5D IV underexposure banding

which Canon forums?
If you watch the Video, he's pretty specific

--
My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)
Yes, I started a thread on this, didn't realize this was being talked about or I wouldn't have.... I knew the issue was being talked about but not the fact that he has taken it upon himself to get enough people to provide evidence, along with his evidence, in order to get Canon to officially respond.

If we can get enough people.... and it wouldn't take many, Canon would respond and probably issue a recall or fix it however they fix it.

I would go to his video, go to his website and post your findings.

I know Michael is a reliable guy, this is not some trick to get you to go to his website, Canon did take his post down. He has a big enough audience that he can actually get Canon to officially respond if enough people will go there.
 
After seeing how this banding seems to present adjacent to highlight areas in a frame, I wonder if it it some sort of streaking from the dual pixel sensor architecture?

Any thoughts?
 
So ... is the "problem" only visible in some cameras or all are the same?

This is the only question I have left at this point! :)
Mine does it too. I do not personally see it as a "problem". How is 9 stop push even remotely related to any type of photography, I wonder. I would like to see this so call test done with any other camera😛. Today's softwares are able to push the RAW data to levels never seen before and I dare to say, way beyond any practical use.
 
  • Like
Reactions: EG
I really appreciate everyone talking about this topic, quoting my thread on Canon Rumors, and the assorted responses within this discussion topic.

Try as you might to ignore this, write it off, say it's a problem with the "grey matter behind the camera", it still doesn't change the fact the camera seems to have a weird issue in high contrast situations where there is a bright exposure on the left side of the horizontal frame. No other camera I have ever seen does this, nor has Michael the Maven seen in his very intensive testing.

How will this impact you?

I just went through over 10,339 frames I've captured with the Canon 5D Mark IV since last month. This was a race weekend, 6 magazine photoshoots, product photography, a wedding, a portrait session, a hockey game, 2 documentary photojournalism assignments, headshots, etc....I have used the snot out of this camera in 30 days as a working professional.

With that said, I just spent lots of time trying to find files I've shot with the camera that may exhibit the issues. Of the 10,339 frames, maybe only 10 real-world photos have the issue to the point where it impacted the photo...in a lab setting trying to create the issue? I can create it 100% of the time. LOL I can't explain it. I just know that if you have something bright in the left side of the frame and underexpose by as little as 2 or 3 stops...you're done for. Haha

So, on a positive note, this means it shouldn't have any real impact on your photography. I've probably done more variety of work in the last 30 days than some 5D owners will the entire time they own the camera.

Remember when the Nikon D750 had it's weird issue that people simply wrote off? Well this, I feel, is the same situation. It won't impact every photo you take or every kind of image, but it will most certainly be an unwanted guest in the ocasional image file you maybe need to push beyond your normal 2-stops.

These threads have taken over too much time out of my life, so I'm going to say I won't be spending much time on the subject anymore unless something new gets discovered with it. My camera is being sent into CPS next week since I should have a light week and will use the 5D3.

Let's see what Canon has to say about this and does in the future.

- Kevin
 
Well wait a second... Shouldn't it have showed up in the 5-stop push that I posted here:
https://www.dpreview.com/forums/post/58562830

That is a 5 stop push, with over exposure in the left of the frame
Never mind, I do see it. It's just a larger area slightly less defined banding.
I think you're onto something. It must have to do with the sensor read out in my opinion.
 
Well wait a second... Shouldn't it have showed up in the 5-stop push that I posted here:
https://www.dpreview.com/forums/post/58562830

That is a 5 stop push, with over exposure in the left of the frame
Never mind, I do see it. It's just a larger area slightly less defined banding.
I think you're onto something. It must have to do with the sensor read out in my opinion.
I don't see the issue at all. You would see a distinct highlight streak from the bright portion of the frame.





Underexposed 4 stop image pushed 3.5 stops and highlights 100%. Pushing the Highlights makes the problem more noticeable.
Underexposed 4 stop image pushed 3.5 stops and highlights 100%. Pushing the Highlights makes the problem more noticeable.
 
Well wait a second... Shouldn't it have showed up in the 5-stop push that I posted here:
https://www.dpreview.com/forums/post/58562830

That is a 5 stop push, with over exposure in the left of the frame
Never mind, I do see it. It's just a larger area slightly less defined banding.
I think you're onto something. It must have to do with the sensor read out in my opinion.
I don't see the issue at all. You would see a distinct highlight streak from the bright portion of the frame.

Underexposed 4 stop image pushed 3.5 stops and highlights 100%. Pushing the Highlights makes the problem more noticeable.
Underexposed 4 stop image pushed 3.5 stops and highlights 100%. Pushing the Highlights makes the problem more noticeable.
Is this the same problem outlined in a video by a guy named Michael where 1/3 of the frame is overexposed and the rest underexposed?

--
Once you've done fifty, everything else is iffy.
 
Worrying about most of the issues brought up in this discussion is analogous to the owner of a Ferrari 488 GTB worrying because if he attaches his sports car to a 53 foot semi trailer with a 70,000 pound load and runs it in first gear for an hour the engine overheats.
Exactly. I read this whole thread, and still picked up a 5D MIV yesterday. I traded in my A7RII to get it.

I must be the anti-Christ

--
Some favourite pics:
http://garyp.zenfolio.com/p518883873/
 
Last edited:
Well wait a second... Shouldn't it have showed up in the 5-stop push that I posted here:
https://www.dpreview.com/forums/post/58562830

That is a 5 stop push, with over exposure in the left of the frame
Yes, your camera doesn't appear to have the problem and you should be really happy. Not every camera has this problem.
You need to move the Shadow slider to 100% on top of 5 EV push. Than, I am pretty sure, it will show the same "banding" like everybody elses 5DIV😉
 
I lose sleep over whether Trump will win the election, whether Britain will go for Brexit, when we defeat ISIS, whether we can depose Assad or Putin will take over the whole Ukraine..

I do not lose sleep over whether, when I push my exposure slider all the way to the right and push my shadow slider all the way to the right, I will see some banding.

I have seen Sebastao Salgado's photographs enlarged to take up a wall. On older DSLR cameras. I fear little in that regard.
 
Well wait a second... Shouldn't it have showed up in the 5-stop push that I posted here:
https://www.dpreview.com/forums/post/58562830

That is a 5 stop push, with over exposure in the left of the frame
Never mind, I do see it. It's just a larger area slightly less defined banding.
I think you're onto something. It must have to do with the sensor read out in my opinion.
I don't see the issue at all. You would see a distinct highlight streak from the bright portion of the frame.

Underexposed 4 stop image pushed 3.5 stops and highlights 100%. Pushing the Highlights makes the problem more noticeable.
Underexposed 4 stop image pushed 3.5 stops and highlights 100%. Pushing the Highlights makes the problem more noticeable.
Is this the same problem outlined in a video by a guy named Michael where 1/3 of the frame is overexposed and the rest underexposed?

--
Once you've done fifty, everything else is iffy.
Yes, Michael and I were in contact discussing the issue. If you have a large exposure variance on the left portion of the frame and then bring up the exposure, this is what happens. In as little as a 2-stop push.
 
Exactly. I read this whole thread, and still picked up a 5D MIV yesterday. I traded in my A7RII to get it.

I must be the anti-Christ

--
Some favourite pics:
http://garyp.zenfolio.com/p518883873/
Not really.

(...With a strong benefit to your actual photography...) you're just out-of-touch on how every single Canon camera is treated by DPR forums crowds, these days.

1. Pre-launch: it won't possibly be able to match whichever other brand/model is cherry-picked for a particular capability.

2. Launch: It will be "meh!" to the extreme.

3. After-launch: The body won't even be able surviving being tossed in a washing machine and the files will be demonstrated not being able to withstand a digital treatment akin to rolling over film with a bulldozer and then some ice skates.

PK

--
“Loose praise may feed my ego but constructive criticism advances my skills”
************************************************************
-------------------------------------------------
http://www.humbertoborgesfotografia.com/
http://www.pbase.com/photokhan
(PBase Supporter)
-------------------------------------------------
 
Last edited:
...by heavily compressing levels, shows some more lines.

I don't think that any camera wouldn't show something, if a file was brutally treated like this one was.
Shooting a flat, solid color at -X stops is not the same thing as shooting a dynamic scene at -X stops, because the scene has shadows stops below metered gray.
 
PhotoKhan wrote

(...With a strong benefit to your actual photography...) you're just out-of-touch on how every single Canon camera is treated by DPR forums crowds, these days.
Ha ha. Not so much out of touch as out of caring . Particularly now I have tried one of the most touted alternatives.

As the thread I linked to above suggests, you can do it with any camera, even Nikon, if you torture test enough.

This may really affect some people. Others are certainly working pretty hard to construct a mountain out of that molehill.

--
Some favourite pics:
http://garyp.zenfolio.com/p518883873/
 
Last edited:
One thing that doesn't seem to be mentioned is that we're assuming all sensors are created equal. They aren't.

There's sample variation across sensors, same as with cpu's. Some can be clocked higher than others. Some will fall over sooner when pushed.

Just because people are pushing files 5 stop and then brutalising the shadows doesn't mean it's a practise endorsed by the manufacturers and I can't remember any official line from any manufacturer ever saying this sort of sensor abuse is a feature.

I personally think that people testing stuff beyond it's recommended usage until it breaks then calling it a design flaw / issue / cause for complaint is moronic.
 
Last edited:

Keyboard shortcuts

Back
Top