First Impressions: Metabones Speed Booster
|The Metabones Speed Booster is shown here mounted on a Sony NEX-6 with a Canon EF 40mm F2.8 lens.|
We've had the Metabones Speed Booster in the office for a few days and wanted to share some initial thoughts on the possibilities and challenges presented by such a unique product. While Metabones has announced plans to support a variety of lens/mount combinations, its first offering is a Canon EF to Sony NEX Speed Booster, so we put some Canon full frame lenses on our Sony NEX-6 and did some comparisons alongside the Canon EOS 6D we're currently reviewing.
What is the Speed Booster?
As we reported recently, the Speed Booster is an adapter that allows for a full frame lens to be mounted on APS-C or Micro Four Thirds mirrorless cameras. There are two reasons why this is very big news, and both revolve around the optics inside the adapter. Firstly, Metabones claims that by reducing the image magnification, the Speed Booster gives an overall focal length multiplier of around 1.09x, promising a nearly full frame field of view on APS-C. The second claim, as its name implies, is that the Speed Booster increases the lens's speed by a stop, turning an F2.8 full frame lens into an F2.0 optic that allows for low light shooting at a lower ISO sensitivity.
How can an adapter change the focal length and aperture of a lens?
The Speed Booster is most easily understood as being the opposite of a teleconverter – in that it reduces your focal length. This, then, has a domino effect on the F-number of the lens. Because the F-number is the focal length divided by the width of the entrance pupil, reducing the focal length without changing the entrance pupil increases the F-number. So, adding the speed booster to a 50mm F1.4 lens turns it into a 35.5mm F1.0. But, since we’re then going to mount it on an APS-C camera, we have to multiply these numbers by the camera's 'crop factor' if we want to compare this lens-plus-adapter combination to a full frame camera. The end result is a 53.25mm equivalent lens that gives the same depth-of-field as an F1.5 lens would on a full frame camera.
|The Metabones Speed Booster contains optics, designed by highly regarded lens designer Brian Caldwell, that match the image circle of a full frame lens to a smaller format sensor.|
And, if making lenses faster sounds like witchcraft, don’t forget that when you normally mount a full-frame lens on an APS-C camera, it illuminates an area larger than the sensor. With the Speed Booster, rather than losing this light, you’re effectively condensing it all onto your sensor. Again, it's doing the exact opposite of what a teleconverter - which loses light - does. The end result is that in terms of light capture and depth-of-field - as well as field of view - your full-frame lens behaves essentially as it would on a full-frame body.
The Canon EF to Sony NEX version of the Speed Booster supports autofocus. But before you get your hopes up, here's a quote from Metabones' own product page:
'Autofocus speed is very slow and inadequate for most moving subjects. The autofocus speed is unfit for professional use for sure, and it would disappoint most enthusiasts.'
Our experience confirms this. Essentially you need to think of the Speed Booster as a manual-focus adapter. Even in good light, you'll get much faster results by focusing manually, particularly on a camera like the Sony NEX-6, which offers focus peaking. This is not surprising if you think about it, as the adapter is mating a lens optimized for phase-detection autofocus (PDAF) to a camera that uses a contrast-detection autofocus (CDAF) system.
It's also worth noting that the Canon EF to Sony NEX mount model is currently the only Speed Booster version to support AF at all. And the choice of AF-compatible lenses is limited to Canon EF models introduced in 2006 or later. Older (and third party) lenses operate in manual focus only.
As you can see from the examples below, Metabones' claim of 'almost full-frame coverage' is no idle boast. We've shot our resolution chart using the same lens on both the NEX-6 with Speed Booster adapter, and the EOS 6D; each camera was shot with its sensor plane located the same distance from the resolution chart. Based on our image comparisons, we calculated a focal length multiplier of about 1.07x, which is virtually identical to the company's assertions.
|Sony NEX-6 + Canon EF 50mm F1.4 USM with Speed Booster: 1/50 sec. F8.0, ISO 100||Canon EOS 6D + EF 50mm F1.4 USM:
1/50 sec. F8.0, ISO 100
The Speed Booster combines with the Sony NEX-6 to provide a slightly narrower field of view compared to the full-frame Canon EOS 6D. The raw files were processed in ACR 7.3 beta to match white balance and brightness, with all other settings, including sharpening at their defaults.
As the 100% crops below attest, the center sharpness performance of the Speed Booster is very impressive. You'd be hard-pressed to find significant difference between the two files in the center of the image, other than the slightly higher resolution of the 20MP Canon EOS 6D versus the 16MP NEX-6. In the second set of crops, however, taken from the lower left corner of the chart, you can see softer corner performance and more prominent chromatic aberration (CA) from the Speed Booster.
|Sony NEX-6 + Canon EF 50mm F1.4 USM with Speed Booster. 100% crop||Canon EOS 6D + EF 50mm F1.4 USM. 100% crop|
|100% crop||100% crop|
With such impressive results, we thought it would be interesting to see how optical performance fares at a very wide aperture. The image below was shot with the same Canon EF 50mm F1.4 USM lens on the Sony NEX-6, and the Canon EOS 6D, but this time at an aperture of F1.4. Raw files were again, processed with ACR 7.3 beta.
|Sony NEX-6 + Canon EF 50mm F1.4 USM with Speed Booster: 1/40 sec. F1.4, ISO 100|
In the image above, you can see more obvious lens distortion at F1.4, as you'd expect. (Click the image for a full resolution version.) In the 100% crops below you can see that image quality suffers somewhat in the corner of the frame compared to the full frame EOS 6D.
|Sony NEX-6 + Canon EF 50mm F1.4 USM with Speed Booster. 100% crop||Canon EOS 6D + EF 50mm F1.4 USM. 100% crop (Click for full resolution image).|
|Top right corner: 100% crop.||Top right corner: 100% crop.|
We've also found that the Speed Booster introduces significantly more vignetting in image corners with the Canon EF 85mm F1.8 lens than you find on the full frame Canon EOS 6D, presumably due to a mismatch between the size of the lens's rear element and the Speed Booster’s front element.
|Sony NEX-6 + Canon EF 85mm F1.8 USM with Speed Booster: F8, ISO 100||Canon EOS 6D + EF 85mm F1.8 USM:
F8, ISO 100
Still, you can't help but be impressed with the sharpness performance of the Speed Booster at an optimum aperture. And while vignetting, CA and large aperture corner performance are concerns, we must stress that 100% crops of test charts can overstate the practical implications of how the lens performs out in the real world. So we took the Speed Booster out of the testing studio for some shooting.
|Sony NEX-6 + Canon EF 85mm F1.8 USM with Speed Booster: 1/160 sec. F4, ISO 1600. ACR 7.3 beta raw conversion.|
There's no denying the pleasure of putting your SLR lens on a mirrorless body without significantly limiting your field of view, particularly with wide angle primes. And the ability of the Speed Booster to allow more light to reach the sensor should mean that you can capture equivalent exposures while shooting at an ISO setting that is one stop lower than using the same lens without the adapter.
|Sony NEX-6 + Canon EF 40mm F2.8 STM with Speed Booster: 1/30 sec. F2.0, ISO 3200||Sony NEX-6 + Canon EF 40mm F2.8 STM with Speed Booster: 1/60 sec. F2.0, ISO 1000|
In the low-light scenarios above, using the Speed Booster allows the Canon EF 40mm F2.8 STM lens to behave as an F2.0 optic, thus letting you shoot handheld at ISO 3200 and ISO 1000, respectively.
We did find, however, that shooting at the widest possible aperture (so F1.0 on a nominal F1.4 lens, for example) consistently resulted in a darker exposure. We found this to be true whether we allowed the Sony NEX-6 to meter the scene or set exposure manually. As you can see in the comparison below, maintaining an equivalent exposure between F1.4 and F1.0 actually results in a darker scene.
|Sony NEX-6 + Canon EF 35mm F1.4 L with Speed Booster: Av mode, 1/500 sec. F1.4, ISO 100||Sony NEX-6 + Canon EF 35mm F1.4 L with Speed Booster: Av mode, 1/1000 sec. F1.0, ISO 100|
The most likely explanation is not due to any flaw of the Speed Booster. Rather, with the phenomena known as pixel vignetting, in which at very wide apertures, the light gathering ability of the sensor's photosites diminishes slightly. If you've never heard of this, it's because characteristics like these are often corrected automatically in-camera based on the information the camera has about the lens mounted on it. And of course, this chain of information is affected with the use of the adapter. We want to stress, however, that on all of the EF lenses we've used with the Speed Booster it is only at the maximum aperture (enabled by the Speed Booster) that we've seen this occur.
There's a lot to be excited about with a product that so transforms lens behavior and addresses such a long-standing issue for photographers who long for the wide-angle and light-gathering performance of their older full frame lenses on their newer APS-C body. Image quality is very impressive, with results that compare exceptionally well against a full frame sensor. You do pay a small price in terms of corner softness and CA.
The biggest remaining challenge, however, with the Canon EF to Sony NEX mount is AF speed, a direct consequence of pairing a PDAF-optimized lens on a CDAF sensor. And to their credit, Metabones makes it very clear that AF performance is going to lag far behind what you'd get by using the Sony NEX's conventional AF system.
This does effectively make the Speed Booster an even more niche product, as it's best suited for manual-focus shooters. We'd be keen to see a camera manufacturer license this now-proven technology and offer this optical quality and near-full-frame performance mated with an effective AF system. But for now, we tip our hats to Metabones for releasing a product that accomplishes what many had thought nearly impossible.
Jan 20, 2016
Jan 18, 2016
Jan 15, 2016
Jan 15, 2016
- Fujifilm X-T223.6%
- Nikon D50025.4%
- Nikon AF-S 105mm F1.4E8.2%
- Olympus M.Zuiko 12-100mm F47.5%
- Panasonic Lumix DMC-G857.2%
- Sigma 85mm F1.4 Art6.7%
- Sigma 50-100mm F1.8 Art5.1%
- Sony a63006.4%
- Sony Cyber-shot RX10 III3.7%
- Sony Cyber-shot RX100 V6.3%