Canon EF-S 10-18mm & 10-22mm on FF compared w/ TC & MSC

EF Mount conversion

Now, at least with the Sigma MC-11 adapter I use to mount Canon EF lenses on my Sony E mount cameras, you cannot attach an EF-S lens. This will probably vary depending on the adapter you are using. So I purchased EF mounts for both these lenses to convert them to EF mount. I thought that possibly these may also allow for a slightly wider image circle than the native EF-S mounts.

Vignetting pre versus post mount conversion

This test is to see if the EF mount conversion actually reduces vignetting at all, or if it makes no improvements other than allowing use of the lens on an adapter that is not compatible with EF-S lenses. These test shots are all full frame, not APS-C crops.

We can see from these that, yes, the EF mount conversion does reduce vignetting. It is not simply just making the lens so it can use an EF mount.

Using with dumb adapters / avoiding forced APS-C crop on Canon

Rather annoyingly, on Canon bodies when it detects an EF-S lens is attached, it will force an APS-C crop, with no way to turn this off. To avoid this you can dismount the lens slightly (just enough so it isn't locked). This breaks the electrical contact between the lens and body, and so the body no longer knows what lens is mounted and will not force an APS-C crop.

However, both lenses have electronic aperture, so you have no way of setting the aperture when used like this. And you have the same issue when you are using the lenses on a 'dumb' adapter with no electronic contacts like the Laowa MSC. And to make matters worse, the 10-18mm is a focus-by-wire lens, meaning you cannot adjust the focus without any electrical contact.

To work round this, you need to get the lens in a state where the aperture and focus are actually set at the correct values you want before you disconnect it from the camera body. This should be possible to do by dismounting the lens while holding the DoF preview button. However, on my R5 the DoF preview button doesn't seem to do anything, I'm not sure if this is normal or its just something wrong with my camera.

Instead you can set the focus and aperture, then remove the lens while taking a photo (meaning you'll need to set a few second exposure to give you enough time to disconnect the lens during the exposure). Or you can put the camera in video mode, and then disconnect the lens.

When using with a 'dumb' adapter, you will need a non-dumb adapter or another camera you can mount the lenses to so you can set the aperture and focus. For example, when taking the test photos with the MSC and 10-18mm lens, I first mounted the lens on my Sony camera with Sigma MC-11 adapter. I set the aperture and focus, then disconnected the lens while taking a 4s photo. I then moved the lens to my Canon camera with the MSC (as my MSC is RF mount), checked the focus was okay, then took the photo. If the focus was off, I would have to re-mount the lens on my Sony and repeat the process.

Focus drift when changing focal length

For these tests the lenses were focused at 10mm and then zoomed to their maximum focal length to see if the image would remain in focus or not. The purpose of this test is to see if you focus the lens near infinity could you then use it on a dumb adapter at any focal length without worrying about having to refocus the lens. (Assuming you're happy having the lens focused near inifinity, which I would be for the majority of what I would use these lenses for).

Please note the images shown for this test are all grabbed from video, they were not shot as stills.

Image quality

All the following test images were shot at f/8 and focused near infinity. The shots using the MSC and teleconverter were also shot at f/8, which is an effective aperture of f/11. Photos were shot in RAW and converted in Adobe Camera RAW with Chromatic Aberration correction turned on.

Versus APS-C / Lack of lens profile corrections

Now let's look at images shot full frame and then cropped in post by 1.6x versus images shot in-camera with an APS-C crop. The full frame images were shot on my Sony camera and the in-camera cropped ones were shot on Canon, so there are some colour differences between the two. But mainly I wanted to look at how the application of the lens profile corrections applied to the in-camera cropped images compare to the uncorrected manually cropped FF images.

When comparing native APS-C shots from the lenses versus full frame shots cropped to APS-C in post I don't see much of a difference.

I assume that applying one of the lens profiles to a FF image taken with the lens applies an APS-C lens profile. And so distortion and vignetting corrections would not be correct. However, unlike some modern lenses designed for mirrorless, these lenses do not appear to rely heavily on software corrections. I can't detect things getting overly warped or stretched wrongly when looking at these test images and adding a lens profile.

10-18mm @ 14mm FF no lens profile

10-18mm @ 14mm FF with lens profile (presumably meant for APS-C crop)

10-18mm @ 14mm shot in-camera in APS-C mode (lens corrections applied)

10-18mm @ 14mm shot in-camera in APS-C mode, lens profile turned off

10-22mm @ 14mm FF no lens profile

10-22mm @ 14mm FF with lens profile (presumably meant for APS-C crop)

10-22mm @ 14mm shot in-camera in APS-C mode (lens corrections applied)

10-22mm @ 14mm shot in-camera in APS-C mode, lens profile turned off

Laowa Magic Shift converter

This adapter allows mounting an EF lens on on a Canon RF mount camera. They also offer the same adapter with a Nikon F lens mount, and on the camera mount side it is also available for Sony E and Nikon Z mounts. What is special about this adapter is that it includes a 1.4x teleconverter and allows shifting the lens.

The idea is that you use a full frame lens with it, and the 1.4x magnification enlarges the image circle of the lens enough to allow you to shift the lens without really bad vignetting.

When using it with a lens designed for APS-C instead, what happens is that it enlarges the image circle so that it just about covers full frame. (Canon APS-C is a 1.6x crop compared to the 1.4x magnification of the MSC). While I did also take test shots with shift on the MSC, I am not including these below for each focal length as the image quality when shifted is so bad you would not want to use these lenses like that.

The MSC has no electrical contacts. Using the 10-18mm lens on it was a nightmare to try and get the focus set correctly.

Both the 10-18mm and 10-22mm cannot be used at their widest focal lengths as the rear glass of the lens physically touches the glass of the MSC. For the 10-18mm this is at 11mm, so 12mm onwards should be safe to use, and for the 10-22mm this is at 12mm, so 13mm onwards should safe to use.

The Laowa MSC has a noticeable cool colour shift compared to the same lens on a standard adapter, however we see the colour temperature warm back up when shifting the lens. This could potentially be down to the filter I was using on the lens? (My Canon R5 this test was done on is full spectrum converted and so requires a filter to get normal visible light images).

EF-S 10-18mm lens at 18mm on Laowa MSC with max shift - a much warmer colour balance than the unshifted images below.

Note with these images a step-up ring and 82mm filter was mounted on the lens, and so vignetting may be worse than you would get without a filter.

If you were to use shift, one potential problem is that lateral CA becomes more difficult to correct. Since the lens is no longer centred, the CA will be stronger on one side of the image than the other. However I don't see any bad CA with these lenses and the MSC, and as I mentioned previously, they are too soft at the edges to use with shift anyway in my opinion.

Also note that you lose a stop of light with the MSC.

10-18mm + Laowa MSC vs 10-18mm APS-C comparison crops

12mm MSC vs 10mm APS-C - same level of pixel zoom, not size normalised

12mm MSC vs 10mm APS-C - size normalised

(The rest of these comparisons are all size normalised).

13mm MSC vs 11mm APS-C

14mm MSC vs 12mm APS-C

15mm MSC vs 13mm APS-C

16mm MSC vs 14mm APS-C

17mm MSC vs 15mm APS-C

18mm MSC vs 16mm APS-C

FoV on the MSC is wider than same focal length APS-C cropped. This makes sense since the MSC is a 1.4x magnification while Canon APS-C is a 1.6x crop. I would say 11mm on the MSC is a bit wider than 10mm APS-C, which matches up to the calculated effective focal lengths (11 x 1.4 = 15.6, 10 x 1.6 = 16). The MSC looks softer at the pixel level than APS-C, but looks quite a bit better when both are matched to the same display size.

10-22mm + Laowa MSC vs 10-22mm APS-C comparison crops

12mm MSC vs 10mm APS-C - same level of pixel zoom, not size normalised

12mm MSC vs 10mm APS-C - size normalised

(The rest of these comparisons are all size normalised).

13mm MSC vs 11mm APS-C

14mm MSC vs 12mm APS-C

15mm MSC vs 13mm APS-C

16mm MSC vs 14mm APS-C

17mm MSC vs 15mm APS-C

18mm MSC vs 16mm APS-C

19mm MSC vs 17mm APS-C

20mm MSC vs 18mm APS-C

21mm MSC vs 19mm APS-C

22mm MSC vs 20mm APS-C

The results here are pretty much the same as with the 10-18mm. The lens used normally with an APS-C crop has better pixel level sharpness than when shot full frame with the MSC. But when normalised to the same display size the images with the Laowa have more detail. With the possible exception of the very edges and corners of the frame, which I think look a bit better in the APS-C image.

EF-S lenses + MSC vs 16-35 FF lens comparison crops

I took test shots with my Canon 16-35mm f/4 lens so we can compare against a lens that is designed for full frame. I took test shots with this lens on both my Sony and Canon cameras, however some of the 16-35mm images on the R5 were quite front-focused, which is probably down the to EF adapter I used. So comparisons here are the EF-S lenses with the Laowa MSC on my Canon R5 versus the EF 16-35mm on my Sony A7R2 with Sigma MC-11.

10-18 + MSC at 11mm vs EF 16-35 at 16mm

10-18 & 10-22 + MSC at 12mm vs EF 16-35 at 17mm

10-18 & 10-22 + MSC at 13mm vs EF 16-35 at 19mm

10-18 & 10-22 + MSC at 14mm vs EF 16-35 at 20mm

10-18 & 10-22 + MSC at 15mm vs EF 16-35 at 22mm

10-18 & 10-22 + MSC at 16mm vs EF 16-35 at 23mm

10-18 & 10-22 + MSC at 17mm vs EF 16-35 at 24mm

10-18 & 10-22 + MSC at 18mm vs EF 16-35 at 25mm

10-22 + MSC at 19mm vs EF 16-35 at 26mm

10-22 + MSC at 20mm vs EF 16-35 at 28mm

10-22 + MSC at 21mm vs EF 16-35 at 29mm

10-22 + MSC at 22mm vs EF 16-35 at 30mm

The 16-35mm on the Sony is sharper than the 10-22mm + MSC. It also seems sharper than the 10-18mm + MSC, but it is more difficult to compare there due to the difficulty of focusing that lens.

EF-S lenses with a teleconverter

What about if we simply use a teleconverter to increase the image circle of the lenses to cover full frame? Unfortunately my Sony TC is not compatible with my E mount EF lens adapters. However, I do have an EF mount Kenko C-AF 1.4x DGX teleconverter, which is specially designed for use with EF-S lenses. Note that you lose a stop of light with a 1.4x TC.

With the TC, images are sharp in the centre but soft at the edges, particularly with the 10-22mm. There is also more CA.

Images with the MSC are better than with the TC.

Comparison between all

Here I'm comparing uncorrected images with lens profile corrections turned off and Chromatic aberration correction off.

Various lens, crop, and adapter combinations compared that all give a similar FoV

Crops from left edge

From the crops you can see neither lens has particularly bad lateral CA. Adding the Kenko DGX teleconverter does cause quite a bit of CA, but all other combinations look pretty good.

Flare

The 10-18mm seemns to have more flare than the 10-22mm. However, at least with my copy of the 10-22mm it has ugly flare that looks a bit like the lens is covered in raindrops. I can't see anything on the front or rear elements, so I assume this is dust inside the lens causing this effect.

Lens hood

As you might imagine, the lens hoods for both lenses are designed for an APS-C image circle. And so using the lens hoods on full frame will just give you a lot of vignetting. Note that these images were taken on my Canon camera, so they are at a slight angle due to the lens being not quite fully mounted in order to avoid the forced APS-C crop Canon does.

Conclusion

The lenses can be used full frame, but they are a bit soft at the edges for my liking. You could use something like an APS-H crop and get a wider FoV than you would in APS-C mode for a given focal length, while retaining decent image quality. However, at least with my cameras, framing images for a 1.3x crop would be difficult as there is no way to do this in-camera. And other than increased resolution there is little benefit over just using the lens in APS-C mode at a wider focal length.

The 10-18mm is a better lens in terms of image quality, weight, and size than the 10-22mm. But if you want to use the lens on an adapter without electronic contacts, the focus by wire 10-18mm is extremely difficult to use, and the 10-22mm lens is a much better choice.

The Laowa MSC is okay, but you're much better off just getting a full frame lens than using an APS-C lens with the MSC to make it cover full frame. Image quality is better with an FF lens, and for electronic lenses they are much easier to use when connected to the camera. (Laowa do market their adapter for use with their fully manual lenses, and using APS-C lenses on full frame is not its intended purpose either).

Image quality with the teleconverter is not good enough to be worthwhile.

So while you could technically use an EF-S wide-angle zoom lens for both APS-C and full frame, it does not work out as well in practice as I had hoped.