marc aurel
Active member
Hi Marc -
It's very important to stress the sensor here. With the 24 TSE at least, there is a significant difference on FF MFDB when you go from f/8 to f/11, and another improvement from f/11 to f/16.
The diagonal of the MFDB is 14mm longer than that of the A7R, so it's perhaps not surprising that if you're "only" shifting 10mm on the A7R you're not getting into the zone where stopping down starts to make a big difference.
Looking at your test image, I'm assuming that you're shifting up 10mm in portrait orientation - that means the image circle required to cover that shift is actually just 61mm.
The TS-E image circles are a lot larger than that - they have to be a minimum of 67.2mm to cover the IQ180 used in my earlier linked example.
It's great to see these lenses getting used on a wide range of cameras and formats - they are incredibly versatile - but care needs to be taken not to interpret results of tests on one format as being relevant to another. I wouldn't dream of shooting these lenses at f/8 on FF MFDB unless I wasn't at all worried about the corners.
Kind regards,
Gerald.
Hi Gerald,
of course you are right - you can not simply take results from one sensor size to another. I didn't mean to do that. I was just trying to give an impression of what happens when you use the outer parts of the image circle and present the results in a way that can be easily consumed.
But the pixel size of the A7R is very close to that of the IQ250 that I referred to (4,9 vs. 5,3). Both have no AA-filter and are made by Sony. So resolution on pixel level should be similar if you use the same lens and aperture. The image that the lens projects is the same, you just use a larger part of it.
With that in mind - I admit you're right. My recommendation for f8 unshifted would be wrong for a MF sensor - because you use parts of the sensor that are further away from the centre. The corner of an unshifted IQ180 sensor is 33,6mm from the centre, the corner of an unshifted IQ 250 is 27,5mm from the centre - very close to where I took my crop from. And for that crop - 30mm from the centre - yes, f16 looks even better than f11. But the centre gets a bit worse already at f11 because of diffraction, and a bit more at f16. So there is a tradeoff. From what I have seen - the Contax 35mm for 645 holds sharpness better and does not need to be stopped down to f16 to do this. As much as I love the TS-Es - they are great and offer extreme wide angles, but they are not perfect for a MF sensor in my opinion.
I would be really interested to see a well done comparison between the Contax 35mm for 645 versus the HR Digaron 32mm with an IQ 250.
Best regards - Marc
Last edited: