Well, interesting read. I'm not so sure that I would spend so much money on a 24MP camera if my intention was to always use 11MP files from it personally. Per pixel noise at the native resolution output is not as good as a D3 as he says.
My favorite remark from that blog is this:
"Now if you doubt just how extremely sharp the Alpha 900 gets using normal ISO speeds, and exporting to 11.2 megapixel size, click this one and view full size. Look at the structure of the ‘foam’ below the Alpha camera and you’ll see moiré patterns. The camera, here, is outresolving the subject."
Interesting take on the problem of moire. I guess we can all just learn to accept the moire problem knowing that it results because our cameras are "outresolving the subject." I don't quite buy that explanation. Outresolving the subject? The moire in that particular photo was simply created from the pattern/frequency of the dot screen in the ad. This occurs when you are photographing detail that has a higher spatial frequency than the sensor can record.
Agreed. I think David was hasty with that moire comment.
Tariq, I don't think you're grasping the concept of sizing down to 12MP from 24MP. It's the best of both worlds. It's basically having a high resolution, studio camera at 24mp, with a D3 built-in for low light shots at 12mp. Barring Canon completely obliterating the detail of the 5D II, the 5D II will more than likely only compete with the D3 when sized down, as well. Or do you actually believe the 5D II will compete with the D3 on a per pixel noise at native resolution basis? I believe both the A900 and 5D II will compete quite favorable with the D3/D700 at high ISO when sized down to 12MP, but not at full native res.
I do believe that the 5DII will compete with the D3 at the per pixel level. I understand the idea of having a high MP camera for studio and using downsampled high ISO images for other purposes. I just have the gut feeling that even at its native output of 21MP, the 5DII will end up performing far better at high ISO's than the Sony is capable of. Personally, I will most likely still get the A900 because high ISO performance really does not matter to me.
douglasf13 wrote:
Agreed. I think David was hasty with that moire comment.
Tariq, I don't think you're grasping the concept of sizing down to 12MP from 24MP. It's the best of both worlds. It's basically having a high resolution, studio camera at 24mp, with a D3 built-in for low light shots at 12mp. Barring Canon completely obliterating the detail of the 5D II, the 5D II will more than likely only compete with the D3 when sized down, as well. Or do you actually believe the 5D II will compete with the D3 on a per pixel noise at native resolution basis? I believe both the A900 and 5D II will compete quite favorable with the D3/D700 at high ISO when sized down to 12MP, but not at full native res....Show more →
Funny we're both talking about high ISO, and neither of us will use it much
The 1DsIII at 21MP does not compare to the D3 in noise, so, unless Canon dramatically leapfrogs the 1Dsiii noise with the 5Dii, I don't expect it to compare, either. However, like I said, I do expect the 5Dii to compete well at 12MP.
douglasf13 wrote:
Funny we're both talking about high ISO, and neither of us will use it much
The 1DsIII at 21MP does not compare to the D3 in noise, so, unless Canon dramatically leapfrogs the 1Dsiii noise with the 5Dii, I don't expect it to compare, either. However, like I said, I do expect the 5Dii to compete well at 12MP.
Those saying that it's a 24 megapixel camera at low ISO's and a 12 megapixel hi ISO camera seem to be ignoring the fact that the 24 megapixels resolve less than the 21 of the 1Ds mkIII (and the 5D mkII I assume) and that the low ISO noise seems to be on par with the original 1Ds (which I used to own). They say that the base iso is 200 so I can't expect iso 100 noise, ignoring the fact that other cameras in the megapixel range have the ability to shoot with more resolution and less noise period - at any iso. I really hate the idea of the 5D mkII, it's a studio or amatuer camera only IMO when compared to the Nikon offerings which is why I'm sticking with my 5D's at present. However I do believe that if you want a high resolution and low noise camera then the Sony is the wrong choice - whatever the iso.
I was really interested in the Sony, wanted to get it with 2 primes and use it for all my high rez work, seemed to me a better choice given the glass than the 5D mkII. I'm afraid though that I hated the noise of my 1Ds and am not volunteering to go through that again with a 35mm form DSLR. Seriously thinking of a Mamiya AFD II and P21 instead, given the recent Hasselblad price plummet and the axing of all the MFDB upgrade programs, I expect to see 'low' megapixel MFDB's at very cheap prices within 6 months. If I'm going to have lots of megapixels then I want better megapixels, not just smaller. For that I'm willing to put up with the noise.
Tariq Gibran wrote:
Well, interesting read. I'm not so sure that I would spend so much money on a 24MP camera if my intention was to always use 11MP files from it personally. Per pixel noise at the native resolution output is not as good as a D3 as he says.
My favorite remark from that blog is this:
"Now if you doubt just how extremely sharp the Alpha 900 gets using normal ISO speeds, and exporting to 11.2 megapixel size, click this one and view full size. Look at the structure of the �foam� below the Alpha camera and you�ll see moir� patterns. The camera, here, is outresolving the subject."
Interesting take on the problem of moire. I guess we can all just learn to accept the moire problem knowing that it results because our cameras are "outresolving the subject." I don't quite buy that explanation. Outresolving the subject? The moire in that particular photo was simply created from the pattern/frequency of the dot screen in the ad. This occurs when you are photographing detail that has a higher spatial frequency than the sensor can record. ...Show more →
Per pixel noise will never be the same since the larger pixel collects more light, thats why you scale the image, bacause the whole sensor collects the same amount of light, but spread out on more pixels.
douglasf13 wrote:
Agreed. I think David was hasty with that moire comment.
Tariq, I don't think you're grasping the concept of sizing down to 12MP from 24MP. It's the best of both worlds. It's basically having a high resolution, studio camera at 24mp, with a D3 built-in for low light shots at 12mp. Barring Canon completely obliterating the detail of the 5D II, the 5D II will more than likely only compete with the D3 when sized down, as well. Or do you actually believe the 5D II will compete with the D3 on a per pixel noise at native resolution basis? I believe both the A900 and 5D II will compete quite favorable with the D3/D700 at high ISO when sized down to 12MP, but not at full native res....Show more →
I think he gets it. It is not a difficult. I however, think the problem with your agreement is that you seem to believe that all sensors of roughly equivalent size and MP will have the same level of noise. It is not necessarily true. "All other things" are NOT "equal". Companies can improve the state of the art, and some companies will then spend time catching up.
To address some of your points:
1) Yes, the 5D MKII does indeed leap frog the 1Ds MkIII.
2) Yes, the 5D MKII appears (from everything I've seen to date) inherently less noisy than the a900.
3) The new 5D MkII down sized to 12mp is better than the original 5D. And, on a pixel basis, appears as good as the original 5D, in spite of all the additional pixels (that's technology advancing).
4) The new 5D MkII down-sized to 12mp will be better than the D3/D700.
5) Downsizing the a900 to 12 mp may indeed make it noise characteristics competitive with the original 5D and D3/D700. That's not a huge incentive to switch.
If noise is your concern, why get a 24mp a900 when you can only get the benefit of all those mp at ISO 100, when the 5D MkII can give you it's full benefit of its 21 mp all the way up to 800 ISO (if not more) -- and it doesn't "obliterate" detail at all (unlike the a900 which, as is apparent from the imaging resource samples, starts destroying detail at 400 ISO, if not at 200 ISO.
The refusal to accept that this is even possible is very surprising. i can see where one might disagree on the extent of any actual difference, but to reject it out of hand as being essentially impossible, is just not at all reasonable.
Again, and let my underline this once again, this is based entirely on these early samples that are showing up on review sites and elsewhere. These are not controlled comparisons, and these early sample may not be representative. However, the differences as so marked that it is safe to conclude that it is at least directionally correct. We need to see more samples, and have controlled comparison to to fully understand the magnitude of the differences. I firmly believe, that in the fullness of time, that most of the noise assessment that I have laid out above, will be borne out.
moire wrote:
Per pixel noise will never be the same since the larger pixel collects more light, thats why you scale the image, bacause the whole sensor collects the same amount of light, but spread out on more pixels.
This makes the huge implicit assumption that aside from the pixel count, the embodied technology of the 2 sensors is the same. With technology continuing to improve as, and with different proprietary technology and manufacturing methods employed by different sensor manufacturers, there is absolutely no reason to assume that this assumption is true or even reasonable.
Please show us examples of the A900 destroying detail. I'm assuming you're not talking about jpegs, right? Now that some 3rd party RAW converters are accepting A900 RAWS, there have been much better results. Besides, I never said that high ISO was the reason to switch from Canon. AF Zeiss lenses and in-body IS are the reasons to do that.
You know, I have never once said that the IQ of the A900 is far better than the 5Dii, but rather competitive with it. It's funny how Canon users always seem to say their IQ is far better.
Here is a sample from the dpreview.com preview of the a900. It is at 800 ISO. For argument's sake, I down-sampled it to the original 5D image size (2912x4368). Here is a 100%. This is clearly worse than the original 5D and the D3/D700 cameras. I'll try to post more later tonight. http://boncratious.com/images/a900-sample800ISO-12mp.jpg
That is a JPEG! It has been gone over many times that the jpeg engine in this camera is not very good, and I firmly believe that. Don't post JPEGs for this discussion, and if you're a jpeg only shooter, the A900 isn't for you. You're swift boating, here.
Those that think that the d3/d700 is better than the 1ds mark III (when resized), check out brainiacs analysis on the topic, that goes for the 1ds mark III being worse than the 5d mkII too :P
I'm looking forward to more user pictures from the a900, skilled raw-conversions. I have no doubts that it will rock at iso 100-400, but I think it will do ok at higher isos too.
Yeah, please don't get me wrong. I'm not saying that the A900 is better than the 5Dii, I just don't think there will be a wide expanse between them, like some have been implying. Therefore, I doubt many Canon users will jump to Sony, unless they're really interested in IS, af zeiss, VF, etc. I know I wish the A900 had movie mode.
