Its not really just that simple either. Simply raising the ISO also increases noise in an underexposed picture. Its a balancing act of aperture, shutter time, and ISO without increasing the noise.
Quote:
Originally Posted by mercury7
ok...so here is the scoop, the difference you see is based on movement, what apple has done is used the accelerometer as an image stabiliser somehow. ie taking the photo when no motion is detected.....so my bet would be this plus a software treak on iso in low light.
but it is mostly movement that is flooring everyone. basic photography, slow shutter equals blurry photos, low light equals slow shutter so by tweaking the iso in lower light they are allowing a faster shutter speed.
Depends what. The type of blur? Motion blur can be excluded with the probability close to 1.
That the picture was post-processed on PC? No, one can not say this for sure; but I have already a lot of small inklings being pieces of the puzzle of my guesswork, which suggest the probability of that is well above 0.
I'd be seriously disappointed if Apple would use a 3.2 megapixel camera in the new iPhone. That would be an "update" not an "upgrade". A min. of 5 megapixels would put the iPhone at the same level as the upcoming Samsung I7500. And shouldn't the iPhone 3.0 video output "support" iMovie's HD format?
A complete waste.
In another two years a 5 MP camera phone might produce a good image.
This will happen and its largely because of improved software that will render pictures free of artifacts and noise. Essentially the software will become better at hiding defects.
But its all a sliding scale. As camera phones become better, all cameras will be better, so a camera phone will always be at the bottom of acceptability.
Quote:
Originally Posted by melgross
In another two years a 5 MP camera phone might produce a good image.
This will happen and its largely because of improved software that will render pictures free of artifacts and noise. Essentially the software will become better at hiding defects.
But its all a sliding scale. As camera phones become better, all cameras will be better, so a camera phone will always be at the bottom of acceptability.
Hopefully the sensors will be better too.
But as you say, they will always remain at the bottom of the barrel.
What will possibly happen, should they get "real" zoom lenses rather than this digital cropping garbage, is that the very bottom tier of compacts will go away.
What will possibly happen, should they get "real" zoom lenses rather than this digital cropping garbage, is that the very bottom tier of compacts will go away.
Is that even feasible with how the thin the iPhone is as an optical zoom would require moving parts?
This will happen and its largely because of improved software that will render pictures free of artifacts and noise. Essentially the software will become better at hiding defects.
But its all a sliding scale. As camera phones become better, all cameras will be better, so a camera phone will always be at the bottom of acceptability.
Actually, I think the same thing will happen with camera phones as has happened with music players-- you hit a point of "good enough" that balances storage and bandwidth concerns against quality, at which point the fact that there is more sophisticated equipment available becomes irrelevant.
Sure, the dedicated cameras will continue to spec whore their 10, 20, 30, ad infinitum megapixel sensors, but a low noise 5MP camera with a half decent lens and good processing firmware is going to produce as good a picture as most people would ever want or need. Unless making prints larger than 8x10 or heavy cropping become the norm amongst casual snapshot takers, I would guess we're rapidly approaching the point were camera phones will provide 90% of what people want cameras to do, and any further refinements will be around software and lens materials technology.
Admittedly I'm versed more in digital video than digital stills, but I find their is quite a bit of over lap.
In the digital video world sensors are improving but not nearly at the pace that software improves. Everyone basically has access to the same sensor technology. The real differentiating secret sauce is the software algorithms that processes and render the image. Most companies give general details and use marketing terms to describe their software process. No one gives any real detail about how their software actually works. They keep it completely secret.
The race between software algorithms is essentially the ability to brag about the largest recording resolution that compress video down to a size that is manageable with currently affordable computer workflow, while limiting artifacts and noise. Video codec's are quickly advancing and are far more efficient today than they were a few years ago. Sensors themselves are somewhat better but not much better than a few of years ago.
Quote:
Originally Posted by melgross
Hopefully the sensors will be better too.
But as you say, they will always remain at the bottom of the barrel.
What will possibly happen, should they get "real" zoom lenses rather than this digital cropping garbage, is that the very bottom tier of compacts will go away.
Is that even feasible with how the thin the iPhone is as an optical zoom would require moving parts?
There's some new lens tech out there in experimental designs that will revolutionize these tiny lenses.
One that Phillips has been working on is a liquid lens that changes shape with the application of a very small electrical signal. Others use small "muscles" to distort the lens. There are even designs that are further out than those. One of them changes the optical properties of the material to change not only the focus, but the focal length of the lens.
None of these approaches are suitable for large lenses.
Some of these are being looked into as lens replacements for human eye lenses too.
I did read about that somewhere. But this is Apple and using untested tech is usually not their modus operandi, so should we expect an optical zoom next month?
Actually, I think the same thing will happen with camera phones as has happened with music players-- you hit a point of "good enough" that balances storage and bandwidth concerns against quality, at which point the fact that there is more sophisticated equipment available becomes irrelevant.
Sure, the dedicated cameras will continue to spec whore their 10, 20, 30, ad infinitum megapixel sensors, but a low noise 5MP camera with a half decent lens and good processing firmware is going to produce as good a picture as most people would ever want or need. Unless making prints larger than 8x10 or heavy cropping become the norm amongst casual snapshot takers, I would guess we're rapidly approaching the point were camera phones will provide 90% of what people want cameras to do, and any further refinements will be around software and lens materials technology.
Those of us with high rez cameras ARE printing to large sizes. 8 x 10 needs about 7 MP for a really sharp image. But the lens must be good too. To get the same sharpness in a 16 x 20, twice the linear size, requires 28 MP, and so it goes.
Admittedly I'm versed more in digital video than digital stills, but I find their is quite a bit of over lap.
In the digital video world sensors are improving but not nearly at the pace that software improves. Everyone basically has access to the same sensor technology. The real differentiating secret sauce is the software algorithms that processes and render the image. Most companies give general details and use marketing terms to describe their software process. No one gives any real detail about how their software actually works. They keep it completely secret.
The race between software algorithms is essentially the ability to brag about the largest recording resolution that compress video down to a size that is manageable with currently affordable computer workflow, while limiting artifacts and noise. Video codec's are quickly advancing and are far more efficient today than they were a few years ago. Sensors themselves are somewhat better but not much better than a few of years ago.
There are differences in the approaches between the manufacturers of sensors that make a difference. I hear of some new tech that may be here in a year or two that is better than whatever is out now, but we'll have to wait for it and see.
Software will continue to increase its presence. More and more, software is making digital photography a far better place than film ever was.
I did read about that somewhere. But this is Apple and using tested tech is usually not their modus operandi, so should we expect an optical zoom next month?
I wish I could get my friends there to be specific about things, but they rarely are.
I agree the use of camera phones will grow. But I don't believe camera companies will cede their point and shoot market to phone manufacturers. They will adjust and figure out how to out market phones.
Quote:
Originally Posted by addabox
Sure, the dedicated cameras will continue to spec whore their 10, 20, 30, ad infinitum megapixel sensors, but a low noise 5MP camera with a half decent lens and good processing firmware is going to produce as good a picture as most people would ever want or need. Unless making prints larger than 8x10 or heavy cropping become the norm amongst casual snapshot takers, I would guess we're rapidly approaching the point were camera phones will provide 90% of what people want cameras to do, and any further refinements will be around software and lens materials technology.
Especially the firmware update for that random blue line on the right side of the picture. Or the firmware update that fixes noise exhibited at 5800 kelvin more than at 2800 kelvin.
For filmmaking, digital acquisition is still the wild wild west of non-standardization and experimental workflows. Film is still largely used because it has a standard workflow and predictable outcome.
Quote:
Originally Posted by melgross
Software will continue to increase its presence. More and more, software is making digital photography a far better place than film ever was.
Especially the firmware update for that random blue line on the right side of the picture. Or the firmware update that fixes noise exhibited at 5800 kelvin more than at 2800 kelvin.
For filmmaking, digital acquisition is still the wild wild west of non-standardization and experimental workflows. Film is still largely used because it has a standard workflow and predictable outcome.
We said the same thing in the publishing industry around 1990.
The big fight back then was making color seps using film, or by scanning.
This is the same problem with your industry.
The expression for us was this, and it's the same for you:
Early on, if you worked digitally, and you were VERY good, you would be fit into the workflow.
Later, if you worked in analog, and you were VERY good, you would be fit into the workflow.
The biggest problem in filmmaking is that everyone competes in a total sense. Every camera manufacturer has their own codec and HD format. Sony and Panasonic have around 12 different HD formats between the two of them. Then every other camera manufacturer has their own codec/format.
Post Production houses then have acquire equipment and software to deal with all of those formats. On top of eventually most of these formats are eventually going to be abandoned and some companies are not going to survive.
The only resolution I can see to this problem is one standard post production format that is supported by everyone regardless of which camera is used. This would go along way to the adoption of digital moviemaking, but no one seems motivated to do this.
Funny enough in the lower budget realm the one format that is filling the role of this universal format is Apple's Pro Res 422. A growing number of cameras and acquisition equipment can record data natively in Pro Res 422.
Quote:
Originally Posted by melgross
We said the same thing in the publishing industry around 1990.
Just wanted to point out that the current iphone under ideal circumstances will
take "keeper" photos, here is an example....note the photo it self is not that great but it is a good photo to blow up and analyze ,when you look at it full size, the sharpness, purity of the chrome fixtures etc is really pretty great for a 2 megapixel camera phone. Sure there is noise everywhere especially in the darker parts but I can't complain overall.
Those of us with high rez cameras ARE printing to large sizes. 8 x 10 needs about 7 MP for a really sharp image. But the lens must be good too. To get the same sharpness in a 16 x 20, twice the linear size, requires 28 MP, and so it goes.
Of course, but that certainly isn't going to be the mass market norm, any more than "audiophile" home stereo systems are going come roaring back against cheap digital players and docking systems.
The home audio industry has been an object lesson in how, once the tech reaches a certain point, the up-sell gets harder and harder and further advances are just churning a small percentage of the market.
Comments
ok...so here is the scoop, the difference you see is based on movement, what apple has done is used the accelerometer as an image stabiliser somehow. ie taking the photo when no motion is detected.....so my bet would be this plus a software treak on iso in low light.
but it is mostly movement that is flooring everyone. basic photography, slow shutter equals blurry photos, low light equals slow shutter so by tweaking the iso in lower light they are allowing a faster shutter speed.
Very unlikely. This is not directional blur. I still suspect the picture having been taken with iPhone was "spoiled" artificially.
You cannot tell this from that picture.
Depends what. The type of blur? Motion blur can be excluded with the probability close to 1.
That the picture was post-processed on PC? No, one can not say this for sure; but I have already a lot of small inklings being pieces of the puzzle of my guesswork, which suggest the probability of that is well above 0.
I'd be seriously disappointed if Apple would use a 3.2 megapixel camera in the new iPhone. That would be an "update" not an "upgrade". A min. of 5 megapixels would put the iPhone at the same level as the upcoming Samsung I7500. And shouldn't the iPhone 3.0 video output "support" iMovie's HD format?
A complete waste.
In another two years a 5 MP camera phone might produce a good image.
But its all a sliding scale. As camera phones become better, all cameras will be better, so a camera phone will always be at the bottom of acceptability.
In another two years a 5 MP camera phone might produce a good image.
This will happen and its largely because of improved software that will render pictures free of artifacts and noise. Essentially the software will become better at hiding defects.
But its all a sliding scale. As camera phones become better, all cameras will be better, so a camera phone will always be at the bottom of acceptability.
Hopefully the sensors will be better too.
But as you say, they will always remain at the bottom of the barrel.
What will possibly happen, should they get "real" zoom lenses rather than this digital cropping garbage, is that the very bottom tier of compacts will go away.
What will possibly happen, should they get "real" zoom lenses rather than this digital cropping garbage, is that the very bottom tier of compacts will go away.
Is that even feasible with how the thin the iPhone is as an optical zoom would require moving parts?
This will happen and its largely because of improved software that will render pictures free of artifacts and noise. Essentially the software will become better at hiding defects.
But its all a sliding scale. As camera phones become better, all cameras will be better, so a camera phone will always be at the bottom of acceptability.
Actually, I think the same thing will happen with camera phones as has happened with music players-- you hit a point of "good enough" that balances storage and bandwidth concerns against quality, at which point the fact that there is more sophisticated equipment available becomes irrelevant.
Sure, the dedicated cameras will continue to spec whore their 10, 20, 30, ad infinitum megapixel sensors, but a low noise 5MP camera with a half decent lens and good processing firmware is going to produce as good a picture as most people would ever want or need. Unless making prints larger than 8x10 or heavy cropping become the norm amongst casual snapshot takers, I would guess we're rapidly approaching the point were camera phones will provide 90% of what people want cameras to do, and any further refinements will be around software and lens materials technology.
In the digital video world sensors are improving but not nearly at the pace that software improves. Everyone basically has access to the same sensor technology. The real differentiating secret sauce is the software algorithms that processes and render the image. Most companies give general details and use marketing terms to describe their software process. No one gives any real detail about how their software actually works. They keep it completely secret.
The race between software algorithms is essentially the ability to brag about the largest recording resolution that compress video down to a size that is manageable with currently affordable computer workflow, while limiting artifacts and noise. Video codec's are quickly advancing and are far more efficient today than they were a few years ago. Sensors themselves are somewhat better but not much better than a few of years ago.
Hopefully the sensors will be better too.
But as you say, they will always remain at the bottom of the barrel.
What will possibly happen, should they get "real" zoom lenses rather than this digital cropping garbage, is that the very bottom tier of compacts will go away.
Is that even feasible with how the thin the iPhone is as an optical zoom would require moving parts?
There's some new lens tech out there in experimental designs that will revolutionize these tiny lenses.
One that Phillips has been working on is a liquid lens that changes shape with the application of a very small electrical signal. Others use small "muscles" to distort the lens. There are even designs that are further out than those. One of them changes the optical properties of the material to change not only the focus, but the focal length of the lens.
None of these approaches are suitable for large lenses.
Some of these are being looked into as lens replacements for human eye lenses too.
Actually, I think the same thing will happen with camera phones as has happened with music players-- you hit a point of "good enough" that balances storage and bandwidth concerns against quality, at which point the fact that there is more sophisticated equipment available becomes irrelevant.
Sure, the dedicated cameras will continue to spec whore their 10, 20, 30, ad infinitum megapixel sensors, but a low noise 5MP camera with a half decent lens and good processing firmware is going to produce as good a picture as most people would ever want or need. Unless making prints larger than 8x10 or heavy cropping become the norm amongst casual snapshot takers, I would guess we're rapidly approaching the point were camera phones will provide 90% of what people want cameras to do, and any further refinements will be around software and lens materials technology.
Those of us with high rez cameras ARE printing to large sizes. 8 x 10 needs about 7 MP for a really sharp image. But the lens must be good too. To get the same sharpness in a 16 x 20, twice the linear size, requires 28 MP, and so it goes.
Admittedly I'm versed more in digital video than digital stills, but I find their is quite a bit of over lap.
In the digital video world sensors are improving but not nearly at the pace that software improves. Everyone basically has access to the same sensor technology. The real differentiating secret sauce is the software algorithms that processes and render the image. Most companies give general details and use marketing terms to describe their software process. No one gives any real detail about how their software actually works. They keep it completely secret.
The race between software algorithms is essentially the ability to brag about the largest recording resolution that compress video down to a size that is manageable with currently affordable computer workflow, while limiting artifacts and noise. Video codec's are quickly advancing and are far more efficient today than they were a few years ago. Sensors themselves are somewhat better but not much better than a few of years ago.
There are differences in the approaches between the manufacturers of sensors that make a difference. I hear of some new tech that may be here in a year or two that is better than whatever is out now, but we'll have to wait for it and see.
Software will continue to increase its presence. More and more, software is making digital photography a far better place than film ever was.
I did read about that somewhere. But this is Apple and using tested tech is usually not their modus operandi, so should we expect an optical zoom next month?
I wish I could get my friends there to be specific about things, but they rarely are.
Sure, the dedicated cameras will continue to spec whore their 10, 20, 30, ad infinitum megapixel sensors, but a low noise 5MP camera with a half decent lens and good processing firmware is going to produce as good a picture as most people would ever want or need. Unless making prints larger than 8x10 or heavy cropping become the norm amongst casual snapshot takers, I would guess we're rapidly approaching the point were camera phones will provide 90% of what people want cameras to do, and any further refinements will be around software and lens materials technology.
For filmmaking, digital acquisition is still the wild wild west of non-standardization and experimental workflows. Film is still largely used because it has a standard workflow and predictable outcome.
Software will continue to increase its presence. More and more, software is making digital photography a far better place than film ever was.
Especially the firmware update for that random blue line on the right side of the picture. Or the firmware update that fixes noise exhibited at 5800 kelvin more than at 2800 kelvin.
For filmmaking, digital acquisition is still the wild wild west of non-standardization and experimental workflows. Film is still largely used because it has a standard workflow and predictable outcome.
We said the same thing in the publishing industry around 1990.
The big fight back then was making color seps using film, or by scanning.
This is the same problem with your industry.
The expression for us was this, and it's the same for you:
Early on, if you worked digitally, and you were VERY good, you would be fit into the workflow.
Later, if you worked in analog, and you were VERY good, you would be fit into the workflow.
Post Production houses then have acquire equipment and software to deal with all of those formats. On top of eventually most of these formats are eventually going to be abandoned and some companies are not going to survive.
The only resolution I can see to this problem is one standard post production format that is supported by everyone regardless of which camera is used. This would go along way to the adoption of digital moviemaking, but no one seems motivated to do this.
Funny enough in the lower budget realm the one format that is filling the role of this universal format is Apple's Pro Res 422. A growing number of cameras and acquisition equipment can record data natively in Pro Res 422.
We said the same thing in the publishing industry around 1990.
take "keeper" photos, here is an example....note the photo it self is not that great but it is a good photo to blow up and analyze ,when you look at it full size, the sharpness, purity of the chrome fixtures etc is really pretty great for a 2 megapixel camera phone. Sure there is noise everywhere especially in the darker parts but I can't complain overall.
http://public.fotki.com/Mercury7/iph...s/img0031.html
p.s. you have to click get original photo underneath the picture to see the full size version
Those of us with high rez cameras ARE printing to large sizes. 8 x 10 needs about 7 MP for a really sharp image. But the lens must be good too. To get the same sharpness in a 16 x 20, twice the linear size, requires 28 MP, and so it goes.
Of course, but that certainly isn't going to be the mass market norm, any more than "audiophile" home stereo systems are going come roaring back against cheap digital players and docking systems.
The home audio industry has been an object lesson in how, once the tech reaches a certain point, the up-sell gets harder and harder and further advances are just churning a small percentage of the market.