R5 Dead Pixels

[Chris.Chapterten]

I recently switched from using Adobe Camera Raw to Canon DPP to process my RAW photos from the R5 (much better colour!). Unfortunately DPP doesn't seem to automatically remove dead or stuck pixels like ACR does.

Scanning through a couple of photos at 200% magnification I have so far found two dead pixels... one is right at the very edge of the frame and is not much of a concern. The second is unfortunately right in the middle of the frame and is a distinct purple colour.

This single, problematic dead pixel is obviously easy to clone out in post when shooting stills. My question though, is how likely is this to show up in Video? It will be more of a pain removing it from video footage I'm guessing.

Do any other R5 owners have some dead pixels, and if you do video work regularly how often do they show up?

Thanks!
Chris

 

[Mt Spokane Photography]

Its NR that removes them as I understand it. Do you have DPP NR off? I haven't looked for dead pixels from mine, I'd probably never look at a 45 mp photo at 200%, I don't have a use that critical.

I should take the time to check those things out while I'm still in the 30 day return period. If it was really bad, I'd return it.

 

[Chris.Chapterten]

Its NR that removes them as I understand it. Do you have DPP NR off? I haven't looked for dead pixels from mine, I'd probably never look at a 45 mp photo at 200%, I don't have a use that critical.

I should take the time to check those things out while I'm still in the 30 day return period. If it was really bad, I'd return it.

I see, I have noise reduction turned off... I'll try a mild amount of NR and see if that does anything, thanks.

It's time consuming to scan through a 45mp image like that, but worth it for piece of mind.. my main concern is them being visible in video footage.

Let us know how you get on with yours if you do end up checking the sensor. Cheers!

 

[Mt Spokane Photography]

I had my camera right here and took photos with lens cap on. two of them seemed appropriate to check for stuck or hot pixels. I took one at f/4, 0.5 seconds, ISO 51200 and one in bulb mode for 17 seconds at ISO 400.

The first one in both Adobe LR and in DPP 4 showed no hot spots.

The 17 second exposure had 26 of them of various colors, red, blue, green, and white. They did not show in LR but did in DPP 4 and NR made no difference.

I'm not sure of the significance of that, certainly, astro photographers who use long exposures don't need 26 stars showing up, I think they have software that first takes a image to locate the hot pixels, then they are subtracted from the final image. For me, a normally exposed or even badly over exposed like the 0.5 sec at ISO 51200 will not see them.

I was surprised that LR was able to remove them but DPP Did not.

I went thru the procedure to remove hot pixels by putting the camera in sensor cleaning mode for a minute and repeated the 17 second exposure. I did that twice. Each time, there were 26 hot pixels, maybe 1 or 2 very dim ones that I wasn't positive about.

My conclusion is that it would not affect me. If you are concerned about video, try putting the lens cap on, setting what might be a mild overexposure and then look for hot pixels.

If you are finding stuck pixels with a normal exposure with lens cap on, I'd call Canon since they will show in video's.

 

[Chris.Chapterten]

I had my camera right here and took photos with lens cap on. two of them seemed appropriate to check for stuck or hot pixels. I took one at f/4, 0.5 seconds, ISO 51200 and one in bulb mode for 17 seconds at ISO 400.

The first one in both Adobe LR and in DPP 4 showed no hot spots.

The 17 second exposure had 26 of them of various colors, red, blue, green, and white. They did not show in LR but did in DPP 4 and NR made no difference.

I'm not sure of the significance of that, certainly, astro photographers who use long exposures don't need 26 stars showing up, I think they have software that first takes a image to locate the hot pixels, then they are subtracted from the final image. For me, a normally exposed or even badly over exposed like the 0.5 sec at ISO 51200 will not see them.

I was surprised that LR was able to remove them but DPP Did not.

I went thru the procedure to remove hot pixels by putting the camera in sensor cleaning mode for a minute and repeated the 17 second exposure. I did that twice. Each time, there were 26 hot pixels, maybe 1 or 2 very dim ones that I wasn't positive about.

My conclusion is that it would not affect me. If you are concerned about video, try putting the lens cap on, setting what might be a mild overexposure and then look for hot pixels.

If you are finding stuck pixels with a normal exposure with lens cap on, I'd call Canon since they will show in video's.

Thanks for following up. Great news that you aren't seeing any at normal exposures. In my case it was iso 400 1/250s that they were showing up.

I have tried the manual sensor cleaning trick with other Canon cameras but it has never worked for me, I might give it another go before contacting Canon.

If I hadn't switched to DPP I would have never of known the dead pixels were there. I'm not sure why Canon doesn't just use the same algorithm as Adobe to get rid of them.. it's a very handy feature.

My EOS R had 4 or 5 dead pixels and Canon performed what they called a 'Pixel compensation' to fix it.. strangely the dead pixels of the EOS R showed up in Adobe as well as DPP.

I was hoping I wouldn't habe to send the camera away. but might be no other option... hmm..

 

[Ericsmith]

While they don't mention stuck/dead pixels for the sensor, Canon does mention in the R5's manual that up to 0.01% of the pixels in the EVF or rear LCD may be stuck or dead, and is normal.

I just noticed by coincidence that my rear LCD has at least one, and the EVF has several stuck red pixels when I turned on the R5 with the lens cap on. I'm going to test a few frames later to see if there's any stuck pixels on the sensor itself.

I'm a bit surprised because I wasn't even looking for them and seen it by coincidence. Never noticed them before on my Panasonic and Sony bodies.

Does anyone know if this is common? (stuck pixels on EVF/LCD/Sensor) I'm interested in taking an informal poll, as I'm still in the return window for my camera, but would hate to get an exchange only to have the replacement have the same issue, or worse! ...If it is that common...

 

[SteveC]

Does anyone know if this is common? (stuck pixels on EVF/LCD/Sensor) I'm interested in taking an informal poll, as I'm still in the return window for my camera, but would hate to get an exchange only to have the replacement have the same issue, or worse! ...If it is that common...

Common? I don't know. But I saw some red pixels in my EVF when I tried to do some wide field astrophotography.

 

[Ericsmith]

Common? I don't know. But I saw some red pixels in my EVF when I tried to do some wide field astrophotography.

Thanks for your input. Yes, maybe common is the wrong word for it (although Canon felt it warranted mentioning in the user manual). I have a feeling even if I got an exchange, I'd potentially be looking at the same "problem"...

 

[SteveC]

Thanks for your input. Yes, maybe common is the wrong word for it (although Canon felt it warranted mentioning in the user manual). I have a feeling even if I got an exchange, I'd potentially be looking at the same "problem"...

If it's in the viewfinder I can ignore it. Especially since it's only a nighttime issue.

 

[mkamelg]

Scanning through a couple of photos at 200% magnification I have so far found two dead pixels... one is right at the very edge of the frame and is not much of a concern. The second is unfortunately right in the middle of the frame and is a distinct purple colour.

Please read my post posted on another thread.

R5/R6 Camera defect reports (no hype or wish lists please)

I had an older battery in my R6 and for about a minute, the evf started Behaving Badly. It would just stutter and jerk like crazy even moving the camera slowly. I turned the camera off and on and everything was fine.

www.canonrumors.com