Very close? I would think not. The resolution of the 70-200f4IS is quite poor (i.e. zero) at f2.8, f3.2 & f3.5. It does not generate any decent background blur unless the subject is a good distance from the background and the focal length is approaching 200mm. Virtually useless for this purpose near the 70mm end.
Simply not true. After extensive bokeh testing at different focal lengths and distances my observation is that while the f/2.8 provides slightly more background blur at equal distance/focal length, the difference is very small and would rarely, if ever, change the character of the shot enough to make or break the image. Either it's a good image or it's not. The difference in amount of blur between f/2.8 and f/4 at these focal lengths is quite small to my eye.
The f2.8LIS II has no such limitations. This apart, the f2.8 gives a brighter viewfinder and activates additional sensitivity in the AF system.
Wow, not true. The f/2.8 (like all lenses) has limitations. They are just one stop less. And this extra stop comes at the trade-off of slightly less depth of field (which is often precious at these focal lengths, particularly for close facial shots, shots at 200mm and any shot that is taken at close range). I often shoot my 70-200 f/2.8 IS II at f/4 when I want a little less blur on the portions of my subject that are different distances. I've never liked the look of eyes in focus, end of nose blurred. It can make peoples nose look larger than it is. Very unflattering. I also like to shoot tightly framed images of enthused dogs running toward (and looking at) the viewer. For maximum impact, it's important to get the entire face in focus so f2.8 is often not the best solution.
Brighter viewfinder? LOL! The difference is barely (if at all) detectable. This is true whether comparing f/2.8 to f/4 in full daylight or in the dimmest room imaginable. I've done the comparison on my 7D and my 40D, almost no difference. The reason why gets into the optical properties of the viewfinder systems on our DSLR's. Namely that the viewscreens are not ground glass but instead are primarily sensitive to light coming from the center of the image cone and have very little sensiitivity to light coming from the portion of the lens between f/2.8 and f/4. In the real world the f/2.8 has no advantage in terms of viewfinder brightness.
Additional sensitivity in the AF system? Only if you believe Canon's marketing hyperbole. Don't misunderstand, I have great respect for Canon but their marketing department knows how to portray a small technical truth as if it were some kind of important difference in the real world.
I've done extensive testing over three days exploring the difference in AF between these two lenses. They are both exceptional in both speed and accuracy of AF. In the real world you will not notice any advantage of one over the other (either in terms of speed or accuracy). In fact, the f/4 IS is actually MORE capable in speed AND accuracy than the older 70-200 f/2.8 IS (version I). The misconception comes from Canon's claim that f/2.8 lenses are required to activate the higher precision AF sensors typically found at the center AF point. What they do not tell you is that these higher precision sensors cannot function at low light levels or on subjects lacking clear contrast. And, when the high precision sensors can't get a measurement, the system reverts to the same regular precision sensors all lenses use. Under bright light conditions with good target contrast there is a tiny accuracy advantage of the center AF point when used with f/2.8 lenses but the difference requires extreme pixel peeping and many shots to notice the statistically slightly better performance. Any lack of accuracy of the f/4 lens is almost irrelevant (because of the slightly greater depth of field of f/4 compared to f/2.8). In other words, when the conditions are such that you are likely to have a lower AF hit rate, both lenses are using the same AF sensors. This includes all shots taken in Servo AF mode, even under good lighting conditions, because the higher precision AF point requires an extra step and Servo AF does not allow enough time for the use of the HP AF sensor.
I've published the results here on DPR of the testing that supports the above conclusions. These are the type of misrepresentations that can be put to rest when one actually bothers to spend time with both lenses in question rather than just repeating "common sense wisdom".
To me, if truth be told, the f4 IS is little more than a slow alternative to the 70-200f2.8L IS II for consideration only if the budget won't stretch.
Not true, having both lenses shows me I prefer the 70-200 f/4 IS 90% of the time (even though my copy of the 70-200 f/2.8 IS II is marginally sharper at f/4 than it's f/4 IS brother). This difference is so tiny it requires extreme pixel peeping to see. However, with both lenses wide open, the difference is reversed and more apparent, the f/4 IS is sharper. But that's not why I tend to prefer the f/4 IS. The f/4 IS just has a little better color and contrast (particularly when my subject is backlit) and, being smaller and lighter provides better handling and makes it easier to get the shot. I only reach for the f/2.8 IS II when I know I'll want to shoot at f/2.8 or I'll be making extensive use of the 2X TC to shoot out to 400mm.
My advice to the OP is to buy the f2.8L IS II. Costs more and weighs a little more but is a much better lens.
Not "much better". Much more expensive, yes. Able to shoot at f/2.8, yes. Much better, not really. The ONLY advantage of the f/2.8 is it's one stop advantage (which for some applications is enough to justify it's cost and weight). But it's not a BETTER lens, for me the "better lens" is the one I reach for 90% of the time.
--
Mike Mullen
