While I am impressed with the 7d, i don't feel "match" describes the files, look at the difference in the highlights and text on the bottle. I would say about a "stop" of extra noise is present in the picture, including the highlights.
I will give it a bonus for the lack of banding!
Also, 9 months is an eternity in the world of electronics.
With that said it looks pretty damn good for apc. This bodes very well for the new 1d and 1ds.
brainiac wrote:
It's such a shame they crippled the 5D2 like this. If an APS-C camera can match it 9 months later, they really dropped the ball.
I suspect it was a case of cost cutting (perhaps quality of components or number of readouts, etc...) since the 1Ds3 doesn't exhibit the same kind of pattern noise.
Not to get too far off topic, but something very interesting on that "technical" section caught my eye. Notice how on page 8 they start to mention how larger AF sensors were thought to be better, but that resulted in problems at high temp. Hmm, sounds a lot like the 1dm3......
Here's the excerpt (emphasis added):
"When designing the sensor, a larger size was considered however _research_ has shown that when _all_ factors are considered, a larger AF sensor is not necessarily better. Larger sensors are more susceptible to changes in temperature rises and falls. This in turn has an impact on the accuracy of the AF system overall. In order to maintain consistency and accuracy the size of the sensor has been optimized to ensure that changes in temperature have little or no affect on AF performance"
timbop wrote:
Not to get too far off topic, but something very interesting on that "technical" section caught my eye. Notice how on page 8 they start to mention how larger AF sensors were thought to be better, but that resulted in problems at high temp. Hmm, sounds a lot like the 1dm3......
Here's the excerpt (emphasis added):
"When designing the sensor, a larger size was considered however _research_ has shown that when _all_ factors are considered, a larger AF sensor is not necessarily better. Larger sensors are more susceptible to changes in temperature rises and falls. This in turn has an impact on the accuracy of the AF system overall. In order to maintain consistency and accuracy the size of the sensor has been optimized to ensure that changes in temperature have little or no affect on AF performance"
timbop wrote:
Not to get too far off topic, but something very interesting on that "technical" section caught my eye. Notice how on page 8 they start to mention how larger AF sensors were thought to be better, but that resulted in problems at high temp. Hmm, sounds a lot like the 1dm3......
Here's the excerpt (emphasis added):
<snip>
---- smoking gun, perhaps?
Yeah, I noticed that too. Hopefully they've significantly advanced their understanding of some of the critical factors that go into AF performance.
dolina wrote:
Thanks for posting this. This pretty much seals my opinion to wait for the 1D3 replacement.
I doubt any camera coming out in the next two years will have iso 12,800 images much better , its a rediculously high sensitivity and , coming from 20 years of film shooting, I can only laugh in incredulence at people moaning about performance at this iso.
The fact that this , or any , camera, can actually get usable images of any sort (and for small web images or small prints they would be fine after processing) at all is a minor miracle.
As an aside, for the 'affordable FF' crowd in mourning that this camera has an aps-c sensor when Sony can do FF.
Here in the UK, the 'affordable' Sony 850 is currently £70 more expensive than a 5DII.
Mark Bishop wrote:
I doubt any camera coming out in the next two years will have iso 12,800 images much better , its a rediculously high sensitivity and , coming from 20 years of film shooting, I can only laugh in incredulence at people moaning about performance at this iso.
The fact that this , or any , camera, can actually get usable images of any sort (and for small web images or small prints they would be fine after processing) at all is a minor miracle.
As an aside, for the 'affordable FF' crowd in mourning that this camera has an aps-c sensor when Sony can do FF.
Here in the UK, the 'affordable' Sony 850 is currently £70 more expensive than a 5DII....Show more →
+1 on shooting at ISO 12,800. This will comprise less than 1% of images taken with this imaging device (camera).
Not to steal any thunder from all of the pixel-peeping, but here are some early examples of AF performance for the 7D:
I realize these are Canon samples, but until we get more independent field test reports we have to start somewhere.
To me, notwithstanding its superb video imaging capabilities, this imaging device (camera) is pointed squarely at the action/sports/wildlife photographer.
brainiac wrote:
It is not theory that DPR compares noise between cameras at different magnifications. Go and look at any of their stamp comparisons. When cameras have different numbers of pixels, the size of the Queen's head in the crop varies in proportion.
(1) DPR compares 100% crops in order to make judgements about noise.
(2) that heavily biases the result in favour of the lower resolution camera.
(3) DXO also shows per pixel comparison by default.
(4) most people don't understand why that makes comparison entirely bogus.
Those are facts, not theory.
There are a lot of shooters who don't really need more than 12MP. Wedding shooters, for example . For this category, the lowest possible noise at high iso's wins against more megapixels. They don't need to mess around with huge raw files of 20+ MP, and maybe 30+ in the future. Most of the times there is somebody who tells you that you can shoot sRAW, but that sRAW will have the same noise as the full size RAW, becouse the pixel binning is not happening yet...
So, while i agree in part with your arguments above, i still would prefer maybe 10% less noise per square inch of print from a 12MP 7D, than from a 18MP 7D.
Mark Bishop wrote:
I doubt any camera coming out in the next two years will have iso 12,800 images much better , its a rediculously high sensitivity...
...and , coming from 20 years of film shooting, I can only laugh in incredulence at people moaning about performance at this iso.
I agree - it's a great facility, but let's not pretend its perfect yet. All improvements in low light image quality will be gratefully received by this photographer. We concentrate on iso 12800 because that's the front line, the limit of what can be handed over to a client. Hopefully in a few years that limit will be 25600 or even 51200. Why stop when we're loving it?
brainiac wrote:
But you would not get less noise per square inch of print from a 12 Mpixel 7D. Noise per square inch of print is largely independent of pixel pitch.
I was just hoping that a hypothetical 7D with only 12 MP should have bigger pixel pitch than the 18MP version. We are talking about the same sensor age and technology...
Ransome wrote:
Nail on the head. The 7D will be a replacement for my 30D, so it and the 5D II will make for a good combination. The ideal combination would be the 7D and the mythical 3D (FF 7D)... but then there's always next year.
I agree here. I don't see the 7D as a replacement for the 5D / 5DII. It certainly appears to be a great "next step" upgrade for XXD owners though. I see 5 levels now.. Rebels, XXD's, 7D's, 5D's, and 1D's. That's a lot of market coverage. Throw the G9, 10, 11's in to the bottom of that pile, and Canon is covering a lot of the market with a great selection of quality cameras. Smart move.
I'm a 20D owner, and I would certainly give the 7D a LOT of serious thought had I not spoiled myself with a 1DmkIIn several months ago. My thoughts are now leaning toward retiring the 20D or handing it over to my wife, and getting a 1DsII (or maybe a 5DII) for my one-two punch. All in all though, the 7D hits a sweet spot.
Had a look at the 7D RAW files from Imaging Resource. I downloaded only the NR OFF files. iso 400 is not as good as my 5D, iso 800 and 1600 look remarkably similar though which is a shock (in a good way!). iso 3200 is as bad as my 5D which is again incredible for a crop camera with that pixel density. I can't tell anything about DR from ACR 5.5 which only has the 7D as beta, I remember how awful the 5D files looked from the beta version of ACR which supported it at the time. I have no doubt that they will also get a better handle on the noise once it's properly supported in ACR, there seems to be inconsistency, I think iso 400 should be better than that given how good the next two iso's are and I hereby blame ACR totally .
The tonality and probably DR are always going to be better with the bigger pixels but the idea that a crop camera with so many megapixels can come head to head for noise with the benchmark 5D is truly astounding. That and a body worthy to be enough but all for the most demanding user makes this quite a camera (on paper, few have ever held one, hope it's less cheap feeling than the mkII).
If only it was FF....
Oh and to head off Braniac at the pass - I don't buy cameras to downrez the files and have to apply noise reduction. The above is my personal opinion based on looking at files the way that they are relevant to me.
Beni wrote:
I don't buy cameras to downrez the files and have to apply noise reduction. The above is my personal opinion based on looking at files the way that they are relevant to me.
Beni wrote:
Oh and to head off Braniac at the pass - I don't buy cameras to downrez the files and have to apply noise reduction. The above is my personal opinion based on looking at files the way that they are relevant to me.
Which makes the results even more astounding. I really think Canon pulled off an absolute winner (even if it is 2 years late), and I am in line to be beta tester as soon as my name comes up on the preorder list
Just did some playing comparing the (RAW) files from the 7D vs the mkII, at iso 400 the 7D looks like the mkII at 1600. I think Adobe have got some work still to do. I can't believe that. DR is also a lot less.
First time I've looked at those Imaging Resource test shots from the mkII, holy cow! A clear one stop advantage over my 5D and the noise is so fine and clean looking even when it is there. Don't know what it is though with canon cameras, every single one I've seen, iso 1600 is great but 3200 falls apart. Same with the mkII, 3200 colour noise gets to the seriously blotchy when dealt with whereas 1600 is so perfect. No doubt C1 will be better than ACR/LR for this but it's worse at a lot of other things.
Interestingly when renewing my camera insurance today they told me to put the value of a mkII for my 5D's as that is the most accurate 'like for like' replacement. Oh woops, both of my 5D's suddenly fell under the wheel of a truck! Oh dear...
I know it's OT but however hard I pushed those mkII files from IR, I couldn't get banding or cross hatching in the shadows. I have no doubt whatsoever that what people are reporting is real, is it that their examples are shot in 'good' light?
. For this category, the lowest possible noise at high iso's wins against more megapixels. They don't need to mess around with huge raw files of 20+ MP, and maybe 30+ in the future. Most of the times there is somebody who tells you that you can shoot sRAW, but that sRAW will have the same noise as the full size RAW, becouse the pixel binning is not happening yet...
.