RF 14mm f/1.4L VCM

[neuroanatomist]

The RF 14mm f/1.4L VCM is now Canon's widest current L-series prime for the RF mount (supplanting the RF 20mm f/1.4L VCM), and is a significant engineering achievement with excellent optical quality (and a lot of glass) packed into small, light package.

For many people interested in this lens, the 20/1.4 and its slightly older fraternal twin, the RF 24mm f/1.4L VCM, will also be under consideration. All are very similar in IQ and size/weight, and the price is inversely related to the focal length.



Working on a 'first impressions' mini review that I will add over the next day or two, time and weather permitting (particularly for night skies, since that will be a primary use case for this lens).

 

[neuroanatomist]

As is obvious from the above image, the RF 14mm f/1.4L VCM is very similar to its VCM siblings, and also something of a slimmer version of the RF 10-20mm f/4L STM lens. The build quality of all of them is similar, though the 14/1.4 feels a bit more dense.

One of the useful features of a fast, ultrawide lens is the ability to blur out a background with a very wide field of view. Given the 40 cm of snow on the ground in my area, getting close to an interesting feature of the landscape without putting on my snowshoes required something contrived like this. The fieldstone wall and planter on it are about 4 m from the leaf, but already strongly blurred.


EOS R1, RF 14mm f/1.4L VCM, 1/1600 s, f/1.4, ISO 100 (in-camera JPG)

Case in point, there's a cemetery nearby that I walked through for initial test shots with the 24-105/2.8L Z and the 20/1.4L VCM, but this time it would have been into knee-deep snow so I stayed in the car (and was fortunate that the Subaru X-Mode pulled me through the deep mud). With a 14mm lens, the mud splatter on the snow and the glow of my headlights (this was after sunset) are visible in lower right corner of the frame, illustrating the challenges of ultrawide compositions.


EOS R1, RF 14mm f/1.4L VCM, 1/60 s, f/3.2, ISO 2500 (in-camera JPG)

When I tested the 20/1.4, I was surprised to find that wide open mechanical vignetting was a bit less on the 20/1.4 than on the 24/1.4. Turns out that's a trend, since it's even less with the 14/1.4. It's still present after stopping down to f/8, somewhat reduced but not as much improvement as with the 20mm lens. The images in the composite below are the upper left corners of RAWs exported from DxO PhotoLab 9 with no corrections applied.



Obviously, starscapes are going to be a primary use case for this lens. The full moon in the shot of the cemetery above hints that it's not the best time of the month to shoot such images. and my usual test subject (Ursa Major) was very close to that moon. Instead, here is a 100% crop of the extreme upper left corner of the image, composed with Orion's Belt right next to that corner in the camera's LCD.


EOS R1, RF 14mm f/1.4L VCM, 13 s, f/1.4, ISO 100 (in-camera JPG, 100% crop)

Coma is very well controlled in this lens, although there is some tangential astigmatism (with virtually no sagittal astigmatism). Here's just Orion's Belt at 400%:



I did take one shot with the full moon, and I see that the lens does exhibit some flare. I think this is the first time I've seen moon flare...

EOS R1, RF 14mm f/1.4L VCM, 20 s, f/1.4, ISO 100 (in-camera JPG)

One other bit of trivia about this lens. For those into lens cap photography, this is not the lens for you...but the light leak with the cap in place is rather interesting.


EOS R1, RF 14mm f/1.4L VCM, 1/4 s, f/1.4, ISO 51200 (NR in DxO)

 

[neuroanatomist]

Quick update – after comparing the stars in the initial test with shots posted by others, the astigmatism in my lens looked stronger than I think it should have. I compared the upper left corner (with Orion's Belt) to the upper right corner, and the astigmatism appeared worse in the upper left (that comparison also ruled out other possibilities, such as star trailing or bumping the tripod, since the 'stretching' of the stars was pointed toward the center of the image at both corners).



I was fortunate to find a second copy of the lens in stock at a small brick-and-mortar retailer with an online storefront (the same one at which I found an RF 10-20/4 when no one else had them in stock). I tested the two lenses head-to-head, and found the second lens performs slightly better. On ISO 12233-type charts, center performance was the same but the second copy (bottom) was slightly sharper in the mid-frame and the corners (these are 100% crops of the lower right corner, first copy on top); the second copy was a bit better in all four corners.



Last night I took a few star images, here are the extreme upper right corners (better corner of my first copy of the lens) of the uncorrected RAW images, this time the star in the corner is Alkaid (the star at the end of the handle of the Big Dipper). 100% (top) and 400% crops of the first (left) and second copies of the lens.



At the end of the day, in real-world use both copies of the lens are fine. But, since I have them both and I don't need two...the first copy will go back to B&H.