Wonder if this lens will be eligible for Sigma's "mount conversion" service?
I'm surprised that so far nobody has said DxO's results are meaningless because they don't represent bokeh in their tests results anywhere! But at least the first post on this thread doesn't disappoint with the expected putting down of DxO.
When DxO get a Nikon mount copy of this lens, I think we'll see a much better representation of its capabilities. In at least one score, the "megapixel" thing, the scores are obviously limited to what Canon cameras can provide.
Compared to the 50/1.2L
|Name||Canon 50/1.2L||Sigma 50/1.4A|
Wow, that's about as cherry picked and biased a comparison as I've ever seen. Do you even try
to be objective?
Here is a more reasonable comparison:
However, this is highly skewed, because DXO uses their T-stops "measure" to determine what the "best" aperture is...and they chose f/1.2 on the 50mm as it's "best". That is about as close to the WORST aperture the 50/1.2 has...it gets far sharper and eliminates a ton of CA and vignetting when you stop down a bit. I would have chosen f/1.4 or f/1.8, both of which are definitely better than f/1.2 on the 50L, however in all their great and wonderful BIAS, DXO has conveniently not offered those as options.
When I choose f/2.8 for the Sigma, Otus, and 50L, the sharpness plots norm up quite nicely. The falloff in the corners of the 50L is due to the spherical aberration...the same spherical aberration that Canon EXPLICITLY LEFT IN BY DESIGN, for aesthetic purposes. Ironically, at f/2.8, the Sigma beats the Otus corner to corner...you can see a bit of falloff on the Otus at f/2.8 and f/4, where as the Sigma is sharp through and through:
Similarly, stop down the lenses a bit, and the vignetting issues clear right up as well. The 50L actually has better vignetting than the new Art 50 at f/2.8:
I would share the CA fields maps, however again, DXO, in all their biased wisdom, only seems to have produced CA data for the maximum aperture. CA DEFINITELY drops in the corners as you stop down, since the narrower aperture is blocking light from the periphery of the lens where the most CA occurs. As such, it is only possible to compare the wide open performance, where, once again, the 50L is at it's worst (although again, much of it's wide open performance is explicitly by design, in order to achieve a specific aesthetically pleasing result.)