brainiac wrote:
Perhaps I didn't explain myself very well.
It is not a problem of explanation - your initial example was wrong: you won't have (same technology) the same relative noise in a pixel and a twice-bigger pixel (your "10%" for each).
So it is not necessarily the case that "the smaller the pixel, the more likely noise will bias the colour accuracy". That dogma is not physics, but a myth whose popularity derives from the widespread failure to view crops at equal magnification.
Of course it is true. Of course this is a per-pixel approach. Physics is all about pixels and photosites.
And there is the downscaling approach, which is more practical for the photographer.
Two approaches. And we shouldn't talk only about the downscaling one. And we shouldn't talk only about the pixels one. Depends on context.
This is not going to help you much as an argument: p&s have a pixel density soooooo much higher than a dslr OR MF back, that pixel density is the greatest candidate for their lower DR.
How on earth would multiplying the same pixel more and more on a bigger and bigger surface make them together have a bigger DR?
I would have never thought of myself getting into this kind of arguments; it's obviously Canon's fault! I'm so glad it's almost over - more than building up the hype, they risk of getting people tired of this cr@p!
brainiac wrote:
Do you accept that point&shoots generally have less good dynamic range than DSLR's? Considering that we have many examples of sensors of a particular size with different densities but similar dynamic range, and a tendency for larger sensors to show more of it, what makes more sense: to attribute dynamic range to pixel density or to sensor size?
Although I understand and agree with Braniacs point that comparison of noise between two camera can only be with the larger image downsized to make a like for like comparison, I would say that 100% per pixel noise data is very relevant given the apparent end use of the higher megapixel camera. I want to print larger from my 1Ds mkIII than my D3, it's the only reason I would have one and not the other. As such the noise at those larger print sizes is relevant to me as a consumer. As a comparison it isn't fair, in itself, I think that is still useful.
Brainiac never said that per pixel noise isn't relevant. His point is that to fairly compare the noise characteristics of two different cameras, you need to adjust both images to the same resolution. Otherwise it isn't a fair comparison.
Kind of like comparing the sharpness of two different lenses - you want to do the comparison at the same aperture(s), otherwise it isn't a fair evaluation. My 18-55mm kit lens is sharper at f/8 than my 24L wide open. Does this mean that the 18-55 is a sharper lens?
Brainiac's (valid) point is that 1Ds3 files when properly downrezzed to 12MP have less noise than 12MP files from other cameras that have been touted for their high-ISO noise performance.
The point I am trying to make is that even if it's 38 megapixels, a new full frame Canon is going to have far better dynamic range than a point&shoot. Equally, even if it's 8 megapixels, it still won't have the DR of a medium format camera.
As far as noise goes, today there seems to be practically no noise penalty at high iso for very high density DSLR sensors, as shewn by the superb high iso performance of the 1Ds3.
So in reality there is little to lose by having lots of pixels in the 5D2, and there is plenty of sharpness and detail to be gained. mRaw means you don't have to shoot at the full resolution. Those wishing for a 15 megapixel 5D2 instead of a 24 megapixel one would gain very little for the sacrifice, except blur.
The sentiment that pushing up the numbers of pixels is significantly going to compromise DR or noise is a fallacy based, on the unfair comparison of 100% crops at varying magnification. We will be hearing much more about that as people start pixel peeping their new 21 megapixel cameras at 100% and are thoroughly disappointed. Those who take the trouble carefully to use the best glass and go so far as to make comparison prints, or resize before examining crops will be pleasantly surprised. The differences between 12, 21 and 24 megapixel are incremental rather than gaping, but no less worthwhile for that.
Leica came out with new 21mm f1.4, 24mm f1.4 and 50mm f0.95 ASPH, I hope that Canon will introduce new redesigned prime lens for 5D MKII. The leica 21mm f1.4 is so sweet.
www.l-camera-forum.com/leica-forum/vbcms_areas/vbcms_custom_content/images/original_images/photokina2008.jpg
xichlo wrote:
Leica came out with new 21mm f1.4, 24mm f1.4 and 50mm f0.95 ASPH, I hope that Canon will introduce new redesign prime lens for 5D MKII. The leica 21mm f1.4 is so sweet.
...until you
- see the price
- realise that it occludes half your viewfinder
- find that your M8 still only gives you something akin to an 8 megapixel iso 3200 file due to the poor noise performance and the camera's in-built mandatory moire-reduction algorithm
- find there is no way of adapting it to our favourite sensor maker's cameras
rsg_1 wrote:
Ok. Its the 15th now, so where is the new camera?
First it was the 12th, then the 15th and now it is said it will be the 17th. Reminds me of the 'announcement' dates that were told for April this year.
My bet: a few days before Photokina.
Perhaps so. My guess is that we should see something credible by Friday. The anticipation is killing me. I can't sleep & have butterflies in my stomach. This thread hasn't done much for me either.
That's why I hope to see Canon will come out with the new prime lens. The 35mm L is good , but the 24mm f1.4 MK II or 21mm f1.4 will be great.
Talking about Leica, If I ever shoot Leica, I will pick the M6/M7 to go with those leica lenses. The M8 is nothing better than the Digital Rebel series.
But who knows the Leica will come with the new Full frame M9 very soon.
brainiac wrote:
...until you
- see the price
- realise that it occludes half your viewfinder
- find that your M8 still only gives you something akin to an 8 megapixel iso 3200 file due to the poor noise performance and the camera's in-built mandatory moire-reduction algorithm
- find there is no way of adapting it to our favourite sensor maker's cameras
So if the camera is actually named the 5D mark II, I'll bet we'll never see the 3D. With the shifts that are taking place in the market, the 1D will stay the next level, but likely at a lower price with a FF sensor. (and what if the 1Ds is eliminated completely?)
The big question with the rumoured 24L is this, is it a warm over or an full on redesign? The 14IIL was a completely new optical design and they (correct me if I'm wrong) added weather sealing. I'm kind hoping that this new 24L will have updated optics, coating and get weather sealing. I think it's bizarre that the f2.8 L zooms get weather sealing but these more costly primes don't. The 50mm f1.2 L (which is a new design and it's debatable if it's a replacement fot the 50mm f1.0 L). The 85 II L was mostly an AF warm over but with new optical coatings.
I was hoping to see a 5D II with a 24/1.4 Mk II, and I will order them both at the same time .... but if Zeiss is launching a 21/2.8 Distagon in the ZE mount for EOS, I will have a problem in which one to go with. Having said that, if the 24L Mk II is as good as the 35L, the decision would be very easy for me.
