EOS R5 Mark II When to use Distortion Correction?

[boelens218]

Distortion correction is disabled by default on the r5 mark2. In what situations would you want to enable it?

I plan to only use the single lens I have which is the 24-70mm F2.8. 

I use Affinity Photo 2 to edit my cr3 images, should I use the "lens correction" (profile for my lens) and the "remove lens vignette" features? They are both enabled by default. 

Thank you,

[boelens218]

Without Lens Correction (24mm)With Lens Correction (24mm)

[boelens218]

Again, all the sample photos were taken with the distortion correction on the camera disabled. Then, I opened the raw files with Affinity Photo and exported them as jpegs, with and without the lens correction feature. I'm asking if I should use it or not. 

[jrhoffman75]

---

@boelens218 wrote:

Again, all the sample photos were taken with the distortion correction on the camera disabled. Then, I opened the raw files with Affinity Photo and exported them as jpegs, with and without the lens correction feature. I'm asking if I should use it or not. 

---

If you see an improvement in the image use it. If you don't save yourself a step.

[Peter]

Personally, I’ve turned off lens correction in the camera due to the performance loss and because I shoot in raw format. Canon has, unlike Sony and Olympus, chosen the right approach and, as far as I know, doesn’t bake any correction into the raw image. Instead, the correction data is stored in the metadata so that programs like Iridient C-Transformer can make use of Canon’s corrections. The program is, by the way, worth trying out.

In darktable, which just like Affinity uses Lensfun for Canon’s raw files, I almost always activate the lens correction profile to avoid curved lines and to remove the effects of chromatic aberration.

In your examples, it looks like only distortion correction has been applied. I don’t see any brightening from vignetting correction, even though the profile you have should correct that. There is no TCA profile at all, which you can see in the corners. You therefore need to contribute to Lensfun yourself, wait for another user to do it, or manually correct chromatic aberration in Affinity.

Whether you should use distortion correction in Affinity is something only you can decide, but as guidance I think you should photograph something with straight lines, like a fence or a window section.

Without distortion correction

[boelens218]

This is what Affinity said about the lens profile for my lens,

"No, we wouldn’t class the LensFun lens profile in Photo V2 as incorrect in a functional sense. It just misidentifies your lens."

I don't even understand what this means. Do you recommend I use the lens correction feature in Affinity after seeing this? You are correct, I turned off the remove lens vignette feature in Affinity for my lens correction examples. I only wanted to show the distortion correction effect by itself since this is the topic.

[SignifDigits]

OK, I may be an illuminary, but do not actually claim to be an expert - especially compared with all of the talent here.   First, I'm so sorry for your months of frustration.  It must be difficult to be sitting on photos and have to be waiting to process them.  I would absolutely hate that.  Next, I don't know what to make of the Affiinity response either.  Sounds like they may still think that they are choosing the EF lens perhaps?  I can't tell.  Sounds like "well, V2 is working fine but it's just using the wrong lens".  LensFun at https://lensfun.github.io/lenslist/ shows your lens on the list, but with no chromatic aberration correction.  Next, I've looked at all 4 of your pics.  I'm sorry we haven't acknowledged them.   To be fair none of them seem super stressful on barrel, pincushion or vignetting.  The differences are nearly indiscernible to me on a 4k monitor.  I might see a SLIGHT improvement in vignetting, so, if it were me, I'd use the V2 correction just for that.   And, it speaks well of your ability in that you may not be pushing the lens to to it's limits.  It also speaks well of that lens - it's a super great lens so needs very little in the way of correction to start with.  Peter's suggestion of shooting strong horizontal and vertical lines to help you to better see effect of corrections is a good one.   I would also suggest taking a blue sky or bright white wall photo with the lens wide open to stress the lens vignetting.  Then try turning the lens correction on and off to see if you see a difference.  If it's not obvious that V2 is making a correct on photos like those, then there still could be an issue with the V2 software not loading in the lens correction for your RF lens.  

I'm also sorry to hear using DPP was frustrating for you.  That is the best tool for my use.  I shoot RAW and I know I'm getting 100% approved lens and camera data straight from Canon with DPP.  I know it doesn't help with what you've already shot, necessarily, and you're probably more used to V2, but if you decide to try DPP again, there are some really good Canon videos to use it at https://www.usa.canon.com/learning/training-articles/training-articles-list/digital-photo-profession....  

I don't know that any of that really helps you in making your decision, and I would definitely defer to others more experienced than I am here on any opinions that they have.  I hope at least you feel that you've been heard and your pictures looked at, etc. in the forum.  We certainly do NOT want to add to the frustration you've already been going through over the past months.

[Peter]

Matching a lens profile with the correct lens can be difficult for developers. Some lenses share the same number, especially third-party ones.

If the lens profile in your case only contains correction for distortion and vignetting and missing TCA, it appears to be the correct profile. The name of the lens profile should be Canon RF 24-70mm F2.8L IS USM.

[boelens218]

Can someone please tell me if they would use the distortion correction for the 4 sample images below? I am unable to determine which is better or more correct. I understand that distortion correction is both objective and subjective. 

Thank you!

With Lens Correction (50mm)Without Lens Correction (50mm)

With Lens Correction (35mm)Without Lens Correction (35mm)With Lens Correction (24mm)Without Lens Correction (24mm)With Lens Correction (50mm)Without Lens Correction (50mm)

[SignifDigits]

I vote yes.   I ran the series zoomed in on a 32" OLED monitor and scrolled back and forth on each image to notice changes.  I notice that it corrects some pincushion.  That's all that I can see.  Correction appears to me to be most noticeable on the horizontal galvanized fence rail behind the motorcycle and the dog closeup.  The ccorrection does some cropping around the edges.  But the correction levels out the horizontal fence rail that appears slightly bent without it.  I think the dog benefits from the more accurate distant surroundings, which I find a bit more pleasing to my brain, though I would not really ever notice the pincushioning in that shot.  I don't notice anything significant in the lake scene and don't notice the horizontal or vertical lines changing enough to make much of a difference in the monument picture.

The only times that I would personally ever NOT use the correction would be a) I wanted a slightly pincushioned area surrounding a center focal area (and I cannot really think of an example of what that might be, but I can imagine intentionally pincushioned dark lines on a light background somehow being a possible desirable artistic effect) or b) the slight cropping of the edges that comes from correcting the pincushioning removed some edge detail that mattered to me more than the pincushion.

I cannot discern any other correction occurring.

[boelens218]

Thank you, your message was very helpful. Hopefully, maybe 1-2 other people can give their opinion also. I just wrapped up a very long email exchange with Affinity about the distortion correction feature in the software. As I said before, they told me the lens profile in Affinity misidentifies my lens (RF 24-70mm F2.8L IS USM) with the EF version. Therefore, the distortion correction is not the way it should be. I used Canon's DPP software, applied the distortion correction and compared the results with Affinity. It's a bit hard to tell for certain since one image is more cropped than the other, but it looks to me that the distortion correction is about the same. If anything, the correction is slightly more aggressive in the DPP. What do you think and do you think the profile in Affinity misidentifies my lens with the EF?

Affinity (With Distortion Correction)Canon (With Distortion Correction)