5D4 Sensor Defect Discovered
- Thread starter LSXPhotog
- Start date
the capacitors or whatever charge store they are using is being overloaded.
it could also be that LR is still not using floating black offset values, so this is "mis reporting" based upon not decoding the raw data fully.
since you can boost midtones in DPP well above +5EV .. has anyone actually tried it with DPP?
it could also be that LR is still not using floating black offset values, so this is "mis reporting" based upon not decoding the raw data fully.
since you can boost midtones in DPP well above +5EV .. has anyone actually tried it with DPP?
Upvote
0
Sporgon said:
Gotcha.
Yes, that was a 100% crop, and it was pushed beyond what was necessary - it wasn't nearly that bright in real life. I was just messing around to see what I could "recover" (not really the right word in this context). It's pretty silly; there is absolutely no depth when pushing such tiny signals. ND filtering or multiple exposures remain the appropriate tool.
In the final image, which I don't particularly like anyway, I reeled it in to realistic levels, and I suspect I could have done the same shooting with my 5D3.
Upvote
0
^
Yes indeed, but your exposure is spot on, you've pushed those highlights in the clouds to the limit. Now as you were shooting a Sony camera I can only assume that the reason you're still able to expose correctly is because you still shoot Canon too !
Yes indeed, but your exposure is spot on, you've pushed those highlights in the clouds to the limit. Now as you were shooting a Sony camera I can only assume that the reason you're still able to expose correctly is because you still shoot Canon too !
Upvote
0
There's some interesting stuff posted in the thread that a lot of people are simply being dismissive about. As I've previously posted in some of my other comments - just because you don't see the value in a 4 stop exposure push it doesn't mean that others don't either. People like Sam Hurd, Ryan Brenizer and Ross Harvey - who are some of the most in demand wedding photographers in the world at the moment, regularly push files 4 stops or more for specific techniques. So if you're saying it's wrong and there’s no value in it please also go and tell these guys. In some of the situations where they would be using a 4 stop push you could say the same or a similar effect could be achieved using off camera flash. However; this adds further complexity to the shot, increases setup time which means you could be missing moments and lugging lighting gear around reduces your mobility. Some of these techniques are about trying to make the job a little easier, which is always much appreciated when your 12 hours into a wedding!
As some of you have already said - pushing a file 5 stops and then pushing the shadows is extreme. Would you ever do that in the real world? No . . . probably never. However; LSXPhotog and jakdaniel1975 have said that they are doing this to make it easier to see the issue.
I came across this phenomenon when I was testing the sensor out on my mkiv to find out how far I can push the files. I know I can't push the mkiii files more than 2 stops so I shoot accordingly. From my initial tests I've found it looks like 4 stops is the limit on the mkiv. Is this enough for most real world situations . . . maybe. Would I like to be able to push the files more? Of course. The more malleability in the files the easier my job becomes. I'll say it again, this will not apply to everyone, you won't all see the value in these types of manipulation as it depends on what and how you shoot.
Ok so some of the wording in LSXPhotog original post is way off the mark but there's still some interesting stuff in the information that he's posted for those of us that are interested in finding the limits of the sensor.
What I find really interesting is jakdaniel1975 post which has this file attached . . .
https://www.dropbox.com/s/ftcnduq69wx8stu/_H9A3340.CR2?dl=0
I can see from this that his camera (in that particular situation) shows banding with a 3 stop push.
Mikehit posted a link which shows a clean 4 stop push
http://www.birdphotographers.net/forums/showthread.php/137014-5D-Mark-IV-Shot-from-Today-(With-4-stop-push)
Also Labdoc provided a file which doesn't appear to have the issue after a 5 stop push and a 100% shadow push . . . I'd be interested if Labdoc could provide the original RAW file so that I could have a play too.
So when comparing to LSXPhotog and jakdaniel1975 do the files in Mikehit's and Labdoc's links show that there is some variability in the sensors or is this due to something else such as the lighting conditions or the contrast in the scene or something else? From the information posted so far I'd say it's too early to tell.
Some people have raised the issue of if it's a Lightroom problem. I can confirm that it isn't I've tested this with my own files and also the one provided by jakdaniel1975 by loading them into DPP and I still see the issue. Others have said it could be that you're trying to read from a charge site with a very low or zero value. This could hold some merit but the banding doesn't always appear in the darkest parts of the frame. In jakdaniel1975’s file it looks like the banding is present or at least smeared into the lighter areas.
This thread is now quite long so some of you might have missed one of my earlier postings where I put a link to my review of the mkiv and compared exposure pushed files to the Nikon D750. What was apparent was the D750 files could be pushed further. Feel free to take a look if you wish. . .
http://simonbrettellphotography.co.uk/canon-5d-mk-iv-review/
My personal opinion (which is worth not much) is that the limits of the mk iv sensor are around a 4 stop push. It's not a defect, just the limits of what's capable. There may be some mkiv cameras that are a little better and some a little worse due to manufacturing variation. It may be that some particular scenes enhance the issue. We need to do a little more analysis to find out. Is a 4 stop push enough for most real world situations? Maybe . . . I’ll find out when I’ve got a few more weddings under my belt with the mkiv.
So if you've got a mkiv what would be interesting would be if you could take some 4 or 5 stop under exposed shots in a couple of different situations (high contrast, low contrast, backlit, frontlit etc) and post them here as RAW files so others can take a look.
As some of you have already said - pushing a file 5 stops and then pushing the shadows is extreme. Would you ever do that in the real world? No . . . probably never. However; LSXPhotog and jakdaniel1975 have said that they are doing this to make it easier to see the issue.
I came across this phenomenon when I was testing the sensor out on my mkiv to find out how far I can push the files. I know I can't push the mkiii files more than 2 stops so I shoot accordingly. From my initial tests I've found it looks like 4 stops is the limit on the mkiv. Is this enough for most real world situations . . . maybe. Would I like to be able to push the files more? Of course. The more malleability in the files the easier my job becomes. I'll say it again, this will not apply to everyone, you won't all see the value in these types of manipulation as it depends on what and how you shoot.
Ok so some of the wording in LSXPhotog original post is way off the mark but there's still some interesting stuff in the information that he's posted for those of us that are interested in finding the limits of the sensor.
What I find really interesting is jakdaniel1975 post which has this file attached . . .
https://www.dropbox.com/s/ftcnduq69wx8stu/_H9A3340.CR2?dl=0
I can see from this that his camera (in that particular situation) shows banding with a 3 stop push.
Mikehit posted a link which shows a clean 4 stop push
http://www.birdphotographers.net/forums/showthread.php/137014-5D-Mark-IV-Shot-from-Today-(With-4-stop-push)
Also Labdoc provided a file which doesn't appear to have the issue after a 5 stop push and a 100% shadow push . . . I'd be interested if Labdoc could provide the original RAW file so that I could have a play too.
So when comparing to LSXPhotog and jakdaniel1975 do the files in Mikehit's and Labdoc's links show that there is some variability in the sensors or is this due to something else such as the lighting conditions or the contrast in the scene or something else? From the information posted so far I'd say it's too early to tell.
Some people have raised the issue of if it's a Lightroom problem. I can confirm that it isn't I've tested this with my own files and also the one provided by jakdaniel1975 by loading them into DPP and I still see the issue. Others have said it could be that you're trying to read from a charge site with a very low or zero value. This could hold some merit but the banding doesn't always appear in the darkest parts of the frame. In jakdaniel1975’s file it looks like the banding is present or at least smeared into the lighter areas.
This thread is now quite long so some of you might have missed one of my earlier postings where I put a link to my review of the mkiv and compared exposure pushed files to the Nikon D750. What was apparent was the D750 files could be pushed further. Feel free to take a look if you wish. . .
http://simonbrettellphotography.co.uk/canon-5d-mk-iv-review/
My personal opinion (which is worth not much) is that the limits of the mk iv sensor are around a 4 stop push. It's not a defect, just the limits of what's capable. There may be some mkiv cameras that are a little better and some a little worse due to manufacturing variation. It may be that some particular scenes enhance the issue. We need to do a little more analysis to find out. Is a 4 stop push enough for most real world situations? Maybe . . . I’ll find out when I’ve got a few more weddings under my belt with the mkiv.
So if you've got a mkiv what would be interesting would be if you could take some 4 or 5 stop under exposed shots in a couple of different situations (high contrast, low contrast, backlit, frontlit etc) and post them here as RAW files so others can take a look.
Upvote
0
All sensors are a compromise and regardless of how sensors & processing have improved, some are better at dynamic range whilst others provide better color or better resolution. Sony market three different A7 cameras, if one was perfect for every situation they would not be making three. Canon make two different 5 series cameras, the 5D MKIV and the 5DS/R, again both do slightly different jobs.
The problem is no matter what engineers produce someone will want more or push the design to a point it was not designed to do. I would argue pushing any camera 5 stops is really asking for trouble and unrealistic and whilst sensors are not film no one would have pushed film 5 stops without expecting issues.
Cameras are giving anything from the equivalent 12-15 stops of dynamic range that's pretty incredible but pushing a camera to give effectively 17-18 stops is asking too much, just because you can do something doesn't mean its right or the product is defect.
The problem is no matter what engineers produce someone will want more or push the design to a point it was not designed to do. I would argue pushing any camera 5 stops is really asking for trouble and unrealistic and whilst sensors are not film no one would have pushed film 5 stops without expecting issues.
Cameras are giving anything from the equivalent 12-15 stops of dynamic range that's pretty incredible but pushing a camera to give effectively 17-18 stops is asking too much, just because you can do something doesn't mean its right or the product is defect.
Upvote
0
Going through loads of images I took with a rental 5D Mk IV last month, the images DO NOT show this behavior on any file when pushed while my personal 5D IV does. I have concluded there is a variance in sensors and mine suffers from an issue not all 5D IV sensors do.
Also:
de·fect1
ˈdēfekt/
noun
a shortcoming, imperfection, or lack.
Understanding the definition of the word helps.
Also:
de·fect1
ˈdēfekt/
noun
a shortcoming, imperfection, or lack.
Understanding the definition of the word helps.
Upvote
0
Well, I just got off the phone with Canon. The representative I was speaking to was pretty adamant that what I'm experiencing is not normal. We are going to be exchanging photo files this afternoon so they can look into it further, but the conclusion was I will probably need to exchange the camera body. I'll keep the thread posted, but it sounds like this indeed shouldn't be happening, at least not to this level.
Upvote
0
neuroanatomist said:
Dunno if OP changed the wording from the original post, but the current post talks about "my 5D".
Hence the post would have been perfectly right and spot on...
Upvote
0
romanr74 said:
The only changes made to the post were the order of the images and not the phrasing. It's the result of people getting upset and not actually reading the post before they comment... which people admitted to. LOL
Upvote
0
Since my abilities and competence as a photographer was questioned in this thread a few times, I'd just like to share some images I've captured with the 5DIV in the last 3 week. My experience with files I captures at a wedding this past weekend were positive, but images of high contrast/dynamic range are unacceptable in some limited cases... obviously not what you're looking for after dishing out $3499.
Upvote
0
Kevin, thanks for bringing this to everyone's attention. I've noticed this with my camera as well and I'm glad you started a thread for all of us to compare notes on a new camera with unknown problems and limitations.
It's not your responsibility to write a post that is perfectly worded and diplomatic. You didn't say anything racist, bigoted or offensive. But, you did bring to attention what may or may not be a defect in some of the new 5D's. As we get more feedback from 5D owners, we can get a better perspective on what's going on.
It's not your responsibility to write a post that is perfectly worded and diplomatic. You didn't say anything racist, bigoted or offensive. But, you did bring to attention what may or may not be a defect in some of the new 5D's. As we get more feedback from 5D owners, we can get a better perspective on what's going on.
Upvote
0
Similar threads
- Deleted member 399245
- EOS R