A typical 8 Mpix sensor would be much better as a 2 Mpix sensor of the same total detector size and sensitivity... So if you have a crappy electron well that can hold 10^4 photoelectrons...
That's all fine, but what really matters how many more electrons the well will hold if you make the sensor pixel 4 times larger in area and how many more photoelectrons you can generate at the same exposure settings. If both of them are scaled by a factor 4, you can just as well average four adjacent small pixels to create a lower-resolution image.
Of course this averaging should be done in the camera/phone firmware before the JPEG compression step.
Not so; the averaged pixel would be noisier than a larger pixel of the same area as the aggregate even if the number of detected photoelectrons scaled perfectly.
I mentioned only photon shot noise, but other sources of noise exist in CCD and CMOS detectors, and their effects are additive (in a mean-square fashion). In particular, readout noise and bandwidth noise will make the situation worse if the same number of photoelectrons are divided among more pixels. Bandwidth noise increases with the bandwidth at which the chip operates; with four times as many pixels, the bandwidth must be higher to get an image in the same time, so the bandwidth noise will be higher per pixel. Similarly, readout noise depends on the number of pixels to be read, and is higher for four times as many pixels.
In aggregate, the noise in detectors which are equal in area and sensitivity will be rather higher if that area is divided among more pixels. Aggregating four pixels into one after readout does not help much. The aggregated pixel will always have a worse signal to noise ratio than a single pixel of the same size as the aggregated pixel. You could try reading this to get a basic grip on imaging noise. Or use your google-fu.
### This line should go close to the start, near where
### the current Acrobat Reader macro is defined
define(EVINCE, [repeat swallow(evince) fill needs_xembed: evince "$file"])
### this group should go in the Documents section, possibly
### replacing the defective existing group for PDF
application/pdf:pdf:PDF file
application/x-pdf:pdf:PDF file
text/pdf:pdf:PDF file
text/x-pdf:pdf:PDF file
EVINCE()
Firefox is pretty much the last browser to finally get PDF viewing support.
Although they appear to have got it right (based on my Android phone). Unlike Chromium and Google Chrome on Ubuntu/Xubuntu, where you have to fiddle with the/etc/mozpluggerrc file to get it to work right (and avoid the broken reference to Acrobat Reader).
Here's a hint. In/etc/mozpluggerrc, add the lines:
### This line should go close to the start, near where the current Acrobat Reader macro is defined
define(EVINCE, [repeat swallow(evince) fill needs_xembed: evince "$file"])
### this group should go in the Documents section, possibly replacing the defective existing group for PDF
application/pdf:pdf:PDF file
application/x-pdf:pdf:PDF file
text/pdf:pdf:PDF file
text/x-pdf:pdf:PDF file
EVINCE()
Lower res pictures with bigger pixels? That sounds more like "we've put in a lower resolution camera, and that's better".
No, it's better - provided they have made the pixels bigger. I'm sick of phones with so-called multi-megapixel cameras, which give noisy photos in the best of circumstances. A typical 8 Mpix sensor would be much better as a 2 Mpix sensor of the same total detector size and sensitivity, and sometimes, they should have even fewer pixels with the the same total detector area and sensitivity.
Here's the essential info: shot noise is unavoidable - it's intrinsic in the physics of photon arrival at the detector. The sigma of the output noise is the square root of the number of photoelectrons.
So if you have a crappy electron well that can hold 10^4 photoelectrons when full (a "decent" cellphone camera), the signal to noise ratio is barely 100 (10^4 divided by square root of 10^4). Similarly, the photon flux per pixel in good lighting will rarely exceed 10^5 photons per second per pixel, due to the tiny lens aperture and small pixel size. It's unsurprising that the images are utter crap, as the output gain must be cranked up (amplifying noise as well as signal) to get any shot in less than 1/100 second. People downsize their images in almost all circumstances, unless they're happy with blurry and/or noisy images. FWIW, this is borne out by my experience with my own Samsung Galaxy S3 and Nokia E70, my daughter's HTC Desire Z, a colleague's Nokia 920, a friend's Samsung Galaxy S2, another colleague's iPhone 4, and various other Nokia, Samsung, and Siemens phones belonging to family and friends They are all crappy in nearly all circumstances[*], unless downsized 2:1 or more (i.e. at most one quarter of the pixels).
In a DSLR, the much larger electron well means that a pixel can hold up to 10^6 photoelectrons, so the signal to noise ratio is closer to 10^3. Similarly, the larger aperture (there's a reason for those big lenses) and larger detector pixels mean that it gets a flux of more than 10^8 photoelectrons per second per pixel in typical lighting. That's why even action shots in 1/1000sec exposure can be sharp and have relatively low noise.
[*] Exception: a relatively long exposure shot of a still life scene, or a deliberately extended exposure shot of running water or similar (with hand support to improve steadiness), say 1/15 sec or thereabouts. Not what cameras in phones are commonly used for...
It is illegal when the big compagnies are loosing money....
Companies in which countries?
Consider the fate of AllOfMP3. The web site was operating in an ostensibly legal way according to Russian laws and paid royalties through ROMS, but were sued for trillions of dollars by the RIAA and IFPI (who had apparently refused royalties). Eventually, AllOfMP3 ceased operations, but its sister sites such as AllTunes are still going strong, and are apparently quite legal in Russia. The RIAA and others would claim that they are illegal in the US and in many other places.
So, what would be the result of payment processing being halted on a site which is operating legally in a large country?
And Twitter, Foursquare, and the rest of the so-called "social" web. Anyway, if they're interested in finding terrorists and whatnot, they should probably look elsewhere. If they're interested in picking up stuff to use against their own citizens (Stasi-style), then they're probably on the right track.
Next time you have a headache, please remember, that the problem is that you have a faulty head and should be using another, better one. So please have someone chop it off, and replace it with a new one.
And we'd really like to hear about your experiences afterwards. Especially if this problem keeps recurring despite your solution...
I currently use "11111111", and Deloitte says I should use at least 9 characters?
Easy peasy, I'll buy some time by making it 12 characters long: "111111111111".
I have Facebook's known IP ranges blocked incoming and outgoing at my router. It's actually quite a lot of ranges, as they've purchased IP ranges from others (but 69.171.224.0/19 and 66.220.144.0/20 will stop most of Facebook's crap). As a result, I see a surprising number of frames and whatnot which simply have "forbidden" text in them. They occur on a wide selection of sites.
No, of course not. I draw the line at having an actual keyboard (which makes my daughter's HTC Desire Z phone more of a PC than a typical tablet is). TFA is just self-serving bullshit. And shame on TFS for publicizing the rubbish.
Or, since corporations are "persons", why not tax them in an analogous fashion?
For instance, divide their income by the number of full-time employees they have (averaged over the year, not just on a particular date), and determine their tax rate based on that metric. For this purpose, a full-time employee would be one to whom they pay a salary which exceeds the local minimum and additionally receives full social benefits. Social benefits in the U.S. would mean health insurance and suchlike; in much of Europe, it would mean the extra social taxes which often amount to an additional third to half of the pre-tax salary.
Part-time employees would count as fractions of a full-time employee. The fraction being determined as the lesser of the fraction of hours worked and the fraction of social benefits received. No benefits would mean they are not employees at all.
Seriously, if you have one rough rounded number you can't do an exact convert and add false precision to the statement...
At least they didn't quibble about the difference between the UK Statute mile and the US Survey mile (the US mile is longer by 3.2mm), or even the rounding error of over a third of a km in their conversion.
Would you rather buy an ultrabook/laptop without trackpad to save money?
Yes. Emphatically yes, and I'd even pay a bit extra to avoid the things.
I disabled the trackpad on every laptop I've had which was handicapped with one. They are an ill-conceived nuisance which prevent natural positioning of the hands and lead to unexpected and unintended mouse movement. On ThinkPads, the keyboard clit works quite well, but on other brands it's generally a disappointment (and hence usually gets disabled also). Since work lumbers us with Dell laptops which have a rotten keyboard clit and an abominable trackpad, I carry a small wireless mouse and have a wireless receiver permanently in one of its USB ports. Even if there are no usable level surfaces around, it's still far better than any trackpad.
Lord Ballmer: "We would have sold more of them if more people had bought them!"
Unwise Minion: "Uh, Lord Ballmer, sir, isn't that almost a tautology?"
[brief pause]
Lord Ballmer: "Get the cleaners in here. Some minion just died while eating a chair."
How do you get your cut of that? Say you are a random person who picks up a guitar, learns how to play it and releases an album (self published). How do you claim your share of this tax?
You can apply for an exemption on the tax for whatever media you decide to publish your work on. This assumes that the media would be used exclusively for works you own, such as CD or DVD, and you'd avoid the few cents of tax on each (at least, on -R media, but obviously not on -RW media).
If you were to distribute digitally via the net or on pre-recorded memory sticks or suchlike, then you'd have to join one of these organizations to get compensated. Note, also, that you'd be unlikely to get very much unless you have a major hit on your hands. The total amount distributed is fairly small and almost half of it goes to "promotional" organizations rather than to the copyright royalty collectors.
Unless you get a horse that has been in the US racing circuit and has been fed steroids, antibiotics and bog knows what for most of it's life. I'd bet that the horsemeat in the British burgers was never tested for same.
Inthe EU, every horse is required (by EU law) to have a record of medications given to it in its entire lifetime. In particular, the record must state whether any particular item given would render it "unfit" for human consumption. There are several such medications, but they are not commonly used; all horses I know of would be fit for consumption on that basis. Of course, inspection of the horse's carcase by a veterinarian is required before it is deemed suitable for use as meat, since eating something which had died from certain diseases might be unpleasant, whatever the animal.
Actually horse meat is pretty good; I like it more than beef, and around here the price is comparable. It's tastier than beef, and also has fewer adverse consequences for your health. Horse meat becomes more tender as the animal ages - unlike cattle - and a rather larger percentage of the animal is good meat (although each horse eats more than cattle yielding similar meat mass). Of course, horses are often though of as companion animals, (disclaimer: I own and ride a horse) and it's not customary to eat any animal you gave a name to. Our horse has a name, and the kids would not tolerate any discussion of eating him...
Chapter 12: Unless you're one of the old fogies with the 360kB disks. Forget that noise, we've got COLOR in our.tiffs now!
Who are you calling "old", sonny?
I'm not that old (far from retirement age), and worked with brand new 140kB and 160kB 5.25" floppy disks on a brand new PC, several years after graduating. Earlier I worked with PDP-8, PDP-11, IBM-360, and DEC-20, which were floppy-free, and cassette-tape systems such as the PET. Even those who recall 80kB 8" floppies, or subsequent 100kB and 110kB 5.25" ones might not be retired yet.
What would that do to a Wine session, or to a VM running Windows on a Linux host?
If it does what I think (Linux-only here, no VM or Wine), there could be confusing issues in figuring out palettes afterwards in the VM or in a Wine session. The OP mentioned using Office and Visual Studio, so it's likely to be a Windows user.
Entrpreneurs have a corpus callosum, which enables both hemispheres to communicate, and even to cooperate in problem solving. Presence of a corpus callosum (even in vestigial form) precludes a successful career in management. Presence of an anterior commisure does not prevent one from rising to middle management, as it is not involved in higher thought, but may prevent entry to the executive levels.
Some meetings are necessary obviously... I'm in my first senior developer position now and I've instituted hard limits on meetings because it frustrates me to no end when idiots discuss minutiae for valuable hours of my team's time.
Agreed, but the actual time spent in meetings exceeds the required time by a significant factor at some workplaces.
Then there's the pay structure. It's almost as if some places want to pay as much as possible for work (paid overtime for panic fixes) or intend it to take as long as possible (long unproductive unpaid overtime). A better approach would be to pay for results, and damn the work time spent to achieve them, provided they meet the schedule. A 20 hour week can be as productive as a 40 hour week, and more productive than a 60 hour week.
"even when the analysis is limited to non-Hispanic whites and people with relatively high incomes and health insurance, nonsmokers, or people who are not obese."
That would not preclude an effect of inequality in the society on the health of the society. If such an effect were to exist, it would affect the well off in addition to (and plausibly more than) the not so well off.
Slashdot Mirror
That's all fine, but what really matters how many more electrons the well will hold if you make the sensor pixel 4 times larger in area and how many more photoelectrons you can generate at the same exposure settings. If both of them are scaled by a factor 4, you can just as well average four adjacent small pixels to create a lower-resolution image.
Of course this averaging should be done in the camera/phone firmware before the JPEG compression step.
Not so; the averaged pixel would be noisier than a larger pixel of the same area as the aggregate even if the number of detected photoelectrons scaled perfectly.
I mentioned only photon shot noise, but other sources of noise exist in CCD and CMOS detectors, and their effects are additive (in a mean-square fashion). In particular, readout noise and bandwidth noise will make the situation worse if the same number of photoelectrons are divided among more pixels. Bandwidth noise increases with the bandwidth at which the chip operates; with four times as many pixels, the bandwidth must be higher to get an image in the same time, so the bandwidth noise will be higher per pixel. Similarly, readout noise depends on the number of pixels to be read, and is higher for four times as many pixels.
In aggregate, the noise in detectors which are equal in area and sensitivity will be rather higher if that area is divided among more pixels. Aggregating four pixels into one after readout does not help much. The aggregated pixel will always have a worse signal to noise ratio than a single pixel of the same size as the aggregated pixel. You could try reading this to get a basic grip on imaging noise. Or use your google-fu.
### This line should go close to the start, near where
### the current Acrobat Reader macro is defined
define(EVINCE, [repeat swallow(evince) fill needs_xembed: evince "$file"])
### this group should go in the Documents section, possibly
### replacing the defective existing group for PDF
application/pdf:pdf:PDF file
application/x-pdf:pdf:PDF file
text/pdf:pdf:PDF file
text/x-pdf:pdf:PDF file
EVINCE()
Firefox is pretty much the last browser to finally get PDF viewing support.
Although they appear to have got it right (based on my Android phone). Unlike Chromium and Google Chrome on Ubuntu/Xubuntu, where you have to fiddle with the /etc/mozpluggerrc file to get it to work right (and avoid the broken reference to Acrobat Reader).
Here's a hint. In /etc/mozpluggerrc, add the lines:
### This line should go close to the start, near where the current Acrobat Reader macro is defined define(EVINCE, [repeat swallow(evince) fill needs_xembed: evince "$file"])
### this group should go in the Documents section, possibly replacing the defective existing group for PDF application/pdf:pdf:PDF file application/x-pdf:pdf:PDF file text/pdf:pdf:PDF file text/x-pdf:pdf:PDF file EVINCE()
Lower res pictures with bigger pixels? That sounds more like "we've put in a lower resolution camera, and that's better".
No, it's better - provided they have made the pixels bigger. I'm sick of phones with so-called multi-megapixel cameras, which give noisy photos in the best of circumstances. A typical 8 Mpix sensor would be much better as a 2 Mpix sensor of the same total detector size and sensitivity, and sometimes, they should have even fewer pixels with the the same total detector area and sensitivity.
Here's the essential info: shot noise is unavoidable - it's intrinsic in the physics of photon arrival at the detector. The sigma of the output noise is the square root of the number of photoelectrons.
So if you have a crappy electron well that can hold 10^4 photoelectrons when full (a "decent" cellphone camera), the signal to noise ratio is barely 100 (10^4 divided by square root of 10^4). Similarly, the photon flux per pixel in good lighting will rarely exceed 10^5 photons per second per pixel, due to the tiny lens aperture and small pixel size. It's unsurprising that the images are utter crap, as the output gain must be cranked up (amplifying noise as well as signal) to get any shot in less than 1/100 second. People downsize their images in almost all circumstances, unless they're happy with blurry and/or noisy images. FWIW, this is borne out by my experience with my own Samsung Galaxy S3 and Nokia E70, my daughter's HTC Desire Z, a colleague's Nokia 920, a friend's Samsung Galaxy S2, another colleague's iPhone 4, and various other Nokia, Samsung, and Siemens phones belonging to family and friends They are all crappy in nearly all circumstances[*], unless downsized 2:1 or more (i.e. at most one quarter of the pixels).
In a DSLR, the much larger electron well means that a pixel can hold up to 10^6 photoelectrons, so the signal to noise ratio is closer to 10^3. Similarly, the larger aperture (there's a reason for those big lenses) and larger detector pixels mean that it gets a flux of more than 10^8 photoelectrons per second per pixel in typical lighting. That's why even action shots in 1/1000sec exposure can be sharp and have relatively low noise.
[*] Exception: a relatively long exposure shot of a still life scene, or a deliberately extended exposure shot of running water or similar (with hand support to improve steadiness), say 1/15 sec or thereabouts. Not what cameras in phones are commonly used for...
It is illegal when the big compagnies are loosing money....
Companies in which countries?
Consider the fate of AllOfMP3. The web site was operating in an ostensibly legal way according to Russian laws and paid royalties through ROMS, but were sued for trillions of dollars by the RIAA and IFPI (who had apparently refused royalties). Eventually, AllOfMP3 ceased operations, but its sister sites such as AllTunes are still going strong, and are apparently quite legal in Russia. The RIAA and others would claim that they are illegal in the US and in many other places.
So, what would be the result of payment processing being halted on a site which is operating legally in a large country?
Yet Another Reason Not To Do FaceBook.
And Twitter, Foursquare, and the rest of the so-called "social" web. Anyway, if they're interested in finding terrorists and whatnot, they should probably look elsewhere. If they're interested in picking up stuff to use against their own citizens (Stasi-style), then they're probably on the right track.
Next time you have a headache, please remember, that the problem is that you have a faulty head and should be using another, better one. So please have someone chop it off, and replace it with a new one.
And we'd really like to hear about your experiences afterwards. Especially if this problem keeps recurring despite your solution...
I currently use "11111111", and Deloitte says I should use at least 9 characters?
Easy peasy, I'll buy some time by making it 12 characters long: "111111111111".
I have Facebook's known IP ranges blocked incoming and outgoing at my router. It's actually quite a lot of ranges, as they've purchased IP ranges from others (but 69.171.224.0/19 and 66.220.144.0/20 will stop most of Facebook's crap). As a result, I see a surprising number of frames and whatnot which simply have "forbidden" text in them. They occur on a wide selection of sites.
Do tablets really count as a "PC"?
No, of course not. I draw the line at having an actual keyboard (which makes my daughter's HTC Desire Z phone more of a PC than a typical tablet is). TFA is just self-serving bullshit. And shame on TFS for publicizing the rubbish.
Or, since corporations are "persons", why not tax them in an analogous fashion?
For instance, divide their income by the number of full-time employees they have (averaged over the year, not just on a particular date), and determine their tax rate based on that metric. For this purpose, a full-time employee would be one to whom they pay a salary which exceeds the local minimum and additionally receives full social benefits. Social benefits in the U.S. would mean health insurance and suchlike; in much of Europe, it would mean the extra social taxes which often amount to an additional third to half of the pre-tax salary.
Part-time employees would count as fractions of a full-time employee. The fraction being determined as the lesser of the fraction of hours worked and the fraction of social benefits received. No benefits would mean they are not employees at all.
more than 1,000 miles away (1609 kilometers)
Seriously, if you have one rough rounded number you can't do an exact convert and add false precision to the statement...
At least they didn't quibble about the difference between the UK Statute mile and the US Survey mile (the US mile is longer by 3.2mm), or even the rounding error of over a third of a km in their conversion.
Would you rather buy an ultrabook/laptop without trackpad to save money?
Yes. Emphatically yes, and I'd even pay a bit extra to avoid the things.
I disabled the trackpad on every laptop I've had which was handicapped with one. They are an ill-conceived nuisance which prevent natural positioning of the hands and lead to unexpected and unintended mouse movement. On ThinkPads, the keyboard clit works quite well, but on other brands it's generally a disappointment (and hence usually gets disabled also). Since work lumbers us with Dell laptops which have a rotten keyboard clit and an abominable trackpad, I carry a small wireless mouse and have a wireless receiver permanently in one of its USB ports. Even if there are no usable level surfaces around, it's still far better than any trackpad.
Lord Ballmer: "We would have sold more of them if more people had bought them!"
Unwise Minion: "Uh, Lord Ballmer, sir, isn't that almost a tautology?"
[brief pause]
Lord Ballmer: "Get the cleaners in here. Some minion just died while eating a chair."
How do you get your cut of that? Say you are a random person who picks up a guitar, learns how to play it and releases an album (self published). How do you claim your share of this tax?
You can apply for an exemption on the tax for whatever media you decide to publish your work on. This assumes that the media would be used exclusively for works you own, such as CD or DVD, and you'd avoid the few cents of tax on each (at least, on -R media, but obviously not on -RW media).
If you were to distribute digitally via the net or on pre-recorded memory sticks or suchlike, then you'd have to join one of these organizations to get compensated. Note, also, that you'd be unlikely to get very much unless you have a major hit on your hands. The total amount distributed is fairly small and almost half of it goes to "promotional" organizations rather than to the copyright royalty collectors.
I first read your post as "AT&T bought that part of the government...". Then I realized that was the case.
But that's old news...
This story doesn't end with a monkey desperately trying to put the can back... does it?
That's no monkey!
It's a Thetan desperately trying to avoid being in the volcano when it gets plugged...
Unless you get a horse that has been in the US racing circuit and has been fed steroids, antibiotics and bog knows what for most of it's life. I'd bet that the horsemeat in the British burgers was never tested for same.
Inthe EU, every horse is required (by EU law) to have a record of medications given to it in its entire lifetime. In particular, the record must state whether any particular item given would render it "unfit" for human consumption. There are several such medications, but they are not commonly used; all horses I know of would be fit for consumption on that basis. Of course, inspection of the horse's carcase by a veterinarian is required before it is deemed suitable for use as meat, since eating something which had died from certain diseases might be unpleasant, whatever the animal.
Actually horse meat is pretty good; I like it more than beef, and around here the price is comparable. It's tastier than beef, and also has fewer adverse consequences for your health. Horse meat becomes more tender as the animal ages - unlike cattle - and a rather larger percentage of the animal is good meat (although each horse eats more than cattle yielding similar meat mass). Of course, horses are often though of as companion animals, (disclaimer: I own and ride a horse) and it's not customary to eat any animal you gave a name to. Our horse has a name, and the kids would not tolerate any discussion of eating him...
You were going for a "Funny" mod, weren't you?
The only plausible alternative would be a "Stupid Beyond Belief" mod, and they don't exist...
Chapter 12: Unless you're one of the old fogies with the 360kB disks. Forget that noise, we've got COLOR in our .tiffs now!
Who are you calling "old", sonny?
I'm not that old (far from retirement age), and worked with brand new 140kB and 160kB 5.25" floppy disks on a brand new PC, several years after graduating. Earlier I worked with PDP-8, PDP-11, IBM-360, and DEC-20, which were floppy-free, and cassette-tape systems such as the PET. Even those who recall 80kB 8" floppies, or subsequent 100kB and 110kB 5.25" ones might not be retired yet.
Invert the colors in X11 itself: xcalib -i -a
What would that do to a Wine session, or to a VM running Windows on a Linux host?
If it does what I think (Linux-only here, no VM or Wine), there could be confusing issues in figuring out palettes afterwards in the VM or in a Wine session. The OP mentioned using Office and Visual Studio, so it's likely to be a Windows user.
Entrpreneurs have a corpus callosum, which enables both hemispheres to communicate, and even to cooperate in problem solving. Presence of a corpus callosum (even in vestigial form) precludes a successful career in management. Presence of an anterior commisure does not prevent one from rising to middle management, as it is not involved in higher thought, but may prevent entry to the executive levels.
Some meetings are necessary obviously... I'm in my first senior developer position now and I've instituted hard limits on meetings because it frustrates me to no end when idiots discuss minutiae for valuable hours of my team's time.
Agreed, but the actual time spent in meetings exceeds the required time by a significant factor at some workplaces.
Then there's the pay structure. It's almost as if some places want to pay as much as possible for work (paid overtime for panic fixes) or intend it to take as long as possible (long unproductive unpaid overtime). A better approach would be to pay for results, and damn the work time spent to achieve them, provided they meet the schedule. A 20 hour week can be as productive as a 40 hour week, and more productive than a 60 hour week.
obviouslt your agenda forbade you to read:
"even when the analysis is limited to non-Hispanic whites and people with relatively high incomes and health insurance, nonsmokers, or people who are not obese."
That would not preclude an effect of inequality in the society on the health of the society. If such an effect were to exist, it would affect the well off in addition to (and plausibly more than) the not so well off.