Speedbooster or other focal reducer for macrophotography

[dickb]

Has anyone used focal reducers for their macrophotography? In theory, lenses with a large image circle relative to the sensor size used could be used with a focal reducer. This would decrease the magnification and ideally increase the resolution. My idea was that this could be useful when using specialised lenses like line scan lenses that tend to be optimised for a specific magnification and sometimes have large image circles.

I did a quick very unscientific test on my Sony A7Riii with a Rayfact VL1.4x, first at around 1.1x and next with a Metabones Speedbooster IV 0.7x focal reducer, so roughly at 0.8x. The image quality of both is pretty good as you would expect from this lens (see Robert's review). The reduced photo has a lot of vignetting, it is designed for use with full frame lenses and aps-c sensors, so while it reduces the image 0.7 times, the image size is limited by the diameter of the optics used in the reducer. The result appears to be almost aps-c sized, but still with black extreme corners.

Focal reducers designed for medium format lenses and full frame cameras do exist (Kipon Baveyes) but are rather expensive.

So has anyone done something similar? I suppose DIY focal reducers could be made and astronomical ones exist.

[Lou Jost]

I have the Metabones Speedbooster for MFT. There is sometimes slight CA in the corners and edges. I don't rememeber now what lens I used for the test, this probably matters. Sharpness was excellent.

There are very high quality focal reducers for astronomy, including some made entirely of fluorite or fluorite glass (not sure which). I have a lot of medium format lenses and might give one of these a try sone day. I also want to try a real telescope with one of these astro focal reducers. But I have to see how my finances go as the year progresses.

[chris_ma]

I have some experiences with speedboosters (although not macro) and have read some of the papers about the optical properties.
here's an interesting one by the designer:
https://www.metabones.com/assets/a/stories/The%20Perfect%20Focal%20Reducer%20(Metabones%20Speed%20Booster%20ULTRA%20for%20M43)%20-%20Whitepaper.pdf
(note that they show the graph for a MFT size sensor with a 0.7x speedbooster, so the image circle mathematically should be about 15mm radius but they plot only 11mm.)

it sounds plausible to me that one can possibly increase MTF by shrinking down the image circle by an additional lens group, specially if the master lens is not perfect optically.

I suspect that it gets harder and harder if the master lens is corrected very well. maybe it's possible to get a bit more center resolution, but almost certainly other errors will creep in, like corner softness, field curvature, CA etc.

so most likely it's not the ideal way to get more resolution out of top of the line lenses. if it were, then lenses would be simply designed that way ;)

on the other hand, if you only need a small image circle and have a good lens with a large image center, it might be worth trying, specially if you're interested in the performance at 0.7x of the original magnification

[rjlittlefield]

it sounds plausible to me that one can possibly increase MTF by shrinking down the image circle by an additional lens group, specially if the master lens is not perfect optically.

Remember that a speedbooster works by shrinking the image circle while simultaneously broadening the ray cones by the same factor. In other words, it shrinks the entire image, including the diffraction blurs.

So, if a 0.7x speedbooster is perfect by itself, and is not specially matched to the master lens, then what we'd expect is that the MTF with the speedbooster would be the same at 28 lp/mm and 5 mm image height as the MTF without the speedbooster at 20 lp/mm and 7 mm image height.

Unfortunately for this discussion, the graphs in Caldwell's paper do not make it possible to do that comparison, because the MTF curves are shown only at 10, 20, and 40 lp/mm for both cases.

Doing the best I can to interpolate, it looks to me like the speedbooster is performing about as well as we'd expect based on the assumption that it's perfect.

Certainly that is very Very good, but it's not a case of getting something for nothing.

What the speedbooster is doing, as suggested by dickb's original post, is to shrink a larger image circle to fit a smaller sensor, without introducing any significant degradation in the scaled image quality. Resolution on subject would be the same in both cases.

--Rik

[Scarodactyl]

It's basically analogous to using a shorter focal length tube lens on an objective if everything is ideal, right? Seems like it could be very handy.
I wish there were a nice variable one for some non-vignetting zoom action but it sounds unlikely.[/url]

[chris_ma]

What the speedbooster is doing, as suggested by dickb's original post, is to shrink a larger image circle to fit a smaller sensor, without introducing any significant degradation in the scaled image quality. Resolution on subject would be the same in both cases.

That‘s a good point, so the main use for macro photography would be to use is to slightly demagnify a lens, i.e. if I had a lens which was best at 1.4x with large image circle and wanted a mag of 1.0x on a small sensor, I could probably get a better image with a speedbooster instead of just shortening extension as long as I‘m not worried about distortion or absolute best corner performance.