Hi, I just compared my f/5.6 400 L to the new 100-400 L II.
Image quality difference that I perceived is more or less like displayed at "the digital picture". In short: In the center, the 100-400 L outperforms the 400 L by just a bit, in the corners, the 400 L is better and also, the distortion of the 100-400 L is more pronounced.
But that's nothing new. However, there is another striking difference: The reach at "400 mm". I have no means of testing the absolute focal length of both, but assuming that the focal length of the 400 L is exactly 400 mm, then the longest focal length of the 100-400 II is only 363 mm. If found that by measuring lengths in images of testcharts, taken with both lenses from the same spot. Distance to the test chart was 6 m.
I already knew, that the indications of focal lengths of zooms are often rounded up quite a bit. For instance, my 200 L II + 1.4 extender (= 280 mm) gives practically the same reach as the long end of my 70-300 L.
But in the case of the 100-400, I find the rounding a bit over the top. I didn't examine the wide end, but let's assume that the 100 mm are correct, the we are talking about a 100 - 363 mm lens.
And anyone wants to add reach with a 1.4 x extender must know that he doesn't get 560 mm (which could maybe still be described as "almost 600 mm") but in reality, he gets 509 mm, which of course is something very different.
Again, my calculations assume that the 400 L is 400 mm. If its real focal length were 440 mm, then the long end of the 100-400 would really be 400 mm, but I heavily doubt that.
Does anyone have measured the real focal lengths? Also, I know that the focal lengths sometimes change with distance. I could do the same test with targets at infinity ... but maybe someone of you has already done that?
When Canon states that a focal length is 400mm - whether that's a zoom or prime - it WILL BE 400mm at infinity.
In practice "infinity" may typically be as close as a couple of hundred metres but at 6m there will be a big focal length loss probably more so with zooms than primes i.e. some parabolic type relationship where image size changes more rapidly at close quarters
At app 200m I compared the image size of the 400/5.6 and the 100-400 and found they were almost
exactly the same i.e. @200m if the prime was actually 400mm the zoom was also the same.