Now that the RF system has a crop sensor option with the newly-released R7, many people are curious whether speed booster converters can be exploited to fully take advantage of full frame lenses. The quick answer: Yup.
The primary disadvantage to a crop sensor camera is the smaller light-gathering capacity at the sensor. Speed boosters are EF-to-RF mount adapters that have optical elements in them to squeeze down the lens’s image circle to the crop sensor size. The benefit to doing that is that the squeeze intensifies the amount of light falling on any given pixel, boosting the effective aperture a full stop. All other things being equal (of course, they aren’t) a crop sensor with a speed booster should theoretically provide about the same light gathering capacity as a full frame camera on a per-pixel basis. This can render your f/2.8 lens into an f/2 lens
So we went and acquired three brands of speed booster and tested them to determine how they might be useful, and which ones worked best. There is one stand-out brand, though: Viltrox. It seems to work seamlessly with Canon firmware, even reporting the correct effective aperture. Its image quality is the best of those tested at the center of the frame, although the edges suffer relative to the Canon booster.
The Canon speed booster – really designed for adapting to an even-smaller video sensor format – is nerfed in software to prevent users from getting maximum aperture in cameras such as the R5. You can get another stop of light from the booster, but the software won’t let you set the aperture wider than the lense’s specced maximum. Its image quality appears to be better at the edges than the others, but is not quite as sharp in the center as the Viltrox model.
The Laowa speed booster we tested is marketed specifically for its 24mm f/14 probe lens. Attempting to mount it to most normal EF lenses did not work, as the mount tolerances were extremely tight. We did find that our much-loved 15mm f/4 macro Laowa lens took the speed booster quite nicely. And, lucky for us, we have that probe lens, and have been wanting to boost it ever since we bought it. F/14 is a bit squinchy. The image quality of the Laowa booster appeared to be roughly in between that of the Canon and Viltrox versions. Which is to say, it was the second best – behind Viltrox – for center sharpness and second best – behind Canon – for edge sharpness.
When comparing image quality among all of the speed boosters, there were some pretty interesting results. Testing speed boosters quickly devolves into a game of “Wait, can we do this with it?” So, to satisfy the inevitable forum questions before they are asked, here is a list of situations we tested, just because we could…
1) Can I use a speed booster on the R7 to exploit the large image circles of my supertelephoto lenses to be concentrated down to a higher aperture, offsetting some of the light gathering disadvantages of the crop sensor?
– Yes, yes you can. The reason I personally started fooling with speed boosters was to take an ancient EF 300mm f/2.8 and make it into an f/2 light bucket for the purpose of catching woodcock during evening calling. The little buggers start calling here right about when you need to push the ISO to 3200, even if you’re shooting at 1/8th of a second.
– A note on the Canon booster: it prevented the setting of the aperture below the maximum aperture of the lens itself. So boosting the 300 f/2.8 meant that I could shoot the lens at f/2.8 or above, and it would provide an extra stop of light. I could not shoot it at f/2, as I could the Viltrox booster. This nerfing appears to be accomplished in firmware designed for the purpose, as when the camera turns on with the Canon speed booster, a dialog box appears indicating that the booster is not supported. The Viltrox just works, even reporting the correct aperture that is wider than the lens’s maximum. The benefit is still there for the the Canon speed booster, just only at the lens’s normal aperture range.
2) Do the optical elements of the speed booster reduce image quality?
… and below is the equivalent using the Viltrox speed booster…
You can see the extra stop of light having an effect, but the very fine detail is slightly fudged. Note that this is at a 1:1 pixel ratio, so the loss of detail seen here is unlikely to be well seen with normal cropping.
– Is it worth it? Yes, if you are light-limited and not reach-limited in your shooting. Close-in woodcocks benefit mightily from speed boosting, but boosting your 600 f/4 to an f/2.8 to shoot small birds at a distance would likely be frustrating, unless you were specifically trying to address a low-light problem.
3) Just because we can, what happens if you stack a 1.4x teleconverter and then a speed booster to give you the normal focal length?
– Glad you asked, Beavis. As you’d expect, it reduces the image circle to the crop size, but also boosts the aperture within that circle, reverting it to the lens’s native aperture. So, essentially, you get the same effective focal length and aperture, but over a smaller area. The weird part: the image quality wasn’t a lot worse than the bare lens. The teleconverter (an EF 1.4x Mark III) did not appear to goof up the image quality noticeably, to our surprise.
4) Does speed boosting decrease diffraction at high apertures?
– The explanation for this may forever confound us but, yes, using the speed booster does lessen the negative effects of diffraction. This was easily shown by using the Laowa probe lens at f/40 with and without the speed booster, compensating for the different effective apertures with shutter speed.
5) What speed booster is most recommended? The Viltrox speed booster is superior in several ways. It works across all EF mount glass; it has the best center sharpness (more important for most applications that would be employed with a booster); and it does not suffer firmware nerfing. It costs about $270, versus the Laowa’s $250 and Canon’s $600.
So, how does this work? With the Viltrox booster, it’s pretty simple. After affixing the booster/adapter to the EF lens, you pop that on the camera, and then set your aperture to whatever you like. One foible we found was that if you are using a dial for aperture, it will initially not go wider than the lense’s specced maximum f/stop. If you use the touch screen, though, you’ll be able to set it a full stop wider. After doing this once, the dial will work to bring the aperture down even lower than the specced maximum aperture.
It is interesting that the Viltrox adapter is reporting through to the camera the information about the lens, yet providing it a wider set of aperture options. This works relatively seamlessly on, for instance, the R5.
The Canon speed booster is a bit more finicky. It will initially force a dialog box up on the screen indicating that you are using an unsupported piece of hardware, you rebel, you. After that, it passes along only the lens’s specced information, not giving the camera the option to set the aperture wider than normal. Of course the lens elements in the booster still have the effect of increasing the light by a stop. You just can’t take advantage of that to get wider than the normal maximum aperture. You can, though, take advantage of the added light to decrease your ISO at the lens’s normal apertures, or increase your shutter speed.
This is not as useful as the full Viltrox functionality. If the problem you are trying to solve is getting a faster shutter speed in a slightly light-limited situation, this might suit fine, but if you’re looking to keep out of 16,000 ISO territory, and only an extra stop of aperture is going to help, you’ll need a different booster. We recommend the Viltrox.