400d wrote:
Did the moon take a break over the weekend?
shhhhh. it's sleeping now.
anyway it might be general knowledge, all the details, by maybe 2am EST tomorrow night (or at least the following night after)??
although the moon teaser doesn't at all seem to be progressing fast enough for the supposed 16th or 17th release....
Geoff Costello wrote:
And the old circle of diffraction comes in limiting resolution.. Apparently somewhere around 65mp on a FF it doesn’t make much sense ot have more pixels as they will each generally be resolving too much for a lens at F2 or smaller apertures.
Good morning everybody in here
Moon again does not change. But it can not need much more time now.
I wich you a healthy and successfull new week, Canon a great outcoming and all of us that we get, what whe are dreaming about.
Hey Ralph, good "night " for me it's 11:51pm right now here in California...hopefully Canon will release the much much much awaited 5D2 and hopefully it will be a great camera.
I understand the 'per surface area' noise argument. But what about dynamic range? That should be reduced at the 'per pixel' level with smaller pixels, and not made up by having more of them...
I don't care much about noise, but better DR would be nice.
Geoff Costello wrote:
I think the resolution here (sorry for the pun but I can’t resist) is to think of noise for a given square area of the picture in printed or in screen form, not per pixel. Once you get to gapless micro lenses, the sensor is capturing all the light anyway. While on a 24mp sensor the individual pixels will be noisier at a given low light level than a larger 12mp pixel on the same sensor size (which is getting twice as much light. However for a image at th same size, the visible noise per unit area will be basically the same. This shows in prints form the 1Ds III which are very good noise characteristics, even when quite noisy at a pixel level. Printed at the same size several pixels are averaged, smoothing away the noise. Canon's new sRaw technology will average away the noise from a 24mp sensor to a 12mp image giving the equivalent (broadly) of the same sensor with larger photo sites.
So a 48Mp gapless micro lens sensor would (of course be even nosier at a per pixel level, but have the same level of noise per unit area. as a 24 and 12mp sensor using the same CMOS, DIGIC and A/D technology.
So, from a simplified perspective, yes for a given sensor size and CMOS, DIGC and A/D technology, pixel size is close to irrelevant for high ISO noise on prints. There are some minor details that actually favour the higher pixel count sensors up to a certain point. For example the impact of a 'hot pixel' is smaller and can be more easily averaged out at a high pixel count, but then you get more of them per image... Equally with the blurring of most images on high resolution sensors (especially anything outside the critical focus, having multiple sensors per area getting the same colour values allows easier noise averaging without loss of details. And the old circle of diffraction comes in limiting resolution.. Apparently somewhere around 65mp on a FF it doesn’t make much sense ot have more pixels as they will each generally be resolving too much for a lens at F2 or smaller apertures.
I'm sure the more technical will pick holes in my wording.. However I am confident that the principle is the basis of Canons current and next 3-4 years sensor strategy. You can see the evidence in the 50D high ISO pics now emerging. And I’m sure he images from the 5D II will prove the point even more, with ISO 25600 looking like 12800 on the 50D, which looks like 6400 on the 40D. ...Show more →
For those who haven't followed this thread all the way, one of the more believable rumors is that the press release will come Sept 17th 6am european time (if I got it correct, in the time zone which includes germany), which is Sept 16th 9pm Pacific Standard Tme (US West Coast). A press conference (as opposed to "release") will be held 3 hours later.
I understand the 'per surface area' noise argument. But what about dynamic range? That should be reduced at the 'per pixel' level with smaller pixels, and not made up by having more of them...
I don't care much about noise, but better DR would be nice.
Geoff Costello wrote:
I think the resolution here (sorry for the pun but I can’t resist) is to think of noise for a given square area of the picture in printed or in screen form, not per pixel. Once you get to gapless micro lenses, the sensor is capturing all the light anyway. While on a 24mp sensor the individual pixels will be noisier at a given low light level than a larger 12mp pixel on the same sensor size (which is getting twice as much light. However for a image at th same size, the visible noise per unit area will be basically the same. This shows in prints form the 1Ds III which are very good noise characteristics, even when quite noisy at a pixel level. Printed at the same size several pixels are averaged, smoothing away the noise. Canon's new sRaw technology will average away the noise from a 24mp sensor to a 12mp image giving the equivalent (broadly) of the same sensor with larger photo sites.
So a 48Mp gapless micro lens sensor would (of course be even nosier at a per pixel level, but have the same level of noise per unit area. as a 24 and 12mp sensor using the same CMOS, DIGIC and A/D technology.
So, from a simplified perspective, yes for a given sensor size and CMOS, DIGC and A/D technology, pixel size is close to irrelevant for high ISO noise on prints. There are some minor details that actually favour the higher pixel count sensors up to a certain point. For example the impact of a 'hot pixel' is smaller and can be more easily averaged out at a high pixel count, but then you get more of them per image... Equally with the blurring of most images on high resolution sensors (especially anything outside the critical focus, having multiple sensors per area getting the same colour values allows easier noise averaging without loss of details. And the old circle of diffraction comes in limiting resolution.. Apparently somewhere around 65mp on a FF it doesn’t make much sense ot have more pixels as they will each generally be resolving too much for a lens at F2 or smaller apertures.
I'm sure the more technical will pick holes in my wording.. However I am confident that the principle is the basis of Canons current and next 3-4 years sensor strategy. You can see the evidence in the 50D high ISO pics now emerging. And I’m sure he images from the 5D II will prove the point even more, with ISO 25600 looking like 12800 on the 50D, which looks like 6400 on the 40D. ...Show more →
a 60mp FF sensor is MTF50 diffraction limited at f5.6.
As stated below, in the same article, this might only apply to a few lenses of legendary quality. In 10 years a FF 60MP camera may cost $1500, but the lenses needed to take advantage of the 60MP will cost big big bucks.
Only for highly corrected lenses (with better performance at f/5.6 than f/8) do higher sensor resolutions make sense..........Sensors for larger formats are approaching the diffraction limit of real lenses, and it is more difficult to get high levels of aberration suppression for them. The point is that you cannot fully exploit the resolution potential of high-resolution sensors with regular mass-produced lenses, particularly for larger formats.
thw2 wrote:
The 7 MP Canon G7 has smaller pixels but better performance (high ISO as well as dynamic range) than the 4 MP G3 and 5 MP G5. The 10 MP Nikon D80 has smaller pixels than the 6 MP D70 but improves on its performance. With the swtich to CMOS sensors, the 12 MP D90 is even more promising.
Given a technology, the smaller the pixel, the more likely noise will bias the color accuracy - this is physics.
Now, talking relatively is a different matter.
E.g, relative to the time/maker/technology: the camera X is newer and has smaller pixels than Y but has less noise per pixel (pp).
Or relative to the resolution: X has more noise pp but has much more pixels globally than Y, so shrinking X pic to Y size shows less noise pp. But in this case it is not correct to only say that X has less noise than Y without mentioning "if the pic is scaled"
Welcome this mo(o)nday! The new week makes me feel like EOSfun
To some of the guys above: you're all still denying the progress of technological development. No way the cameramanufacturers, and definitely not Canon, are going to degrade image quality for just more megapixels. More Megapixels brought us better IQ anyway. Every single step of improvement: resolution, noise, dynamic range. If you disagree, then go back to your EOS 30D's, Nikon D70's or other obsolete DLSRs. They are cheap now, and they have little megapixels like you seem to wish
I can't repeat it often enough: there is no substitute for megapixels! Go ahead, to the 100MP camera asap!
I'd guess it also depends on the sensor technology. If the photosites have relatively more noise at lower light levels, then a higher resolution may cause more noise even overall,per-image. If the noise is proportional (or almost), over a large range, then the per-image noise doesn't change. Another possibility, depending on the sensor technology, is that in dark shadow areas, or ultra low-light, there will be more noise (even after downscaling), but otherwise, it will be similar (after downscaling). That would need to be tested, and it seems that currently, this question is not tested very much.
Well, i'm using a 10D still
I'm not saying that more MP DEGRADES IQ, but maybe a smaller increase in MP might increase other aspects (like DR) MORE than it would increasing MP count.
But then, what do i know? I'm still only getting EOS...wait...wait...wait..fun (maybe) with my old 10D
eosfun wrote:
To some of the guys above: you're all still denying the progress of technological development. No way the cameramanufacturers, and definitely not Canon, are going to degrade image quality for just more megapixels. More Megapixels brought us better IQ anyway. Every single step of improvement: resolution, noise, dynamic range. If you disagree, then go back to your EOS 30D's, Nikon D70's or other obsolete DLSRs. They are cheap now, and they have little megapixels like you seem to wish
I can't repeat it often enough: there is no substitute for megapixels! Go ahead, to the 100MP camera asap!
But it can not need much more time now.
