While the 85/1.4 Art is just about perfect, the wider 35/1.4 Art shows some weakness when trying to cover a 645 medium format sensor. The result is some vignetting at wide open, throughout the focusing range. When stopped down, a bit of hard vignetting appears at the extreme corners. Ultimately, less than 10% of the pixels will be cropped out, ensuring a 45+ MP final result. There’s also the option of using a 5:4 or 3:2 ratio crop. Compared to first party Canon lenses, this is acceptable to me.
Lens updated to latest firmware.
Autofocus works quickly.
Continuous AF is epileptic and useless.
Eye AF is inconsistent, often dropping back into face detect only.
I’ve tried the 24/1.4 and the vignette is even more pronounced to the point of being useless to me. It would be better to wait for the GF 23/4 to go on sale or pick up the Laowa 17/4 if I want an ultra-wide angle field of view. I haven’t been able to try out the Sigma 28/1.4, which I feel might be the lowest acceptable limit of vignetting on wide-angles able to cover the full 645 sensor. This is purely speculation as I’ve only been able to try the Zeiss Distagon 28/2 ZE and found it to be acceptable by my standards.
Where the Sigma 85/1.4 is a home run, the 35/1.4 is a bit of a mixed bag. For most people, it should be fine. For some, like me, it’s borderline acceptable. For those few, it will be unacceptable. In 35mm crop mode, it works just fine, but for any other mode, it will require a bit of cropping and correction. At wide open, there is a bit more smearing in the extreme corners that sharpen up by ƒ/4 to 5.6.
This lens isn’t ideal but in some ways it exemplifies that “medium format look,” with the shallow depth of field while pulling in a lot of the background. For some, the vignetting at wide apertures may even be desirable with its ability to frame the subject. While the vignetting is less desired for landscapes, it’s weirdly good for environmental portraiture. Altogether, the lens manages to give the output an “old timey” look. Warm it up with a sepia tone or faded, low contrast color and the output has true character… that’s why you’re adapting lenses onto the GFX in the first place, right?