5D3 same max dynamic range as the 5D2???

[LetTheRightLensIn]

Assuming the side masking area can be used for the measurement then:

Same DR for 5D3 and 5D2 is what I get. Zero improvement for ISO 100 DR.
Wow I had really expected a big improvement. Canon still had a lot of room there to get better.

I think the D800 will quite a lot more maximum dynamic range unless they really messed up the D7000/D3x->D800 Exmor transition.

I just hope that it will turn out the side masking area on the 5D3 images is not valid for this purpose (the top masking area definitely appears to be invalid for this purpose with the 5D3, I just hope same can be said for the side, but it doesn't really look like that will be the case).


On a side note, the SNR might be 1/2 to 2/3rd stops better than the 5D2 though, but that is based off of using ACR which is a sketchy way to compare. High ISO banding appears to be cleaned up on the 5D3 compared to the 5D2, if so and if it is 2/3 better SNR than that is a pretty decent improvement on the high end, but I wonder if Canon has the tech to improve ISO100.

 

[XanuFoto]

How is the D800 selling? Have they have more preorders then the 5D mkIII?

 

[LetTheRightLensIn]

How is the D800 selling? Have they have more preorders then the 5D mkIII?

I have no clue.

 

[Astro]

How is the D800 selling? Have they have more preorders then the 5D mkIII?

I have no clue.

guessing from amazon.. the 5D MK3 was way ahead of the D800.

 

[LetTheRightLensIn]

Actually I did find a way to begin to compare the low ISO banding between the 5D2 and 5D3 and maybe has been improved.

So even though the measured dynamic range is the same , the usable amount might be very slightly to somewhat better than on the 5D2 since the deep shadows seem as if though they will not have as much nasty cross-hatching. I haven't quantified how much yet. And I really would need a full black frame from the 5D3 anyway.

The far edge on the 5D2 also has less fixed pattern banding so it might not be fair to compare the far left masked edge of the 5D3 to the center of the 5D2 frame. Doing that, the 5D3 appears to have noticeably less banding, but trying to compare just the edge it looks much closer. So too early to say for sure. I could see it having anywhere from just slightly (almost meaninglessly less) to considerably less.

 

[LetTheRightLensIn]

How is the D800 selling? Have they have more preorders then the 5D mkIII?

I have no clue.

guessing from amazon.. the 5D MK3 was way ahead of the D800.

Maybe, maybe not but that has nothing to do with how the sensor performs. What in the world does it have to do with this thread? Is this the which camera is selling better thread? Which camera base has the most loyal fans?

If more people liked a sprinter other than Usain Bolt at the last Olympics does that mean that Usain Bolt was not the fastest??? It doesn't matter who has the best, most loyal, or most fans, that is a different matter entirely.

And once again, no I am not some Nikon fanboy. I have never owned a Nikon camera of any sort in my entire life.
I've owned many Canons. That doesn't mean I need to lie and make up BS to make myself feel better about Canon.

It's kinda sad that the fanboys here are so extreme that they just smite anyone and go la-la-la-la I can't hear you rather than look at facts. But go ahead, be my guest if you feel happier for doing it.

 

[Astro]

It's kinda sad that the fanboys here are so extreme that they just smite anyone and go la-la-la-la I can't hear you rather than look at facts. But go ahead, be my guest if you feel happier for doing it.

maybe some here are just bored to see the all this guessing and wait until there are some facts worth to talk about?!

 

[LetTheRightLensIn]

It's kinda sad that the fanboys here are so extreme that they just smite anyone and go la-la-la-la I can't hear you rather than look at facts. But go ahead, be my guest if you feel happier for doing it.

maybe some here are just bored to see the all this guessing and wait until there are some facts worth to talk about?!

Fair enough, although this is getting to be more solid than guessing now.

Looking at it more, the side area appears to be valid, in which case my measurement will be valid. It's not like DxO will do anything different when they calculate DR they use the exact same procedure. I get pretty much same numbers for my 5D2 as they get for the 5D2. (keep in mind DxO pops up Print Plot which is normalized to 8MP and my figures above are not)

And this is the rumors part of the forum.

 

[LetTheRightLensIn]

It is a sad state of affairs when presented with direct evidence people just stick their head in the hand and shoot the messenger if the results don't mesh with their hopes, internal belief system, what have you. And not just in this forum.

I present a careful and reasonable calculation that doesn't tell the story that all the little fanboys had expected (or that I had expected myself) and smite, smite, smite the messenger. ;D It kind of makes me laugh, in this context, talking about just cameras. In more serious realms the same sort of behavior is kind of sad though.

If I had made up numbers and said it had 3 stops better dynamic range than the 5D2 and 6 stops better than the D800 I'd probably have gotten +30 and tons of praise for being creative with the numbers to make big bad awful 5D2 or Nikon/Sony look bad.

While it seems to measure better for high iso performance, nobody has measured anything hinting at any better low ISO dynamic range. Seems we improved only on one side of the range. (and if you look at all the Canon press they kept talking about it being better.... at mid and high isos. I was hoping they just forget to mention improved low iso, but i guess they didn't mention it because they didn't do much there this time. at least the high iso stuff should be better though.)

 

[3kramd5]

How do you calculate the DR performance of a sensor without information about the actual light characteristics of a scene? Isn't it an input/output game?

What's a masking area?

How can I compute the DR of my camera from a random image?

 

[Loswr]

It is a sad state of affairs when presented with direct evidence...

I present a careful and reasonable calculation...

Where, exactly, is this evidence you've presented? Looking above, I see none. Calculation? Aside from model numbers, the only numerical part of your statement is "zero improvement."

If you'd like present a well-reasoned case, backed up by actual evidence, fine. Else, it's just opinion and hot air. I don't shoot messengers who bring bad news, as long as that news is accurate. But an unsupported (and knowingly inflammatory, at that) conclusion deserves the bullet.

 

[unfocused]

It is a sad state of affairs when presented with direct evidence...

I present a careful and reasonable calculation...

Where, exactly, is this evidence you've presented? Looking above, I see none. Calculation? Aside from model numbers, the only numerical part of your statement is "zero improvement."

If you'd like present a well-reasoned case, backed up by actual evidence, fine. Else, it's just opinion and hot air. I don't shoot messengers who bring bad news, as long as that news is accurate. But an unsupported (and knowingly inflammatory, at that) conclusion deserves the bullet.

+1000000000

I know I'm not the most technical-minded person, but I have no 'effin clue what this is all about. Is the "side masking area" a real thing? Does the OP actually have the camera in-hand? Is this some theoretical calculation based on published specifications? And, before you accuse me of being a "fanboy" (whatever the heck that is supposed to be other than just a schoolyard taunt) I don't even care about the 5D III. It's not in my budget and I don't intend to buy one. I don't even care about full-frame for that matter. I just enjoy trying to learn some things, but this doesn't seem like anything but some goofy troll.

 

[LetTheRightLensIn]

It is a sad state of affairs when presented with direct evidence...

I present a careful and reasonable calculation...

Where, exactly, is this evidence you've presented? Looking above, I see none. Calculation? Aside from model numbers, the only numerical part of your statement is "zero improvement."

If you'd like present a well-reasoned case, backed up by actual evidence, fine. Else, it's just opinion and hot air. I don't shoot messengers who bring bad news, as long as that news is accurate. But an unsupported (and knowingly inflammatory, at that) conclusion deserves the bullet.

Load the RAWs into IRIS (it recognizes 5D2 files completely and masks off their masking area so you need to use a black frame from the 5D2). For the 5D3 select a big chunk of the left masking area (the top masking area appears to have some data or something stored in it and won't work) and hit the statistics option. You find that the avg value is 2048 and that is the black level. You then see a standard deviation of about 5.95 or so (I was actually a little generous if anything since a patch higher up that I sampled said 6.0). Then you draw a box around the blown highlights in one of the IR ISO 100 images and notice that the raw saturation level at ISO 100 is 15283.

With those three numbers you do this Log[2] ((15283-2048)/5.95)) and get 11.1 stops.

For the 5D2 you find values 15760,1024,6.1 and get 11.2 stops.

banding looks worse on the 5D2 black frame but it's hard to compare to the tiny area of the 5D3 masking, the 5D3 looks to have noticeably less banding there, but the area is too small to really be able to much for sure, it shows promise, but it remains to be seen.

You can set the pop theshold and base values to a narrow range around 1024 for 5D2 and 2048 for the 5D3 to get a good look at the dark current banding and noise characteristics between the two in a fair way (other than nobody has provided a 5D3 black frame so the area is too small to be able to count as representative, the banding might show up more if you got to see a large chunk of the main area).

The masking area that I talk about is the portion of the sensor that does not get exposed to light but that does get recorded to the RAW files, the actual RAW files are slightly larger dimension than the exposed area size that most people are familiar with. The top masking area is unfortunately not quite as large as would be helpful to better aid RAW converters in removing banding.

Anyway, that is how I did it.
On DPR some others have also done it (using different software in some cases) and they got the same things to a tenth of a stop or two.

There is always a chance that the masking area will not match the main area, but looking at it more closely I see no sign of anything weird in the left masking area and I would be surprised if it turns out to not be representative. I hope I am wrong though since who wouldn't want the real result to shot a stop or two improvement here.

 

[t.linn]

If you'd like present a well-reasoned case, backed up by actual evidence, fine. Else, it's just opinion and hot air. I don't shoot messengers who bring bad news, as long as that news is accurate.

There's actually a number of long threads over at DPR regarding the issue of banding at low ISOs on Canon's post-1D3 sensors. It was a well documented problem with 5D2's and it appears to be resolved in the horizontal direction but unaddressed in the vertical on the 5D3. One discussion is here:

http://forums.dpreview.com/forums/readflat.asp?forum=1032&thread=40825115&page=1

Folks can rightly quibble that the comparison between the two images is invalid because conditions were not equivalent but you can't deny that the banding still exists. The bottom line is that the analog design of the Canon sensor significantly limits the DR of Canon DSLR's at low ISO. The Sony sensors have a more modern design that completely eliminates this phenomenon. Check out the detail Horshack is able to pull out of the shadows on this D7000 image. It's worth noting that the pixel pitch of the D800 is about the same so anyone thinking that image quality is going to be lousy at 36 megapixels had better take a look.

From the DPR forum (same thread as above, p7):

To address those that believe any image whose shadows are raised this much would otherwise be unusable even without banding, here is a D7000 single exposure, exposed for the highlights near the sun, with the shadows raised in LR3 w/fill light +100 plus a +1.17 stop GND additional shadow push for the foreground.

Naturally there is noise in this image from lifting the shadows...but there is still lots of detail and color retention, and this is the full 16MP. It generated a rather nice 8x10 print.

Before: http://horshack.smugmug.com/photos/1173945922_SFC9E-O.jpg
After: http://horshack.smugmug.com/photos/1173945328_YMZ2G-O.jpg

 

[t.linn]

Check out the detail Horshack is able to pull out of the shadows on this D7000 image. It's worth noting that the pixel pitch of the D800 is about the same so anyone thinking that image quality is going to be lousy at 36 megapixels had better take a look.

Shoot! The links are down now. I don't know if that's temporary or not. I looked at the before and after yesterday and felt quite envious of Nikon owners who are able to recover impressive shadow detail out of total darkness. I've tried that with my 5D2...and recovered banding. :

 

[3kramd5]

Load the RAWs into IRIS (it recognizes 5D2 files completely and masks off their masking area so you need to use a black frame from the 5D2). For the 5D3 select a big chunk of the left masking area (the top masking area appears to have some data or something stored in it and won't work) and hit the statistics option. You find that the avg value is 2048 and that is the black level. You then see a standard deviation of about 5.95 or so (I was actually a little generous if anything since a patch higher up that I sampled said 6.0). Then you draw a box around the blown highlights in one of the IR ISO 100 images and notice that the raw saturation level at ISO 100 is 15283.

With those three numbers you do this Log[2] ((15283-2048)/5.95)) and get 11.1 stops.

For the 5D2 you find values 15760,1024,6.1 and get 11.2 stops.

So... what are you doing? Blowing the highlights and measuring, blowing the shadows and measuring, and computing the magnitude between them? If you repeat the same calcs for every 5D3 sample do you get the same thing? If you repeat for every 5D2 image do you get the same 11.2? Or are you just measuring the DR of a single image?

How do you explain the difference between your 5D2 numbers and DXO's who measured it using hardware (known lighting values and a variety ND filters to stop the light down in known intervals) and determined a maximum sensor potential DR of 11.9?

I don't have a dog in the fight, just trying to understand the calcs and whether they are repeatable and meaningful.

 

[wockawocka]

I've never seen banding on my files unless I've screwed up the exposure and have had to push the image.

It doesn't surprise me that dynamic range hasn't improved either. I mean, there are some drawbacks to using a CMOS sensor. Until they change the sensor type there will always be limits.
They could use a CCD sensor but then they'd be limited to a max ISO of 1600, but you'd get 12 stops of range.

 

[Curmudgeon]

And once again, no I am not some Nikon fanboy. I have never owned a Nikon camera of any sort in my entire life.
I've owned many Canons. That doesn't mean I need to lie and make up BS to make myself feel better about Canon.

It's kinda sad that the fanboys here are so extreme that they just smite anyone and go la-la-la-la I can't hear you rather than look at facts. But go ahead, be my guest if you feel happier for doing it.

As a matter of principle, I never use that silly applaud/smite button, but if I did you'd get a +1 for figuring out a preliminary way to get a reading on the 5D3's dynamic range. I find that there's a lot of useful information to be gleaned from many people on this forum--too bad it has to be filtered out from a noisy minority who take personal offense to anything other than praise for Canon.

Asking about the specifics of your methodology is legitimate. Crowing about the 5D3's early sales performance is indeed irrelevant.

On another thread about the pricing of the 5D3 I made the point that the sales domination of the 5D2 over competing full frame models depended largely on its prowess as a video camera. I then noted that a number of leading videographers--including Vincent LaForet--were already on record as being unimpressed by the video features of the 5D3, and that the disappointment of this important market segment might translate to lower-than-expected sales--which in turn might lead to a fairly quick drop in price.

I was assailed immediately. Laforet, it was explained, was a nouveau riche photojournalist turned camera snob by his giddy success as a videographer. And then, having acknowleged LaForet's negative response to the camera by attempting to discredit him, my accuser turned on me and charged me with bias for saying that LaForet was unimpressed with the 5D3's specs. How many ways can you have it?

My point wasn't even about the merits of the camera; it was about the expectations of videographers and what that might do to sales. No matter. It could be construed as an indirect attack on darling 5D3 and needed to be struck down immediately.

The clear fact is that at least initially videographers are not embracing the 5d3 the way they did the 5D2--it's all over the Internet. And the equally clear fact is that Canon cameras have a number of class-leading qualities but they have a problem with dynamic range compared to Sony/Nikon. Photographers know it. On the eve of the introduction of the 5D3, when all of the talk should have been about megapixels and frame rate, etc., this forum sprouted a 5-page thread of people worrying that Canon might not have addressed the problems of dynamic range. Likewise,all the testing organizations find that Canon underperforms in this area. How are we going to explain that away? DPR and DXO rig their tests because because Sony/Nikon own a controlling interest in them?

And while nothing is definite yet, there are reasons to wonder if Canon has been able to improve the performance of the 5D3 in this area. The company would have to be deaf and blind not to know that its customers are worried about this issue. And yet we've heard a lot of talk about the improved SNR, two-stop better high-ISO performance in jpeg, as well as the multi-frame HDR feature, but not a single official claim that I can find about an actual increase in dynamic range. DR is important to my kind of photography and I am increasingly concerned that either Canon's engineers can't figure it out or that Sony holds some patents that Canon can't circumvent.

But raise the possibility that Canon has a problem in this area and you're a troll. Any questioning of Canon's perfection is construed as a personal attack. I'm thinking what's needed is a banner--24 pt bold--across the top of every page in this forum: A reminder: the latest scientific evidence confirms that there is absolutely no correlation--repeat, absolutely NO correlation--between your choice of camera equipment and the length of your phallus.

 

[LetTheRightLensIn]

Load the RAWs into IRIS (it recognizes 5D2 files completely and masks off their masking area so you need to use a black frame from the 5D2). For the 5D3 select a big chunk of the left masking area (the top masking area appears to have some data or something stored in it and won't work) and hit the statistics option. You find that the avg value is 2048 and that is the black level. You then see a standard deviation of about 5.95 or so (I was actually a little generous if anything since a patch higher up that I sampled said 6.0). Then you draw a box around the blown highlights in one of the IR ISO 100 images and notice that the raw saturation level at ISO 100 is 15283.

With those three numbers you do this Log[2] ((15283-2048)/5.95)) and get 11.1 stops.

For the 5D2 you find values 15760,1024,6.1 and get 11.2 stops.

So... what are you doing? Blowing the highlights and measuring, blowing the shadows and measuring, and computing the magnitude between them? If you repeat the same calcs for every 5D3 sample do you get the same thing? If you repeat for every 5D2 image do you get the same 11.2? Or are you just measuring the DR of a single image?

How do you explain the difference between your 5D2 numbers and DXO's who measured it using hardware (known lighting values and a variety ND filters to stop the light down in known intervals) and determined a maximum sensor potential DR of 11.9?

The 11.9 from DxO is for their 8MP print normalized plot. Since the 5D2 and 5D3 have the same MP I didn't normalize them against each or anything else. If you normalize my 5D2 numbers then you get 3.8 normalized read noise and match DxO's 8MP normalization value 11.86 vs 11.9. Also note if you hit the screen button on their chart, non-normalized values, then they happen to list the exact same 11.2 dynamic range I listed above.

Also here is a guy who made a comparison tool, he does things a bit differently so his charts don't quite match others entirely but they fit the same general pattern:
http://home.comcast.net/~NikonD70/Charts/PDR.htm

Segue to:
Some other data he has also appears to hint at a 1/2 to 3/4 stop improvement for 5D3 vs 5D2 high ISO performance, same ball park as most seem to be estimating for that.

 

[3kramd5]

Thanks for trying, but can you dumb it down? How does resolution affect DR?