Color theory dictates a gamma of about 3.0 for our eye's perception of color; i.e. the cube root of voltage changes appear to be about the same distance apart in color space. The L*A*B* color space reflects this.
All output devices except CRT's are more or less linear, gamma about 1, thus the DAC's need LOTS of bits to represent differences near black without contouring/banding - or without lots of dithering noise added in. The good old CRT has a native gamma of about 2.2, better than square root, but not quite the cube root our eye sees. As a result many fewer bits in a DAC produce excellent results. Most good CRT's operate flawlessly with 10 to 12 bit DAC's, while at least 16 bits would be needed to equal this in a linear gamma display.
On another topic, CRT's can be scanned at the native rate of the video source, 720p or 1080i for HDTV; or, if desired, upsampled/deinterlaced by an INTEGER factor 2, 3 or 4 to 1. Fixed pixel displays require all kinds of fancy DSP chips to resample by odd factors and still don't look as good.
I had the good fortune(?) to work with Kai in the late 1990's. He's the kind of guy who always tells a good story and enjoyed the fact a graphic artist portrayed him in a Superman-like suit. He lived near Santa Barbara, CA for years; but now he resides in his dream abode castle on the Rhine back in his native Germany.
Although he undoubtedly was an influential User Interface designer: the whole rounded, weird shaped, non-rectangular look for windows and slide-out expanded functions comes from Kai. He is also the best and most convincing bullshitter I ever met!
To quote his first paragraph from the article above: "I always felt, but can't prove outright: Zen is wrong. Then is right. Everything is not about the now, as in the "here and how", "living for the moment" On the contrary: I believe everything is about the before then and the back then."
I think he wishes he was clever enough to answer the question "Do you do drugs?", like Dali: "I am a drug." Unfortunately, he has to settle for "Yes."
Game developers should only use OpenGL for newly written rendering code. It is a high performance, advancing standard. Microsoft often "borrows" from it for Direct X (n+1).
OpenGL makes Linux, BSD, MacOS X and other ports practical as well as not having to deal with Microsoft's arbitrary API's.
You are correct, if not somewhat pedantic. Most people refer to "WiFi" as a IP stack on top of IEEE 802.11b/g, this is why I phrased my comment in that way.
The Nintendo DS uses a non-standard WiFi protocol, the PHYS and MAC layers are standard, but it's different above that. How else could they seamlessly allow up to 16 DS's to automatically network without IP addresses, DHCP, etc.?
The people hacking it have apparently modified their Linux WiFi device driver so it can communicate to some extent - it is therefore possible that a DS firmware update, if the DS firmware CAN be updated, could make it fully WiFi compatible. Or, perhaps, full WiFi compatibility is already there as an "Easter Egg" waiting for Nintendo to announce how to enable it. Supposely, Nintendo is going to make an announcement in March, 2005 - right before the Sony PSP ships in the US.
What happens if Eolas wins on appeal AND the USPTO wins its appeal and invalidates the patent? Is the judgement against Microsoft retroactively vacated?
If not, this seems an extreme miscarriage of equity in law. Why not put the appeal of Eolas vs. Microsoft on hold pending final resolution of the patent reconsideration litigation? This has been done for lesser SCO vs. Linux suits pending the granddaddy SCO vs. IBM verdict.
Actually tubes produce some 4th order and higher harmonics, too, but you're right that it's mostly 2nd harmonic distortion. The more unique part of this issue is that tubes behave differently than transistors when over-driven. Tubes reach saturation somewhat gradually and the distortion just increases. Transistors tend to clip suddenly with nasty sounding odd order harmonics.
So to simulate tube guitar amps with transistors not only would some non-trivial level sensitive DSP be needed, but you'd need a big ballsy transistor amp that never would clip.
I'm surprised no one realizes nearly ever radio or TV transmitter outputting 1,000's of watts or more uses tubes, albeit ceramic tubes rather than glass ones in most cases. And this is today in late 2004. Although in the most technical sense they are "amplifiers", they aren't audio amps. The transmitter tubes often have metal fins to radiate heat and a "chimney" to air cool the tube.
This technology could be adapted to "hot" guitar amps, although they would be pricey.
Whoever said it... Khan, Quentin Tarantino; they captured an essential part of the concept of revenge. The victim (or deserving target) of revenge is caught unsuspecting and off-guard. It cannot just be another parry to an immediately preceding punch in a fight.
That is why revenge is so sweet, and evil.
The mainstream and tech press is always implicating Russian crackers or links to.ru sites...
What's the real deal? Someone is feeding us disinformation with a shovel.
Well, it looks nobody, including Intel and IBM, had any idea how difficult moving to the 90nm process would be. What about 60nm? I think Moore's Law has finally run out of steam.
I agree with Gutmans completely. Richard Stallman's GPL is free like Henry Ford's quote: "You can have any color as long as it's black." You can link anything with GPL'ed code as long as it's other GPL (or GPL-equivalent) code.
I'll take the BSD license anytime. Code migrates from BSD to Linux (but not Linux to BSD) because of GPL.
Right after registering a domain, you'll often get a few spam's hawking hosting services, ect. Verisign (no flames please!) does allow you to opt out of their bulk sale of whois data - although why are they doing it in the first place?
Also for $9 a year you can buy a redirected e-mail address that changes every 10 days that appears as your whois contact.
Slashdot Mirror
Color theory dictates a gamma of about 3.0 for our eye's perception of color; i.e. the cube root of voltage changes appear to be about the same distance apart in color space. The L*A*B* color space reflects this.
All output devices except CRT's are more or less linear, gamma about 1, thus the DAC's need LOTS of bits to represent differences near black without contouring/banding - or without lots of dithering noise added in. The good old CRT has a native gamma of about 2.2, better than square root, but not quite the cube root our eye sees. As a result many fewer bits in a DAC produce excellent results. Most good CRT's operate flawlessly with 10 to 12 bit DAC's, while at least 16 bits would be needed to equal this in a linear gamma display.
On another topic, CRT's can be scanned at the native rate of the video source, 720p or 1080i for HDTV; or, if desired, upsampled/deinterlaced by an INTEGER factor 2, 3 or 4 to 1. Fixed pixel displays require all kinds of fancy DSP chips to resample by odd factors and still don't look as good.
I had the good fortune(?) to work with Kai in the late 1990's. He's the kind of guy who always tells a good story and enjoyed the fact a graphic artist portrayed him in a Superman-like suit. He lived near Santa Barbara, CA for years; but now he resides in his dream abode castle on the Rhine back in his native Germany.
Although he undoubtedly was an influential User Interface designer: the whole rounded, weird shaped, non-rectangular look for windows and slide-out expanded functions comes from Kai. He is also the best and most convincing bullshitter I ever met!
To quote his first paragraph from the article above:
"I always felt, but can't prove outright: Zen is wrong. Then is right. Everything is not about the now, as in the "here and how", "living for the moment" On the contrary: I believe everything is about the before then and the back then."
I think he wishes he was clever enough to answer the question "Do you do drugs?", like Dali: "I am a drug." Unfortunately, he has to settle for "Yes."
Game developers should only use OpenGL for newly written rendering code. It is a high performance, advancing standard. Microsoft often "borrows" from it for Direct X (n+1).
OpenGL makes Linux, BSD, MacOS X and other ports practical as well as not having to deal with Microsoft's arbitrary API's.
You are correct, if not somewhat pedantic. Most people refer to "WiFi" as a IP stack on top of IEEE 802.11b/g, this is why I phrased my comment in that way.
The Nintendo DS uses a non-standard WiFi protocol, the PHYS and MAC layers are standard, but it's different above that. How else could they seamlessly allow up to 16 DS's to automatically network without IP addresses, DHCP, etc.?
The people hacking it have apparently modified their Linux WiFi device driver so it can communicate to some extent - it is therefore possible that a DS firmware update, if the DS firmware CAN be updated, could make it fully WiFi compatible. Or, perhaps, full WiFi compatibility is already there as an "Easter Egg" waiting for Nintendo to announce how to enable it. Supposely, Nintendo is going to make an announcement in March, 2005 - right before the Sony PSP ships in the US.
What happens if Eolas wins on appeal AND the USPTO wins its appeal and invalidates the patent? Is the judgement against Microsoft retroactively vacated?
If not, this seems an extreme miscarriage of equity in law. Why not put the appeal of Eolas vs. Microsoft on hold pending final resolution of the patent reconsideration litigation? This has been done for lesser SCO vs. Linux suits pending the granddaddy SCO vs. IBM verdict.
Actually tubes produce some 4th order and higher harmonics, too, but you're right that it's mostly 2nd harmonic distortion. The more unique part of this issue is that tubes behave differently than transistors when over-driven. Tubes reach saturation somewhat gradually and the distortion just increases. Transistors tend to clip suddenly with nasty sounding odd order harmonics.
So to simulate tube guitar amps with transistors not only would some non-trivial level sensitive DSP be needed, but you'd need a big ballsy transistor amp that never would clip.
I'm surprised no one realizes nearly ever radio or TV transmitter outputting 1,000's of watts or more uses tubes, albeit ceramic tubes rather than glass ones in most cases. And this is today in late 2004. Although in the most technical sense they are "amplifiers", they aren't audio amps. The transmitter tubes often have metal fins to radiate heat and a "chimney" to air cool the tube. This technology could be adapted to "hot" guitar amps, although they would be pricey.
Whoever said it... Khan, Quentin Tarantino; they captured an essential part of the concept of revenge. The victim (or deserving target) of revenge is caught unsuspecting and off-guard. It cannot just be another parry to an immediately preceding punch in a fight. That is why revenge is so sweet, and evil.
The mainstream and tech press is always implicating Russian crackers or links to .ru sites...
What's the real deal? Someone is feeding us disinformation with a shovel.
Well, it looks nobody, including Intel and IBM, had any idea how difficult moving to the 90nm process would be. What about 60nm? I think Moore's Law has finally run out of steam.
I agree with Gutmans completely. Richard Stallman's GPL is free like Henry Ford's quote: "You can have any color as long as it's black." You can link anything with GPL'ed code as long as it's other GPL (or GPL-equivalent) code.
I'll take the BSD license anytime. Code migrates from BSD to Linux (but not Linux to BSD) because of GPL.
Right after registering a domain, you'll often get a few spam's hawking hosting services, ect. Verisign (no flames please!) does allow you to opt out of their bulk sale of whois data - although why are they doing it in the first place?
Also for $9 a year you can buy a redirected e-mail address that changes every 10 days that appears as your whois contact.