It's like using the condenser end of an air conditioner to heat your home in the winter. They call it a heat pump. But in this case, instead of heat output, it's light output, which can be more convenient.
Just being picky here... actually, it's not exactly 100%. A very small fraction of the electrical power is emitted as electromagnetic radiation and acoustic vibration. It is so small as to be not take much from that 100%, but it is there.
This fails for two reasons. The solar panel efficiency so still so low that it would produce less power than went in to produce the light. And the additional energy comes from ambient heat.
Think of an air-conditioner. You put energy in through the power cord. More energy comes out the condenser in the form of heat. That's because the evaporator is getting cold, drawing in heat. The output energy is the combination of all sources of input heat. Now if you find a means to convert that output heat to electrical energy to power the device, you can have some real fun. But even then you won't achieve free energy since the best you can do is move it around. But I do think there is some small potential to improve the process.
Scaled up, we might have LEDs that can pump heat away for some practical purpose, like cooling CPUs. If that can work, then some day in the future you might have a CPU (more likely a million SoCs on one chip) that is emitting an intense light from the top of its surface. That would be the new heat sink. Now you need to get rid of the light, or the heat from that light hitting another surface. But this would be a better way of cooling since that is easier to convert back to something useful.
If you are saying some of the power comes from the heat present in the atomic lattice, then don't we have a cooling mechanism going on here? if it takes energy out of that atomic lattice, then either its temperature goes down, or heat is drawn in from the environment, or some combination of these.
My air conditioner emits more energy outside than comes in to it through the power cord.
Touch screens replacing physical buttons is a whole LOT older than even this. I first saw them in use in other devices in 1989. Any patents from back then are run out now.
If you would just f---ing quit referring to the issue as "software patent" (as in "software patents are bad"), and focus on the real issue of patents... that they are ALL all bad (except those that meet the original patent justification of innovation that would never have happened were it not for the possibility of a patent), regardless of whether software, hardware, or anything else, then maybe we can actually start to get some traction on the issue. Instead, with people merely claiming "software patents" are what is bad, people just won't see the real issue (that 99.9% of patents are not innovation, and are just corporations trying to grab a government sanction to own something that can let them destroy competition).
Just because the rise of non-innovative patents is happening concurrently with the rise of software methods being patented (or rather, methods that are typically implemented in software), does not mean that software itself is the reason for something to not be patented. The few and rare cases of genuinely innovative ideas could very well be implemented in hardware or software, in many of those cases.
I'm not proposing the scrapping of the patent system. It's still justified. We just need to operate it strictly for the innovations that justify it.
Isn't this what a boycott is? I'm already doing that, now, south of the border. And if anyone you encounter whines about "then I'd have to be illegal", point them at Magnatune.com. And there are new new indie artists giving away their own stuff. I do think we can make a Creative Commons music arena.
I agree, but it is not just the editing/mixing that benefits from full quality.
Doing the compression from 24/192 (or even my preferred 32/480) will be better than doing it from 16/48, even when compressed to the same bit level (though that difference will converge as you push the bit level down). The compression logic will have a cleaner source to work with if the high sample rate and resolution is handled properly. The end user will get slightly better audio in just the same space.
By all means do the studio mixing at as high a sample rate and resolution (uncompressed or non-lossy compressed) as you possibly can. Even video should be edited uncompressed or non-lossy compressed with the best video sourcing you can get, before you crunch it down to satellite, cable, and broadcast limitations.
This article is bunk. I wasted 15 minutes reading through it. And they didn't even cover multi-tone and complex waveforms (which would have shown it to be bunk). Pure sine waves actually do well with digital sampling. But as you reach the edge of the Nyquist limit, you reach a point where the number of waveform states (how many sines waves of various values can be mixed) that can be rendered by the sample converges to unity. E.g. it can only support ONE sine wave at that point. Raise the sample rate and then you have the capacity to render multiple sine waves at the same frequency and many others.
A higher sample rate at say 192kHz is NOT done for the purpose of being able to encode sinusoid components up to 96kHz. It can do that (with that one sine wave limit that point). But is is appropriate to sample after a low pass filter (for example at 18 kHz) that limits the signals to only what you want. And then after conversion back to analog, clean it up with the low pass filter (again, at 18 kHz).
Listen to speech filtered with a 4kHz lowpass filter in an all-analog path. You will be able to tell it is filtered if your hearing is normal. Now digitize that filtered speech with an 8kHz sample rate. Convert it back to analog, and filter it again. The highs (up to 4kHz) will still be there (Nyquist says so, and this is valid). But there will also be new intermodulation products all over the place, especially among the high frequency components. It will give the audio a tinny or metallic sound quality.
Looking at it as combinations, a 44100 Hz sample rate at 16 bits is enough to render a 22050 Hz tone at any of 32768 intensity levels. However, if you have a 2nd tone of 22000 Hz, with each at 16384 intensity levels to avoid an overload, there are now 268435456 level combinations to be encoded. Now the 16 bits isn't enough. You need to double it. That can be done by either 32 bit sampling (hard to do) or doubling the sample rate (still 32 bits but now done as a pair of 16 bit samples). Fortunately you won't have mixed signals that high very often. However, you can easily have many signal components at lower frequencies. You will need plenty of bits for each. Even 192 kHz sampling is not enough to render 4 full range sine components at around 4 kHz. One or even a few levels of inaccuracy won't be heard. But these combinations rise very rapdily with just a few components.
For wider band audio with a higher sample rate, because most people hear weakly at higher frequencies, the effects will be less perceived, if at all. But they will be there, and a small portion of the population (including myself) can hear it.
Personally I'd rather they would go with 32 bits and 480kHz sampling.
Slashdot Mirror
So what does that say about my prospects of getting hired by a dumb employer? Oh wait!
Except you are wrong. They pick the dump ones.
Basically, they have invented a 133 picoBTU/hr air conditioner would a "cool blue" effect.
It's like using the condenser end of an air conditioner to heat your home in the winter. They call it a heat pump. But in this case, instead of heat output, it's light output, which can be more convenient.
Just being picky here ... actually, it's not exactly 100%. A very small fraction of the electrical power is emitted as electromagnetic radiation and acoustic vibration. It is so small as to be not take much from that 100%, but it is there.
This fails for two reasons. The solar panel efficiency so still so low that it would produce less power than went in to produce the light. And the additional energy comes from ambient heat.
Think of an air-conditioner. You put energy in through the power cord. More energy comes out the condenser in the form of heat. That's because the evaporator is getting cold, drawing in heat. The output energy is the combination of all sources of input heat. Now if you find a means to convert that output heat to electrical energy to power the device, you can have some real fun. But even then you won't achieve free energy since the best you can do is move it around. But I do think there is some small potential to improve the process.
Scaled up, we might have LEDs that can pump heat away for some practical purpose, like cooling CPUs. If that can work, then some day in the future you might have a CPU (more likely a million SoCs on one chip) that is emitting an intense light from the top of its surface. That would be the new heat sink. Now you need to get rid of the light, or the heat from that light hitting another surface. But this would be a better way of cooling since that is easier to convert back to something useful.
... it would add a new more truthful meaning to "cool lights".
If you are saying some of the power comes from the heat present in the atomic lattice, then don't we have a cooling mechanism going on here? if it takes energy out of that atomic lattice, then either its temperature goes down, or heat is drawn in from the environment, or some combination of these.
My air conditioner emits more energy outside than comes in to it through the power cord.
Oh yeah, LFS is fun. I ported mine to my old Sun Sparc box.
Yeah, like school was a waste of time, too. I didn't want to learn all that stuff.
... which will hopefully have an open platform version somewhere so I can run my own stuff on it.
And I was looking for a blog hoster this week, and specifically at WordPress. Anyone got a list of free blog hosters (moving away from blogspot)?
... just sue them for getting it wrong. Besides, suing is the national sport.
Sorry, but they already have the patent on that.
Better yet, crush it, or even shoot it. Put the video proof online.
Touch screens replacing physical buttons is a whole LOT older than even this. I first saw them in use in other devices in 1989. Any patents from back then are run out now.
If you would just f---ing quit referring to the issue as "software patent" (as in "software patents are bad"), and focus on the real issue of patents ... that they are ALL all bad (except those that meet the original patent justification of innovation that would never have happened were it not for the possibility of a patent), regardless of whether software, hardware, or anything else, then maybe we can actually start to get some traction on the issue. Instead, with people merely claiming "software patents" are what is bad, people just won't see the real issue (that 99.9% of patents are not innovation, and are just corporations trying to grab a government sanction to own something that can let them destroy competition).
Just because the rise of non-innovative patents is happening concurrently with the rise of software methods being patented (or rather, methods that are typically implemented in software), does not mean that software itself is the reason for something to not be patented. The few and rare cases of genuinely innovative ideas could very well be implemented in hardware or software, in many of those cases.
I'm not proposing the scrapping of the patent system. It's still justified. We just need to operate it strictly for the innovations that justify it.
There doesn't have to be an actual lawsuit. The screwed up patent system creates its own FUD.
Proxy through international servers so you are coming from different IP addresses.
Sure ... in the minds of motion picture executives, where lots of false facts abound.
... but do want a job involving Facebook or such, there's always moderating the reported photos.
Isn't this what a boycott is? I'm already doing that, now, south of the border. And if anyone you encounter whines about "then I'd have to be illegal", point them at Magnatune.com. And there are new new indie artists giving away their own stuff. I do think we can make a Creative Commons music arena.
And at that time, I don't want to have to redo all my music.
And at that time, I don't want to have to re-pirate all my music.
There, fixed it for ya.
I agree, but it is not just the editing/mixing that benefits from full quality.
Doing the compression from 24/192 (or even my preferred 32/480) will be better than doing it from 16/48, even when compressed to the same bit level (though that difference will converge as you push the bit level down). The compression logic will have a cleaner source to work with if the high sample rate and resolution is handled properly. The end user will get slightly better audio in just the same space.
By all means do the studio mixing at as high a sample rate and resolution (uncompressed or non-lossy compressed) as you possibly can. Even video should be edited uncompressed or non-lossy compressed with the best video sourcing you can get, before you crunch it down to satellite, cable, and broadcast limitations.
This article is bunk. I wasted 15 minutes reading through it. And they didn't even cover multi-tone and complex waveforms (which would have shown it to be bunk). Pure sine waves actually do well with digital sampling. But as you reach the edge of the Nyquist limit, you reach a point where the number of waveform states (how many sines waves of various values can be mixed) that can be rendered by the sample converges to unity. E.g. it can only support ONE sine wave at that point. Raise the sample rate and then you have the capacity to render multiple sine waves at the same frequency and many others.
A higher sample rate at say 192kHz is NOT done for the purpose of being able to encode sinusoid components up to 96kHz. It can do that (with that one sine wave limit that point). But is is appropriate to sample after a low pass filter (for example at 18 kHz) that limits the signals to only what you want. And then after conversion back to analog, clean it up with the low pass filter (again, at 18 kHz).
Listen to speech filtered with a 4kHz lowpass filter in an all-analog path. You will be able to tell it is filtered if your hearing is normal. Now digitize that filtered speech with an 8kHz sample rate. Convert it back to analog, and filter it again. The highs (up to 4kHz) will still be there (Nyquist says so, and this is valid). But there will also be new intermodulation products all over the place, especially among the high frequency components. It will give the audio a tinny or metallic sound quality.
Looking at it as combinations, a 44100 Hz sample rate at 16 bits is enough to render a 22050 Hz tone at any of 32768 intensity levels. However, if you have a 2nd tone of 22000 Hz, with each at 16384 intensity levels to avoid an overload, there are now 268435456 level combinations to be encoded. Now the 16 bits isn't enough. You need to double it. That can be done by either 32 bit sampling (hard to do) or doubling the sample rate (still 32 bits but now done as a pair of 16 bit samples). Fortunately you won't have mixed signals that high very often. However, you can easily have many signal components at lower frequencies. You will need plenty of bits for each. Even 192 kHz sampling is not enough to render 4 full range sine components at around 4 kHz. One or even a few levels of inaccuracy won't be heard. But these combinations rise very rapdily with just a few components.
For wider band audio with a higher sample rate, because most people hear weakly at higher frequencies, the effects will be less perceived, if at all. But they will be there, and a small portion of the population (including myself) can hear it.
Personally I'd rather they would go with 32 bits and 480kHz sampling.