I was just taking a look around Canons Web site at the new IS telephoto lenses and while being impressed by their potential a few things really struck me funny. I noticed that all the new IS telephotos have 17 optical elements in 13 groups. Now thats one PILE of glass elements. I know that the IS lens group comprises a significant portion of these new elements, so they are necessary. For a comparison, the non-IS 600mm f/4 only has a 9 element construction, 8 less than the new one. My general concern is: how can a super-telephoto of this length deliver images as sharp and contrasty as its predecessor (which has an excellent reputation) with 8 more elements getting in the way. Now 17 elements doesnt seem totally ridiculous but adding the 1.4X TC you add another 5 elements to the mix for a whopping total of 22 pieces of glass (24 if using the 2X as Art Morris had mentioned doing). I was wondering if anyone had made a comparison between their non-IS lens and a newer IS version of the same lens and what have you found, aside from the obvious benefit of IS itself. Its just something about a lens with that many elements that has got me wondering. I also noticed that the tripod collars of especially the 600mm f/4 and 400mm f/2.8 seem to be placed way at the back of the lens, with the front element hanging way out in mid air. I read Art Morris preliminary review of the 600 on his site and he did mention it was front heavy, has anyone found this to be a problem? Thanks.

-- Rob Landry (landryj@nbnet.nb.ca), January 16, 2000

Answers

I'm in the process of comparing the 300/4L to the 300/4L IS (8 vs 13 elements if I remember right). I won't comment on sharpness yet since I don't have enough data, but under severe conditions (i.e. with the sun actually in the frame), I see more flare (lower contrast) with the IS lens. No big surprise. Whether increased flare and lower contrast would be noticable in normal use I wouldn't care to say yet.

-- Bob Atkins (bobatkins@hotmail.com), January 16, 2000.

I sold my normal 300 f4 L in favor of the IS model after I tested the two extensively and convinced myself that the IS lens was better during practical day to day use. Yes the non-IS lens is marginally sharper than the IS lens and lower in flare, but so what? The IS lens allowed be to bring back more usable slides than the non-IS lens. The IS made this possible, as well as the closer focusing of the IS lens. Do not mislead yourself by trying to split hairs based on test slides shot under circumstances other than your day to day use of the lens. IS is the way of the future - don't look back. All of the new super teles with IS is lighter and focuses closer (and faster if you believe Canon). The only serious limitation of the new IS lenses is their price.

-- Nico Smit (nico@anp.co.za), January 18, 2000.

Rob: Nico makes some very good points. I would add that while I've never owned the 300/4 for comparison, my 300/4 IS is one of my most cherished lenses. In fact, I sold my EOS 400/5.6L in favor of the 300/4 IS and an EOS 1.4 teleconvertor, and I feel I made a very wise decision whenever I compare the sharpness of my slides from the 400/5.6 to those from the "420"/5.6 IS, and I believe the IS feature of the latter is largely responsible for this. As you likely know, the drawback to the 420/5.6 IS compared to the 400/5.6 (using an EOS 1n or other non-EOS 3 body) is that the 420 is not a great bird flight lens, because the extender slows down target acquisition and autofocus speed more when used with the earlier Canon bodies (I have not had an opportunity to use the 420 IS with an EOS 3, but I have heard that performance is enhanced). However, I have captured a few great flight shots with the 420 IS, so this is certainly possible, but not nearly as likely when compared to the 400/5.6. FWIW, I too initially got caught up in the "theoretical" musings over the "number of elements" issue, and at least for my tastes, I (in retrospect) made a blunder by opting for the 400/5.6L over the 300/4 IS and a 1.4 teleconvertor--primarily because the former was less expensive than the latter option, but also because of the lens element issue. But as it turned out, I lost money when I sold the 400, and it would have cost me less (overall) to hang on and save for the 300 IS and the 1.4 TC. In summary, if there is in fact a slight decrease in optical sharpness of the 300/4 over the 300/4 IS (or the 400/5.6 vs. the "420"/5.6 IS, it is likely to be subtle indeed, and as Nico suggests, the advantages of IS seem to greatly overshadow any disadvantages, and in the "real world", the IS lens wins by a large margin, in my opinion.

-- kurt heintzelman (heintzelman.1@osu.edu), January 19, 2000.

I'll post the results of my tests on my web page when I finish them, but as has been said above, there's more to a lens than resolution testing. Under real world conditions off a tripod there's no dispute about the IS lens giving sharper images.

I'm a little more concerned about flare and ghosting. I've done a lot of sunsets with my 300/4L and even under REALLY adverse conditions with the bright sun in the frame the contrast is excellent and ghosting is almost undetectable most of the time. My impression is that the IS lens might not be so good under such conditions, but time (and testing) will tell.

-- Bob Atkins (bobatkins@hotmail.com), January 19, 2000.

Addendum: This last Summer, I had an opportunity to shoot blazing sunsets at the beach with the 300/4 IS (no extender, and with the B&W UV-H MC filter removed), with the following conditions: tripod mounted, IS shut off, MLU, and remote shutter release. In addition, I was at a decidedly oblique angle to the sun(set), and I metered well to the side of the sun, but bracketed (quickly!) in both 1/3rd and 1/2 stop increments. In summary, I saw no flaring or ghosting in the resulting images (Velvia rated at 50 ISO). I found this to be very encouraging with respect to this lens. I would also like to add that I do a good deal of railroad photography, and my experience here is potentially quite relevant because I frequently encounter scenes containing INTENSELY bright locomotive headlights (or other relatively bright light sources amidst a darker background, such as railroad signals at night), and I've found that the 300/4 IS often (but not always) renders these scenes without any discernable ghosting. (FWIW, I have also used the 300/4 IS on the same type of subjects, but "hand-held", with the IS engaged, and I still see the same results). While I certainly can't explain the optical physics of why this seems to be the case, I wonder if two factors might be operative here: First, in looking at the simplified diagrams of the 300/4 IS, I notice that the relative diameter of the IS gyro stabilized lens element(s?) is much narrower than the surrounding elements--could this serve to shunt most or "all" of any unwanted, reflected light beyond the critical cone of light that strikes the film? Secondly, it could be that Canon has found a way to cost-effectively and substantially improve the multi-coating of the most critical lens elements in such a way that ghosting and flaring is impressively suppressed or even eliminated in some circumstances (?). Obviously, I'm merely speculating here!

Finally, I would like to stick my neck out a bit here and comment on something I may, at times, be noticing when using the 300/4 IS (with IS engaged) when shooting, for instance, a bird that is in very close range RELATIVE TO "INFINITY". In such circumstances, I've noticed that the BACKGROUND can occasionally appear slightly "smeared" or "brush stroked". If what I'm occasionally perceiving here is actually "real", I've tended to explain it to myself as being due the IS system tending to stabilize a foreground main subject at the occasional expense of the background "bokeh". If my inferences here prove worthy enough to be transormed into "observations", I would still stand by my love and affection for the 300/4 IS, because I've never been unhappy with the "bokeh" of this lens, but in fact find it to be pleasingly "painterly"--a nice change from the more usual, diffused circles of confusion that one normally sees with even the finest "L" lenses. BTW, in my limited experience with EOS lenses (both "L" annd non-L lenses), I've encountered the worst ghosting and flare with the 28-70/2.8L, and the 100-300/4.5-5.6 ranks second.

-- kurt heintzelman (heintzelman.1@osu.edu), January 20, 2000.

FWIW - I use a 35-350 Canon which has (I think) 21 elements. That is a lot of glass. However it performs very well against the light, which I do often. These will not usually be 'bare sun' images but more likely lightly diffused sun with some haze or low-horizon mush. The lens performs very very well in these conditions - flare free, contrasty and quite impressive for a lens of this type. ANY lens will likely flare and ghost to some degree in 'bare sun' shooting, irrespective of the number of elements. I think the number of elements the lens contains is certainly an important factor, but has to be judged against the degree of internal reflection-control and the quality of the multi-coating and general optical design. <

-- John (John.MacPherson@btinternet.com), February 23, 2000.