DxOMark results are...

[etherial]

Another "interesting" result I noticed last week (and just remembered to post) was their ratings of the Canon 70-200f2.8 IS II against the older version of the lens. The scoring system they use rates the older version better for all except one parameter, yet in their "user case scores" they give the new one more stars.

http://www.dxomark.com/index.php/Len.../(camera2)/436

Then they write an article comparing the two:
http://www.dxomark.com/index.php/New...nts-and-review

I wonder what Canon thinks of such results? Hard to imagine they would spend all that money improving a lens only for it to go backwards? And just about every other review I've read on the version 2 says it is an improvement over the (allready excellent) version 1.

[arthurking83]

If you read all(but one) of the reviews of the Nikon 70-200/2.8 VR lens comparions, everyone agrees that the VRII version is sharper, yet Photozone's 'numbers' clearly show that the VRII version is actually sharper at 200mm(in ultimate raw numbers).
Look deeper into this strange assessment tho and the difference is trivial(a single digit or two) .. but it is still technically sharper.
The newer VRII lens is better overall in all other aspects, other than raw values of an almost meaning test procedure.

It can be normal for a newer lens to be 'worse' than the predecessor lens it's supposedly replacing, but there are other gotcha's to be weary of in this situation too.

Sample variation!!

We all know it exists, and maybe DxO got a dud.. or didn't sample many lenses .... or if they did use many samples, they just happened to luck right out and get all duds!!

I remember that when PZ finally posted their results for the 70-200 VR(1st version) way back when, they made a note that explained that there was a delay in the posting of the results, and used two samples of the lens because the first results were quite average(for such a high end lens).
I'm sure they had a link to post the differences, but can't see that now.
The second sample provided figures, graphs and numerical values that look more like what a super expensive pro lens should produce!


DxO's scoring system for lens tests is ridiculous to say the least!!
Whether they're accurate or not is not really part of the problem as I see it,

That they think they can assess a diverse range of lenses into a lowest common denominator manner with a single figure score is simply stupid!

[Bennymiata]

I think that one thing that must be taken into consideration, DXO are French and anyone who has dealt with French companies will know what I mean.

[arthurking83]

..... DXO are French and anyone who has dealt with French companies will know what I mean.

C'est ... LOL!

"we are right, and we fart in you silly English knigget's general direction!"



But it also must be noted, that it was them that invented this photography thing too.
.... and not only did they invent it, their government of the day purchased the patents and declared the earliest photography processes patent free and for everyone to use as they wished.
I can't imagine that having happened on another (unnamed)continent on the other side of the world tho!

[etherial]

I think that one thing that must be taken into consideration, DXO are French and anyone who has dealt with French companies will know what I mean.

I work for a French company, and I know exactly what you mean!