Keith Green wrote:
> 300 f4 non IS
> Weighted MTF    @f4.0 = 0.84    @f8.0 = 0.84
> Average Weighted MTF = 0.84 Grade: 4.3
> 
> 300 f4 IS
> Weighted MTF    @f4.0 = 0.71    @f8.0 = 0.80
> Average Weighted MTF of 0.76 Grade 3.4
> 
> 100-400 IS
> Weighted MTF    100 mm @f4.5 = 0.78    @f8.0 = 0.80
>                               180 mm @f5.0 = 0.77    @f8.0 = 0.82
>                               300 mm @f5.0 = 0.75    @f8.0 = 0.79
>                               400 mm @f5.6 = 0.75    @f8.0 = 0.76
> Average Weighted MTF of 0.78 Grade 3.6
> 
> Can a 100-400 zoom really be better than a 300 prime?
> (That statement should get some reaction!)

Sure it can, though in this case the data don't necessarily say that - Note
that the wide-open 300 mm scores for the 100-400 are at f/5 - The 300/4L 
IS might clean up a bit between f/4 and f/5. The f/8 scores are about the 
same.

Also, it's useful to understand what "weighted MFT" means in the photodo 
scores - It's about 70% driven by center results if I recall correctly. My 
experience with the 100-400 (the one time I rented it) was that it's good 
in the center but degrades a bit towards the edges, so photodo's weighting 
would tend to give it higher marks than I might (the two 300s on the other 
hand perform very uniformly center-to-edge - They're not enjoying such a 
boost).

> One last set of figures from Photodo, for the 600 f4 IS.
> 
> Weighted MTF    f5.6 0.81    f8 0.82
> Average weighted MTF 0.81 Grade 4.0
> 
> So the old 300 f4 non IS was better than the latest 600, and the 100-400
> (at 180 mm) is the same at f8.0.

First off, the difference in the weighted scores between the 300/4L and 
600/4L IS is very small - 0.84 vs 0.81 at f/4, and 0.84 vs 0.82 at f/8.
This is probably down in the realm of measurement error and sample-to-
sample variation, so the only thing you can really safely say is that
they're in the same league.

That having been said, if the difference in scores were real then that would
still be believable. The 600/4L IS is expensive not because it's necessarily 
a better lens than its shorter brethren, but because:

a. There's a lot more glass in such a lens (the front group is 4X as large
   in area as that of the 300/4L, for example)

b. Long, fast lenses require extraordinary optics simply to avoid being 
   complete dogs. As an extreme example, my $300 50/1.4 easily rivals
   my $4K 300/2.8 IS at equivalent apertures. All that really says is that
   it's a lot easier to design and build a great normal lens than a great 
   long tele. Likewise 300 mm lenses are easier to design and build than
   600 mm lenses. This is why I'm not even going to address your patently
   bogus comparisons involving the 100-400 at 180 mm ;-).

Getting more concrete, I owned both the 300/4 non-IS and the 300/2.8L IS 
simultaneously for a while. The 300/4 was slightly softer at f/4 than the
300/2.8, but at f/5.6 and smaller it was not easy to tell the difference,
even under a microscope. I've rented the 600/4L IS, and thought it a tiny 
bit softer than the 300/2.8, so yes, I'd expect it to test in the same 
league as the 300/4L.

-- Patrick
*
****
*******
***********************************************************
*  For list instructions, including unsubscribe, see:
*    http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************

Reply via email to