Canon’s Ultra Rare Lenses – Using the Canon 50mm f/1.0L

A month or so ago, I was able to get my hands on the Canon 200mm f/1.8L Lens and test it for a review on It was an experiment of an article, as often, we try to discuss gear that we can stock and rent out to our customers. With the Canon 200mm f/1.8L being wholly discontinued and unserviceable, it’s not something we could rent out. However, the readers seemed to enjoy reading about this relic of a lens, so we decided to find another ‘Holy Grail’ of lenses, and I got my hands on the Canon 50mm f/1.0L.

Canon 50mm f/1.0L Review

This lens was graciously loaned to me by Julian Chen out of Santa Monica, CA.

History of the Canon 50mm f/1.0L

The Canon 50mm f/1.0L was introduced in 1989 and is the fastest AF lens available in EF mount, and one of the fastest lenses in the world. At 1018 grams, the Canon 50mm f/1.0L is also an incredibly well-constructed metal-bodied lens, and considered to be one of the best built Canon 50mm’s in the world. Sadly, however, because of its razor-thin depth of field and slow focusing (by today’s standards), the Canon 50mm f/1.0L was discontinued in 2000 and has been hard to find ever since.

The Canon 50mm f/1.0L works on all EF mount camera systems, and uses the Focus-by-Wire system found on the Canon 85mm f/1.2L II, meaning while accurate, the focus is slower when comparing it to more modern designs. Priced at ~$2,500 during the majority of its production run, the Canon 50mm f/1.0L was considered way too expensive for the average photographer, leading to its inevitable production end. However, if you have to have one, many can still be found on eBay for $3,800 – $4,500.

Canon 50mm f/1L Example Photo

Canon 50mm f/1L at f/1.2

Comparing the Canon 50mm f/1.0

When it came down to comparing this lens, I figured the most obvious comparison to make would be against the Canon 50mm f/1.2L, right? Well, wrong. In fact, the Canon 50mm f/1.2 and f/1.0 versions are entirely different by design, and the Canon 50mm f/1.0 is based more on the Canon 85mm f/1.2L I than anything else. So while I do not have a version 1 of the Canon 85mm f/1.2L, being that it was discontinued in 2006 and replaced with the Mark II lens of the same name – a lens I do in fact, have.

Canon 50mm f/1.0L Review

Comparing the Canon 50mm f/1.2L to the Canon 50mm f/1.0L

Canon 50mm f/1.0L Review

From a visual standpoint, the Canon 50mm f/1.0L looks nearly identical to the Canon 85mm f/1.2L II, with its front-heavy design, leaving you with a large front element, and asymmetrical design from front to back (as seen below). The images are very much similar from one to the next as well. I’ve put together a small table of these three lenses below to give you an idea of how it compares to the competition.

Canon 85mm f/1.2 Canon 50mm Comparison

LensPriceMin Focusing DistanceAperture RangeWeight
Canon 50mm f/1.2L$1,300.45mf/1.2 - f/16590g
Canon 50mm f/1.0$4,000 (Used).6mf/1.0 - f/161,017g
Canon 85mm f/1.2L II$1,850.95mf/1.2 - f/161315g

Build Quality

The Canon 50mm f/1.0L has a lot to it that makes it quite a bit different than the other options in the same focal length. As mentioned above, the Canon 50mm f/1.0L is based more on the Canon 85mm f/1.2L design than it is of it’s younger brother, the Canon 50mm f/1.2L. Because of it’s larger dense body, the Canon 50mm f/1.0L no doubt feels premium to the touch. If you’ve ever held the Canon 85mm f/1.2L, you’ll know what I’m talking about. But it’s dense body is well balanced, and feels good on the camera. The added cuff at the base of the body makes it easy to hold when mounting, and just further mimics the feeling of the Canon 85mm f/1.2L.

But with the larger body, also comes a few little-added things that many people might not know about the Canon 50mm f/1.0L. For one, people will often assume that the f/1.0 version has a 1/3rd stop over the Canon 50mm f/1.2L, but that’s actually false. In fact, it has 2/3rds of a stop, being able to implement both f/1.1 and the f/1.0 f-stops.

Canon 50mm f/1 Test Photo

Canon 50mm f/1L at f/1.0

A second surprise is the focusing system on the Canon 50mm f/1.0. Using the same focus-by-wire system as the Canon 85mm f/1.2L, the focusing system is slow by comparison. To help counteract this, Canon has added two focus distances on the focusing switch, to help speed along the process. giving you the option of focusing in two different ranges (0.6m – infinity and 1m – infinity), this option should both speed up the focusing of the lens, as well as provide better accuracy. That said, for the interest of my testing, I kept it in the 0.6m – infinity mode for the entire duration of my testing.

Canon 50mm f/1L Test Photo

Canon 50mm f/1L at f/1.0

Image Quality

I took plenty of photos with the Canon 50mm f/1.0L during my week with it, but I figured it was best to give a more scientific approach to its image quality. In short, it’s pretty mediocre. It vignettes a lot at f/1.0, and its sharpness is pretty lackluster, especially when compared to the competition. Below are some test photos, comparing the Canon 50mm f/1.0L to the Canon 50mm f/1.2L and Canon 85mm f/1.2L II. All of these were shot on a tripod, 55 inches from the stem of the lemon (conveniently pulled from my lemon tree), with the focus point being where the stem meets the body of the lemon.

Canon 50mm f/1L at f/1

Canon 50mm f/1L at f/1.1

Canon 50mm f/1L at f/1.2

Canon 50mm f/1.2L at f/1.2

Canon 85mm f/1.2L II at f/1.2

In addition to the vignetting, I also had a number of sharpness issues with the lens. First, this certainly has to do with it being a lens I was pretty actively shooting at f/1.0, giving you a razor-thin focus plane. But after showing some photos to Roger, he was also able to assess that the copy of the Canon 50mm f/1.0L I had, looked to be decentered. Not exactly a surprise, given the copy I had was 25+ years old and hadn’t been serviced in 20 years. At f/1.0, the depth of field is less than .8 of a centimeter, and when shooting handheld, it’s easier to just say that the Canon 50mm f/1.0 doesn’t really have a focus plane wide open, and many of the photos I took were slightly out of focus.

Canon 50mm f/1L Test Photo

Canon 50mm f/1L at f/1.4

When shooting at the widest apertures (f/1.0, f/1.1, and f/1.2) there seemed to be quite a bit more bokeh cutoff from the mirrorbox than I’ve seen before in lenses. This most often happens at lenses with wider apertures and transforms your circular bokeh into trapezoidal or semicircular bokeh balls. I believe this has to do with and is more apparent with the extremely wide aperture capabilities of f/1.0, but I don’t know the science behind it enough.

Canon 50mm f/1.0L Example Photo

Canon 50mm f/1L at f/1.4


So is the Canon 50mm f/1.0L worth seeking out and owning? Well, probably not. It’s sharpness and usability pales in comparison to the Canon 50mm f/1.2L, and is priced more for rich lens collectors than working photographers. However, leading up to this review, people have asked me what I thought of the Canon 50mm f/1.0, and I’ve been calling it ‘The best worst lens I’ve ever used”, because, well that is what it is. Are you going to get gloriously sharp images from it? No. Are you going to get a nonflaring workhorse? No. But are you going to get a giddy, excited feeling when spinning that dial and seeing f/1.1 and then f/1.0 on that top digital screen? Yeah, probably. So in short, the Canon 50mm f/1.0L is flawed, and it shows its age. But it still has elegance in its imperfections, and it still has a certain charm that sways people into paying $4,000 just to experience it.


Author: Zach Sutton

I’m Zach and I’m the editor and a frequent writer here at I’m also a commercial beauty photographer in Los Angeles, CA, and offer educational workshops on photography and lighting all over North America.

Posted in Equipment
  • I kind of disagree with the verdict. Portraits look phenomenal. The ultra-strong bokeh is genuinely one of a variety. It has this unique, blurry but very picturesque effect that cannot be easily mimicked with technically better lenses. Those are the photos that your fellow photographers will ask you about “how did you do this? What lens it was?”.

  • There is something to be said for the CLOSENESS that you get with a medium / normal focal length lens that can still produce this amount of background blur. In other words, sure you could match this level of blur with a super-telephoto lens, but it would look totally different. At 50mm, you totally feel the nearness of the subject.

    Personally, I’m not obsessed with bokeh enough to spend $4K on it, but at the same time I appreciate the beautiful look that it offers for flattering portraiture. I’d definitely consider buying an old manual focus ~50mm f/1.0, if one existed. I’m also definitely considering the 35mm f/1.2 that I think is available on Sony FE. Alas, I’ll have to settle for an old Nikon 50/55/58mm f/1.2 AI-S etc in stead. Still plenty of background blur for me, and you can’t beat the prices which can be as low as $400 if you get the right lens!

  • CheshireCat

    People buy this lens for its unique character (because, yes the lens is uniquely “flawed”).
    Apart from the shallower depth of field, and the discussed aberrations wide open, colors are rendered with more pastel tones, and the flare has a multi-rainbow signature that I wish you had shown in your examples.
    The lens shines on Sony mirrorless full-frame bodies, as there is no mirror-box to spoil the bokeh.
    There’s more to a photo than sharpness, yet the very few good copies around are more than sharp enough for any artistic usage (arguably sharper than the EF 50/1.2), and savvy post-processing can produce stunning results on real subjects.
    Is it worth as much as a Zeiss Otus ? That is very subjective.
    But stay away from this lens if you need to take photos of fast moving subjects or test charts 😉

    P.S. The generic “unusable depth of field” argument is nonsense. Some people don’t seem to understand that DoF varies with focus distance.

  • CheshireCat

    Correct. About 2/3 of a stop faster.
    It may not seem a big deal, but it is a challenge for optical engineers (especially 30 years ago).

  • Walter Lysenko

    So the actual ratio of f-numbers between the f/1.2 and f/1 lenses is 1.27. This means the f/1 lens is 2*log(1.27)/log(2)=0.69 stops faster. Very impressive.

  • CheshireCat

    It is supposed to be a true f/1, using the best technologies available back then (a super-complex design with 11 elements, 2 of which aspherical).
    As reported in other posts, the lens was a flagship item for Canon to show off their superiority. It was never meant to sell well. It seems that only about 4500 copies were ever made; fewer are still functional, and even fewer are still in pristine optical condition (and thus not often on the market).

  • CheshireCat

    Not really. The clipped bokeh is due to the mirror box.

  • CheshireCat

    There are 3 totally different versions of the Noctilux.
    The Noctilux f/1 is – subjectively – a much better lens, but you can’t use it on a reflex.

  • CheshireCat

    Don’t know about the M “mirrorless box” :-)… but certainly the bokeh is not clipped on the Sony A7.

    I would not use the 50/1 on anything other than full-frame. You would lose a lot of the overall character of the lens.

  • Walter Lysenko

    And what is the actual f-number for the f/1 lens?

  • CheshireCat

    The apparent DoF difference is because the comparison shots have been focused differently (and the decentered copy of the 50/1 did not help with this).

  • CheshireCat

    The Canon 50/1.2 and the 85/1.2 are actually f/1.27, but Canon marketing prefers the f/1.2 labels than the more proper f/1.3. Guess why ? 😉

  • CheshireCat

    Can’t adapt the Zeiss 50/0.7 even on a Sony mirrorless camera, certainly not with infinity focus.
    Besides, the lens does not cover full frame.

  • CheshireCat

    Lens is much better on A7. No bokeh cutoff.

  • Rob Crenshaw

    Yes, thanks, by the time I’d gotten to the end of that initial review it was just a tossoff number from an internal debate. I actually thought about that 2/3 stop comment before I hit Post, and wondered whether that was true, but was too lazy to look it up. I thought probably it was a full stop. But instead of looking it up, I snuck in the word “usable”, which gave me leeway to argue that the lens isn’t actually an f/1.0, that its measured light transmission would be less than a full stop. Practical versus theoretical. I then thought about changing it to a generic and unarguable “usable slight increase in speed over a 1.4”, but for reasons of laziness left it quantified because I didn’t realize my experience would generate such a long discussion. 🙂

  • Brandon Dube

    “Mechanical vignetting” is a term invented by the photography community. I would recommend not using it, since there is no distinction between that and just “vignetting.” Vignetting is, by definition, a mechanical process.

  • Peter Boorman

    I think I recall it was the British photo press that gave it that name at the time the lens came out, and it stuck. There are also versions of it made for broadcast TV, and it may be those that originally got the name (those broadcast versions sell for a bit less than the ones for the Canon 7s rangefinder because they can’t be converted to M-mount: I don’t know whether they can be converted for use on mirrorless, but if so they might be a cheaper option for use on Sony Alphas etc.)

  • Peter Boorman

    There are lenses where the lens itself clips the full beam from the wide open aperture: this the usual cause of the “cats’-eye bokeh” seen with some, especially, FSU, lenses and is known as mechanical vignetting. You also see it often using large format lenses wide open – but then they’re usually expected to be focused wide open and then shot closed down at least a couple of stops, so it’s not generally an issue in use. (That’s why some older LF lenses used to be described as “covers 4×5 from f5.6” or whatever, because wider than that you get mechanical vignetting from the body of the lens itself, which produces strong vignetting in the image as well as the odd shaped OoF highlights.)

    The ‘clipped bokeh balls’ seen with the Canon 50/1.0, though, seem to have straight lines along the edges, and right angles in the corners, so I’m pretty sure it’s the mirror box here, as you say, rather than the lens itself (unless the lens has a rectangular baffle at its back end, which I don’t know but think unlikely.)

    Assuming you’re right about the mirrorbox being the cause, and the 1D bodies having more space in there than the 5D series do, then the lens will vignette less on the 1D bodies as well.

    (Now I want to put one on a GFX to see how it behaves with no mechanical vignetting, and what the usable image circle is in that situation…)

  • Peter Boorman

    Never used this lens (and have little desire to) so I’m sure your experience is accurate – but someone ought to point out that f1.0 is a full stop faster than f1.4, not 2/3 of a stop, and f1.8 is a stop and two thirds slower than f1.0, not two stops.

    I completely agree with you that the 50/1.0 was all about Canon wanting to be seen to ‘trump’ Nikon, especially when they knew they had to address the bad PR impact of all the FD users who had had their investment rendered obsolete at a stroke by the introduction of the EF mount. History has shown just how successfully Canon did weather that potential storm, and thrive thereafter.

  • Jestaplo Photography

    I am thinking of buying this lens as I shoot a lot of weddings but its quite costly.

  • NSU67
  • ????? ??????

    Could you try this lens on Sony’s A7 with simple adapter? It’s interesting to chek THE bokeh without mirror’s cutoff.

  • Arthur Meursault

    Of course it is.

    USING is a ‘use’.

    Use as an ashtray = use.

    Use as a toilet = use.

    What part of use don’t you get? You’re without use or better said – you are useless.

  • brett turnage

    I agree with you. Unless it’s a really beat up lens, it should not be super out of spec.

    The aperture that I choose really depends on the type of photography that I’m doing in the article and whether or not it is my solo article or if it is an article directed by one of the editors. If it is the former I have more leeway to determine the artistic direction, but if it is for the Senior technical editor, then I have to shoot it the way that he likes it. For technical articles everything is usually small apertures f/16-f/32 because they want everything in frame completely clear, and they don’t care about diffraction. However, if I’m shooting a scored cylinder wall, then I will put on the 50mm f/2.5 macro and shove it right down the cylinder and shoot at f/8 so that they can zoom into the picture to show the tiny knicks and other damage. Most of my technical shots I shoot are done with my 70-200 f/2.8 and the exposures can be pretty long, but with the wifi adapter installed, I’m across the room planning the next shot, so I’m not waiting for the exposure to finish or worrying about camera shake. For panning shots its all 70-200 f/2.8, usually set at f/11 because they want the car perfectly clear.

    For my own articles I will play around with f/1.2 for some shots. I might be 15-20 feet back to elongate the DOF, but I find that the subject isolation of the larger aperatures adds a feeling to an image that would not exists if most or everything was clear in the shot. For instance, a shot I took last weekend of a technician spray painting a flex plate to check for lateral runout. I took two versions of the same shot, one at 1.2 and another at 5.6. The 5.6 image looked average, but nothing jumped out at you because everything clear it was not apparent what you are supposed to focus on. The same picture at 1.2 was drastically different, your eyes were immediately drawn to the hand, the paint can, and the spray going on to the flex plate, and it was clear not relying on a caption to explain what was happening (although it would still have a cation because it’s a tech article).

    I had 3 shoots last week, and I used it to test out the 85mm. I primary shot the first car at 1.2 and like before I used the distance between me and the car to expand the DOF. I think I was back about 30-40 ft, which for a car still filled the frame. I loved the pictures that I got at that aperture, but as I’m sure you know if you are shooting cars that are not yours, you have to assume that you will never have the opportunity to reshoot. So getting duplicate shots that are at different apertures is key. I don’t want to only have 1.2 because blown up there might be something annoying that might bug the editor in chief who is the main decider of all content. So better safe than sorry.

  • NSU67

    Cool effect for sure!

  • Rob Crenshaw

    I’ll post a portrait and you can see for yourself. This was a grab shot over lunch, so please don’t judge! Same 2.8 aperture disk. Lensbaby behavior is different from normal: it’s got a very sharp sweet spot, which transitions quickly into massive blurring. The smaller apertures widen the sweet spot and reduce the blur. The lens collar is flexible, allowing some built-in tilt, so plane of focus can be razor thin. This shot is slightly OOF, I just missed it.

  • NSU67

    Wow that’s neat.. don’t think that’s really suitable for portraits though : POrtrait with Sony A7R III and 50mm 1.0 wide open.

  • Rob Crenshaw

    I think one-trick pony is an apt description, and the reason for the love and hate: your work either encompasses that trick or it doesn’t. Yours does in spades, mine doesn’t. If I’m doing blur everything except my focus point I use a bendable Lensbaby. It’s exceptionally sharp and, depending on aperture, blurs OOF areas to unrecognizable. It’s got a nice look that suits my subject, but is way too blurred for some!

  • NSU67

    Yes I see your point about the 1.8 STM. And I’m inclined to agree. the 50 1.0 IS a one trick pony, but that trick is really great. It’s only good for a wide view and OOF stuff. But that’s what I enjoy.

  • Rob Crenshaw

    The 6D is noisy at high ISO, I shot with one for years. But it didn’t bother me, I don’t mind noise, in fact I like noise more than I like the intentional blurring cameras perform to smooth it out, along with the corresponding loss of detail. It just reminds me of grain. That said, I just did a similar test to what you’ve done.

Follow on Feedly