Some time ago I was ruminating about wanting to rationalise my lenses now that I had bought a second hand 7D. Based on the high praise on many review sites, I bought the Tamron 17-50 f2.8 as a general walk-round lens, and hoped it would allow me to replace my Minolta 28 f2.8 and Minolta 50 f1.7 lenses with little appreciable loss of image quality – and to then sell them to part fund the Tamron. A few people here and in other places told me I’d need to keep the primes, while others said the Tamron should be a more than acceptable replacement.
Well, I finally got organised to do a simple comparison of the lenses.
Here is how I did the test. Tripod mounted, with a focus on a cityscape with plenty of detail, and in excellent late afternoon light. I expect focus was effectively infinity, although I left it to the autofocus to sort it out. After all, this is how I would typically do my shooting. All shots were at ISO 100.
Firstly I tested the Tamron at 28mm and then 50mm, at a full range of apertures starting at f2.8 running out to f22. I then repeated the process with the Minolta 28mm lens and then the Minolta 50mm lens.
To do the comparison, I put together test sheets with pairs of crops from each subsequent file for each focal length and aperture of both the Tamron and the relevant Minolta. The sheets are laid out so that the comparable Tamron and Minolta crops are placed directly side by side. Rather than draw conclusions from screen images, I then had the files printed at my local processor as 10 by 8s, and did the comparison from the print. No post processing was done on the individual files. I also had a print made of each test sheet with some moderate sharpening applied – this should mean that all crops had exactly the same sharpening applied.
My conclusions, from both the sharpened and non-sharpened sheets, are that at 50mm in particular, the Tamron is significantly softer and lower contrast than the Minolta until f8, which is a worse performance than I had anticipated. I was not expecting a performance as good as the primes, but I was expecting a little better than I actually get.
The shop manager agreed with my conclusions and generously sent the files on to the local Tamron supplier who responded by saying there is nothing wrong. He also suggested I should have tested on a calibrated camera using test targets – which kind of missed the point that I wanted a comparison on my equipment doing the sort of shooting that I do.
Anyway, I’ve put compressed versions of the tests here:http://www.railgrafx.id.au/Shaw/Sundries/Lenstest28mm.html http://www.railgrafx.id.au/Shaw/Sundries/Lenstest50mm.html
A guide to reading the comparisons is on the left of the page when you open it.
I’d be interested in your views – maybe this is all the lens is capable of and these results are what should be expected (and so perhaps my expectations were too high), or maybe despite what Tamron say I do have a dodgy copy.
The issue for me is less that the Tamron performs worse then the primes (which I expected) but whether the results should be regarded as acceptable at all.
It’s also possible I have a focus problem, maybe associated with the camera, but I tend to doubt that as the subject was effectively at infinity and at lower apertures depth of filed would have more than covered it.
It is also possible that my test procedure is completely invalid!
As it is though I am now having second thoughts about using the Tamron to replace my primes and may need to continue searching for a suitable walk-round lens.
Thanks in advance.