Canon 50mm f/1.2
(Continued from first page)
© 2010 KenRockwell.com

Intro   Specs   Performance    Compared   Recommendations

Canon EF 50mm f/1.2L (72mm filters, 20.9 oz. / 592g). I'd get mine here, here or here. It helps me keep adding to this site when you get yours from these links, too.

< < Back to first page of review

COLOR RENDITION

back to Performance, to Introduction and Index or to top.

I see no differences from my other Canon lenses.

CONSTRUCTION QUALITY

back to Performance, to Introduction and Index or to top.

It has a rubber gasket on the lens mount to keep crud out of your camera.

Exterior: Plastic.

Filter Threads: Plastic.

Focus Ring: Ribbed rubber.

Markings: Silver paint.

Switches: Plastic.

Mount: Metal.

Internals: Seems like plastic and metal.

Noises when shaken: Plenty of klunking. This is normal.

Made in: Japan.

DISTORTION

back to Performance, to Introduction and Index or to top.

Distortion is typical, to a little bit better than normal, for an ultra-speed lens.

It's trivial to correct in Photoshop if it bothers you.

Wall of Shame. Roll mouse over to see correction in Photoshop CS2's lens distortion filter.

Plug +1.4 into Photoshop CS2's lens distortion filter to correct the distortion. It takes me hours to calculate this stuff and its all © and registered, so you'll need permission to use these figures for anything else. Thanks! Ken.

ERGONOMICS

back to Performance, to Introduction and Index or to top.

Canon 50mm f/1.2L on a black Canon Rebel XTi.

It handles well. Controls and manual focus are fine. It feels good.

Manual focus is reasonably firm and very damped as the other good USM lenses. You can't flick it around as you could with Nikon's AI-s manual focus lenses from the 1970s and 1980s.

It feels good and solid on a 5D. It's not nose-heavy as the gigantic Canon 85mm f/1.2L II feels to me.

Oddly it feels very good on an XTi, since you hold the lens and the body is only along for the ride. It's a solid little package.

FALLOFF (Darkened Corners)

back to Performance, to Introduction and Index or to top.

For portraits and many other subjects, falloff makes a better picture by concentrating the viewer's attention on the subject and away from the corners. I like the falloff in the f/1.2 gazebo shot above.

For photos of blank sky or brick walls it's bad, so who cares?

At f/1.2 on a full frame camera it's got a lot of falloff.

Here are shots of an Expodisc on my full-frame 5D and my 1.6x sensor Rebel XTi. There is less falloff on the XTi since the XTi only sees the center of the image. (see Crop Factor.)

Shooting flat fields and laying them on another flat field is a tough test which exaggerates even the slightest falloff. You'll never see it this bad in normal photography.

The minor exposure variations between frames was seen in real photography, which is good. The f/1.2 shots exposed the center correctly and let the sides get dark, which is perfect. The last thing you want is to be shooting at f/1.2 in the dim light and have the camera mistakenly average the exposure to bring up the corners, and overexpose the center. If anything, we want f/1.2 to underexpose to keep shutter speeds short.

USE ON FILM CAMERAS

back to Performance, to Introduction and Index or to top.

Perfect; the Canon 50mm f/1.2L is designed for them, too.

USE WITH FILTERS

back to Performance, to Introduction and Index or to top.

Canon EF 50mm f/1.2L

Easy! The 72mm thread is much bigger than the glass. Throw any filter or filters you want on it - it's not going to vignette.

FLARE AND GHOSTS

back to Performance, to Introduction and Index or to top.

It's great. You'll go blind trying to get it to ghost or flare.

You'll notice in the examples below that the sky is completely wasted - it's much brighter than it appears and this lens is doing a great job.

This Canon 50mm f/1.2L is is better than the 85mm f/1.2L II, which isn't bad either.

I see no difference with or without a Hoya HMC multicoated filter.

These examples are with a very bright sun - it isn't sunset yet!

Ghosts at f/1.2: a slight blob in the lower left.

Ghosts at f/5.6: a slight green ghost just inside of the the sun radially.

Use with Built-in Flash

back to Performance, to Introduction and Index or to top.

Built-in flashes are often so close to the camera that wide lenses can see far enough down to see their shadow cast by a built-in flash.

No problems with this lens, even at 1.5 feet (0.45m) with my XTi.

MACRO

back to Performance, to Introduction and Index or to top.

Close focus distance is typical for a fixed 50mm lens. No news here. This example is from my 1.6x Rebel XTi. It will be a much looser composition on a 1.3x or full-frame camera.

The optics are great at close focus distances.

At closest focus at f/2.8, full image from 1.6x Rebel XTi.

Unsharpened crop from above image at 100%, no extra sharpening.

If you print the entire image at the same 100% magnification as the crop above, you'll have a 38" (1m) wide print!

PROGRAM EXPOSURE MODE

back to Performance, to Introduction and Index or to top.

Every camera chooses its favorite combinations of apertures and shutter speeds. These may or may not be your favorites. Canon bugs me because even though every modern SLR lets you shift these preferences, Canons revert to their own preferences every time the meter turns itself off. Nikons stay where you put them unless you turn off the camera yourself.

My 5D and XTi choose:

f/1.4 @ 1/60

f/2 @ 1/125

f/2.8 @ 1/250

f/4 @ 1/500

f/5.6 @ 1/1,000 etc.

I hate this and always have to boink the knob to revert it to my preference of:

f/1.4 @ 1/8

f/2 @ 1/15

f/2.8 @ 1/30

f/4 @ 1/60

f/5.6 @ 1/125 etc.

No big deal, but I thought you'd like to know that these bodies default to very high shutter speeds with this lens.

SERIAL NUMBER

back to Performance, to Introduction and Index or to top.

The serial number is engraved in the lens mount and filled with black paint. You have to remove the lens from a body or remove the rear cap to see it.

SHARPNESS

back to Performance, to Introduction and Index or to top.

This lens is very sharp. This Canon 50mm f/1.2L is the sharpest high-speed 50mm I've ever used.

Who cares about sharpness?

This is how much isolation (shallow depth-of-field) you get at f/1.4 on a 5D, one of the main reasons people shoot fast lenses.

No real photographer buys an f/1.2 lens for shooting subjects that care about sharpness. You buy it to throw almost everything out of focus, or to shoot in crappy available darkness where you're happy to see anything at all.

These strong caveats aside, this Canon 50mm f/1.2L is so good that it's as sharp as it can get by about f/2.8.

To see how sharp it is you need to be shooting flat test targets, Walls of Shame or landscapes at infinity.

If you're shooting landscapes at infinity and want sharp, use the 50mm Macro (or 60mm Macro on the 1.6x cameras). If you're doing astronomy and want insane sharpness edge-to-edge at f/1.2, get the Canon 85mm f/1.2L II instead.

Focus is everything with an f/1.2 lens. This has been the gotcha with these lenses ever since they were invented, and modern digital science still hasn't solved it. Mechanics are the limitation. I remember 25 years ago when I had to take butter knives to my Nikons' mirror screws to dial in the focus offset so my shots at f/2 were in focus.

Most AF systems, including those on my personal 5D and XTi, have some offset. It will vary from sample-to-sample of the same lens.

An offset is a constant error. Mine tended to focus a little in front of the subject. To get the sharpest results at large apertures I had to point my AF sensor at something just a tad behind my subject (see the Color Fringes section for examples). Canon's new 1D Mk III addresses this with a user adjustment for this for the first time in history, so maybe digital science will be solving this. Hooray!

Offset is relatively invisible with real 3D subjects like people. I see it only when I deliberately photograph tests to look for it.

Unlike the common focus offset issues, a real problem is that I couldn't get 100% reliable autofocus on either my 5D or XTi. This means one out of every batch of shots wasn't in proper focus, wasting that one frame. Say goodbye to the Pulitzer if that was the one frame you needed.

My focus confirmation lights came on before releasing the shutter and my 5D reiterated which AF zones were locked-on in playback. I wish that this was me doing something stupid, but it probably isn't. This would drive me absolutely crazy, so I probably wouldn't use AF.

If you are silly enough to be shooting tests, sharpness is limited by the ability to focus accurately. Since the depth of field is zero at f/1.2, the only way you'll see the sharpness of this lens is by focusing very carefully, probably with a special manual focus screen. In actual use for the purposes intended for this lens, most of your subject will be out of focus anyway. Focus on the closer eye.

Diffraction is usually the sharpness-limiting factor for all lenses, especially digital SLRs, at about f/11 and above. See also How Sharpness Varies with Aperture.

Here are my observations, presuming unreasonably perfect focus each time. Letters correspond to center - side - corner, or center - corner. Possible ratings are E - VG - G - F - P. These are subjective. Don't fret over differences between adjacent grades. You're never going to see how sharp this lens is in most actual use, since subjects and realities of focus will obscure it.

* The last millimeter suffers a little.

Here's a comparison to the inferior combination of a Zeiss 50mm f/1.4 on my Nikon D200 (Sorry, I didn't have them both at the same time):

Crop from center of 100% 5D image with Canon 50mm f/1.2L at f/1.2.

Crop from center of Nikon D200 image with Zeiss ZF 50mm f/1.4 at f/1.4.

Sorry that I didn't have a Nikon F -> Canon EOS converter and my 5D back when I had the Zeiss in my hands. I shot the Zeiss lens on my D200, which has higher linear resolution than my 5D, so I reduced the D200 image to 75% to match magnification. The difference would have been more obvious if I enlarged the 5D image instead by 30%.

Most lenses, including the Zeiss, are common spherical designs. They have high resolving power wide open, but the spherical aberration gives rise to a veiling haze, which gets really bad at the corners. The Canon 50mm f/1.2L uses an aspherical element to eliminate most of the spherical aberration, which clears away the haze at large apertures.

All these differences go away stopped down. By f/2.8 to f/4 all 50mm lenses look the same on these cameras, if they're in perfect focus.

SUNSTARS

back to Performance, to Introduction and Index or to top.

Sunstars will be ordinary 8-pointed stars.

IS (Image Stabilization)

back to Performance, to Introduction and Index or to top.

see also Why IS is Important.

There is NO image stabilization, however I tested the slowest speeds at which I could hold it.

This is the percentage of sharp shots that I get at each speed, hand held without bracing any part of my body above my waist.

Percent perfectly sharp shots held free-hand:

This data tells me that 50% of the time I get a perfectly sharp shot at 1/20 of a second.

I wouldn't take any of the slight differences between cameras seriously. I was getting bored by the time I ran the XTi test. They are the same; I was looking to see if there was any significant difference between the two. There isn't.

By "sharp" I'm applying the highest standard. Sharp for the purposes of this test is perfect tripod-sharpness when viewed at 100% magnification, which is equivalent to a 40' (1m) wide print. We're all usually printing smaller, in which case slower speeds work even better. For instance, here's what I flunked for sharpness shot at 1/2 second on the XTi, which has higher liner resolution and magnification than my 5D:

Full XTi image, shot free-hand at 1/2 second, which looks perfect here.

Seen at 100% I consider this blurred.

The bad shots at faster speeds were even better. When I say I got 33% of my shots sharp at 1/15, that means one out of three were absolutely perfect. The other two weren't bad either. That means you can make a three-shot burst and usually pull one absolutely perfect one out of it.

Since marketing people use the old-wives-tale speed of 1/focal length as a minimum sharp speed, we can Lie with Statistics as they do and claim that, even without IS, I get a 1-1/3 stop improvement with this Canon 50mm f/1.2 lens.

Compared     top

Intro   Specs   Performance    Compared   Recommendations

RECOMMENDATIONS     top

Intro   Specs   Performance    Compared   Recommendations

This is the sharpest 50mm high-speed lens I've ever used, when it's in perfect focus.

I'm not all that hot on this 50mm f/1.2 lens because I didn't get proper autofocus results 100% of the time. I'm used to that on my Nikons.

If you can get over that, or you have better luck with autofocus or focus manually with a special screen, go get one!

If you're a newsman and want a fast 50mm, go get one today. This is a tough little lens which handles and focuses much better and faster than any 85mm f/1.2L II. Just be sure that you can get solid, consistent focus. I got more consistent results with the slower-focusing 85mm.

If you're an astronomer you'll prefer the insane flat-field and corner performance of the 85mm f/1.2L II over the merely very good performance of this 50mm f/1.2, and you won't mind the slow AF of the 85mm. The 85mm f/1.2L II autofocuses more accurately than this 50mm f/1.2L and is even better optically, although the 85mm is a pig to use and carry.

If you're a collector, or just want the best lens regardless of practical application, spend the extra $300 for the 85mm f/1.2L II which doesn't cost much more, but is in a completely different optical and mechanical class. The 85mm uses a ground aspheric element while the 50mm uses a molded one, and the huge 85mm uses a lot more metal in its construction.

If you're a regular photographer like me, personally I rarely use middle focal lengths and prefer the convenience of a zoom. I prefer wider lenses. I want to try the Canon 35mm f/1.4L and 24mm f/1.4L, which can be hand-held at even slower speeds because of their lower magnification.

At normal aprtures, Canon's 50mm f/1.4 or 50mm f/1.8 are as good, and the f/1.8 model is the sharpest, too. The extra $1,200 you spend on the f/1.2 version goes towards resistance to weather and better performance f/1.2; it's unlikely you'll ever see any performance difference in real photos.

For instance, this shot was made at f/11 with the f/1.2 lens, where all these lenses look alike:

Rancho Santa Fe, California.

The $310 f/1.4 has better MTF (sharpness) at f/8 than the f/1.2 version, and the f/1.4 offers the same excellent "just grab the ring" instant manual focus.

The $80 f/1.8 has better MTF than any of these three 50mm lenses at f/8, but requires you to jack a switch around to focus manually. For $80 you won't be disappointed, so long at the need to flip a switch between auto and manual is acceptable.

Help me help you         top

I support my growing family through this website, as crazy as it might seem.

If you find this as helpful as a book you might have had to buy or a workshop you may have had to take, feel free to help me continue helping everyone.

If you've gotten your gear through one of my links or helped otherwise, you're family. It's great people like you who allow me to keep adding to this site full-time. Thanks!

If you haven't helped yet, please do, and consider helping me with a gift of $5.00.

The biggest help is to use these links to Adorama, Amazon, B&H, Calumet, Ritz, J&R and when you get your goodies. It costs you nothing and is a huge help to me. eBay is always a gamble, but all the other places have the best prices and service, which is why I've used them since before this website existed. I recommend them all personally.

Thanks for reading!

Home   Donate   New   Search   Gallery   How-To   Books   Links   Workshops   About   Contact