From my perspective, if Y'CbCr RGB conversions are so bad, just keep everything in Y'CbCr.
For any display of any size, you aren't going to see a difference between 4:2:2 and 4:4:4. And of course, lots of apps to Y'CbCr 4:4:4 processing internally, and only do the downsampling on import and export.
I agree. This guy isn't a professional in this space. He makes a lot of the assumptions and errors I made five years ago. But they're smart errors and assumptions, and I'm sure he'll find the Slashdot experience a great way to get all your errors pointed out quickly:).
He really hasn't done the math to realize the advantages of Y'CbCr are more about bandwidth reduction, interoperability with existing color video technology, and easier signal processing. While we certainly could build an RGB DVD player for roughly the cost of a Y'CbCr one, it'd only get half the run time, and would look worse on standard television (only VGA is natively RGB for signal).
Full agreement on Macrovision. Like most CD audio protection, it's the kind of copy protection that irritates users without providing any effective deterrent against skilled pirates.
Oh, I think we've got decades of Y'CbCr to look forward to. it's really an advantageous format in a lot of ways.
First, a 4:2:0 Y'CbCr is half the bandwidth of 4:4:4 RGB. We're a long way away from having half the processing power required, bandwidth, storage, etcetera simply not mattering. My RAID is 2 TB formatted, but I regularly have projects that take up over 50% of the space.
Second, Y'CbCr is a better native space for video processing, since the channels align better with what we want to filter. Luma filters like gamma or contrast are more than 3x faster in Y'CbCr than in RGB, since only one channel needs to be processed. Saturation is more than 6x faster in 4:2:0, since only two channels, each at 25% bandwidth, need to be processed. Plus a lot of filters have to convert from RGB to another color space to run. Y'CbCr is closer to those other spaces, and often doesn't require any conversion. You can say whatever you will about Moore's law, the difference between 4 and 8 real-time layers will matter for a while. Even the audio guys, who have it a lot easier, still run into performance limits with enough simultaneous tracks and such.
Lastly, our entire video infrastructure is build around subsampled Y'CbCr. Never underestimate the lock-in of standards like this. If computer people couldn't kill interlaced video in HDTV, they're never going to kill subsampling for lots of applications. Color video has always been Y'CbCr, and that's how everyone works and thinks for decades now.
That said, Hollywood is likely to pick a >8-bit RGB solution for digital projection. For digital projection, bandwidth is a non-issue, and quality, and quality like that of film. Film guys live in RGB. Plus, it's a win for that industry to have digital cinema be as INCOMPATIBLE with consumer digital video tech as possible, in order to reduce the ease of piracy, and to maintain an advantage of the theatrical experience over home theater.
FWIW, I'm a member of the SMPTE groups working on both video compression and digital cinema.
MacWorld will be my first real road test of it, but it's great so far. Handles when I want to look professional, shoulder strap for when I'm walking around, and hidden backpack straps for racing through the airport. Lots of internal zippered pockets. Room for my Wintel laptop in there as well. It isn't 100% of what my old Petral bag did, but that didn't fit the 17".
Yeah, and 1024x768 is sub HD. You really want 1280x720 at a minimum. And in order to get as many pixels out of the screen as are being broadcast, you really need 1920x1080.
I've been doing a ton of HD video authoring the last couple months. It's a startling realization to discover that a 1920x1200 23" LCD is the SMALLEST monitor that can be used for this kind of stuff! And for quality grading, I'm sitting there with my nose eight inches from the screen for a couple of hours at a time, looking for minute compression artifacts.
Wow, we're decades away from having AI good enough to do that! Detecting watermarks in a rectangular video stream is hard, sure. But in the real world? That's basically asking for reliable object and feature extraction from real-world images. This is canonically one of the things humans are incredibly better at than computers. Any watermark visible enough to be detected by a moving camera, at an angle on several axes, in variable lighting, etecetera, would be so blindingly obvious it wouldn't qualify as a watermark.
Pixlet is great if you need an intraframe-only codec for playback on Mac OS X 10.3 systems only (no Jaguar, no Windows). Pixlet is meant for film production workflows, since it offers frame-accurate rough cuts at half D1 resolution on a fast G4. It is in no way useful as a codec to distribute content to end users.
Think of it as a souped up version of Photo-JPEG, or a complexity constrained version of JPEG2000. It no more competes with Sorenson Video than a Freightliner competes with a Ferrari.
Apple's excellent marketing has actually backfired on them a bit here. Pixlet is very useful for what it's meant for, but if you're not a film professional, it isn't meant for you.
Actually, I think Apple quit trying to fight Windows Media quite a while ago. They're not pushing or enhancing the.mov file format anymore - the best codec available is still Sorenson Video 3, which shipped two and a half years ago.
QuickTime's importance to Apple today is much more as a digital media SDK, and hence the foundation of their very successful products like iMovie, iDVD, Final Cut Pro, DVD Studio Pro, iTunes, iPhoto, etcetera.
QuickTime is also the best authoring architecture out there, with wonderful features like reference movies.
But as a delivery format, Apple has been letting it linger, but isn't putting much effort into enhancing it. They've talked a lot about MPEG-4 being the future of the file format, but haven't done much technically to make that viable either.
Unfortunately, QuickTime as architected can't deal with out of order data in frames, and so can't natively deal with MPEG-1, MPEG-2, or Advanced Simple MPEG-4. I suppose you could make export components for those, but it'll take some serious refactoring before QuickTime could really treat those formats as first class citizen.
As far as QT Pro for $100, I assure you that the user base would go ballistic. Apple does install MPEG-2 encoding (fast, but not very flexible) as a QuickTime component for owners of Final Cut Pro or DVD Studio Pro.
I have to say, this article is the exception that proves Waggoner's eleventh law: "Any reference to 'toxins' as a generic class is in the service to someone selling a bogus pseudo-medical treatment."
So, if we assume for the moment that this theory is true, and that new oil, etcetera is being made all the time, that doesn't necessarily mean that we aren't going to run out.
Would new petrochemicals be formed faster than we can use them up? Certainly the levels in proven reserves drop as they're being pumped up, so any regeneration is on a slower scale than our current consumption.
The problem is, one can die without ever having bought a movie or a computer, since next year it's going to be so much better.
If I buy a DVD, it's because I figure I'll watch it enough during one technological cycle to make it be worth it (five years or so). Now that I've got two kids under 4, I don't buy many DVDs:).
It's nice we have the announcement, but this is Slashdot!
Have ANY technical details of the new standard been published anywhere yet? I can't find any public resources I can link to.
100% success with G5 SuperDrive and Reitech media
on
DVD-Rs go 8x
·
· Score: 1
I've burned about 60 discs in the last week at 4x with the SuperDrive in my G5 with Reitech media. Not a single failure the whole time. And the discs perform as well - I've been able to play back digital media files with 20 Mbps peaks without a hitch. That's about twice the performance required for DVD video.
720x576 is PAL DV resolution. Of course, it's non-square pixel, so playing back on many players will result in a distorted image (about 10% too tall).
It's 6 Mbps! That's enough bitrate for HD encoded content. 640x480 at 1.5 Mbps would have been just fine.
It wasn't deinterlaced! Thus ugly, hard to compress horizontal lines throughout the image.
Basically, the guy did a dump of the PAL DV master into WMV, without any processing to make it appropriate for computer playback.
Big picture is that it'd be trivial to do a better looking version of this clip at a quarter the data rate. And he could have used WMV8 for MPlayer compatibility. WMV9 is a great codec, but for something posted on Slashdot...
Reencoding from one compressed delivery format typically results in a big, big quality hit compared to encoding to the same final format from a high quality mastering format, like DV.
Think of it as giving you the artifacts from BOTH codecs at the same time. But worse.
Also, why is it.zip compressed! Any decent video and audio compression includes its own entropy encoding, so that a.zip or whatever will produce virtually no size savings.
If you had a file where.zip worked, that'd tell you that you did something wrong!
Yeah, considering its age, MPEG-2 scales remarkably well. And the MIPS per pixel required are a lot lower than more modern codecs. Of course, if bandwidth is at a premium, MPEG-2 isn't a good option. If you want to get a 1080p 2.5 hour movie onto DVD-9 media, Windows Media Video 9 can handle the compression without breaking a sweat, while MPEG-2 wouldn't look that good even with the best encoder. While you might spend more on an ASIC for decoding the WMV9, that could be more than made up for by being able to use conventional media
As in all things, it really depends on what is your limiting factor.
First off, VCD isn't really any better quality wise than VHS. VCD is digital, so you don't get analog errors and wear, but it has only half the temporal and vertical resolution.
VCD didn't fail in the states for any reason other than that it didn't provide any better customer value than VHS. VCD won in Asia since it's a cheaper medium to counterfeit. In general, Asian audiences also seem more willing than US audiences to accept lower quality for lower price for video. Have you looked at many of those audience-shot Hong Kong bootlegs? Awful.
DVD is way better than VHS, and so won in the US.
And the DVD Forum is hard at work on a next generation HD DVD standard, and US companies are part of the Forum.
China might be trying an end run around this format, but the US-Japan-Europe industry axis will have their own format coming.
No dis on VP3/4/6, but MPEG-2 does HD fine just well. US HD TV broadcasts are all MPEG-2 (1920x1080 or 1280x720). I'm sitting next to two computers encoding HD MPEG-2 RIGHT NOW for a Kiosk application.
While MPEG-2 doesn't have cutting edge compression efficiency anymore, it isn't bad, and decoder chips are dirt cheap. And you can actually play back 1920x1080 60i MPEG-2 on a fast P4.
Er, sounds good, but you actually don't really know what you're talking about.
A well encoded DVD has very high quality, certainly on par with Betacam SP, the high end analog broadcast production format before digital took over. A well encoded disc won't show significant artifacts.
HDTV resolution goes up to 1920x1080, which is about 6.5x the pixels of DVD (720x480). How high do you want to go? The cheapest displays that can meaningfully do more than 1920 lines wide on a largish screen are awesomely expensive.
Today's displays are crappy? Compared to what? A tapestry? IMAX? We're really at the beginning of a golden age for consumer video technologies. The quality you can buy for $5000 is vastly improved in the last three years, let alone the last 30. Most people don't have eyes good enough to appreciate anything beyond a good 1920x1080.
Lastly, fractals are really the Grassy Knoll of video compression. Yes, Iterated was created to make products on them. No, fractals didn't work. I spent a lot of the mid-late 90's working with Iterated's stuff in different forms. Bitrate scalability was interesting (you could truncate the file at any point, and the more bits you grabbed, the better an image you got). Compression ratios were somewhat better than JPEG. They scaled pretty well. But the net gains were too small to overcome the market share advantages of lowly JPEG.
Iterated simply couldn't make a business around fractal compression. They sold their stuff to AltaMira, who still are selling their fractal compression stuff. Iterated morphed into a company providing image management solutions for the prepress industry. There was still some fractalish stuff underneath, but that wasn't where the value was really added.
The big thing about "fractal" compression is that it wasn't really fractal - its ability to take advantage of self-symmetry was very limited. Heck, even with today's computer power, a "true" fully automatic fractal compressor would take unbelievable amounts of CPU power - many orders of magnitude beyond what realistic video codecs do today. You're basically extending motion search into so many axes.
The only fractal video codec was ClearVideo, which was interesting I suppose, but was roundly stomped by both DCT H.263 derived codecs, and VQ derived codecs like Sorenson Video 1.
Almost everything good about fractals has been inherited by wavelets. And wavelets have also inherited fractal's difficulty in handling motion estimation. That's why DCT and DCT-derived codecs still rule the roost today. Wavelets are great for still image, but no one has come close to devising a really competitive wavelet motion codec.
Maybe someday we'll have a revolution in codecs, but DCT-based codecs like WMV9 and AVC keep on trucking in providing excellent compression efficiency, scalability, and decoder performance.
Back when I paid attention to Dr. Who in detail (late 80's), there were lots of missing episodes from the early years, especially the first two Doctors. The BBC had somehow lost all copies of them internally.
Have all or at least some of these been found in the last ~15 years?
I think "New Ideas from Dead Economists" is a very readable, juicy introduction to the history of the field. Good mix of cleanly presented theory, and biographical information. No particular ideological bent, which is a nice antidote to a lot of the other stuff going on out there. Think a mix of "Quicksilver" and a one of the better O'Reilly books - can't get much more hacker friendly than that!
Reading the Economist every week is also a great way to keep up to date on what's going on in the world, and in the world of money, with an nicely non US, non Dem v. Rep view of things. It's very classic Liberal - pro free trade, and pro civil rights.
Estimates of the Bush tax cuts is that we get something like $500,000 of deficit spending per new job created. I'm not sure what they wanted these tax cuts for, but they clearly weren't tuned to provide maximum job creation at minimum cost.
I don't know that the war is causing that much economic activity. Deficit spending is my big worry. Deficits lead to higher inflation, which causes the Fed to raise interest rates. Higher interest rates tend to push housing down. Since people buy houses based on the sum of the mortgage, interest and taxes, higher interest rates reduce the value of houses people can buy, having a negative effect on house prices. Of course inflation in general has a raising effect, but you can get a situation where the true value of homes is declining, with prices rising, but not as fast as inflation.
I wish more hackers would get interested in economics - it's really very similar to a lot of technical topics. It has become painfully clear to this economics nerd that the USA simply doesn't have a competent economics policy or staff under the current administration.
I'm admittedly in a weird niche (compressed video consulting and training), but the demand for my services has been ramping up nicely since February, and just blasting off in the last month. I expect to bill more Q4 of this year than I billed in all of 2002.
And it isn't just one client. It's coming from a lot of different directions, from a lot of different companies and industries. And nice, juicy, interesting jobs too. It was like the outsourcing switch just got turned on.
Of course, independent consultants like me are often a good six months ahead of the rest of the economy. When things get tight, consultants get cut first, and when things are looking up, consultants get hired before full-time employees, since if things turn out to be not THAT up, we're easier to get rid of.
Slashdot Mirror
From my perspective, if Y'CbCr RGB conversions are so bad, just keep everything in Y'CbCr.
For any display of any size, you aren't going to see a difference between 4:2:2 and 4:4:4. And of course, lots of apps to Y'CbCr 4:4:4 processing internally, and only do the downsampling on import and export.
I agree. This guy isn't a professional in this space. He makes a lot of the assumptions and errors I made five years ago. But they're smart errors and assumptions, and I'm sure he'll find the Slashdot experience a great way to get all your errors pointed out quickly :).
He really hasn't done the math to realize the advantages of Y'CbCr are more about bandwidth reduction, interoperability with existing color video technology, and easier signal processing. While we certainly could build an RGB DVD player for roughly the cost of a Y'CbCr one, it'd only get half the run time, and would look worse on standard television (only VGA is natively RGB for signal).
Full agreement on Macrovision. Like most CD audio protection, it's the kind of copy protection that irritates users without providing any effective deterrent against skilled pirates.
Oh, I think we've got decades of Y'CbCr to look forward to. it's really an advantageous format in a lot of ways.
First, a 4:2:0 Y'CbCr is half the bandwidth of 4:4:4 RGB. We're a long way away from having half the processing power required, bandwidth, storage, etcetera simply not mattering. My RAID is 2 TB formatted, but I regularly have projects that take up over 50% of the space.
Second, Y'CbCr is a better native space for video processing, since the channels align better with what we want to filter. Luma filters like gamma or contrast are more than 3x faster in Y'CbCr than in RGB, since only one channel needs to be processed. Saturation is more than 6x faster in 4:2:0, since only two channels, each at 25% bandwidth, need to be processed. Plus a lot of filters have to convert from RGB to another color space to run. Y'CbCr is closer to those other spaces, and often doesn't require any conversion. You can say whatever you will about Moore's law, the difference between 4 and 8 real-time layers will matter for a while. Even the audio guys, who have it a lot easier, still run into performance limits with enough simultaneous tracks and such.
Lastly, our entire video infrastructure is build around subsampled Y'CbCr. Never underestimate the lock-in of standards like this. If computer people couldn't kill interlaced video in HDTV, they're never going to kill subsampling for lots of applications. Color video has always been Y'CbCr, and that's how everyone works and thinks for decades now.
That said, Hollywood is likely to pick a >8-bit RGB solution for digital projection. For digital projection, bandwidth is a non-issue, and quality, and quality like that of film. Film guys live in RGB. Plus, it's a win for that industry to have digital cinema be as INCOMPATIBLE with consumer digital video tech as possible, in order to reduce the ease of piracy, and to maintain an advantage of the theatrical experience over home theater.
FWIW, I'm a member of the SMPTE groups working on both video compression and digital cinema.
I just got one of these new bad boys last week:
_ ps bc.html
http://www.macally.com/new/carrying_case/new_cb
MacWorld will be my first real road test of it, but it's great so far. Handles when I want to look professional, shoulder strap for when I'm walking around, and hidden backpack straps for racing through the airport. Lots of internal zippered pockets. Room for my Wintel laptop in there as well. It isn't 100% of what my old Petral bag did, but that didn't fit the 17".
Yeah, and 1024x768 is sub HD. You really want 1280x720 at a minimum. And in order to get as many pixels out of the screen as are being broadcast, you really need 1920x1080.
I've been doing a ton of HD video authoring the last couple months. It's a startling realization to discover that a 1920x1200 23" LCD is the SMALLEST monitor that can be used for this kind of stuff! And for quality grading, I'm sitting there with my nose eight inches from the screen for a couple of hours at a time, looking for minute compression artifacts.
Wow, we're decades away from having AI good enough to do that! Detecting watermarks in a rectangular video stream is hard, sure. But in the real world? That's basically asking for reliable object and feature extraction from real-world images. This is canonically one of the things humans are incredibly better at than computers. Any watermark visible enough to be detected by a moving camera, at an angle on several axes, in variable lighting, etecetera, would be so blindingly obvious it wouldn't qualify as a watermark.
Pixlet? As a distribution codec? RTFM.
Pixlet is great if you need an intraframe-only codec for playback on Mac OS X 10.3 systems only (no Jaguar, no Windows). Pixlet is meant for film production workflows, since it offers frame-accurate rough cuts at half D1 resolution on a fast G4. It is in no way useful as a codec to distribute content to end users.
Think of it as a souped up version of Photo-JPEG, or a complexity constrained version of JPEG2000. It no more competes with Sorenson Video than a Freightliner competes with a Ferrari.
Apple's excellent marketing has actually backfired on them a bit here. Pixlet is very useful for what it's meant for, but if you're not a film professional, it isn't meant for you.
Actually, I think Apple quit trying to fight Windows Media quite a while ago. They're not pushing or enhancing the .mov file format anymore - the best codec available is still Sorenson Video 3, which shipped two and a half years ago.
QuickTime's importance to Apple today is much more as a digital media SDK, and hence the foundation of their very successful products like iMovie, iDVD, Final Cut Pro, DVD Studio Pro, iTunes, iPhoto, etcetera.
QuickTime is also the best authoring architecture out there, with wonderful features like reference movies.
But as a delivery format, Apple has been letting it linger, but isn't putting much effort into enhancing it. They've talked a lot about MPEG-4 being the future of the file format, but haven't done much technically to make that viable either.
Unfortunately, QuickTime as architected can't deal with out of order data in frames, and so can't natively deal with MPEG-1, MPEG-2, or Advanced Simple MPEG-4. I suppose you could make export components for those, but it'll take some serious refactoring before QuickTime could really treat those formats as first class citizen.
As far as QT Pro for $100, I assure you that the user base would go ballistic. Apple does install MPEG-2 encoding (fast, but not very flexible) as a QuickTime component for owners of Final Cut Pro or DVD Studio Pro.
I have to say, this article is the exception that proves Waggoner's eleventh law: "Any reference to 'toxins' as a generic class is in the service to someone selling a bogus pseudo-medical treatment."
So, if we assume for the moment that this theory is true, and that new oil, etcetera is being made all the time, that doesn't necessarily mean that we aren't going to run out.
Would new petrochemicals be formed faster than we can use them up? Certainly the levels in proven reserves drop as they're being pumped up, so any regeneration is on a slower scale than our current consumption.
The problem is, one can die without ever having bought a movie or a computer, since next year it's going to be so much better.
:).
If I buy a DVD, it's because I figure I'll watch it enough during one technological cycle to make it be worth it (five years or so). Now that I've got two kids under 4, I don't buy many DVDs
It's nice we have the announcement, but this is Slashdot!
Have ANY technical details of the new standard been published anywhere yet? I can't find any public resources I can link to.
I've burned about 60 discs in the last week at 4x with the SuperDrive in my G5 with Reitech media. Not a single failure the whole time. And the discs perform as well - I've been able to play back digital media files with 20 Mbps peaks without a hitch. That's about twice the performance required for DVD video.
720x576 is PAL DV resolution. Of course, it's non-square pixel, so playing back on many players will result in a distorted image (about 10% too tall).
It's 6 Mbps! That's enough bitrate for HD encoded content. 640x480 at 1.5 Mbps would have been just fine.
It wasn't deinterlaced! Thus ugly, hard to compress horizontal lines throughout the image.
Basically, the guy did a dump of the PAL DV master into WMV, without any processing to make it appropriate for computer playback.
Big picture is that it'd be trivial to do a better looking version of this clip at a quarter the data rate. And he could have used WMV8 for MPlayer compatibility. WMV9 is a great codec, but for something posted on Slashdot...
Reencoding from one compressed delivery format typically results in a big, big quality hit compared to encoding to the same final format from a high quality mastering format, like DV.
Think of it as giving you the artifacts from BOTH codecs at the same time. But worse.
Really, it's a bad idea.
Also, why is it .zip compressed! Any decent video and audio compression includes its own entropy encoding, so that a .zip or whatever will produce virtually no size savings.
.zip worked, that'd tell you that you did something wrong!
If you had a file where
Yeah, considering its age, MPEG-2 scales remarkably well. And the MIPS per pixel required are a lot lower than more modern codecs. Of course, if bandwidth is at a premium, MPEG-2 isn't a good option. If you want to get a 1080p 2.5 hour movie onto DVD-9 media, Windows Media Video 9 can handle the compression without breaking a sweat, while MPEG-2 wouldn't look that good even with the best encoder. While you might spend more on an ASIC for decoding the WMV9, that could be more than made up for by being able to use conventional media
As in all things, it really depends on what is your limiting factor.
Eh?
First off, VCD isn't really any better quality wise than VHS. VCD is digital, so you don't get analog errors and wear, but it has only half the temporal and vertical resolution.
VCD didn't fail in the states for any reason other than that it didn't provide any better customer value than VHS. VCD won in Asia since it's a cheaper medium to counterfeit. In general, Asian audiences also seem more willing than US audiences to accept lower quality for lower price for video. Have you looked at many of those audience-shot Hong Kong bootlegs? Awful.
DVD is way better than VHS, and so won in the US.
And the DVD Forum is hard at work on a next generation HD DVD standard, and US companies are part of the Forum.
China might be trying an end run around this format, but the US-Japan-Europe industry axis will have their own format coming.
No dis on VP3/4/6, but MPEG-2 does HD fine just well. US HD TV broadcasts are all MPEG-2 (1920x1080 or 1280x720). I'm sitting next to two computers encoding HD MPEG-2 RIGHT NOW for a Kiosk application.
While MPEG-2 doesn't have cutting edge compression efficiency anymore, it isn't bad, and decoder chips are dirt cheap. And you can actually play back 1920x1080 60i MPEG-2 on a fast P4.
Er, sounds good, but you actually don't really know what you're talking about.
A well encoded DVD has very high quality, certainly on par with Betacam SP, the high end analog broadcast production format before digital took over. A well encoded disc won't show significant artifacts.
HDTV resolution goes up to 1920x1080, which is about 6.5x the pixels of DVD (720x480). How high do you want to go? The cheapest displays that can meaningfully do more than 1920 lines wide on a largish screen are awesomely expensive.
Today's displays are crappy? Compared to what? A tapestry? IMAX? We're really at the beginning of a golden age for consumer video technologies. The quality you can buy for $5000 is vastly improved in the last three years, let alone the last 30. Most people don't have eyes good enough to appreciate anything beyond a good 1920x1080.
Lastly, fractals are really the Grassy Knoll of video compression. Yes, Iterated was created to make products on them. No, fractals didn't work. I spent a lot of the mid-late 90's working with Iterated's stuff in different forms. Bitrate scalability was interesting (you could truncate the file at any point, and the more bits you grabbed, the better an image you got). Compression ratios were somewhat better than JPEG. They scaled pretty well. But the net gains were too small to overcome the market share advantages of lowly JPEG.
Iterated simply couldn't make a business around fractal compression. They sold their stuff to AltaMira, who still are selling their fractal compression stuff. Iterated morphed into a company providing image management solutions for the prepress industry. There was still some fractalish stuff underneath, but that wasn't where the value was really added.
The big thing about "fractal" compression is that it wasn't really fractal - its ability to take advantage of self-symmetry was very limited. Heck, even with today's computer power, a "true" fully automatic fractal compressor would take unbelievable amounts of CPU power - many orders of magnitude beyond what realistic video codecs do today. You're basically extending motion search into so many axes.
The only fractal video codec was ClearVideo, which was interesting I suppose, but was roundly stomped by both DCT H.263 derived codecs, and VQ derived codecs like Sorenson Video 1.
Almost everything good about fractals has been inherited by wavelets. And wavelets have also inherited fractal's difficulty in handling motion estimation. That's why DCT and DCT-derived codecs still rule the roost today. Wavelets are great for still image, but no one has come close to devising a really competitive wavelet motion codec.
Maybe someday we'll have a revolution in codecs, but DCT-based codecs like WMV9 and AVC keep on trucking in providing excellent compression efficiency, scalability, and decoder performance.
Back when I paid attention to Dr. Who in detail (late 80's), there were lots of missing episodes from the early years, especially the first two Doctors. The BBC had somehow lost all copies of them internally.
Have all or at least some of these been found in the last ~15 years?
I think "New Ideas from Dead Economists" is a very readable, juicy introduction to the history of the field. Good mix of cleanly presented theory, and biographical information. No particular ideological bent, which is a nice antidote to a lot of the other stuff going on out there. Think a mix of "Quicksilver" and a one of the better O'Reilly books - can't get much more hacker friendly than that!
2 4/ benwaggoner-20
http://www.amazon.com/exec/obidos/ASIN/04522805
Reading the Economist every week is also a great way to keep up to date on what's going on in the world, and in the world of money, with an nicely non US, non Dem v. Rep view of things. It's very classic Liberal - pro free trade, and pro civil rights.
Nah, alas you've roughly nailed it.
Estimates of the Bush tax cuts is that we get something like $500,000 of deficit spending per new job created. I'm not sure what they wanted these tax cuts for, but they clearly weren't tuned to provide maximum job creation at minimum cost.
I don't know that the war is causing that much economic activity. Deficit spending is my big worry. Deficits lead to higher inflation, which causes the Fed to raise interest rates. Higher interest rates tend to push housing down. Since people buy houses based on the sum of the mortgage, interest and taxes, higher interest rates reduce the value of houses people can buy, having a negative effect on house prices. Of course inflation in general has a raising effect, but you can get a situation where the true value of homes is declining, with prices rising, but not as fast as inflation.
I wish more hackers would get interested in economics - it's really very similar to a lot of technical topics. It has become painfully clear to this economics nerd that the USA simply doesn't have a competent economics policy or staff under the current administration.
I'm admittedly in a weird niche (compressed video consulting and training), but the demand for my services has been ramping up nicely since February, and just blasting off in the last month. I expect to bill more Q4 of this year than I billed in all of 2002.
And it isn't just one client. It's coming from a lot of different directions, from a lot of different companies and industries. And nice, juicy, interesting jobs too. It was like the outsourcing switch just got turned on.
Of course, independent consultants like me are often a good six months ahead of the rest of the economy. When things get tight, consultants get cut first, and when things are looking up, consultants get hired before full-time employees, since if things turn out to be not THAT up, we're easier to get rid of.