Should I dump the Sigma 70-200 OS and get the Tamron 70-200?

Started Jun 17, 2013 | Discussions thread
RichyjV
Regular MemberPosts: 375
Like?
Re: Should I dump the Sigma 70-200 OS and get the Tamron 70-200?
In reply to Thorstein Korsvold, Jun 21, 2013

On the strength of one review, if you are happy with what you already have?? No!! When 15 review sites are in and broadly say the same thing then go for it. Just because DXO make lots of different numbers out of measurements doesn't make them more credible than other (good) reviewers. I have worked with plenty of statistics, you can really impact the results far too easily, I can't really trust any statistics to give the overall picture, and especially when DXO are so quick to post big headlines making broad conclusions about tiny differences in new numbers to draw in hits and therefore income - this is the opposite of real science which tends to reserve judgement to make sure they don't look like idiots when some new numbers come in (but DXO doesn't retest over time and bring in more accurate mean measurements, so just on sample variation they can make claims about lenses that are in practice untrue over the whole population of lenses). This is the same site that long delayed a 70-200 f2.8 review because the ones they were testing were not getting the results they expected, so they waited till they got a 'good' sample consistent with widely held experience. Think they apply the same method to every other lens? Not to say the Tamron isn't a brilliant lens and the new standard in quality - I just don't know and won't have the information available to make a judgement until a few more months have passed and more data come in from other reviewers.

Its quite easy to do a piece by piece critic of their numbers on lots of lenses and inconsistencies, however it is easy to criticize most reviewers, and indeed forum comments, and I think its a largely pointless exercise. Suffice to say that when lots of reviewers are saying the same thing its probably likely to be useful, no one reviewer (unless they are testing the average of 10 samples or so which is a practical nightmare and not economical at all) should be seen as the voice of 'truth', especially when tiny numbers altering and big words shouted about it on sites like this and others directly affects their revenue stream.

-- hide signature --

Regards
Rich

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow