5D4 Sensor Defect Discovered

[Loswr]

God forbid anyone discover another weird anomaly with this camera!

Start your car's engine and put the transmission in neutral, then floor the gas pedal for an hour or so. When it overheats and/or seizes, log onto a forum for your car make and post about how your car is defective. I'm sure your opinion will be warmly received.

 

[Ozarker]

Always the same lately. One writes something he/she encountered which points out something. That someting could be interpreted as a negative point on the Mark4 or Canon by some and very soon the defenders are there and the whole bashing starts. And goes on and on and on.
It sometimes seems or reads, that most here in the forum are either payed by canon or go to bed with canon or work at canon. And Canon in general is the best thing since sliced bread and no critique, no pointing out is in order. Almost as if pointing out or putting out a thesis is immediately a general war against canon and the fans are here to the rescue. Who gives really. If this is the getting the norm than improvements are out of the window.
If Dave Dugdale can post a video what he wishes the next Sony iteration should have, then I don't see a lot of people complaining that this is not true and the latest Sony is perfect fault free and there is nothing, really nothing to improve. This here in the forums is utter bull as of late. Mark IV is not perfect as is every other camera on the market. Putting the finger on things can help to improve on that. Praising doesn't help. Ever.

A-a-and you miss the point.
This is not about fanboy at all.

Thank you. I appreciate the support and hate being bashed all the time.

 

[Refurb7]

The purpose of this demonstration was to show the problem easier for readers to see, not to say this is what I do and what you should do. The problem exists at any push and is only exaggerated at this level to show it more clearly. I saw these same issues pushed 3 stops and 4 stops, albeit to a lesser extent.

I appreciate your responses, how off topic and diluting it may be to the thread. It's a shame nobody could explain why this sensor does this when other sensors don't...not that anyone here should be doing this in their everyday photography, it just shows a peculiar flaw that the sensor exhibits that others don't. But, I suppose it's hard to get any answers when the banter becomes a mockery of the topic.

Good day,
Kevin

What real world results could this possibly affect if you have to go to such extreme extremes to show it? You're doing stuff to the image that boggles the mind: "The details on what generated this phenomena: I underexposed by 5 stops and then recovered the RAW file 5 stops and pushed the shadows 100% to intensify the ability to see the problem."

That's just so unrealistic. I have never in my photo career thought to do that (or anything near that) to any photo for any reason. Never.

No one can explain it because no one can take it seriously. With that approach, I can make any camera appear "defective". Just do something far beyond anything it is designed to do, and announce the discovery of a defect.

You said in the first post, "problem doesn't seem to appear when you're 3-stops under." Now you're contradicting that, saying that it does appear 3 stops under. If it does appear at 3 stops, why not show that?

Forget far-fetched mind-numbing tests. Show real-world non-test photos made normally where there is an actual problem, and then people will look more seriously. Otherwise this sounds like a classic case of FUD.

 

[rrcphoto]

OK, I've learned my lesson with this group. They get offended by anything that doesn't talk about the awesomeness of Canon...

Please take note that at no point did I try to backpedal, many of you just simply couldn't read and even admitted to stopping because of the context. God forbid anyone discover another weird anomaly with this camera!

Meanwhile, I've been using my 5DIV daily and love it. Already shot a race weekend, 5 automotive magazine features, a portrait session, a journalism story, a home, and a promotional poster.

the hyperbole in your original post kind of put a target on your back.

 

[Loswr]

Unfivestopliftable. You keep using that word, I do not think it means what you think it means.

Stop correcting others. Nobody cares what you think you know.

Difficult to disagree...

Always the same lately. One writes something he/she encountered which points out something. That someting could be interpreted as a negative point on the Mark4 or Canon by some and very soon the defenders are there and the whole bashing starts. And goes on and on and on.
It sometimes seems or reads, that most here in the forum are either payed by canon or go to bed with canon or work at canon. And Canon in general is the best thing since sliced bread and no critique, no pointing out is in order. Almost as if pointing out or putting out a thesis is immediately a general war against canon and the fans are here to the rescue. Who gives really. If this is the getting the norm than improvements are out of the window.
If Dave Dugdale can post a video what he wishes the next Sony iteration should have, then I don't see a lot of people complaining that this is not true and the latest Sony is perfect fault free and there is nothing, really nothing to improve. This here in the forums is utter bull as of late. Mark IV is not perfect as is every other camera on the market. Putting the finger on things can help to improve on that. Praising doesn't help. Ever.

Oh, the sot has spoken. What happens to the 5DIV is not truly your concern.

I rest my case.

I hear that brandy can help people to lighten up.

 

[jakdaniel1975]

I bought the 5D mk4 a few days and also I have noticed of the sensor defect in some scenes, that can happen not only when searching recoveries 5 stop but also with much less !! Thanks to the odious policy of canon who declined to give the spot metering linked to the af points, reportage where you need it to be fast and use the av mode shooting backlit can happen often in the room measuring errors and find yourself with shots like what you attaching ... that if you want to recover it has an obvious flaw, which is not only the banding but much more!

https://www.dropbox.com/s/ftcnduq69wx8stu/_H9A3340.CR2?dl=0

excuse my language but use google translated.

 

[ivanku]

So, what I've noticed is, if a shot is underexposed by one stop or 1.5, then you can bring the exposure back and also lift shadows without seeing any banding.

However, if the entire shot is underexposed by two stops or more, and to get to a proper exposure histogram you need to add two or more stops to the exposure, you will start to see very distinct, irregular horizontal banding.

is this what you guys are seeing as well?

I'd love to hear more experiences from other 5D IV users

 

[LSXPhotog]

So, what I've noticed is, if a shot is underexposed by one stop or 1.5, then you can bring the exposure back and also lift shadows without seeing any banding.

However, if the entire shot is underexposed by two stops or more, and to get to a proper exposure histogram you need to add two or more stops to the exposure, you will start to see very distinct, irregular horizontal banding.

is this what you guys are seeing as well?

I'd love to hear more experiences from other 5D IV users

You apparently didn't get the memo: the 5D Mark IV is the best camera ever with no flaws at all and if you find an issue with the sensor you are a bad photographer.

The issue isn't simple banding because it doesn't follow a pattern that banding does. However, it is influenced by the brightness of the left side of the frame. I'm seeing the issue 3 stops under but not with 2 stops under. I feel it's a sensor issue, but clearly by this thread I'm wrong. The 5D IV is the greatest thing ever.

 

[Mikehit]

So, what I've noticed is, if a shot is underexposed by one stop or 1.5, then you can bring the exposure back and also lift shadows without seeing any banding.

However, if the entire shot is underexposed by two stops or more, and to get to a proper exposure histogram you need to add two or more stops to the exposure, you will start to see very distinct, irregular horizontal banding.

is this what you guys are seeing as well?

I'd love to hear more experiences from other 5D IV users

This guy doesn't seem to have a problem with a 4-stop push

http://www.birdphotographers.net/forums/showthread.php/137014-5D-Mark-IV-Shot-from-Today-(With-4-stop-push)

 

[ivanku]

This successful and clean 4 stop push may indicate that a) either the original photograph was not underexposed more than 2 stops, or b) this phenomenon does not affect all 5D sensors

 

[LSXPhotog]

This successful and clean 4 stop push may indicate that a) either the original photograph was not underexposed more than 2 stops, or b) this phenomenon does not affect all 5D sensors

I'm trying to find out if all sensors have the issue. Because apparently, according to what I've seen around the web, it doesn't....which is why I started this thread.

 

[jakdaniel1975]

But it does not always occur, but only in units with strong contrast and controluce.In these cases is always presented with my camera.

 

[Labdoc]

I read this thread and almost got a case of dystopia myself. Maybe I am missing the point but I took this picture the other morning at dawn and maybe I'm doing it wrong but I don't see anything. The first pic is obviously underexposed by a lot (1/60, ISO 100, f5.0 , 35mm), taken with the 5Dmk4, 16-35 MK3. The second pic was processed with LR and pushed 5 stops, highlights 100%, shadows 100%, no banding.

P.S. If this is a real issue maybe my camera is not affected.

 

[gowiththeflow]

I know this thread has come full circle a couple of times. But I'll add this comparison as it might appear in another profession/hobby:

**My GE Oven Has a Defect**

I cooked a soufflé in my oven. The correct cooking time should have been 20 minutes. I intentionally undercooked it by 5 stops. (20 minutes divided by 2. Divided by 2. Divided by 2...) for a total of 37.5 seconds. The soufflé was very undercooked, just as I had expected. I then opened the very undercooked soufflé in my microwave and increased the cooking time until it was no longer undercooked. It didn't turn out the same as if I had correctly cooked it in the first place. Can someone please explain, in this day and age, why GE can't make a better oven???

 

[raptor3x]

Please ask those who have the mk4 to do the same backlit tests to see if it's my faulty sensor or are all so!

Sorry i use translate.

DPReview did up to +6EV pushes and didn't see anything like what you're seeing. I would guess that your particular camera is indeed faulty.

 

[3kramd5]

I know this thread has come full circle a couple of times. But I'll add this comparison as it might appear in another profession/hobby:

**My GE Oven Has a Defect**

I cooked a soufflé in my oven. The correct cooking time should have been 20 minutes. I intentionally undercooked it by 5 stops. (20 minutes divided by 2. Divided by 2. Divided by 2...) for a total of 37.5 seconds. The soufflé was very undercooked, just as I had expected. I then opened the very undercooked soufflé in my microwave and increased the cooking time until it was no longer undercooked. It didn't turn out the same as if I had correctly cooked it in the first place. Can someone please explain, in this day and age, why GE can't make a better oven???

haha! That one got me.

 

[Sporgon]

You are seeing the issue because you are trying to lighten under exposed black, that is black, dark, can't see anything, coal hole - you know, it was black and you under exposed it.

I downloaded the CR2 file from dropbox of the graduation. I had to lift 5 stops and then I could just about make out some banding. To get to the same brightness and banding as the other examples that have been shown I had to then lift shadows by about 50%, so its probably a 7 stop lift. What if there is some banding when you have done this ? It's making the information up anyway; it didn't actually record anything !

I'd say there is nothing wrong with any of these cameras. For the record here's what happens when you lift the exmor sensor by 5 stops and then do a 50% lift on the shadows slider. It's not pretty. Do you think that bank really had magenta and green bands running through it ?

 

[ivanku]

Thank you guys for posting more sample photos!

@gowiththeflow, your analogy is funny and makes sense if this is simply the limitation of the sensor. However, what we are trying to figure out in this thread is whether this phenomenon is true for all new 5D sensors, or is a defect or variance, and occurs in only some of the new cameras. I think there's a big difference between those conclusions, right?

 

[bwud]

I'd say there is nothing wrong with any of these cameras. For the record here's what happens when you lift the exmor sensor by 5 stops and then do a 50% lift on the shadows slider. It's not pretty. Do you think that bank really had magenta and green bands running through it ?

I'm not exactly sure what Exmor sensors have to do with this particular discussion, but that doesn't match my experience. Rather, I get flat flat flat images. They aren't banded or disturbingly noisy like your example, they're just flat.

Perhaps your scene was initially even darker than mine.

 

[Sporgon]

Yes, much darker, much later in the day, very thin light etc, but also my subject was black, or at least very dark colour. Mine is also 100% crop which I think yours is too.