5D4 Sensor Defect Discovered

[TimoV]

I start on Monday 8)

 

[AlanF]

sound like a plan! right.. where do we start? 8)

I have a new idea. Everyone who finds this "defect" should immediately return their cameras and ask for a new one. When that one has the same "defect" do it again and again and again. This will build up stock at the Canon refurbished store and enable more photographers who aren't as worried about lens cap photography to buy the 5D IV at a bargain price.

That should make many people happy.

Sending back cameras by the shoot-your-lens-cap-and-push-by-five-stops brigade pushes up the price for the rest of us as the stores and manufacturers strive to keep up their profit margins.

 

[JukkaS]

I start on Monday 8)

Just that you are not planning to get one back Timo . I would join if there would be possibility to have a 5D4 without stripes. Even though I like stripes in some clothing, I would prefer to make the decision of wearing those myself - not the camera.

I thank all of you have contributed to this topic - some even without owning the 5D4. I will now use my limited time differently.

 

[Loswr]

I used to be a happy fool with my 6D. If the picture of the girl in the former pages exposed by the highlights (from DPReview) + Timos Shopping mall frame is everything but normal, I get your point. And if you are not able to see a problem in those frames, I recommend a visit to doctor. It is obvious that Canon does not admit this to be a problem, they just cannot.

You mean the shopping mall frame which I pushed +4 EV and +100 shadows, and posted my results showing essentially no banding? If you can see a serious problem in the images I posted, you may have been inadvertently exposed to hallucinogenic substances.

What would you like Canon to admit? That when you apply a high gain to a negligible signal, you get artifacts? I suppose after that, they should admit that the sun rises in the east, and then hire Captain Obvious to help them 'admit' all the other stuff everyone knows.

I thank all of you have contributed to this topic - some even without owning the 5D4. I will now use my limited time differently.

Thanks, that's your most sensible contribution to this topic so far.

 

[TimoV]

neuroanatomist, you still continue looking that photo by iPhone?

Canon service in Finland admit to see the streaks by +1 boosting and many other also.

I suggest you to go and buy proper monitor and stop arguing with your iPhone results, that makes many people laugh to you only

 

[Orangutan]

Canon service in Finland admit to see the streaks by +1 boosting and many other also.

Please post a link to that statement.

 

[TimoV]

I have proof in my email but I don’t place it here to public, not so pissed off about this things… I show this later at evening to one person here in forum that I know and he may comment it to proof.

 

[JukkaS]

I have proof in my email but I don’t place it here to public, not so pissed off about this things… I show this later at evening to one person here in forum that I know and he may comment it to proof.

I will stretch myself to comment this . Thanks Timo for your email, in which Authoritized Canon service clearly says that they can see the stripes. Then Canon officially do not admit this to be nothing but a limitation (which proves that they see the stripes as well). I a way feels funny to convince somebody about this. I can see it very easily with less than 1 stop increase (no shadows), with even better monitor no lift needed at all to see it. Just do not use the softening in the software!

I have similar shots of a clear night sky, where there is streetlights far away, and same thing shows. This camera is not that capable to perform in those conditions - sadly - 6D at least is/was better. Timo here is a very talented shooter and selling his pics as well. If he to his main purposes gets only stripes in the sky for paying 3 times more for a camera - cannot blaim him for choosing to rather continue with 6D. To somebody else it can be possible to survive with this "feature" - I am still considering if or not. Anyhow not happy.

The spirit and commenting in this thread is something I do not like. The original starter got tired (a pro shooter) of this and me as well. There are wonderful members who really do their best to help. The main idea has all the time been a) to understand if there really is 5D4´s without this phenomena and b) to have more proof towards Canon that we as users do not think this is ok. Too many here are just shooting down opinions instead of photos and the whole thing goes OT compared to the original idea. Especially when too many do not own the camera. With more than 200 frames showing this it feels different. Referring to all these comments about lens cap shooting etc. I have plenty of real life photos which are ruined - before with 6D I still did recover those to some extent at least. In many aspects the camera performs incredibly well, too expensive though to have this problem.

Again thanks for all you who take this as a mutual thing to deal with - that should be the power of social media - not the sick behaviour we see here too much.

I have spoken.

 

[Alex_M]

Was not the issue resolved in firmware version 1.02? there are multiple reports and sample photos posted here by forum members that demonstrate that issue was resolved. Setting emotional canvas aside, are you still seeing streaks following firmware 1.02 upgrade? Thanks!

Again thanks for all you who take this as a mutual thing to deal with - that should be the power of social media - not the sick behaviour we see here too much.

I have spoken.

 

[PixelTrawler]

Based on my two units I wouldnt say 1.02 has solved it. The second body is better. But its still there.
Only one of those bodies had 1.01 to see the change. The change was drastic on that unit. Massive.

Im impressed with the second unit, the limit is there but it seems be a lot of flexibility in it.
It still odd that stripes occur at all.

I think what we are seeing is that 1.02 is better but there is unit to unit variation. So some owner suffer more than others. But Id be hopeful Canon are on top of it. I think Canon are learing to randomise the read noise somehow but each sensor is slighty different.

 

[PixelTrawler]

I have proof in my email but I don’t place it here to public, not so pissed off about this things… I show this later at evening to one person here in forum that I know and he may comment it to proof.

I will stretch myself to comment this . Thanks Timo for your email, in which Authoritized Canon service clearly says that they can see the stripes. Then Canon officially do not admit this to be nothing but a limitation (which proves that they see the stripes as well). I a way feels funny to convince somebody about this. I can see it very easily with less than 1 stop increase (no shadows), with even better monitor no lift needed at all to see it. Just do not use the softening in the software!

I have similar shots of a clear night sky, where there is streetlights far away, and same thing shows. This camera is not that capable to perform in those conditions - sadly - 6D at least is/was better. Timo here is a very talented shooter and selling his pics as well. If he to his main purposes gets only stripes in the sky for paying 3 times more for a camera - cannot blaim him for choosing to rather continue with 6D. To somebody else it can be possible to survive with this "feature" - I am still considering if or not. Anyhow not happy.

The spirit and commenting in this thread is something I do not like. The original starter got tired (a pro shooter) of this and me as well. There are wonderful members who really do their best to help. The main idea has all the time been a) to understand if there really is 5D4´s without this phenomena and b) to have more proof towards Canon that we as users do not think this is ok. Too many here are just shooting down opinions instead of photos and the whole thing goes OT compared to the original idea. Especially when too many do not own the camera. With more than 200 frames showing this it feels different. Referring to all these comments about lens cap shooting etc. I have plenty of real life photos which are ruined - before with 6D I still did recover those to some extent at least. In many aspects the camera performs incredibly well, too expensive though to have this problem.

Again thanks for all you who take this as a mutual thing to deal with - that should be the power of social media - not the sick behaviour we see here too much.

I have spoken.

Well spoken. Thanks to everyone who is helping to understand how our new sensors perform.

To the stripe denier posters, enjoy your good units, you have a wonderful camera. ive no doubt some sensors have very little trace of stripes (monitor calibration could also be a factor in seeing it). But the feeling that we should go away and stop moaning where we do is not a nice aspect to this forum.

People didnt spend 4 grand to buy a camera and then make up problems. They are not imagining. Expectations were very high for this camera driven by Canon themselves. People are allowed to be disappointed if their real world usage is impacted.

 

[Mikehit]

What steps are being taken to make sure that the shadows being pushed by different people are the same EV? This alone would explain why some see artefacts with 1-stop push and others only see it with a 5 stop push.

 

[TimoV]

What steps are being taken to make sure that the shadows being pushed by different people are the same EV? This alone would explain why some see artefacts with 1-stop push and others only see it with a 5 stop push.

…another who don’t have monitor to see streaks properly and don’t even own 5Dmk4.

Take a look back to page 30 where I send link to RAW and comments after that ;-)

 

[Mikehit]

What steps are being taken to make sure that the shadows being pushed by different people are the same EV? This alone would explain why some see artefacts with 1-stop push and others only see it with a 5 stop push.

…another who don’t have monitor to see streaks properly and don’t even own 5Dmk4.

Take a look back to page 30 where I send link to RAW and comments after that ;-)

Hear that sound? it is the point flying over your head.

What has my owning the 5D4 got to do with my question?

 

[TimoV]

What steps are being taken to make sure that the shadows being pushed by different people are the same EV? This alone would explain why some see artefacts with 1-stop push and others only see it with a 5 stop push.

…another who don’t have monitor to see streaks properly and don’t even own 5Dmk4.

Take a look back to page 30 where I send link to RAW and comments after that ;-)

Hear that sound? it is the point flying over your head.

What has my owning the 5D4 got to do with my question?

I don't know your point here but your comments are very useless in this thread, value zero or below that.

 

[Mikehit]

I don't know how I can make it any clearer so I will repeat and then offer an interpretation

What steps are being taken to make sure that the shadows being pushed by different people are the same EV? This alone would explain why some see artefacts with 1-stop push and others only see it with a 5 stop push

You take a picture at -5 exposure compensation.
Another person take a photo at -5 exposure compensation.

If you take a picture of the same outdoor scene, one on sunny day and one on a bright day, the shadows may well have different EV.
Your shadow of the supermarket has RGB values of 1.5%, 1.8% and 4.5%.
If someone else's shadow has 3%, 3.5% and 7% that is almost twice as 'bright' and given information in a digital image is logarithmic, significantly more information and a greater capacity for pushing.

So my questions is what is being done to standardise between the different claims.

Your posting a single image does nothing to answer this - zip, zilch, nada - to answer my question. So before you get all sarcastic and downright insulting I suggest you take time to make sure you understand the question first. Your posting the raw file is commended. Your attitude to someone who dares to challenge you is not.

Oh, and please point to my other 'useless' comments in this thread.

 

[Loswr]

neuroanatomist, you still continue looking that photo by iPhone?

Canon service in Finland admit to see the streaks by +1 boosting and many other also.

I suggest you to go and buy proper monitor and stop arguing with your iPhone results, that makes many people laugh to you only

I suggest you go back and read more carefully my statement that the only place I saw even a hint of this horrible defect you are complaining about was on my iPhone. I didn't see any banding in my hard push of your shopping mall shot on any of my properly calibrated displays.

Since Canon Finland admit to seeing this banding (maybe they're using iPhones, wouldn't that be a laugh), please do let us know their response: will they be replacing the sensor in your defective 5DIV, or will they simply replace your defective 5DIV with a new unit? Or did they just tell you it's not defective?

Captain Obvious says, "You can't create something from nothing: if you amplify a negligible signal, it's not going to turn out well."

 

[tron]

I don't know how I can make it any clearer so I will repeat and then offer an interpretation

What steps are being taken to make sure that the shadows being pushed by different people are the same EV? This alone would explain why some see artefacts with 1-stop push and others only see it with a 5 stop push

You take a picture at -5 exposure compensation.
Another person take a photo at -5 exposure compensation.

If you take a picture of the same outdoor scene, one on sunny day and one on a bright day, the shadows may well have different EV.
Your shadow of the supermarket has RGB values of 1.5%, 1.8% and 4.5%.
If someone else's shadow has 3%, 3.5% and 7% that is almost twice as 'bright' and given information in a digital image is logarithmic, significantly more information and a greater capacity for pushing.

So my questions is what is being done to standardise between the different claims.

Your posting a single image does nothing to answer this - zip, zilch, nada - to answer my question. So before you get all sarcastic and downright insulting I suggest you take time to make sure you understand the question first. Your posting the raw file is commended. Your attitude to someone who dares to challenge you is not.

Oh, and please point to my other 'useless' comments in this thread.

I admit you have a point and I was thinking something similar myself. We could possible make a step towards standarization by: metering strictly from the dark object and then underexposing. Recordiing shutter, aperture (I guess we leave iso at 100) we can translate to EV. But I saw color bands with not extreme settings. The first time it was a 100-400 on 100mm on a tripod which means aperture set to 4.5 or 5.6 (My new camera didn't have an arca swiss compatible plate yet so I used a telephoto with plate). The second time I thought to handhold my 5D4 and I used the 35mm 1.4L. The fact that I could handhold it implies not very low ev values although underexposing increases shutter. I can update this when I have the actual values.

 

[JukkaS]

I don't know how I can make it any clearer so I will repeat and then offer an interpretation

What steps are being taken to make sure that the shadows being pushed by different people are the same EV? This alone would explain why some see artefacts with 1-stop push and others only see it with a 5 stop push

You take a picture at -5 exposure compensation.
Another person take a photo at -5 exposure compensation.

If you take a picture of the same outdoor scene, one on sunny day and one on a bright day, the shadows may well have different EV.
Your shadow of the supermarket has RGB values of 1.5%, 1.8% and 4.5%.
If someone else's shadow has 3%, 3.5% and 7% that is almost twice as 'bright' and given information in a digital image is logarithmic, significantly more information and a greater capacity for pushing.

So my questions is what is being done to standardise between the different claims.

Your posting a single image does nothing to answer this - zip, zilch, nada - to answer my question. So before you get all sarcastic and downright insulting I suggest you take time to make sure you understand the question first. Your posting the raw file is commended. Your attitude to someone who dares to challenge you is not.

Oh, and please point to my other 'useless' comments in this thread.

I don´t want anymore to comment other aspects than related to photography and gear. This is an very important end educational point you make here - thanks for that. Too easily we talk only about shadows without actual values. With most of the monitors the darks seem quite the same even though the variation. What is according to your extensive knowledge a good "rule of thumb" that is needed for RGB-channels for the shadows to be lifted without expecting big problems? In my case 5D4 sometimes do the stripes even with all the channels around 2,5% or even a bit higher (high contrast needed then). Strongly below that, not to mention about near 0% it is what we heard Captain Obvious already telling - the artifacts show more the closer we get to zero .

 

[Orangutan]

So my questions is what is being done to standardise between the different claims.

I'll add two thoughts to the question. First, a camera is not just a sensor. The point is to capture two identical scenes so that they look the same in raw/neutral. It doesn't matter if the pixel values are identical unless you're interested in an engineering test of the sensor. To test overall camera performance it's enough to determine whether you can capture a particular scene with a particular camera to create an acceptable image. Second, there is a way to do a rough calibration, I think. I'd think about introducing a small reference light into the frame, e.g. a weak LED bulb at a distance. I'd mask the scene so only the LED shows through, then use the histogram spike to calibrate exposure on both cameras. That will get you close.