EOS R5 purple banding/haze with strong light source in the picture

[Uneternal]

So, I pulled the trigger on an EOS R5 this Black Friday. I guess somebody at Canon must have listened when I said Canon's sale prices are a bit underwhelming. Anyways I was super excited trying out the camera and loved all the resolution it gives you. But when l tried the dynamic range I was in for a ...

[Removed 3rd party link]

I noticed when I took a photo of my dog lit by a harsh sunlight but surrounded by shadows. When I pushed the shadows, suddenly a weird purple banding started to appear. It only seems to appear with harsh light sources and strong shadow pushes. I did another test with my EOS R vs the R5 and interestingly the R has some slight banding going on but none of that purple issue. And we're talking about an almost 8 year old sensor here (it is basically the sensor from the 5D4)!

ISO 100, f/4, Shadows +100, Exp +1.5 (EOS R5 vs EOS R)
I know this is an extreme push, but this purple banding/haze is definitely an issue.

EOS R5EOS R

I'd appreciate if somebody here could try to take a photo of a light source, push the shadows, and confirm this issue. Or do you think it could be my camera?

[shadowsports]

Greetings,

Congrats on the acquisition.  I don't recall if I sent you a PM or posted on another thread, but the 2 sales on the R5 this weekend were hot. 

$2199 for refurb, $2999 for new.  These were exceptional deals.  Lens deals were also plentiful.  

But let's focus on your question.  I'll start.

Are you using Canon lenses?  If so, which exactly?  Are they RF or are you using an adapter?  If yes, is it Canon brand?

What does this mean exactly:

ISO 100, f/4, Shadows +100, Exp +1.5

You took a picture in incandescent light at ISO 100, f4, shadows +100 (100 what?) And then pushed EV +1.5 and got a purple cast?  

Can you please save the problem image to google drive, Dropbox, one drive, etc and post a link for us so we can review the EXIF data and image(s) ourselves.  

Others may have additional questions once we see the image(s).  🙂

[Uneternal]

Are you using Canon lenses?  If so, which exactly?  Are they RF or are you using an adapter?  If yes, is it Canon brand?

Yes, I tried with RF 24-105mm and with a EF 50mm 1.2L - yes on the Canon adapter. Both times the same issue. It's not lens-related.

What does this mean exactly:

ISO 100, f/4, Shadows +100, Exp +1.5

These are Lightroom edits, sorry if that wasn't clear enough. 

Can you please save the problem image to google drive, Dropbox, one drive, etc and post a link for us so we can review the EXIF data and image(s) ourselves.  

Here have fun: https://drive.google.com/drive/folders/1-KJvYymwVDsal34w4usIxu2ZKBH2kvVd?usp=drive_link

[Peter · Accepted solution]

Seems to be more than 6 EV lifting in post. I got 6,956 EV when I measured the brightness from the right field in your JPEG and applied it to the raw file. No contrast applied.

The purple cast is probably due to the black level. In Exif it is set to 511, but I think 512 looks better when pushed in post.

511 above, 512 below.

I am surprised R handled it better, because I have seen many samples the other way round with R and a light source like that.

With that said your shadows are really down in the noise floor and very close to to the optical black area that should be pitch black, see the black frame left and top.

If you try the same with my raw file here, 4737, from R6 I don't think you will see the same type of artifact because I didn't get near the noise floor.

[Uneternal]

Thanks for your reply. I know the EOS R had this issue before, and they fixed it with a firmware update. I'm not sure why this wasn't already resolved with the R5. If you own a EOS R5 could you do a quick test to confirm this also happens on your camera? 

[Peter]

I know one that owns a R and the firmware update didn't help him.

When R5 was released I saw an astro image with the same artifact /limit as you re showing.

I don't own R5 but R6. Do you want the raw file?

[Uneternal]

Yes please 🙂
Thank you.

[Peter]

PM sent to you.

[shadowsports]

Why would you shoot at 1/1000 of a second at f4 in low light?  What were you expecting?

If you want to test DR, its your ISO that should be increasing progressively.

[Uneternal]

Why would you shoot at 1/1000 of a second at f4 in low light?

Rick, maybe you could try to stay on topic for once and read again what I was asking for. I'm testing the dynamic range. That's what I was trying.

What were you expecting?

Um l dunno, the EOS R result maybe. But how dare I expect better results from an 8 year newer sensor and a $3800 camera? 

What I clearly wasn't expecting somebody sitting in front of their computer and thinking about how they can spin purple haze into a user error. If you want to contribute anything useful to this discussion, go take your EOS R5C, take a picture of a light source in a dark room and tell me if you have the same issues. That's all I'm asking for, if that behavior is normal. Because some people already told me it's unusual and their EOS R doesn't do it.