while high ISO shots from the 40D might look maginally better at 100% crop, when scaled (or printed) to the same size the shots from the 50D will look significantly better.
Don't quite follow you on this point. Seems to be a contradiction. But might only be the way you are using terms?
Also, it seems to be generally said that noise visible on a monitor will not necessarily be visible when printed (?).
So, when comparing the 40D and 50D, while they have essentially the same per pixel noise, the 50D has a stop (1.4 times) more pixels which means 50D sensor has stop more low light sensitivity.
... So, when comparing the 40D and 50D, while they have essentially the same per pixel noise, the 50D has a stop (1.4 times) more pixels which means 50D sensor has stop more low light sensitivity. So, while high ISO shots from the 40D might look maginally better at 100% crop, when scaled (or printed) to the same size the shots from the 50D will look significantly better.
...
Unfortunately, Phil did not write the review. Lars Rehm and Richard Butler wrote the review for DPReview.
Instead of saying that the 50D has the same per pixel noise, they said (according to the following link):
"Despite the fact that the 50D is the newer camera it shows visibly more chroma and luminance noise than the 40D. Considering the 50D's much more tightly packed sensor (4.5 MP/cm² vs 3.1 MP/cm² on the 40D) this comes hardly as a surprise. It would have been unreasonable to expect Canon's engineers to overcome the laws of physics."
It is still my contention that the 50D handles noise in a different way than the linear method employed by the 40D. If indeed the 50D uses a more selective algorithm that targets shadow areas to apply a higher level of NR than highlight areas, that could explain the visible "appearance" of less noise on the 50D images. That could also explain the test results measuring more noise, since the test measures all noise where the human brain is more sensitive to shadow noise.
Otherwise you are right that "binning" pixels will generally reduce the visible effects of noise and that binning will occur during most printing operations until close to a 1:1 print ratio and above. The extra pixels of the 50D should be a benefit for most printing situations over the 40D.
I also agree that it would appear that we have reached the practical diffraction limits for photosite density at the current technology APS-C sized imager with the density of the 50D imager (... and the Pentax K20D imager as well).
It is still my contention that the 50D handles noise in a different way than the linear method employed by the 40D. If indeed the 50D uses a more selective algorithm that targets shadow areas to apply a higher level of NR than highlight areas, that could explain the visible "appearance" of less noise on the 50D images. That could also explain the test results measuring more noise, since the test measures all noise where the human brain is more sensitive to shadow noise.
Otherwise you are right that "binning" pixels will generally reduce the visible effects of noise and that binning will occur during most printing operations until close to a 1:1 print ratio and above. The extra pixels of the 50D should be a benefit for most printing situations over the 40D.
Once again I wonder at the wisdom of buying a new body for the new NR software it uses. And NR that you have only very crude control over at that.
As the quote below says (referring to pixel binning on the sensor-chip, but the principal is the same for printing?), there is a trade-off of binning in loss of resolution. So, camera applied NR + binning ( + conversion to jpeg) = loss of resolution, and more pixels = increase in resolution. Which wins out? Where do we get?
The primary benefit of pixel binning is to improve the signal-to-noise ratio in low light conditions at the expense of spatial resolution. Summation of many charge packets reduces the read noise level and produces an improvement in signal equal to the binning factor. Contributing Authors Mortimer Abramowitz - Olympus America, Inc., Two Corporate Center Drive., Melville, New York, 11747. Michael W. Davidson - National High Magnetic Field Laboratory, 1800 East Paul Dirac Dr., The Florida State University, Tallahassee, Florida, 32310.
Well, I must say after reading this review I started second guessing myself on my purchase of the 50D. I had ordered last week the 50D (had not arrived yet) just because it was the newest, latest, 15 MP, and I just had to have it. I know I reacted just how Canon wanted me to. I'm in marketing and sales myself and still fell for it. SO, I started reading this review very carefully and looked elsewhere for other opinions and they all came to the same conclusions. Btw, I'm upgrading from my Rebel Xti.
I actually cancelled my order on the 50D and went ahead on ordered the 40D and put the 550.00 difference towards a new lens. Thanks for all the discussion guys.
Once again I wonder at the wisdom of buying a new body for the new NR software it uses. And NR reduction that you have only very crude control over at that.
Neil,
forget the whole noise issue for a second, let's assume 50D is marginally the same as 40D in this department (although I do believe Ken is correct speaking of effective one stop improvement, and my own iso1600/iso3200 shots prove that to me more than any reviews in the world).
Much better screen, new and very useable highlight optimization mode..
And by the way, not all the shooting happens in lowlight AND at high ISO at the same time. So, for the 99% of time 50D owners can safely enjoy new 15mp resolution and don't even worry about those pesky effects that seem to draw so much of attention lately..
And, btw, it focuses MUCH better in twililight than 40D ever did
Just seeing that image makes my back hurt.... I love the plane, but it's so small
I don't know about the low light situations, but I did see some shots in daylight (40d vs 50d) and the darker areas where about the same, noise-wise.
Heheh....it was a tad tricky getting in and out, especially with a camera. Regarding the showdowed areas/banding, I think we sometimes expect too much out of high ISO. Sometimes we can nail it, clean it and the images are good to go....sometimes not.
Some of the features of the 50D look way cool....especially calibrating each lens. Also, I'm sure the highlight/shadow retention is better. I haven't seen the new DPP software yet for the 50D but only can assume one could implement some of those features in post if shooting RAW. The 40D is a great camera, the 50D may be a bit better and even more so on some things. I'd be interested to see the focus speed differences and accuracy in AI Servo.
One thing for sure, for those that have an earlier version (xt, 10, 20, 30) will be blown away at the advancements. I certainly was with the 40D. The focus was much better all around in each mode.
One thing for sure, for those that have an earlier version (xt, 10, 20, 30) will be blown away at the advancements.
That's the part I'm wondering about. I have a 30d, but is it really worth upgrading for me? The camera functions fine, I'm not 'missing' anything right now. I'm sure it will be better at some points, but is it worth it for me? If I'm going to spend that much money on something I don't need, I might as well go for the 5dmkII, I think.
Ivar, if you're not missing anything then I say why bother?
Well, I guess what I'm saying is that after reading the review I am not feeling like upgrading. I was actually expecting (hoping?) to feel like upgrading after a decent review, but I'm just not seeing 'the light'
Well, I guess what I'm saying is that after reading the review I am not feeling like upgrading. I was actually expecting (hoping?) to feel like upgrading after a decent review, but I'm just not seeing 'the light'
I hear you. "Gut feeling" is very important, that's what I ultimately go by. And, as I said, if you don't feel like upgrading I don't think you should.
For me I decided it was worth it, so I went on with it, and thus far no regrets...
I read the DPReview of the 50D today and afterward I pulled the trigger on one from B&H. Though I love my 30D, it frustrates me in a number of ways: no live view for tabletop macro, only one cross-type focus point, no auto-dust prevention; I believe the 50D will alleviate if not resolve those frustrations.
I also pulled the trigger on a Tokina 11-16 f/2.8. I borrowed a Tamron 11-18 f/4.5-5.6 on the Keeble- and Tamron-sponsored trip to Alcatraz a couple weeks ago; I fell in love with its wideness, but its slowness frustrated me. The reviews of the Tokina on many sites totally intrigued me.
I can't wait to see what that wonderful 50D sensor will capture when paired with that super-sharp new lens, and my all my other beautiful lenses as well.
I'm used to dealing with an occasionally noisy but wonderful sensor; my favorite camera right now is my G9, which is quite noisy above ISO 400. Noise is manageable when you know how to avoid or deal with it.
I think Nik is right: the 50D is getting a bad rap, and the DPReview seems a bit heavy-handed, if not unfair. I expect the 50D to prove itself to me when its in my hands. I honestly can't to get started with it.
Neil,
forget the whole noise issue for a second, let's assume 50D is marginally the same as 40D in this department (although I do believe Ken is correct speaking of effective one stop improvement, and my own iso1600/iso3200 shots prove that to me more than any reviews in the world).
Much better screen, new and very useable highlight optimization mode..
And by the way, not all the shooting happens in lowlight AND at high ISO at the same time. So, for the 99% of time 50D owners can safely enjoy new 15mp resolution and don't even worry about those pesky effects that seem to draw so much of attention lately..
And, btw, it focuses MUCH better in twililight than 40D ever did
I'm enough of a geek to get aroused just holding the box of a new gadget
So, I'm drooling reading your descriptions of the 50D, which doesn't help my stern perusal of its 'limitations'! (I also drooled over these shots of yours )
How to satisfy my 'need for the newest'? I'm not enthusiastic about the 5DII, and the 60D might have that elephant - video - in its luggage, which I dread. The 50D doesn't tempt me enough to part from my 40D
What do you think of the camera manufacturers attempting to build 'photographers' skills' into cameras? What I mean is, with knowledge and experience photographers develop all kinds of skills which allow them to get keeper shots quickly and successfully without elaborate help from the camera. For example, a photographer I know can manually while looking through the viewfinder quickly set all the parameters to grab an opportunity as it presents.
Seems to me there is a trend to try to build those kinds of skills into hardware, when they really need, in the end, to be painstakingly developed by the person on the end of it.
That's a suspicion I have about some of the latest camera features (and you know I'm no Luddite).
I'd rather R&D go into improving the tool (ISO, dynamic range, focusing, stabilisation ...) rather than into trying to make the tool the photographer as well.
I'm enough of a geek to get aroused just holding the box of a new gadget
So, I'm drooling reading your descriptions of the 50D, which doesn't help my stern perusal of its 'limitations'! (I also drooled over these shots of yours )
How to satisfy my 'need for the newest'? I'm not enthusiastic about the 5DII, and the 60D might have that elephant - video - in its luggage, which I dread. The 50D doesn't tempt me enough to part from my 40D
What do you think of the camera manufacturers attempting to build 'photographers' skills' into cameras? What I mean is, with knowledge and experience photographers develop all kinds of skills which allow them to get keeper shots quickly and successfully without elaborate help from the camera. For example, a photographer I know can manually while looking through the viewfinder quickly set all the parameters to grab an opportunity as it presents.
Seems to me there is a trend to try to build those kinds of skills into hardware, when they really need, in the end, to be painstakingly developed by the person on the end of it.
That's a suspicion I have about some of the latest camera features (and you know I'm no Luddite).
I'd rather R&D go into improving the tool (ISO, dynamic range, focusing, stabilisation ...) rather than into trying to make the tool the photographer as well.
Neil, they are trying... However, different people have different needs and given it's all hardware yet it's hard to come by with a device that would satisfy all. However, I see more smart software in the future, and hopefully one day they put a real OS into these things (afterall, even cellphones have one nowadays) - THEN the real fun begins...
Neil, they are trying... However, different people have different needs and given it's all hardware yet it's hard to come by with a device that would satisfy all. However, I see more smart software in the future, and hopefully one day they put a real OS into these things (afterall, even cellphones have one nowadays) - THEN the real fun begins...
(although I do believe Ken is correct speaking of effective one stop improvement, and my own iso1600/iso3200 shots prove that to me more than any reviews in the world).
So, when comparing the 40D and 50D, while they have essentially the same per pixel noise, the 50D has a stop (1.4 times) more pixels which means 50D sensor has stop more low light sensitivity.
I've thought about this long and hard, and it's not true. To gain 1 stop, means the AREA that the light falls on has to DOUBLE. The area occupied by the pixels remains essentially the same, regardless of the number of pixels. I'll post the math once I figure out a convenient way to do it.
I've thought about this long and hard, and it's not true. To gain 1 stop, means the AREA that the light falls on has to DOUBLE. The area occupied by the pixels remains essentially the same, regardless of the number of pixels. I'll post the math once I figure out a convenient way to do it.
With the new microlens design they saved on unused space. Check out the specs.
I've thought about this long and hard, and it's not true. To gain 1 stop, means the AREA that the light falls on has to DOUBLE. The area occupied by the pixels remains essentially the same, regardless of the number of pixels. I'll post the math once I figure out a convenient way to do it.
You're right, I got that wrong. When you average pixels, the noise decreases by the square root of the number of pixels you average. Its an apropos night to bring that up as that is the same principle which is used to determine the margin of error on polls: poll 100 people and the margin of error is 10%, poll 400 people and the margin of error is 5%.
Applying it to cameras, when you scale from 14MP to 10MP you are (on average) polling 1.4 pixels to generate each new pixel which reduces the noise by the square root of 1.4 or half a stop. To get a full stop improvement in a 10MP image you would (as you say) need to start with 20MP (e.g. 5DmII).
Slightly off topic for this thread, but note that the 5DII video mode is scaling down by a factor of 8 (20MP to 2.5MP) for 3 stops improvement in noise level. No wonder its low light video looks so good.
scratch Call me thick, but I don't get this whole discussion about noise, stops and pixel density. I would think that you need to make an (erroneous) assumption that the number of noise pixels remains constant across the various sensors to draw these conclusions. But isn't the number of noise pixels an empirical question? What am I missing here?
scratch Call me thick, but I don't get this whole discussion about noise, stops and pixel density. I would think that you need to make an (erroneous) assumption that the number of noise pixels remains constant across the various sensors to draw these conclusions. But isn't the number of noise pixels an empirical question? What am I missing here?
Richard, in short: you don't miss a thing Neither do I. Pictures look better and I can do more. That sums it up for me.
scratch Call me thick, but I don't get this whole discussion about noise, stops and pixel density. I would think that you need to make an (erroneous) assumption that the number of noise pixels remains constant across the various sensors to draw these conclusions. But isn't the number of noise pixels an empirical question? What am I missing here?
The asumption I am making is that what you really care about is noise in your output rather than in your capture. In other words, it is noise in your print (or online JPEG) that matters rather than in your RAW file.
As an example, if I print an 8x10 at 360 dpi, I am sending a 10MP file to my printer no matter what the resolution of my camera. If I start with a 15MP RAW file then at some point during my process I am going to be scaling my image from 15MP to 10MP. Because scaling the image involves averaging multiple pixels, the file sent to my printer will have lower noise than my original capture. If you never print larger than 8x10, then the 50% more pixels in the 50D file will not give you a sharper print. However those extra pixels in your RAW file will serve to reduce your shadow noise in a way that will be visible in the final print.
My personal opinion is that comparing 100% crops between cameras of different resolution gives you no real information about image quality because what matters is what the file you send to your printer which has been scaled to the output resolution.
comparing 100% crops between cameras of different resolution gives you no real information about image quality because what matters is what the file you send to your printer which has been scaled to the output resolution.
Yep, this reads better, and is TRUE! ...for printing. What about jpeg conversion?
However, I see more smart software in the future, and hopefully one day they put a real OS into these things (afterall, even cellphones have one nowadays) - THEN the real fun begins...
What? You want to trade ERR99 for a BSOD
More seriously, when does the code already running on these turn into a real OS? You have histograms, Liveview, HD output, JPEG algorithms, all sorts of stuff. Is it some specific level of interlacing of these? You have a processor, I/O, the whole bit already (pun intended). Symbian and Palm in their early days were considered very lean OSs, and yet probably not as sophisticated as what we have on a DSLR today. Just wondering.
And it would be interesting to see what is under the hood of the new 5D MkII.
"Don't ask me what I think of you, I might not give the answer that you want me to. Oh well."
-Fleetwood Mac
More seriously, when does the code already running on these turn into a real OS? You have histograms, Liveview, HD output, JPEG algorithms, all sorts of stuff. Is it some specific level of interlacing of these? You have a processor, I/O, the whole bit already (pun intended). Symbian and Palm in their early days were considered very lean OSs, and yet probably not as sophisticated as what we have on a DSLR today. Just wondering.
And it would be interesting to see what is under the hood of the new 5D MkII.
When they allow user access to the OS? Sort of what CHDK does for compacts.
Canon's dslr cameras have run a version of DOS, VxWorks and lately an in-house developed DryOS.
Comments
Don't quite follow you on this point. Seems to be a contradiction. But might only be the way you are using terms?
Also, it seems to be generally said that noise visible on a monitor will not necessarily be visible when printed (?).
http://www.behance.net/brosepix
I can't figure this one out...
Bugs
Spiders
Flowers
Unfortunately, Phil did not write the review. Lars Rehm and Richard Butler wrote the review for DPReview.
Instead of saying that the 50D has the same per pixel noise, they said (according to the following link):
http://www.dpreview.com/reviews/canoneos50d/page18.asp
"Despite the fact that the 50D is the newer camera it shows visibly more chroma and luminance noise than the 40D. Considering the 50D's much more tightly packed sensor (4.5 MP/cm² vs 3.1 MP/cm² on the 40D) this comes hardly as a surprise. It would have been unreasonable to expect Canon's engineers to overcome the laws of physics."
It is still my contention that the 50D handles noise in a different way than the linear method employed by the 40D. If indeed the 50D uses a more selective algorithm that targets shadow areas to apply a higher level of NR than highlight areas, that could explain the visible "appearance" of less noise on the 50D images. That could also explain the test results measuring more noise, since the test measures all noise where the human brain is more sensitive to shadow noise.
Otherwise you are right that "binning" pixels will generally reduce the visible effects of noise and that binning will occur during most printing operations until close to a 1:1 print ratio and above. The extra pixels of the 50D should be a benefit for most printing situations over the 40D.
I also agree that it would appear that we have reached the practical diffraction limits for photosite density at the current technology APS-C sized imager with the density of the 50D imager (... and the Pentax K20D imager as well).
Moderator of the Cameras and Accessories forums
Once again I wonder at the wisdom of buying a new body for the new NR software it uses. And NR that you have only very crude control over at that.
As the quote below says (referring to pixel binning on the sensor-chip, but the principal is the same for printing?), there is a trade-off of binning in loss of resolution. So, camera applied NR + binning ( + conversion to jpeg) = loss of resolution, and more pixels = increase in resolution. Which wins out? Where do we get?
The primary benefit of pixel binning is to improve the signal-to-noise ratio in low light conditions at the expense of spatial resolution. Summation of many charge packets reduces the read noise level and produces an improvement in signal equal to the binning factor.
Contributing Authors
Mortimer Abramowitz - Olympus America, Inc., Two Corporate Center Drive., Melville, New York, 11747.
Michael W. Davidson - National High Magnetic Field Laboratory, 1800 East Paul Dirac Dr., The Florida State University, Tallahassee, Florida, 32310.
http://www.behance.net/brosepix
I actually cancelled my order on the 50D and went ahead on ordered the 40D and put the 550.00 difference towards a new lens. Thanks for all the discussion guys.
www.hallphotography.smugmug.com
Tool Box:
Canon Digital ReBeL XTi (40d on order)
85mm f1.8, EF-75-300mm f1:4-5.6, 28-55 f3.5
forget the whole noise issue for a second, let's assume 50D is marginally the same as 40D in this department (although I do believe Ken is correct speaking of effective one stop improvement, and my own iso1600/iso3200 shots prove that to me more than any reviews in the world).
There are few other things that 50D sports off.
Highly improved LiveView, with its live focusing and live histogram alone would justify an upgrade for me. I wouldn't be able to get these shots without it
http://nik.smugmug.com/photos/407682120_Tsn7W-XL.jpg
http://nik.smugmug.com/photos/395879464_R4gV8-O.jpg
Much better screen, new and very useable highlight optimization mode..
And by the way, not all the shooting happens in lowlight AND at high ISO at the same time. So, for the 99% of time 50D owners can safely enjoy new 15mp resolution and don't even worry about those pesky effects that seem to draw so much of attention lately..
And, btw, it focuses MUCH better in twililight than 40D ever did
I don't know about the low light situations, but I did see some shots in daylight (40d vs 50d) and the darker areas where about the same, noise-wise.
www.ivarborst.nl & smugmug
Know what you mean. I sat in one once, got out and then went right to a 172 for all my training.
Heheh....it was a tad tricky getting in and out, especially with a camera. Regarding the showdowed areas/banding, I think we sometimes expect too much out of high ISO. Sometimes we can nail it, clean it and the images are good to go....sometimes not.
Some of the features of the 50D look way cool....especially calibrating each lens. Also, I'm sure the highlight/shadow retention is better. I haven't seen the new DPP software yet for the 50D but only can assume one could implement some of those features in post if shooting RAW. The 40D is a great camera, the 50D may be a bit better and even more so on some things. I'd be interested to see the focus speed differences and accuracy in AI Servo.
One thing for sure, for those that have an earlier version (xt, 10, 20, 30) will be blown away at the advancements. I certainly was with the 40D. The focus was much better all around in each mode.
NAPP Member | Canon Shooter
Weddings/Portraits and anything else that catches my eye.
www.daveswartz.com
Model Mayhem site http://www.modelmayhem.com/686552
That's the part I'm wondering about. I have a 30d, but is it really worth upgrading for me? The camera functions fine, I'm not 'missing' anything right now. I'm sure it will be better at some points, but is it worth it for me? If I'm going to spend that much money on something I don't need, I might as well go for the 5dmkII, I think.
www.ivarborst.nl & smugmug
Then again, if 30D works just fine, why considering that path either?
The 5-series is a different tool, in my eyes, and something that I think would go better with my style of shooting.
www.ivarborst.nl & smugmug
For me I decided it was worth it, so I went on with it, and thus far no regrets...
I also pulled the trigger on a Tokina 11-16 f/2.8. I borrowed a Tamron 11-18 f/4.5-5.6 on the Keeble- and Tamron-sponsored trip to Alcatraz a couple weeks ago; I fell in love with its wideness, but its slowness frustrated me. The reviews of the Tokina on many sites totally intrigued me.
I can't wait to see what that wonderful 50D sensor will capture when paired with that super-sharp new lens, and my all my other beautiful lenses as well.
I'm used to dealing with an occasionally noisy but wonderful sensor; my favorite camera right now is my G9, which is quite noisy above ISO 400. Noise is manageable when you know how to avoid or deal with it.
I think Nik is right: the 50D is getting a bad rap, and the DPReview seems a bit heavy-handed, if not unfair. I expect the 50D to prove itself to me when its in my hands. I honestly can't to get started with it.
I'm enough of a geek to get aroused just holding the box of a new gadget
So, I'm drooling reading your descriptions of the 50D, which doesn't help my stern perusal of its 'limitations'!
How to satisfy my 'need for the newest'? I'm not enthusiastic about the 5DII, and the 60D might have that elephant - video - in its luggage, which I dread. The 50D doesn't tempt me enough to part from my 40D
What do you think of the camera manufacturers attempting to build 'photographers' skills' into cameras? What I mean is, with knowledge and experience photographers develop all kinds of skills which allow them to get keeper shots quickly and successfully without elaborate help from the camera. For example, a photographer I know can manually while looking through the viewfinder quickly set all the parameters to grab an opportunity as it presents.
Seems to me there is a trend to try to build those kinds of skills into hardware, when they really need, in the end, to be painstakingly developed by the person on the end of it.
That's a suspicion I have about some of the latest camera features (and you know I'm no Luddite).
I'd rather R&D go into improving the tool (ISO, dynamic range, focusing, stabilisation ...) rather than into trying to make the tool the photographer as well.
http://www.behance.net/brosepix
Neil, they are trying... However, different people have different needs and given it's all hardware yet it's hard to come by with a device that would satisfy all. However, I see more smart software in the future, and hopefully one day they put a real OS into these things (afterall, even cellphones have one nowadays) - THEN the real fun begins...
Oh well, the heck! Bring it all on!!
http://www.behance.net/brosepix
I've thought about this long and hard, and it's not true. To gain 1 stop, means the AREA that the light falls on has to DOUBLE. The area occupied by the pixels remains essentially the same, regardless of the number of pixels. I'll post the math once I figure out a convenient way to do it.
Bugs
Spiders
Flowers
You're right, I got that wrong. When you average pixels, the noise decreases by the square root of the number of pixels you average. Its an apropos night to bring that up as that is the same principle which is used to determine the margin of error on polls: poll 100 people and the margin of error is 10%, poll 400 people and the margin of error is 5%.
Applying it to cameras, when you scale from 14MP to 10MP you are (on average) polling 1.4 pixels to generate each new pixel which reduces the noise by the square root of 1.4 or half a stop. To get a full stop improvement in a 10MP image you would (as you say) need to start with 20MP (e.g. 5DmII).
Slightly off topic for this thread, but note that the 5DII video mode is scaling down by a factor of 8 (20MP to 2.5MP) for 3 stops improvement in noise level. No wonder its low light video looks so good.
The asumption I am making is that what you really care about is noise in your output rather than in your capture. In other words, it is noise in your print (or online JPEG) that matters rather than in your RAW file.
As an example, if I print an 8x10 at 360 dpi, I am sending a 10MP file to my printer no matter what the resolution of my camera. If I start with a 15MP RAW file then at some point during my process I am going to be scaling my image from 15MP to 10MP. Because scaling the image involves averaging multiple pixels, the file sent to my printer will have lower noise than my original capture. If you never print larger than 8x10, then the 50% more pixels in the 50D file will not give you a sharper print. However those extra pixels in your RAW file will serve to reduce your shadow noise in a way that will be visible in the final print.
My personal opinion is that comparing 100% crops between cameras of different resolution gives you no real information about image quality because what matters is what the file you send to your printer which has been scaled to the output resolution.
Yep, this reads better, and is TRUE! ...for printing. What about jpeg conversion?
http://www.behance.net/brosepix
What? You want to trade ERR99 for a BSOD
More seriously, when does the code already running on these turn into a real OS? You have histograms, Liveview, HD output, JPEG algorithms, all sorts of stuff. Is it some specific level of interlacing of these? You have a processor, I/O, the whole bit already (pun intended). Symbian and Palm in their early days were considered very lean OSs, and yet probably not as sophisticated as what we have on a DSLR today. Just wondering.
And it would be interesting to see what is under the hood of the new 5D MkII.
-Fleetwood Mac
When they allow user access to the OS? Sort of what CHDK does for compacts.
Canon's dslr cameras have run a version of DOS, VxWorks and lately an in-house developed DryOS.
http://pyryekholm.kuvat.fi/
And multifunction printers...
I would wager a good beer that the 50D runs a version of DryOS.
40D has VxWorks, there's even a CHDK port in the works for it.
http://pyryekholm.kuvat.fi/