Chromatic Aberration when switching from DSLR (Canon 60D) to Mirrorless (Canon R10)

[BehindTheLens]

After I was unable to find any answers from Canon or Tokina support and searching this forum, I thought I'd start a dedicated thread in the hopes that the gurus out there might be able to help.

I've used my Tokina AT-X Pro SD 11-16mm f/2.8 (IF) DX II for 7 years on my Canon 60D DSLR without any issues, but after upgrading to the Canon R10 mirrorless (FW 1.5.0) with the Canon EF-EOS R adapter I now experience some pretty severe chromatic aberration (CA). Samples uploaded to https://photos.app.goo.gl/tUqadN9B21GrXhcF8 and it's very noticeable in the astrophotography (5702). I don't always want to shoot in RAW and be burdened with the post-processing. Thoughts from the community on the root cause and corrective action?

More details:

1. I did not see any CA issues with the Canon 60D DSLR and I only shot in JPG.
2. My first trip with the R10 was to Antarctica with lots of high contrast white scenery and fortunately I didn't detect any CA (shot in RAW+JPG), but it has since started to appear and increase in severity.
3. CA is present in the on-camera generated JPG. This is true if saving as JPG directly (5363, 5364, 5367, 5609) or on-camera conversion to JPG (5702).
4. CA appears in unprocessed RAW, but then fortunately it will process out (Windows Photo or Canon Digital Photo Professional). Why can't the on-camera conversion correct this?

I'm also new to post-processing and wasn't able to find any lens correction data for the Tokina in Canon Digital Photo Professional. Could that possibly fix the CA better?

I appreciate Tronhard's post in https://community.usa.canon.com/t5/EF-RF-Lenses/Lens-recommendation-for-Iceland-travel-aurora-photog... about Sigma lens, so I'm eyeing the Sigma 10-18mm F2.8 DC DN to replace my beloved Tokina 11-16mm f/2.8 if I can sell that for a reasonable cost to justify the upgrade.

Thanks in advance for the help!

[Tronhard]

Chromatic Aberration is a function of lens optical design and not the camera sensor, but the camera may go some way to mask flaws in a lens through lens correction algorithms, especially one that is of a contemporary design. What is happening with the new generation of MILC sensors is that they are showing flaws in what were acceptable lenses used on DSLRs - especially from other brands.  I had several of the 60D bodies and loved them, but they were a very different kettle of fish compared to modern sensors. Canon (like all manufacturers) have built-in lens correction and optimization for their lenses to correct JPG images, but not for 3rd-party ones: which is why I prefer to stick with OEM gear.  

If you want to work with Sigma, then the lenses they have released for the RF mount are the way to go. While I have used, without issue, the Sigma 150-600c and 60-600s for EF, they are super telephoto lenses and the optical challenges are different for super wide-angle lenses.   (Edited: see my later post on in-camera lens corrections).

Others have done testing on the Sigma RF 10-18 and you might find it instructional to view some of those:
sigma rf 10-18mm f2.8 - YouTube

[Tronhard]

I had a look on Sigma's website and according to the Features section of the page for this lens, the RF cameras do support lens corrections: "the use of in-camera aberration correction further eliminates optical imperfections such as distortion and vignetting" 
Feature | 18-50mm F2.8 DC DN | Contemporary | Lenses | SIGMA Corporation (sigma-global.com)

[BehindTheLens]

Trevor, thanks for the speedy reply! I acknowledge that the optics are inherent in the lens, but was disappointed that my 60D DSLR (circa 2010) could handle that better my R10 Mirrorless (circa 2022, a dozen years newer). A good point about the current sensors being different and showing more flaws. At least working with RAW, I'm able to minimize, but not really sure why I didn't see any CA when I first used it in 2022.

I agree that Canon will not make an ultra-wide fast lenses (I think you mentioned that in another post and that's why I finally got the Tokina 11-16mm f/2.8 instead), nor will they have lens correction data available for non-Canon lenses. I'll have to watch the reviews on the Sigma 10-18mm f/2.8 RF (maybe they'll have lens correction data for camera or DPP) or Tamron 11-20mm f/2.8 to see if I'll try to upgrade. I would get a full frame with a Canon RF15-35mm F2.8 L IS USM (about the same as my APS-C w/ 11mm), but the reality is that my couch doesn't have that much loose change.

[Tronhard]

On a moving train so hard to type on a cell phone.

Old lenses show their weaknesses on newer and better bodies. That is not specific to a brand but just physics. Some work better than others, but the focal range is challenging.

If you were to go FF, there are benefits, especially for landscape where you don't have the effects of crop factor on wide angle lenses. On an APS-C, a10-18 has an equivalent FoV on a FF body of 16-29mm! So huge benefits to FF for you. Also much less noise on FF.

The FF body I would recommend is the R8 and the much more economical but fabulous RF 14-35 f/4.

It is fast, small, compact and light. It also takes normal 77mn filters, which is awesome.

Check out the gallery and look for RF14-35 to see some samples.

Well the good news, as I added in a later post, is that according to sigma their RF 10-18 can be supprted.

You might find a deal on the Canon Refurb lens site.

[BehindTheLens]

Trevor, thanks for battling out a response on a moving train 🙂 While my couch pockets might be deep, they are not filled with money and the dust bunnies won't pay for that full frame body and lens (I appreciate the suggestions though). I'm not that serious of a photographer to justify the budget and have been pretty happy with my crop sensors setups so far. Unfortunately, I really need the f/2.8 for my usage (e.g., inside cultural buildings where flash is not allowed) and I do like that the Tokina does not have an articulating barrel for zoom or focus (makes me think it's less prone to ingress). Thank you again for informing me of the Sigma because I might make an exception there for better compatibility and a smaller package (Tokina + EF-R adapter is big) ... if I find more coins under the dust bunnies.

I'll definitely keep an eye out for refurb or used in good condition (another great suggestion).

Happy Friday!

[Tronhard]

Ah... back in front of a real keyboard again! 🙂

OK, if we stick to the APS-C crop sensor, then I would suggest doing some research on the Sigma 10-18 lens. To get you going here are some reviews from trusted reviewers, Gordon Liang, who compares it to Canon's native cheaper lens.

Certainly, the Canon is significantly cheaper, but comes with a bunch of compromises: 
It is not as well built, especially having a plastic rather than the Sigma's metal mount
It has between 1 and two stops slower aperture.
The Sigma is generally sharper across the range.

Really the question is whether you need the better resolution and faster aperture of the Sigma or can get along with the cheaper, but inferior native Canon lens.  That is the question.

Investing in a lens is a longer term one compared to a camera body, and you get what you pay for, but again I don't know your budget because you have not given a $ value. If your purchase is not urgent it is worth waiting and saving if you value those features.

[BehindTheLens]

Thanks again Trevor for the tips. It's been a long time since I've done any lens shopping, so the recommendations on trusted reviewers is helpful. The Sigma's 659USD price tag is within reason. I'm also curious how the Tamron 11-20mm f/2.8 (https://www.tamron.com/global/consumer/news/detail/b060rf_20240423.html) will compare. I'm hopeful that there will be some Black Friday deals. Sadly, the Canon 15-35mm f/2.8 is out of my budget (do you know how much sushi you can buy with the difference?!?), especially if I need to buy a full frame body (I just got my R10 in 2022 and had my 60D since 2010).

I feel Canon used to dominate the SLR, then DSLR market (I started out with Nikon, but then committed to Canon back in 2010 when I finally bought the 70-200mm f/2.8), but now get the sense that other platforms (Fuji, Sony) are more popular with the mirrorless platform and no one seems to be continuing with the DSLR.

[Peter]

Would you mind sharing some raw files?

It seems like your Tokina is using the same LensType nr as Canon EF20-35 but I don't think Canon has a lens profile for that lens anyway that would interfere (if that even is possible). Your pictures show TCA that is easier than LCA to correct if you have a dedicated lens profile.

[BehindTheLens]

Hey Peter, in my original post, I had a link to Google Photos that should be available to you (https://photos.app.goo.gl/tUqadN9B21GrXhcF8), but let me know if you have any issues. This includes a couple of RAW files to compare against the JPGs. I think the CA is gone in CanonR10_5702_11mm_DPP.jpg that I processed (for the newbie that I am). 

I've heard of some people making their own lens correction data file, but that's way beyond my capabilities and I'm not sure if Canon's DPP will allow for that anyway. If I get more serious with using RAW files, I might switch to DarkTable (https://www.darktable.org/) or ON1 (https://www.on1.com/). While I know that Adobe Lightroom is likely the best solution, I don't like having to pay a subscription for the 1-2 times a year I travel and will want to post-process photos. I'll have to search the forum later to see what people think about post-processing software.