal lengths and points within the frame. Both lenses deliver exposures about 1/3+ stop brighter than usual.

What I'm looking for in a review is a comparison of the 16-35 L II vs the 16-35 L. Maybe I'll try to get ahold of one for a review...

[james]

Thanks for the heads up Eric. I want to see some compelling evidence too - before ditching my 17-40L. But I'd love to have a sharp wide zoom - especially after using the Nikon 17-35 on a FF camera.

Cheers
James

[Paul Kay]

I have now read a few reviews and a lot of comments about this lens. Can I add my own thoughts for it as an underwater lens?

But first .... many people seem to want to compare the 16~35/2.8L with the 17~40/4L lens. Why I am unsure. The smaller aperture lens is likely to be pretty comparable in performance once stopped down a few stops, simply because it is a smaller aperture and relatively recent design. BUT, it is a smaller aperture lens so cannot be used in the same way in low light conditions and, for us underwater photographers, provides a duller view (although I've owned and used a 17~40 I found it not to my liking and am working with fast fixed focal lengths but only starting at 24mm). Its also a lot cheaper though.

The new 16~35/2.8 MkII really needs to be compared with the MkI version and as has been indicated and might be expected it does seem to have improved performance over the MkI although it doesn't sound as if there is a dramatic difference.

However, my own thoughts are that for underwater use, the limiting factor at wider focal lengths than about 20mm and wider apertures (f/8 or wider) will be those imposed by using an ultrawide (weitwinkel) zoom behind a conventional concentric dome port design. So upgrading from a MkI to a MkII lens probably will result in only a minor improvement in central performance with very little if any effect on the edges where degredation is caused primarily by image curvature. Whilst above water photographers are seeking the holy grail of high quality ultrawide (weitwinkel) lenses for use in the digital age, we underwater photographers will have to use a combination of large diameter ports, appropriate diopters and accurate alignment/positioning of port and lens. ultra wide (weitwinkel) zooms remain a compromise and I am far from conviced that this lens will actually prove substantially better for underwater use than its predecessor. In essence I suppose that the problem we face is one of underwater optics as opposed to lens optics and a slight improvement in lens performance will probably be masked by other more substantial factors.

One other comment James. I used a 17~35 on Nikon film cameras and was never that impressed with that lens either. It would be possibly to fit it onto a Canon (adapters are available) but I'd be very surprised if its performance was substantially better underwater than Canon's ultrawide (weitwinkel) zooms. If anyone has tried this I'd be interested in seeing the results though.

[bmyates]

...So upgrading from a MkI to a MkII lens probably will result in only a minor improvement in central performance with very little if any effect on the edges where degredation is caused primarily by image curvature. . .I am far from conviced that this lens will actually prove substantially better for underwater use than its predecessor. In essence I suppose that the problem we face is one of underwater optics as opposed to lens optics and a slight improvement in lens performance will probably be masked by other more substantial factors...

Jeez, Paul, now I'm really starting to wonder if I should have ordered that new "II" version of this lens for $1,599 (from B&H - should arrive any day now) and be selling my old one for $850 (in the Wetpixel classifieds as we speak) after all! I've been quite happy with my old 16-35mm...I just assumed I would be "happier" with the new one (the moss is always greener...).

Now you make it sound like the only thing I may be is $750 poorer!

[echeng]

Well, if Bruce has one, at least I know that I might be able to check it out during our summer shark trip.

[caminu]

I took the 16-35 L II to komodo earlier this month and had several dives with this new lens.

Prior to getting the 16-35 L II, I was using the first generation of this lens. The physical length of the II is about 8mm longer than the 1. As I only got hold of the II just days before my trip, I didn't have time to optimize the port setup for this lens. I could only using the same port setup as the old 16-35 lens (Seacam SD + 35mm port extension).

Here are some of the quick observations from the photos taken by the II:
1/ vignetting occured at 16mm end
2/ corner sharpness didn't seem to be any better than the my older photos taken by the I

I wonder if I use a longer port extension, the result will be better. And I will show you some of the photos later.

On the top side, the edge performance of the II is better than the 1 though.

Edited by caminu, 25 April 2007 - 01:31 AM.

[caminu]

@21mm f/8


@16mm f/8


@16mm f/9


@23mm f/7.1

[james]

Wow, thanks for the samples Caminu!@!! It looks like this lens will need some pool testing w/ the various extension rings to get optimal results. But the CA, color and contrast looks very good on this lens!

Cheers
James

[bmyates]

I agree - thanks for the samples, Caminu! Nice shots, without too much CA, but certainly some noticeable in the corners. As you say, the corners really doesn't look much better than the old version of the lens, but as you also say, hard to know how that will be affected by optimizing the ext. rings (which I assume Stephen F. will do as soon as he gets his hands on the lens). Paul may be right, though; this lens simply may not produce much better results underwater (although it certainly will topside) due to inherent limitations of shooting through a dome...

BTW, I notice that none of those shots you posted were with very large apertures; I'm sure CA would be considerably worse if you used apertures closer to wide open, as is the case with both the old 16-35mm and the 17-40mm. I guess in general, you really do need to use f-stops of 8 or higher with these wide zooms (unless the subject is small enough in the frame that you can crop away big chunks of the corners). As with the predecessor, it looks like most shots taken with this lens on full frame cameras (at least at the wide end of the zoom range) will still need to be cropped to eliminate the soft corners... :glare:

[james]

Hang on Bruce - I don't think it's that bad. The lens clearly wasn't used w/ the right port extension as Caminu said - you can see the edge of the port in one of the photos in the upper right. I bet w/ some pool testing we will see much better results (but I'm an optimist).

Cheers
James

[bmyates]

...I bet w/ some pool testing we will see much better results (but I'm an optimist).

Well, James, having experienced your "pool testing" expertise firsthand (including the patented Wiseman "Find the Sun in Your Viewfinder" test), I'm sure if anyone can figure it out, you can!

Attached Images

• 

[james]

Check out this picture from the review:



You can see that the new lens is longer than the older lenses. I have no idea where the entrance pupil is though.

This is probably why you can see some vignetting in Caminu's shots. I'm guessing that 5-10mm more extension is needed but that's just a guess.

Cheers
James

[bmyates]

...I'm guessing that 5-10mm more extension is needed but that's just a guess.

I'll bet that's right, in part because that would mean I would need to buy a yet another extension ring (the old 16-35mm used a 35mm ext. ring as I recall, and I own a 25, a 35 and a 50). Being able to use one of those (or two of them stacked) would be far too simple!

[StephenFrink]

I agree - thanks for the samples, Caminu! Nice shots, without too much CA, but certainly some noticeable in the corners. As you say, the corners really doesn't look much better than the old version of the lens, but as you also say, hard to know how that will be affected by optimizing the ext. rings (which I assume Stephen F. will do as soon as he gets his hands on the lens).

Bruce - As you've deduced, I'm eager to see how this lens performs as well. The lens is now on its way from Canon and if it arrives this week I'll have pool tests done by end of week. I'll test with Seacam, superdome, and the various extension rings (as well as a diopter or two). Stand by to stand by.

Oops ... guess I won't be testing the diopters at the moment. From the review Eric linked: "The large 82mm filter size is a first for Canon EF Lenses. High quality filters of this size are expensive and at this point, not shareable on other Canon lenses without a stepping ring." The old 16-35 and 17-40 use 77mm diopters & filters. Oh well, I wasn't too happy with the diopter on my 17-40 anyway, so I'll test the various ports and port extensions by trial and error to find the sweet spot.

[hoovermd]

I have no idea where the entrance pupil is though.

This issue is really starting to tick me off. Canon absolutely will NOT reveal the correct entrance pupil location for any of their lenses.
I had to make a jig to find the location on my 14mm Wa and I guess I'll be using that jig again.

We as UW photogs should band together and convince Canon that is it in their best interests to provide this info. what it means to me right now is that I cannot make my decision to purchase this lens because I don't know what extenson ring length I'll be needing.

Come on Canon... the entrance pupil location isn't the formula for Coca-Cola!!!

[Paul Kay]

The entrance pupil is likely to be 5~10mm further away from the camera body as James says. If you add a diopter (+2 for a Superdome I would suggest) then adding 10mm of extra extension tube would be a reasonable starting point. Not forgetting that the entrance pupil will shift slightly during zooming, will be affected by the use of a diopter and that extension tubes (from Seacam at least) are available in 5mm increments, then adding 10mm has to be a good start. However, as I stated in my original post, the corner problems are to do with image field curvature more than anything else and even absolutely accurate lens/dome alignment won't solve this purely optical, underwater problem. Using a large port such as Seacam's Superdome correctly positioned will give state-of-the-art performance - the problem we face is that underwater optics now have their current limitations shown up by high resolution sensors! As I also stated, the MkII 16~35/2.8 will probably have better centre resolution than its predecessor so there should be a (small) benefit. Perhaps of higher benefit to more underwater photographers will be the inevitably suppressed prices of used MkI lenses!!!

[Alex_Mustard]

Looking at those shots the addition of a dioptre should make a positive difference to corner sharpness. You could always try the old BSoUP technique of focusing infront of the subject.

Alex

[dhaas]

Amigos,

I love it The MustardMan weighs in with one quick solution ala' some simple, real world advice instead of endless measurbating......

Man, am I glad I'm a cheap photographer and don't have to worry about such dilemmas with my cropped sensor dSLR cameras..... Even if I HAD a full frame sensor dSLR I'd choose Option "B" below....

Fisheye lenses with inherently sharper corners will outperform any rectilinear zoom, unless you are prepared to spend countless hours testing extension lengths and diopter / no-diopter combinations. And even then, you won't get edge to edge sharpness at all focal length settings....

If you used a Fisheye lens to start with, maybe you'd have to investigate and become skilled using "de-fisheye" software. But it would eliminate a lot of this hand wringing and worrying, and the endless edge sharpness crappola'

But hey, what do I know?

dhaas

[james]

Thanks for your valuable input Dave :-P I use a wideangle lens because I don't want my model's head to be the size of a watermelon :-)

Cheers
James

[Alex_Mustard]

I call it the tadpole effect!

Alex