Canon R5 help...weird swirls in images

Hi,

I have a Canon R5, and during editing, especially when anything over a +8 bump in shadows or small bump up in exposure is done this weird circles pattern engulfs the image making it unusable. I've exaggerated the edit it in the attached image so you can see the pattern, but this doesn't happen on my 5D IV or 5D III when doing the same thing with the same settings. It's pretty much rendered the R5 useless because it shows up doing very minimal editing using Lightroom or ACR. It shows up at most ISOs, and the images across all ranges seem noisy (no matter of how much light I get on the sensor. Is it a noisy camera compared to the 5D IV, even just raising shadows a little seems to bring up noise the 5D IV doesn't have issues with pulling up in the same settings/setup. I can't reproduce the patterns/swirls doing the same on my 5D bodies.

Can anyone help? Is my R5 faulty?

Any help would be great, I'm totally stuck.

Thanks.

Paul

The R5 produces cleaner and less noisy images than the 5DIV.

What you have is something unusual.  Either faulty camera or something in the settings or processing chain.

I actually get something similar but not with every shot. I think it has something to do with Lightroom. If I use DPP and save as tiff and import into Lightroom the issue goes away.

The rings in the sample image look almost like Newton rings ... it could be caused by a defect in the sensor assembly/stack (anti aliasing filter disalignment?).

draacor wrote:

I actually get something similar but not with every shot. I think it has something to do with Lightroom. If I use DPP and save as tiff and import into Lightroom the issue goes away.

First, the circular ring pattern does not look like something a camera could do.

Second, time and again on this forum, people have found that Adobe's raw converter has serious problems, particularly with low light noise.

Thrid, definitely try DPP4 and get a "baseline reading."

Fourth, give DxO PureRAW or PL4 with DeepPrime a try (free trial and often sale).  It is widely recognized as the best RAW converter for reducing noise while preserving detail.  In my tests, DxO can be good for 2 to 3 stops over Adobe Camera RAW, which is just ridiculous. Adobe is not even trying with their RAW converter.

Karl_Guttag wrote:

draacor wrote:

I actually get something similar but not with every shot. I think it has something to do with Lightroom. If I use DPP and save as tiff and import into Lightroom the issue goes away.

First, the circular ring pattern does not look like something a camera could do.

Second, time and again on this forum, people have found that Adobe's raw converter has serious problems, particularly with low light noise.

I haven't experienced such problems with the images from my R5 and LR/ACR.

Thrid, definitely try DPP4 and get a "baseline reading."

That's a good point.

Fourth, give DxO PureRAW or PL4 with DeepPrime a try (free trial and often sale). It is widely recognized as the best RAW converter for reducing noise while preserving detail. In my tests, DxO can be good for 2 to 3 stops over Adobe Camera RAW, which is just ridiculous. Adobe is not even trying with their RAW converter.

You don't get additional dynamic range by using a different raw converter. A stronger noise reduction could give you such impression.

Quarkcharmed wrote:

Karl_Guttag wrote:

draacor wrote:

I actually get something similar but not with every shot. I think it has something to do with Lightroom. If I use DPP and save as tiff and import into Lightroom the issue goes away.

First, the circular ring pattern does not look like something a camera could do.

Second, time and again on this forum, people have found that Adobe's raw converter has serious problems, particularly with low light noise.

I haven't experienced such problems with the images from my R5 and LR/ACR.

Thrid, definitely try DPP4 and get a "baseline reading."

That's a good point.

Fourth, give DxO PureRAW or PL4 with DeepPrime a try (free trial and often sale). It is widely recognized as the best RAW converter for reducing noise while preserving detail. In my tests, DxO can be good for 2 to 3 stops over Adobe Camera RAW, which is just ridiculous. Adobe is not even trying with their RAW converter.

You don't get additional dynamic range by using a different raw converter. A stronger noise reduction could give you such impression.

I might have thought so too before running my own experiments. There are big differences between various Bayer to RAW conversion processes. DxO with DeepPrime will give BOTH better colors and better resolution a the same time. With Adobe RAW you have to trade color saturation and resolution for noise reduction.

DxO with DeepPrime changed the way I shoot is it so good with the R5. I don't mind pushing ISO if that is what is needed rather than say hoping I don't get subject motion blur by going to a slower shutter speed.

Here is a link to Jan Wegener talking about how much DxO (PureRAW which is the also from DxO and uses similar algorithms) improves noise over Adobe RAW: https://youtu.be/q-gMFMd9D7s?t=607

Here is another comparison: review: https://youtu.be/nn520b1mOas?t=167

From: https://www.dpreview.com/forums/post/65418038

Open to 100% to see the dramatic difference

From: https://www.dpreview.com/forums/post/65405499

Some tests I ran about 7 months ago right after getting DxO PL4 compared to DPP4.

https://www.dpreview.com/forums/post/64948762

My shock was that DxO PL4 with DeepPrime was seeming to violate Nyquist by showing single-pixel detail. I think the reason it can do this and why it is so much better at preserving detail is that is has more "scope" in looking at many pixels and making better inferences as to what is color, what is detail, and what is noise.

See: https://www.dpreview.com/forums/post/64949671

Karl_Guttag wrote:

Quarkcharmed wrote:

Karl_Guttag wrote:

draacor wrote:

I actually get something similar but not with every shot. I think it has something to do with Lightroom. If I use DPP and save as tiff and import into Lightroom the issue goes away.

First, the circular ring pattern does not look like something a camera could do.

Second, time and again on this forum, people have found that Adobe's raw converter has serious problems, particularly with low light noise.

I haven't experienced such problems with the images from my R5 and LR/ACR.

Thrid, definitely try DPP4 and get a "baseline reading."

That's a good point.

Fourth, give DxO PureRAW or PL4 with DeepPrime a try (free trial and often sale). It is widely recognized as the best RAW converter for reducing noise while preserving detail. In my tests, DxO can be good for 2 to 3 stops over Adobe Camera RAW, which is just ridiculous. Adobe is not even trying with their RAW converter.

You don't get additional dynamic range by using a different raw converter. A stronger noise reduction could give you such impression.

I might have thought so too before running my own experiments. There are big differences between various Bayer to RAW conversion processes. DxO with DeepPrime will give BOTH better colors and better resolution a the same time. With Adobe RAW you have to trade color saturation and resolution for noise reduction.

DxO with DeepPrime changed the way I shoot is it so good with the R5. I don't mind pushing ISO if that is what is needed rather than say hoping I don't get subject motion blur by going to a slower shutter speed.

In the R5, ISOs up to 6400-12800 are usable, although for action in shots in low light I usually use ISO 800 as the R5 is basically ISO-invariant from ISO 800.

Here is a link to Jan Wegener talking about how much DxO (PureRAW which is the also from DxO and uses similar algorithms) improves noise over Adobe RAW: https://youtu.be/q-gMFMd9D7s?t=607

Here is another comparison: review: https://youtu.be/nn520b1mOas?t=167

From: https://www.dpreview.com/forums/post/65418038

Open to 100% to see the dramatic difference

Overall rendering is better with Adobe - skin tones etc. There's lots of colour noise which is easy to remove in Lightroom, I don't know why it wasn't done in the image in the left.

From: https://www.dpreview.com/forums/post/65405499

Some tests I ran about 7 months ago right after getting DxO PL4 compared to DPP4.

https://www.dpreview.com/forums/post/64948762

My shock was that DxO PL4 with DeepPrime was seeming to violate Nyquist by showing single-pixel detail. I think the reason it can do this and why it is so much better at preserving detail is that is has more "scope" in looking at many pixels and making better inferences as to what is color, what is detail, and what is noise.

See: https://www.dpreview.com/forums/post/64949671

This is clearly some kind of error creating moire, where no competing patterns are visible, that is not something I have ever encountered shooting on my R5 and converting in either Camera Raw in Photoshop or in Lightroom. I can't say what specifically caused the false patterns of the moire effect, but it is clearly not inherent in either the R5 or any "ordinary" Adobe raw conversion.  Perhaps some unusual and extreme combination of things like vignetting control and other erroneous profile corrections have created your issue. Without being able to duplicate the effect, not seen in any of my first approximately 5,000-10,000 R5 images so far converted with Adobe software, I can't give a better answer than that.

I'd just like to thank everyone for the replies and for the help. Sorry for the time it's taken to respond to this.

It seems this was caused entirely by Adobe Lightroom/ACR, which with lens correction turned on it creates the swirls, but the R5 images seem much noisier in Lightroom or ACR still. Seems to be Adobe software causing the issues.

So, I did some tests in Capture One with the same images and couldn't reproduce it, no matter what lens correction, corrections or edits I did.

I just wanted to say a massive thank you so you all, not only did you guys help fix the problem, but you saved 2 sets of images I'd shot and thought were ruined by the camera and managed to keep my clients happy.

The images when editing are certainly less noisy in Capture One compared to ACR, and certainly no swirls. Looks like I'm moving to Capture One from now on and ditching Adobe (which has been getting clunky for the past couple of years now) I just wish Capture One export was faster on my MacBook Pro.

The R5...I still think is noiser than my 5D IV when it comes to needing to pull detail from shadows in low light, but at least the rings/swirls are sorted out.

Thanks everyone!

As you say it is definitely LR. It does it intermittently including adding purple to skies on my 5dsr. I solve it by zooming in and then out again but rather laborious if many images to review!

im guessing that a combination of LR, high mp and potentially older tech may be responsible….

Pmm85 wrote:

The R5...I still think is noiser than my 5D IV when it comes to needing to pull detail from shadows in low light, but at least the rings/swirls are sorted out.

The R5 is much less noisy than 5DIV, if you shoot in the game conditions and settings. I had 5DIV and can compare.

If your R5's images are noisy, there's something wrong with either the camera, or settings, or workflow.

Check if you have something like highlight tone priority or electronic shutter enabled (don't use electronic shutter at ISOs less than 800).

Thanks everyone!

expro wrote:

As you say it is definitely LR. It does it intermittently including adding purple to skies on my 5dsr. I solve it by zooming in and then out again but rather laborious if many images to review!

im guessing that a combination of LR, high mp and potentially older tech may be responsible….

Lightroom is not ideal but it doesn't have the OP's problem. There's something of going on.

With your issue, try playing with preview generation settings in Lightroom.