Lensrentals MTFs for telephoto lenses

[AlanF]

Lensrentals, whom I much admire, and Brandon, who is clearly very smart, have posted MTFs for telephoto lenses http://www.lensrentals.com/blog/2015/07/supertelephoto-mtf-curves They had to modify their equipment for the longest lenses, 300mm and 400mm and use a lash up. The sagittal and tangential values for the Nikon 400mm f/2.8 VR III E are very perturbing as they are significantly different from each other at the centre of the lens for 20-50 lp/mm. Brandon explains this by the discrepancy being caused by a combination of tilt and decentering, which hasn't averaged out over the five samples. I don't want to push this further on their site, and this thread might be a more suitable forum.

Lensfreaks have published data from the Hasselblad factory equipment on the Canon 300mm f/2.8 II that the MTFs for 20 lp/mm are close to flat (~0.92-0.88) as are those at 30 lp/mm (~0.82-0.75) http://www.lensfreaks.com/lens-reviews/canon/canon-ef-300mm-f28-l-is-ii-usm-review-a-nice-little-beast/ which disagree with those from lensrentals, which have serious downward curves for 20 and 30 lp/mm. OK, Hasselblad measured only one and lensrentals several more than five, but the discrepancy is huge. However, I have spent 50 years as a scientist staring at data from very expensive equipment that has been pushed to the limit, and I worry about discrepancies and surprising results - a paranoia that has kept me in good stead and kept my measurements relatively clean of error.

 

[RogerCicala]

I read their methodology and they don't say, but my guess would be they tested MTF at best focus for each point, where we test all points at best center focus. If we tested at best focus for each point our data would be very similar to theirs. Either method is valid, my personal feeling is comparing all points to the center point's best focus is more 'real world', but I realize not everyone feels that way.

We tested 12 of the Canon 300 f/2.8 IS II lenses. Each had a tiny bit of field tilt which makes the off axis points a tiny bit out of focus when 'best center focus' is used for all points. But the inter-measurement variation was about 2% so I'm really comfortable with this data.

Roger

 

[AlanF]

Thanks for the explanation Roger.