5D4 Sensor Defect Discovered

[arthurbikemad]

Oh dear oh dear oh dear

That is all :-X

 

[ksgal]

Hi all!

This picture shows the problem pretty well. This was shot 5 steps under as RAW (ISO 100), then to LR and boosted exposure 5 steps back + shadows up. This is of course extreme, but the picture taken the bright side on the right did not show these "stripes" at all - actually amazing results without these stripes. There is no noice reduction used.

Well, I'm amazed this has gone to 17-18 pages.
Jukka, thank you for the image that explains what people are seeing so well.

In all honesty, I'd take anyones 'defective' 5DIV for half price if an owner is so upset about the 'problem'.
PM me... please!

Back in the day, when we all shot film (I know some of you can't remember that, but it's true, we all did!) if someone underexposed 5 stops and then tried to get something out of that negative, and the result came out horrible.. well, we blamed the operator of the camera. With the latitude of film, usually you could still get something.. but no one expected the results of a 5 stop under exposure to be 'optimum' and usually were grateful we got anything at all.

The film was not defective. The camera was not defective. The Camera Operator had a brain fart and screwed up.

So. the Canon EOS 5D Mark IV is not defective. If you selected the camera and wish to return it because there is another camera out there that can do a 5 stop push and give results to your satisfaction, then you selected the wrong camera here... sell this one and go get the one that can give you the results you want in that situation.

A camera is not defective if it only does 5 fps, or has 11 stops of dr, or can't do a panoramic picture. It means you got the wrong tool for the job.

Saying this again - the Canon EOS 5D Mark IV is not defective.

Go get the camera that fits your needs.

 

[tron]

Hi all!

This picture shows the problem pretty well. This was shot 5 steps under as RAW (ISO 100), then to LR and boosted exposure 5 steps back + shadows up. This is of course extreme, but the picture taken the bright side on the right did not show these "stripes" at all - actually amazing results without these stripes. There is no noice reduction used.

Well, I'm amazed this has gone to 17-18 pages.
Jukka, thank you for the image that explains what people are seeing so well.

In all honesty, I'd take anyones 'defective' 5DIV for half price if an owner is so upset about the 'problem'.
PM me... please!

Back in the day, when we all shot film (I know some of you can't remember that, but it's true, we all did!) if someone underexposed 5 stops and then tried to get something out of that negative, and the result came out horrible.. well, we blamed the operator of the camera. With the latitude of film, usually you could still get something.. but no one expected the results of a 5 stop under exposure to be 'optimum' and usually were grateful we got anything at all.

The film was not defective. The camera was not defective. The Camera Operator had a brain fart and screwed up.

So. the Canon EOS 5D Mark IV is not defective. If you selected the camera and wish to return it because there is another camera out there that can do a 5 stop push and give results to your satisfaction, then you selected the wrong camera here... sell this one and go get the one that can give you the results you want in that situation.

A camera is not defective if it only does 5 fps, or has 11 stops of dr, or can't do a panoramic picture. It means you got the wrong tool for the job.

Saying this again - the Canon EOS 5D Mark IV is not defective.

Go get the camera that fits your needs.

The problem in my (returned) 5D4 could be seen at -3 EV which I believe is a valid selection since Canon supports it (by back dial, etc). Also If you shoot blacks camera meters for zone V and by setting exposure to -3EV you put them in zone II. This is not unreasonable. The post processing (+3 EV) that exhibits the issue maybe a little debatable but there are cases when this is necessary (I have mentioned in previous cycle of this thread so I do not want to repeat). For these cases (plus for some landscapes with no prefect lighting) I manage with my 5D3 using either ISO 50 or I use Magic Lantern's dual iso feature These are no ideal ways so I would like to use 5D4 instead. Again not unreasonable...

 

[JukkaS]

Hi all!

This picture shows the problem pretty well. This was shot 5 steps under as RAW (ISO 100), then to LR and boosted exposure 5 steps back + shadows up. This is of course extreme, but the picture taken the bright side on the right did not show these "stripes" at all - actually amazing results without these stripes. There is no noice reduction used.

Well, I'm amazed this has gone to 17-18 pages.
Jukka, thank you for the image that explains what people are seeing so well.

In all honesty, I'd take anyones 'defective' 5DIV for half price if an owner is so upset about the 'problem'.
PM me... please!

Back in the day, when we all shot film (I know some of you can't remember that, but it's true, we all did!) if someone underexposed 5 stops and then tried to get something out of that negative, and the result came out horrible.. well, we blamed the operator of the camera. With the latitude of film, usually you could still get something.. but no one expected the results of a 5 stop under exposure to be 'optimum' and usually were grateful we got anything at all.

The film was not defective. The camera was not defective. The Camera Operator had a brain fart and screwed up.

So. the Canon EOS 5D Mark IV is not defective. If you selected the camera and wish to return it because there is another camera out there that can do a 5 stop push and give results to your satisfaction, then you selected the wrong camera here... sell this one and go get the one that can give you the results you want in that situation.

A camera is not defective if it only does 5 fps, or has 11 stops of dr, or can't do a panoramic picture. It means you got the wrong tool for the job.

Saying this again - the Canon EOS 5D Mark IV is not defective.

Go get the camera that fits your needs.

Please read the thread and check the original video from Youtube. The point in putting this to extreme is just to highlight the issue. As I say in my other post, a guy managed to get stripes in the sky with his 5D4 just by pushing the exposure 2 stops - nothing else was needed. Actually the problem shows without any boosting with a good monitor.

 

[Alex_M]

It has been mentioned before that the phenomenon potentially could be seen in the corners of normally exposed and corrected for vignetting image taken with new Canon 16-35 2.8 L III lens that exibits strong corner vignetting at levels as much as -4EV. I have no concrete evidence to support this statement but if thats the case, then the argument that the phenomenon could be reproduced only in rear cases involving extreme image manipulation seems to falling apart.
Personally, I believe that Canon will be vocal about this issue once the cause was identified and solution was developed both at technical and business levels. But what do I know, anyway...

 

[unfocused]

I underexposed a scene by five stops and then did a 100 percent shadow lift and I found this strange pattern in the picture. I think I should send the camera back to Canon, but now, wherever I go to take pictures, I find large crowds following me and praying.

Should I return the camera to Canon or should I open up a small chapel in my living room. What do you guys think?

 

[tron]

Open a small chapel, accept donations and when enough go get a 1DxMkII ;D ;D ;D

 

[lkunl]

I also found the problem when i take too much under picture and bring up exposure later in post.
(You can see purple strip when + exp just 1-2 stop but i + 5ev for easy to see it.)

pic 1 ISO 1600
pic 2 ISO 1600 +5 ev purple strip at bottom
pic 3 ISO 3200
pic 4 ISO 3200 +5 ev no strip

I hope this problem is just some raw compression problem or software problem not sensor problem. :'(

 

[JimGemini]

I also found the problem when i take too much under picture and bring up exposure later in post.
(You can see purple strip when + exp just 1-2 stop but i + 5ev for easy to see it.)

pic 1 ISO 1600
pic 2 ISO 1600 +5 ev purple strip at bottom
pic 3 ISO 3200
pic 4 ISO 3200 +5 ev no strip

I hope this problem is just some raw compression problem or software problem not sensor problem. :'(

The C300 Mark II apparently had a similar 'luminous banding'/'CMOS smearing' issue that was resolved or reduced via firmware. Looks to be most likely a sensor issue. The interesting thing is that there are those that can reproduce this issue and those that cannot. If inherit to the sensor then I would think anybody would be able to reproduce this. Hopefully it's not a sensor 'defect' and can be resolved via firmware. I too can reproduce this issue.
http://www.learn.usa.canon.com/resources/articles/2016/c300mkii-firmware-update.shtml

 

[lkunl]

The C300 Mark II apparently had a similar 'luminous banding'/'CMOS smearing' issue that was resolved or reduced via firmware. Looks to be most likely a sensor issue. The interesting thing is that there are those that can reproduce this issue and those that cannot. If inherit to the sensor then I would think anybody would be able to reproduce this. Hopefully it's not a sensor 'defect' and can be resolved via firmware. I too can reproduce this issue.
http://www.learn.usa.canon.com/resources/articles/2016/c300mkii-firmware-update.shtml

Hopefully it can resolve via firmware. Thank for your reply.

 

[ilumo]

for those that have received a body back from repair, I would be interested in hearing what the "fix" was.

 

[romanr74]

for those that have received a body back from repair, I would be interested in hearing what the "fix" was.

would be interesting indeed...

 

[Duckman]

I cant believe this has gone on so long! Is this the same guy under 5 names trolling?!?
I personally am very pleased with what I see in the exposure latitude the 5Div provides. Now, I didn't read every single post so perhaps I missed something but, in my opinion, it seems the only real problem is not taking a proper exposure.
-J

 

[Ozarker]

Can't believe anyone exposes like in the examples and would expect anything else.

 

[Deleted member 7828]

FWIW - took the 5D4 on a shoot of the fall colors, ended up doing some underexposing/pushing to preserve the blue in the sky - clean as a whistle! Amazing really. I've yet to see this issue occur in any of my real-world shoots.

This "phenomenon" may not actually be unique to the 5D4 - one YouTube review of the C300 II video camera from Canon also demonstrated similar streaking when filming high DR scenes (person in dark room in front of window).

 

[Mikehit]

I also found the problem when i take too much under picture and bring up exposure later in post.
(You can see purple strip when + exp just 1-2 stop but i + 5ev for easy to see it.)

pic 1 ISO 1600
pic 2 ISO 1600 +5 ev purple strip at bottom
pic 3 ISO 3200
pic 4 ISO 3200 +5 ev no strip

I hope this problem is just some raw compression problem or software problem not sensor problem. :'(

With a scene like that why would you want to burn out the background so much? Or even lift the shadows that much?

 

[Jopa]

I'm wondering if whoever did this test first has a job and/or a family? The person must be bored to death to start doing this kind of stuff.

 

[lkunl]

I also found the problem when i take too much under picture and bring up exposure later in post.
(You can see purple strip when + exp just 1-2 stop but i + 5ev for easy to see it.)

pic 1 ISO 1600
pic 2 ISO 1600 +5 ev purple strip at bottom
pic 3 ISO 3200
pic 4 ISO 3200 +5 ev no strip

I hope this problem is just some raw compression problem or software problem not sensor problem. :'(

With a scene like that why would you want to burn out the background so much? Or even lift the shadows that much?

I just pick up some of my image that can show the problem. I actually don't use the image.
It will be a problem when extream contrast sittuation like night scence with some bright lighting landmark.
When you have to under exposure to keep some hilight unclipping. I can see it with under exposure imeage even no exposure adjustment in post (especially night scene) .

 

[Alex_M]

I bag to differ. This has nothing to do with person's maritual, job status, gender, age or sexual preferences.
Some people like to push their boundaries quite a bit and we call them explorers. Thanks for your understanding

I'm wondering if whoever did this test first has a job and/or a family? The person must be bored to death to start doing this kind of stuff.

 

[Orangutan]

I bag to differ. This has nothing to do with person's maritual, job status, gender, age or sexual preferences.
Some people like to push their boundaries quite a bit and we call them explorers. Thanks for your understanding

I'm wondering if whoever did this test first has a job and/or a family? The person must be bored to death to start doing this kind of stuff.

I'd reserve the word "explore" for projects that have a legitimate potential for useful results. This is no more "exploration' than trying to dig deep a hole in one's back yard is archaeology.