Following on from my tests last week with the pre-production GH5 and Metabones Speedbooster XL, I’ve done a few more, but this time with the popular Sigma 18-35mm f1.8 lens. The usual caveats apply – because the camera is pre-production things might change with the final version of the camera.
The earlier test shot with Kaiman Wong proved that a full-frame lens like the Xeen 50mm I used will work with the Speedbooster XL, but what about lenses designed for Super35 or APS-C sensors? Because these are not intended to cover a 1.3x crop in a lot of cases they will show black edges. Some are however known to have slightly larger imaging circles that could cover the 1.3x area – at least in 16×9 format.
The most obvious of these is the popular Sigma 18-35mm f1.8. Above is a quick test of the Metabones Speedbooster XL 0.64x and a Sigma 18-35mm lens working on the Panasonic GH5 in UHD mode. You can see that the lens and adapter appear to cover the sensor quite well. The test was not without issues though. The Speedbooster XL with any electronic Canon lens locked up the GH5 when mounted. The only way to make this test was to disconnect the lens slightly from the Speedbooster so that the two were not in contact. A Genus variable ND filter was used to control exposure. This was clearly a firmware issue and I’ve had unconfirmed reports that the latest version of the GH5 firmware does not lock up the camera. Again, more testing needs to be done but it looks promising and Sigma 18-35mm owners look like they are in luck.
To demonstrate the difference in crop between the GH5 with Speedbooster XL and a Super35 camera I shot the pictures below. They are from the same spot with the same Sigma 18-35mm lens with a Speedboosted GH5 and a Sony a6500 with Sigma MC-11 adapter. You can see that the Speedboosted GH5 image has a wider field of view. Of course I could attach a Speedbooster to the a6500 as well, but the point is to show the comparative size of the GH5/Speedbooster image.
The previous tests were done with the in-body stabilisation turned off. That’s because I was concerned that if the sensor shifted around during a shot that you may see the edges of it as black corners when using the Speedbooster XL. For the next test I turned this stabilisation back on and panned the camera quickly from left to right to see if I could get these black corners to show up.
To my surprise they did not seem evident but of course more testing needs to be done. Next time I will move the camera around a lot more and try different lens combinations. Again, it looks promising though, especially for run-and-gun shooters. Also impressive in that last test video was the rolling shutter, which seems incredibly well controlled. Being a small sensor it does have an advantage, but it very good none the less.
All footage was shot in 4K/25P and 150Mbps and recorded internally to SD card. The camera is a pre-production sample and image quality should not be taken as final. The camera did not have V-Log L installed so it was set to Cinelike D. ISO was 200 and the footage is straight out of the camera and ungraded.