DR and RAW bit depth on Canon's forthcoming...

[jrista]

I didn't say that I wouldn't want more DR, quite frankly, any photographer would always be looking for improvements, whatever they are. However, the type of photography I do would probably benefit more from other improvements, such as improved ISO sensitivity (i.e. cleaner). I can already compensate mostly for shortcomings in DR, but noise at higher ISO isn't as easy and if I want to shoot wildlife around dusk, that is a priority for me.

You realize though, that the reason Canon's ISO settings are not that clean is because of their read logic, right? And its also because of their substandard (by todays standards) read logic that they chop off 2.15 stops of DR at ISO 100? The problem is the way they read data off their sensor. If they can fix that properly (i.e. not take some kind of cheap shortcut like they have a couple times in the past), they can fix it all...better high ISO, better low ISO, better dynamic range, etc.

Also, some were offering film as having more DR than Canon's sensors, which certainly isn't the case for slide film and is debatable for negative film. If we want improvements, then we have to be realistic in what we compare it too.

The point about film is not that it has better total dynamic range than a modern digital sensor (although some of the best films do offer a LOT of DR.) The point is that film has much better dynamic range characteristics...with film, its far more difficult to hit a hard cutoff in the highlights than it is with digital sensors. If you know you've over exposed a roll of film or a slide, you can compensate during development with adjusted time or different chemicals, and you have a LOT of freedom to recover highlights. When recovered, you also run a lesser risk of encountering funky color shifts that are so common when recovering highlights with RAW digital images (particularly with Canon, when recovering highlights, yellows and lighter oranges tend to dull and get absorbed into reds, costing you color fidelity.)

 

[LetTheRightLensIn]

I've said before, that lab tests are pretty meaningless, unless they reflect the real world, after all, as photographers, we are judged by what we achieve, not by what could be theoretically achieved.

But they do! Have you seen what a heavily pulled D7000/D3x ISO 100 shots looks like compared to one from 50D or 5D2?

While I've seen the D3, I haven't seen results from those cameras, but have any real comparisons been made? I'm talking about side by side comparisons in the real world, so that as many variables as possible can be ruled out. We can't judge comparisons, if we don't know the conditions at the time of taking. Simple comparisons of unrelated images is simply unscientific and less meaningful than lab tests under artificial conditions (where you can of course control things at least). I'm not saying that there aren't differences, but it's very easy to jump to conclusions. It's very rare, that reviews make direct side by side comparisons, taking near identical shots. Even when you do make comparisons though, Nikon and Canon cameras seem to have different compromises, aiming for more latitude at opposite ends of the spectrum. I've seen too many cases when I was doing research, of papers that put a spin on something that wasn't actually true, when looked at in detail or repeated.

Flake mentioned the use of HDR and grad filters. As a landscape photographer, I know that even with an extra stop of DR, there is no way that I could capture most scenes without using grad filters, unless I make some compromises. In fact, photography is all about making compromises, that is what sets the greats apart from the merely good.

But once you are into nearly 3 stops of usably better DR then it does start to make much more a difference in what things you can capture or not.

Three stops is obviously a different matter, but that is a huge jump, in excess of 30%, so is it realistic?

It's not much use having an extra stop of DR, if your meduim of choice can't show it. I suspect that you'd need an expensive Spectraview type monitor to see any differences.

Not at all. In fact most of them are IPS and don't even have very huge contrast ratio compared to some cheap SPVA screens. But it makes a big difference either way.

And seriously if Canon had 3 usable stops better DR like it did back when, compared to Nikon, would all of you still be finding every excuse in the world why we don't need it and who cares??

Why would anyone, most of all a Canon user, NOT want Canon to fix this up???
I care because I have Canon equipment.

I didn't say that I wouldn't want more DR, quite frankly, any photographer would always be looking for improvements, whatever they are. However, the type of photography I do would probably benefit more from other improvements, such as improved ISO sensitivity (i.e. cleaner). I can already compensate mostly for shortcomings in DR, but noise at higher ISO isn't as easy and if I want to shoot wildlife around dusk, that is a priority for me. Also, some were offering film as having more DR than Canon's sensors, which certainly isn't the case for slide film and is debatable for negative film. If we want improvements, then we have to be realistic in what we compare it too.
[/quote]

I have seen a couple pretty impressive side by side comparisons and it was a very real and impressive difference, almost shocking.

Canon cameras are much closer to the SNR middle gray theoretical limits than they are to low ISO DR range limits so you can't be expecting 2-3 stops better SNR middle gray but you could hope for 2-3 stops better usable low ISO DR.

 

[jrista]

Have you ever manipulated a low ISO D3x or D700 file and then gone an manipulated one from Canon? The difference is every bit as real as DxO says it is.

Yes I have. No it is not. Shoot a transmission step wedge with just about any body and compare your results with DxO. Then you'll understand why I laugh at their DR claims.

You just don't get it.

Wow, he really doesn't. I think we've reached the point where further argument is rather moot...its like talking to that harsh digital highlight cutoff...it ain't listening or moving.

 

[torger]

A problem with DxoMark dynamic range measurements is that they do not show "usable DR". 14 stops measured DR just means that 14 stops down signal is equal to noise, that is how "engineering DR" is measured. But in a real photograph, such noisy information must be black, it is unusable. In fact, those parts of the image that are bright enough to show detail must have pretty good signal-to-noise ratio to look good. So perhaps only 7-8 out of those 14 is usable.

Another problem is that Dxomark does not measure pattern noise, random noise is much more acceptable to the eye than if you see patterns. And this is a weak spot with Canon. So a sensor with random noise with the same DR measurement as say Canon 5D mark II which have some pattern noise will have more dynamic range in practice.

One may argue that Canon sensors are already good enough and that DR problems never occur in normal images. Many photographers have that experience. But that the sensors are actually worse than the competition is easy to prove, just open a raw file from a Canon 7D and from a Nikon D7000 shot at base ISO and push both files 3-4 stops, and watch the noise in the dark parts.

Oh, remember that color negative film, if scanned with HDR on a high quality drum scanner (practically non-existent these days unfortunately), can deliver about 18 stops of dynamic range. There's still improvements to make in digital sensors... People *think* film has poor dynamic range, because they scan it with lousy flatbed scanners, or are used to old analog print workflows when you couldn't really expand the film S-curve. Also slide film while less DR than color negative do have quite good DR, assuming high end scanning equipment.

 

[jrista]

A problem with DxoMark dynamic range measurements is that they do not show "usable DR". 14 stops measured DR just means that 14 stops down signal is equal to noise, that is how "engineering DR" is measured. But in a real photograph, such noisy information must be black, it is unusable. In fact, those parts of the image that are bright enough to show detail must have pretty good signal-to-noise ratio to look good. So perhaps only 7-8 out of those 14 is usable.

Again, I direct everyone to the following example:

http://testcams.com/blog/2011/05/03/nikon-dx-vs-canon-aps-c-dynamic-range/

Based on what you've said, all the "black" in a photograph is unusable because its SNR is too low, and therefor the information is unusable. I think the videos above prove differently. Unusable "blacks" is largely true with Canon sensors, not true in the least with Sony sensors. I think the videos speak for themselves...modern sensors, including canons, offer far more DYNAMIC range than 7-8 stops. Keep in mind, DR and contrast are not the same thing. A final photo may exhibit 7-8 stops of CONTRAST, the measurement of the darkest parts of the scene vs. the lightest parts...but the range of contrast exists within the full dynamic range a sensor is capable of. The range of contrast may start lower and end higher, or start higher and end lower. You may have low-key images that only use the darker range of tones offered by a sensor, or high-key images that only use the lighter range of tones offered by a sensor. You may have something that uses mostly the middle tones. Its called dynamic range for a reason, and its not the same as contrast.

Sensors do indeed provide 12-14 stops of dynamic range, and that full range is usable...either to recover a botched exposure, deal with available light issues, etc. The END RESULT of mucking around with a RAW images exposure in post, shifting highlights and shadows...is a photo that probably has around 7-8 stops of tonal range. Print it, and your probably down to 5-6 stops. The end result does not change the nature of the sensor, though...you still have that original 12-14 stops of DR to work with when taking the shot and working it in post.

 

[torger]

I think the videos speak for themselves...modern sensors, including canons, offer far more DYNAMIC range than 7-8 stops.

May be true, I haven't actually done any subjective tests of how many stops the files actually contains, I have just compared different cameras to eachother and see I sometimes get problems when I push shadows etc. The actual number is not too important, the relative differences are though. What the camera can record in a scene which then can be compressed into a print without visible noise.

The raw files are 14 bit, meaning that 14 stops down the signal is either 0 or 1. At 13 stops down, the range is 0 - 4. Quite coarse quantization. I would be suprised if those last stops are usuable.

The videos are interesting, but I would certainly not use lightroom as some base to check how many stops (the actual number) there are in there, since you can't really trust that program's linearity, so I'd take the "13 or 14 stops of useful dynamic range" with a grain of salt. How much visible noise and dead colors you accept is also subjective, and it also dependent on the picture what works.

 

[Marsu42]

Most DSLR sensors are linear capture so there is no such thing as "at the top of the histogram" and the fact is you can't get out of a 7D image (and yes I have one) wht you can from a D7000 at ISO100.

I won't start an argument over Nikon vs. Canon, but since you say you know about these things: From what I see, a Canon raw image with "highlight tone priority" is not linear - or am I mistaken?

 

[torger]

I won't start an argument over Nikon vs. Canon, but since you say you know about these things: From what I see, a Canon raw image with "highlight tone priority" is not linear - or am I mistaken?

The raw file content is always linear, since the sensor is linear. How Canon DPP interprets a RAW file with highlight tone priority may be a different story. You need a "dumb" raw processor like dcraw or similar to actually be sure that there's no non-linear trickery added to the RAW data in the conversion.

 

[Kernuak]

You should take a look at the videos on this page:

http://testcams.com/blog/2011/05/03/nikon-dx-vs-canon-aps-c-dynamic-range/

A number of things come to mind when watching those videos. He concentrated almost entirely on shadow areas, with very little comment on the highlight areas. DR isn't just about shadow detail, but also highlight detail. Also, it was as much about processing as actual output. Also, they aren't exactly real world images (after all, I don't photograph many beds, with light shining thorugh the window). It was quite disconcerting the way he kept swapping sides, with Canon on the left at the start, then Nikon and back to Canon. Finally, although he didn't show the unprocessed images in detail, the Canon camera was around 1/3rd to 2/3rd stop underexposed in comparison to the Nikon, by my eyes, which will increase noise anyway. No camera (even the same model) will have a guaranteed exposure level for the same ISO, aperture and shutterspeed, due to normal variation. For example, my 7D underexposes by 2/3rd of a stop compared to my 5D MkII, if the same settings are used. To make a true comparison, you would need to compensate for all variations and examine both the highlights and shadows without processing, before trying to extract as much detail as possible. so that you would have before and after results. Without that, testing isn't valid.
It's all pretty moot anyway though, as until anyone sees the results from the 1D X (and the new 5D), nobody can say whether or not the Sony sensors extract more DR, as at the moment, the comparisons are between different sensor generations, with 2+ years difference in release. A more relevant comparison would be between the 5d MkII and D700 or the 7D and D300s. Even the 1D MkIV is now over two years old.

 

[Marsu42]

I won't start an argument over Nikon vs. Canon, but since you say you know about these things: From what I see, a Canon raw image with "highlight tone priority" is not linear - or am I mistaken?

The raw file content is always linear, since the sensor is linear. How Canon DPP interprets a RAW file with highlight tone priority may be a different story. You need a "dumb" raw processor like dcraw or similar to actually be sure that there's no non-linear trickery added to the RAW data in the conversion.

... fyi: adobe raw and the in-camera preview (which you know of course) interpret the htp raw file just like dpp, but I never tried dcraw.

Which brings me again to my good ol' question about how Canon htp mode works at all. I can see that it raises the iso level, expands the top highlights and in return compresses the shadows - but concerning dr, I don't know how much it's really worth, because it certainly isn't a "mini hdr mode".

 

[torger]

the comparisons are between different sensor generations, with 2+ years difference in release. A more relevant comparison would be between the 5d MkII and D700 or the 7D and D300s. Even the 1D MkIV is now over two years old.

The new G1X sensor which is fairly large, does not look very promising in that regard, but perhaps they've made it deliberatly worse than the 7D sensor to not compete within their own lines.

There's also the thing that Canon simply don't have the Sony Exmor technology. The big difference is that the ADC is on chip. Canon does not have that. Sony Exmor is a minor revolution (for low ISO). What you can expect from Canon is a gradual evolution. You'll certainly see "good enough" performance, but I'll (almost) eat my shorts if it will be world-leading.

 

[jrista]

You should take a look at the videos on this page:

http://testcams.com/blog/2011/05/03/nikon-dx-vs-canon-aps-c-dynamic-range/

A number of things come to mind when watching those videos. He concentrated almost entirely on shadow areas, with very little comment on the highlight areas. DR isn't just about shadow detail, but also highlight detail. Also, it was as much about processing as actual output. Also, they aren't exactly real world images (after all, I don't photograph many beds, with light shining thorugh the window). It was quite disconcerting the way he kept swapping sides, with Canon on the left at the start, then Nikon and back to Canon. Finally, although he didn't show the unprocessed images in detail, the Canon camera was around 1/3rd to 2/3rd stop underexposed in comparison to the Nikon, by my eyes, which will increase noise anyway. No camera (even the same model) will have a guaranteed exposure level for the same ISO, aperture and shutterspeed, due to normal variation. For example, my 7D underexposes by 2/3rd of a stop compared to my 5D MkII, if the same settings are used. To make a true comparison, you would need to compensate for all variations and examine both the highlights and shadows without processing, before trying to extract as much detail as possible. so that you would have before and after results. Without that, testing isn't valid.

Very true about the need to normalize exposure, and account for differences in samples. I was not trying to use the videos at that link as a way of objectively measuring DR...its just a way to demonstrate the value if having more DR. Even if the 7D exposures were upped by 2/3rds of a stop, its not going to change the horrible read noise, nor the effect it has on the samples the demonstrator was working with. Yes, it was a contrived example, however it was contrived in order to demonstrate. I encounter problems with Canon's read noise without the need to produce a contrived example. In particular, the color noise on any current Canon body is terrible, and what he mentioned about color fidelity is true even when you expose correctly and that proper exposure has dark shadows. Color noise shows through regardless, and you get that unsightly blotchieness all over.

It's all pretty moot anyway though, as until anyone sees the results from the 1D X (and the new 5D), nobody can say whether or not the Sony sensors extract more DR, as at the moment, the comparisons are between different sensor generations, with 2+ years difference in release. A more relevant comparison would be between the 5d MkII and D700 or the 7D and D300s. Even the 1D MkIV is now over two years old.

As far as I know, the first camera with a Sony sensor with improved read noise was released less than a year after the 7D...I'll have to dig around to see if I can find some specifics. Its not just the D7000 sensor that blows Canon sensors out of the water...its Sony sensors released over the last couple years that do. And its very true...we'll have to wait and see how the 1D X and new 5D sensors fare. They may do superbly well, and if they do, I'm a happy camper. If not, well, thats the reason for this thread.

 

[Aglet]

A problem with DxoMark dynamic range measurements is that they do not show "usable DR". 14 stops measured DR just means that 14 stops down signal is equal to noise, that is how "engineering DR" is measured. But in a real photograph, such noisy information must be black, it is unusable. In fact, those parts of the image that are bright enough to show detail must have pretty good signal-to-noise ratio to look good. So perhaps only 7-8 out of those 14 is usable.

Again, I direct everyone to the following example:

http://testcams.com/blog/2011/05/03/nikon-dx-vs-canon-aps-c-dynamic-range/

Based on what you've said, all the "black" in a photograph is unusable because its SNR is too low, and therefor the information is unusable. I think the videos above prove differently. Unusable "blacks" is largely true with Canon sensors, not true in the least with Sony sensors. I think the videos speak for themselves...modern sensors, including canons, offer far more DYNAMIC range than 7-8 stops. Keep in mind, DR and contrast are not the same thing. A final photo may exhibit 7-8 stops of CONTRAST, the measurement of the darkest parts of the scene vs. the lightest parts...but the range of contrast exists within the full dynamic range a sensor is capable of. The range of contrast may start lower and end higher, or start higher and end lower. You may have low-key images that only use the darker range of tones offered by a sensor, or high-key images that only use the lighter range of tones offered by a sensor. You may have something that uses mostly the middle tones. Its called dynamic range for a reason, and its not the same as contrast.

Sensors do indeed provide 12-14 stops of dynamic range, and that full range is usable...either to recover a botched exposure, deal with available light issues, etc. The END RESULT of mucking around with a RAW images exposure in post, shifting highlights and shadows...is a photo that probably has around 7-8 stops of tonal range. Print it, and your probably down to 5-6 stops. The end result does not change the nature of the sensor, though...you still have that original 12-14 stops of DR to work with when taking the shot and working it in post.

Time for me to jump in on this topic. I generally watch from the wings of non-registered users on this forum but this DR topic is my pet peeve with recent Canon bodies too!

I'm an electronics professional by day, love reading all the tech details from various sites but when it comes down to it, let the images speak for themselves.

I don't completely agree with the usability of DXO's sensors tests; they are a good guideline but they seem to not account for the horrendous pattern noise of Canon's recent sensors at low ISO on some bodies.

Also, if you look at the DR curve vs ISO on Canons sensors from way back until now you'll see they're essentially flat before they drop off into the expected slope at higher ISO. The ISO that this occurs has been increasing with newer generations of their sensors, let's call that "improvement."

I was stunned when I saw the curves for Nikon's D7000/5100. THAT, to me, is how that curve should look; you trade sensitivity for DR in a very linear fashion. Canon's curves are not like that, roughly indicating there's a noise floor problem somewhere in their system. That Canon's raw files exhibit more pattern noise at low ISO than at high ISO is baffling but I'll leave analysis of this to sites like clarkvision and others who can delve into details I don't have available

On the practical side...
All the fuss about HDR as a technique should be nearly moot for most scenes if we have a truly decent 14 bit capture. My first Canon body was the 40D. I deliberately held out from going into DSLR until there was more than 12 bits worth of data to work with. The live-view was a massive bonus.

I've done some shots with the 40D where I expose for the highlites and push the shadows in post for an effectively HDR, but realistic-looking shot, based on one properly exposed raw file. They can look great!

With all the hoopla over the 5D2, I was downright unimpressed when I got mine. I had low ISO banding right in the midtones!
In fact, I think my 5D2 and 7D are both turkeys when it comes to all the banding noise in the shadows at low ISO, something my older bodies do not exhibit.
Latest firmware and upgrades to DPP seem to have somewhat mitigated the severity of the problem but when it comes down to it, it seems like I can push the shadows from my 40D a bit more than I can with the 5D2 or 7D.
In fact, I was just thinking about doing a comparison of sorts amongst all the Canon bodies I have access to for exactly this sort of test. I often like to boost shadows in post. The 5D2 and 7D raw files don't often leave me as much to work with as I'd hoped.

I plan to take some shots of "dark" at low ISO and moderate shutter speeds, the sort I'd use in landscape work. I'll then see how far I can push these files in post before the banding noise becomes apparent.
I have access to the following bodies for this test:
350D, 40D, 60D, 7D, 5D2, G11, G12.

It'll take me a bit of time to get this done, got a few projects in the way at the moment but I'll post the results when I do.

 

[nightbreath]

There's also the thing that Canon simply don't have the Sony Exmor technology. The big difference is that the ADC is on chip. Canon does not have that. Sony Exmor is a minor revolution (for low ISO). What you can expect from Canon is a gradual evolution. You'll certainly see "good enough" performance, but I'll (almost) eat my shorts if it will be world-leading.

The 16 outputs are muxed in 4 ADCs siting on a separate image processor chip Digic 5+

Taken from Image Sensors World. Isn't that the same as "Exmor technology" you are talking about?

 

[LetTheRightLensIn]

Most DSLR sensors are linear capture so there is no such thing as "at the top of the histogram" and the fact is you can't get out of a 7D image (and yes I have one) wht you can from a D7000 at ISO100.

I won't start an argument over Nikon vs. Canon, but since you say you know about these things: From what I see, a Canon raw image with "highlight tone priority" is not linear - or am I mistaken?

What the sensor is grabbing is still linear, regardless and that is what counts. I'm not sure but I think the RAW is still linear anyway. All it doing is underexposing by one stop but applying a different mid-tone point and different roll-off tone curve for the embedded JPG and hints to the RAW processor to do the same. It's not altering the sensor read in any sort of non-linear way. It's not really any sort of special mode at the hardware level.

 

[LetTheRightLensIn]

I don't completely agree with the usability of DXO's sensors tests; they are a good guideline but they seem to not account for the horrendous pattern noise of Canon's recent sensors at low ISO on some bodies.

True and that is why I use both the terms "engineering DR" and "usable DR". When talk "usable DR" the Canon bodies fair worse compared to the Sony exmor sensors than when talking "engineering DR". The 5D2 and 50D lose a noticeable amount of usable performance due to banding. The 1Ds3 and 40D do not lose as much of their engineering DR when comparing it to their usable DR. SOme of the SONY exmor sensors leave even far less.

And it also why all the whining about banding is not just so much nonsense. I've seen people beg for more DR and then whack down all those whining about banding as foolish pixel peepers. I've seen people going on about how some Canon model has 1/3 stop better DR and yet bash down any talk that is has more banding (which ends up making maybe 2/3-1 stop less usable). That doesn't make any sense.

I was stunned when I saw the curves for Nikon's D7000/5100. THAT, to me, is how that curve should look; you trade sensitivity for DR in a very linear fashion. Canon's curves are not like that, roughly indicating there's a noise floor problem somewhere in their system. That Canon's raw files exhibit more pattern noise at low ISO than at high ISO is baffling but I'll leave analysis of this to sites like clarkvision and others who can delve into details I don't have available

On the practical side...
All the fuss about HDR as a technique should be nearly moot for most scenes if we have a truly decent 14 bit capture. My first Canon body was the 40D. I deliberately held out from going into DSLR until there was more than 12 bits worth of data to work with. The live-view was a massive bonus.

I've done some shots with the 40D where I expose for the highlites and push the shadows in post for an effectively HDR, but realistic-looking shot, based on one properly exposed raw file. They can look great!

With all the hoopla over the 5D2, I was downright unimpressed when I got mine. I had low ISO banding right in the midtones!
In fact, I think my 5D2 and 7D are both turkeys when it comes to all the banding noise in the shadows at low ISO, something my older bodies do not exhibit.

Yes, the 50D and 5D2 generation was one of the worst generations for low iso banding while the 40D and 1Ds3 generation had been one of Canons best generations in terms of that (banding). The 7D is sort of in between when it comes to offset banding but added a new issue, gain banding (which can appear as vertical lines, to various degree in different parts of the frame, and which I also found varied considerably in degree from copy to copy, no Canon sensor I've seen has ever varied so much copy to copy in any way as 7D copies due regarding vertical gain banding, granted my sample sizes are modest, less than a dozen), which can also account for the odd kink in the SNR curve seen on some 7D SNR tests. The 7D also uses an odd dual-green color filter array. The initial RAW converters were not expecting this and the 7D files originally suffered from weird looking mazing artifacts. They were later able to fix this while doing something very ingenious and somehow trading away only a tiny bit of resolution and micro-contrast.

Latest firmware and upgrades to DPP seem to have somewhat mitigated the severity of the problem but when it comes down to it, it seems like I can push the shadows from my 40D a bit more than I can with the 5D2 or 7D.
In fact, I was just thinking about doing a comparison of sorts amongst all the Canon bodies I have access to for exactly this sort of test. I often like to boost shadows in post. The 5D2 and 7D raw files don't often leave me as much to work with as I'd hoped.

Very true as soon as I got my 5D2 I noticed it compared to my 40D. As you say the more recent converters do mask it better than the early versions though. Also the 5D2 starts out with more engineering DR than the 40D the so the end result isn't quite as bad relatively speaking as it would be had they started out with the same engineering DR.

I plan to take some shots of "dark" at low ISO and moderate shutter speeds, the sort I'd use in landscape work. I'll then see how far I can push these files in post before the banding noise becomes apparent.
I have access to the following bodies for this test:
350D, 40D, 60D, 7D, 5D2, G11, G12.

It'll take me a bit of time to get this done, got a few projects in the way at the moment but I'll post the results when I do.

Some tests with 1D3,1Ds3,5D2,20D,7D,40D,50D were done a few years ago and as I said above the 1Ds3/1D3/40D generation had usable DR closest to engineering DR and the 50D/5D2 the farthest apart.

Some say the 60D has less banding of both types than the 7D.

Only the Sony Exmor sensors have that really super clean read noise. I hope the new 1DX/5DX can at least get close, we will see. I see some signs that hint that they will only close the gap by half, if at all, but hopefully they will close it by 2/3 or even all the way. Hopefully it is possible to do it without patent violation.

 

[Kernuak]

I have to admit, that I haven't run into problems with pattern noise on either my 7D or 5D MKII, but then I probably don't push the ISO high enough and don't push it in PP either. I tend to keep the both at ISO 1600 or below, although I have pushed the 5D MkII to ISO 6400 on a couple of occasions. Also, when I do shoot at higher ISO, the contrast is lower, so the need for DR is reduced, plus I tend to overexpose slightly, shooting for the highlights, but exposing for the shadows, as much as possible. In addition, when I am in those conditions, I'm not looking for detail in the shadows, as it isn't part of the subject generally. Shooting style plays a large role in the amount of noise visible, as does individual cameras, as they are all different. Despite not seeing banding in my current cameras, I did see it in my old 40D at ISO 1600, which was pushing its limits. That was one thing that I liked about the 7D, as not only did it give cleaner images, the quality of the noise was different.

 

[LetTheRightLensIn]

I have to admit, that I haven't run into problems with pattern noise on either my 7D or 5D MKII, but then I probably don't push the ISO high enough and don't push it in PP either. I tend to keep the both at ISO 1600 or below, although I have pushed the 5D MkII to ISO 6400 on a couple of occasions. Also, when I do shoot at higher ISO, the contrast is lower, so the need for DR is reduced, plus I tend to overexpose slightly, shooting for the highlights, but exposing for the shadows, as much as possible. In addition, when I am in those conditions, I'm not looking for detail in the shadows, as it isn't part of the subject generally. Shooting style plays a large role in the amount of noise visible, as does individual cameras, as they are all different. Despite not seeing banding in my current cameras, I did see it in my old 40D at ISO 1600, which was pushing its limits. That was one thing that I liked about the 7D, as not only did it give cleaner images, the quality of the noise was different.

It's not just about high ISO though, if you need to maximize DR then it's an issue at ISO100 too.
7D tends to have less high ISO banding than the 5D2 or 50D.

 

[Kernuak]

I have to admit, that I haven't run into problems with pattern noise on either my 7D or 5D MKII, but then I probably don't push the ISO high enough and don't push it in PP either. I tend to keep the both at ISO 1600 or below, although I have pushed the 5D MkII to ISO 6400 on a couple of occasions. Also, when I do shoot at higher ISO, the contrast is lower, so the need for DR is reduced, plus I tend to overexpose slightly, shooting for the highlights, but exposing for the shadows, as much as possible. In addition, when I am in those conditions, I'm not looking for detail in the shadows, as it isn't part of the subject generally. Shooting style plays a large role in the amount of noise visible, as does individual cameras, as they are all different. Despite not seeing banding in my current cameras, I did see it in my old 40D at ISO 1600, which was pushing its limits. That was one thing that I liked about the 7D, as not only did it give cleaner images, the quality of the noise was different.

It's not just about high ISO though, if you need to maximize DR then it's an issue at ISO100 too.
7D tends to have less high ISO banding than the 5D2 or 50D.

That was one of the things that was touted in reviews when the 7D came out and was one of the reasons, I went for it (I didn't have the 5D MkII at the time), having passed on the 50D as having worse noise characteristics than the 40D. I've never had problems with noise on the 5D MkII at ISO 100, but then I don't have the patience to spend hours processing an image, preferring to try to get it as right as possible in camera.

 

[jrista]

Well, four more days until the official announcement of whatever is replacing the 5D II! The 1D X announcement did not indicate anything about the DR that camera might be capable of, and only specified two additional stops of native high ISO (which, as it turns out, is only actually possible with JPEG, they won't even guarantee a full additional stop for RAW.) Here's to hoping they announce full specs, and at least offer some insight into dynamic range of the 5D II successor.

All I can really say is...Ima be ubermiffed if they don't fix their read noise banding (fixed pattern noise), and I'll hold onto my cash until the next camera announcement at end of year. I wonder how the speculative 40+mp Canon sensor will fare...and whether it will improve DR and eliminate read noise or not.