1920x1080 pixels 30 frames a second 16 bits per pixel* That's be 949 Mbps, or 118 MB per second.
Or about 70 minutes of uncompressed editing on this at max resolution.
Of course, being FireWire, it'll have a lowly peak data rate of 400 Mbps. We'd need the 1394b 1600 Mbps standard for this to be useful for uncompressed HD editing. This is why honkin' Ultra-160 RAID systems are used for this kind of work!
The good thing is that over the air HD transmissions are a measly 19.2 Mbps. That'd give you 58 hours or so.
* (it's YUV with chroma sampled at 4:2:0, so there is one luma bitmap at 1920x1080, and two chroma bitmaps at 960x520, all at 8 bits per channel).
For smooth motion, capture frame rate should be an integer division of the source frame rate. So, for NTSC, 60, 30, and 15 are good options, but 24 isn't.
Of course, if you do an inverse telecine with film source originally shot at 24p, restoring it to 24, 24 is a fine option.
Why 384x288 for NTSC? That's quarter PAL screen. Since NTSC only has 480 lines, quarter screen is normally 320x240.
DivX 5 handles interlaced video, I believe. 640x480 DivX 5 could give you better quality, and the file size wouldn't be too much worse if you turn on all those Advanced Simple features (you'll need a fast box though).
Plus it's progressive scan, so actually losing 75% of the original 60 frame interlaced NTSC source. This system was definitely designed by a computer guy, not a video guy (as honorable and deep a form of geekhood as any).
He really wants something that can do interlaced capture, like MPEG-2 or some of the higher profile versions of MPEG-4. I believe this is supported in the current CVS of Xvid, at least experimentally. MPEG-4 would give a LOT smaller file sizes than MPEG-2 at the same quality.
Or, since VCD is his goal, he could capture straight to ffmpeg in a VCD compatible profile. Or use SVCD MPEG-2, which would be higher quality, and could still fit a half-hour show on a 800 MB CD-ROM.
Windows Media 9 has great, free, integrated capture that can do interlaced, but that'd be hardly Linux friendly:).
Okay, if each one is double the one before, that means you'll have a 2^100 ratio between lowest and highest data rate. Thus, if your lowest is 1 Kbps, the highest would be... Not going to happen that way.
Also, you assume the sweet spots are 2x the one before. In fact, jumps of more like 1.25 are likely to be optimal (albeit with a lot fewer jumps!).
Scalable techniques like this are very cool, but hardly novel.
MPEG-4's scalable profiles provide a similar effect (albeit in the other direction, with enhancement layers). Some of the higher end audio codecs (beyond CELP and AAC), like ER BSAC (Error Resistant Bit Sliced Arithmetic Coding) do exactly this. The idea in this case is that the server will in real-time only provide as many bits as the connection can currently provide. Very nice for wireless.
QDesign's QDX format does almost exactly what is described for Ogg, with arbitrary bitrate peeling down to the 1 Kbps level. The idea is that you could copy as much data as you want to your mobile player, and it'd dynamically thin to the data rate that would fill up your device.
And still image codecs like JPEG have used progressive modes for years, where additional data adds more detail to the image.
Specifically, MacOS X and PocketPC don't have a decoder for the ACELP.net voice audio codec. Microsoft licensed that from a third party who didn't provide updated binaries. WM9 uses a new voice codec developed internally that they'll be able to deploy cross-platform.
Windows Media for MacOS only supports WM DRM v1, which only supports the older WMV7 codec, not the WMV8 MovieLink is using. Presumably they're using DRM 7.1 (7.0 was cracked). However, MovieLink will run on Windows 98, which doesn't support the Secure Audio Path, so there isn't a huge technical DRM difference here.
Real's subscription service is available for MacOS X with the full functionality of Windows, so their DRM is presumably feature complete cross-platform. And I believe for Linux as well, but I haven't checked.
This is all rather besides the point. Even if Apple doesn't provide system-level DRM, application-level DRM works just fine in the formats MovieLink is using (RealMedia and Windows Media). And Windows, while they talk about system-level DRM eventually with Palladium, doesn't have it either today.
So, whatever MovieLink might claim is their reason, they aren't technical. They probably don't want to do it for marketshare reasons, and are using Apple's DRM statements (which are really rather mild) as an excuse/flogging horse.
Good, focused trade shows, and why they rock
on
The Last Comdex?
·
· Score: 5, Insightful
Me, I love trade shows. More specifically, I love the trade shows I love.
I work out of my house or on client locations most of the year, so trade shows are where I can actually get some networking done. With a focused, industry specific show, a significant portion of my collegues, clients, and potential clients are in one place. Business gets done.
For folks who only get the free exhibit pass and do the show floor, you're missing about 90% of the action, and the 10% you've got left has been rendered somewhat irrevelant by the internet. Sure, trolling the show floor is nice, and you can occassionally see some surprises, or see a product close up and grok it in a way written descriptions didn't work. But, in a four day trade show, I might spend four hours looking at the exhibits.
One thing a good trade show will have is good sessions, taught by people who know what they're talking about. In the dot-com era, there were way too many shows where it was the VP's of marketing up on the stage, but the ones that are left focus much more on people with real-world experience telling their stories and, if they're good, answering questions. War stories can teach a LOT, and an expert can often answer a question in 30 seconds that might take a full day on Google to get straight.
Having a conference pass also helps beyond just doing the conferences. It gets you mingling with the other attendees. There is often free snacks and coffee, and sometimes full meals for attendees as well. Wonderful networking at those. The conference pass also gets you a lot more attention at the booths, because the vendors know that someone with a pass paid money for the show, and immediately consider them a more serious prospect. I've worked both sides of the booth, and the sales managners always stress this point in pre-show prep for those working the booths.
These days, I only go to shows that I'm either speaking at (I do lots of sessions about video compression), or that I have press creditials for (I'm a contributing editor of DV Magazine). Either pass is nice, since you can get into the speaker or press room, which is GREAT networking, plus they normally have broadband, drinks, and snacks, and not just at designated snack time. Wearing a press badge on the show floor can be almost dangerous if the marketing guys see you (the sales guys typically couldn't care less).
The big problem with Comdex is that it is so diffuse, it's hard to imagine it having a focused enough audience to have a good chance of bumping into people into the same stuff you are. The computer industry is so broad, it'd be like having a trade show on "transportation." It underlies so many things, it can't be really treated as a unified whole.
But in general, just going to a show for the exhibit floor is only scratching the surface. Try to get a conference pass, or even try to get a speaking gig if possible. But if you can't swing either, at least try to track down the free vendor classes, and any relevant free Birds of a Feather sessions (generally run in the evening). The one thing the internet can't give us is actually talking to 3D people, so focus on that aspect to get value out of a show.
And if you do go to conferene sessions, ask questions! And it's perfectly expected and accepted to go up to the speaker after the session for followups.
Trade shows I love (being a compression nerd) are:
DVExpo. Lots of classes by practicing video people, very enthusaistic audience. Probably the highest consistant quality of sessions of any show I do.
QuickTime Live: Geeky when it needs to be, but man does Apple know how to throw a party! Also excellent session quality. And catered by Odwalla!
NAB: A huge show for video professionals. Amazing exhibits, and enough different conference tracks to keep things interesting for everybody. The geeky stuff used to be done by DVExpo, who alas don't handle that anymore. Still a fun show.
MacWorld: Verging on diffuse, since people use Macs in so many different ways, but the great Esprit de corps. The Stevenotes really are best experienced in person for maximum RDF impact (and you often get gifts under the seats). I was at the infamous Lou Gestner 3 hour marathon one a few years ago, and man is that a telling contrast!
WEMP: This is put on by the MPEG-4 Industry Forum. I've only gone once, but it was the best in codec nerd love. Truly excellent sessions - it's one thing to read a standard, it's another thing to hear the person who wrote it tell you why it's a certain way.
Well, it would depend on the rights that were assigned to you by the content holder. I imagine a company could assign a "works on all CD Players" right. You might also be able to go to their web site and have them issue you a new rights package for a different machine.
Intertrust doesn't define the business rules, it just enforces them. It's up to the content providers to figure out how much a pain in the arse to be.
As long as it targets MPEG-1 in VCD compatible mode, that means only 352x240 resolution. The only way that's going to look good on a Cinema display is if it just plays in a small window. Even a good 720x480 DVD can show with problems digital scaled up that much. It isn't as bad with CRT, since the video card can lower the resolution, softening out some of the scaling artifacts you get with a fixed resolution display.
Sorenson has been pretty blase about this kind of stuff so far. They couldn't distribute the decoder to other folks due to their contract with Apple. But they make money selling ENCODERS, so they don't have much of a motivation to come down hard on folks who are increasing their audience.
What's wrong with the mail server? My needs aren't complex, but it's been running without a hitch for me for ages.
As for t"just tinker with via & the zone files," that is rather antithetical to whole reason why I'd use MacOS X Server in the first place! Back in the summer of 1989, I worked at a company that did banking software, and the whole company was on SCO. vi was the standard word processor!
Ah, but I bought 10.0 server over a year ago, and am using the $20 10.1 upgrade to that. So I can't really complain. It's just that 10.2 doesn't do enough different than 10.1 for the stuff I do (file server, FTP, email).
I look forward to Apple figuring out compelling resons for me to give them my money (something they have an excellent record of with laptops...).
Apple should put .mac features ing MacOS X Server
on
Fake Your Own .Mac Server
·
· Score: 5, Interesting
There are lots of neat things that.mac does (I paid my $50), but it'd be great if I could do them locally. Backup is a HUGE one - 100Base-T is a lot better than DSL! But being able to use the pretty slick webmail for my own mail domain would be very useful as well, instead of having to forward to my mac.com address. Local iCal would be very nice as well.
I can certainly understand why Apple doesn't want to make these available everywhere for free, but it'd be great if MacOS X Server 10.3 or whatever made it possible to provide some.mac services to my local users. Sure, there are variety of ways to hack it together, but if it all "just worked" that'd be better yet.
This would certainly give me a reason to pay $1K to upgrade from the 10.1.5 server I'm running right now.
And on an unrelated note, 10.3 REALLY should include a graphical DNS admin. It's really jarring to have all these great, simple controls for the whole server experience, except DNS. Webmin works, but still, that's hardly the MacOS X vision!
Of course, if you're implementing a digital stream in the first place, it is MUCH better to just stuff the metadata into the stream as binary data as well. Which is what all the digital formast do. The domain for this solution is only full-bandwidth uncompressed signals, which are going to increasingly be a thing of the past.
Prebinding not needed in 10.2
on
Is Mac OS X Slow?
·
· Score: 3, Informative
Manual prebinding is no longer needed in 10.2. The first time a non-prebound app is launched, the OS will quietly prebind it behing the scenes, so the second launch will be at full speed.
Of course, many installers will still do it on install. This is kind of irritating if you have to do a lot of installs at once, like update a stock install with all the updates.
The thing about perceptual codecs is that they only try to preserve the information that is perceptible. Thus, any imperceptable information, like white noise or watermarking, will tend to get stripped out. And even if a system can carry through a particular compression technique, it might not work with other, future, more advanced techniques.
This whole technology is based around sending full-bandwidth signals, which is definitely NOT the trend in digital communications.
Stegography-like stuff requires lossless compression ala GIF. Doesn't work well with JPEG!
Not only do I think this will be ineffective, I think in many cases it'll be self-defeating.
I've got a toddler in the house, which means that CD cases left in the open get opened and covered in peanut butter fingerprints. C'est la vie, so I went ahead and ripped my library via iTunes to a pair of 80 GB drives, and now I've got a wonderful, searchable, kid-proof music library.
I simply can't imagine going back to having to deal with physical CD media anymore. I'm happy to rip the disc when I get it and put it in the storage room, but that's about it.
So, if I really wanted music that was on a copy-protected format that was effective, I'd HAVE to pirate it to listen to it.
Other folks are in the same boat - everyone who listens to music on systems not compatible with this protection. The presumption behind this copy protection is that users will replace their in-dash CD players with a compatible one. Instead, I think it is MUCH more likely users will return the CD to the store, and download the tracks from a P2P site.
It only takes one user to crack the copy protection to make the content available online. But EVERY case where the copy protection works is a lost sale for the record company.
They need to understand that the effectiveness of a copyright protection scheme is inverse proportion with how difficult the copy protected version is to use compared to a cracked version.
This is one of the reasons dongles have been disappearing in the software industry - users would crack a legit copy just to use the software on a laptop!
Newer codecs certainly do use more CPU cycles per pixel. One of the biggest areas where they get improvements in compression efficiency is through using more efficient but computationally expensive techniques.
They mediate these in a number of ways. First, aggressive SIMD optimization is always used now. RealVideo 9 is highly optimized for AltiVec, which your G3 doesn't have.
Also, they can do optional postprocessing, like deringing and deblocking. These can improve the quality of the decoded video. More expensive but higher quality techniques will be used on faster processors. This means faster machines might not see the CPU load of decode drop much, but the quality of the final video will improve.
Ah, but the fundamental measure of compression efficiency is how few fractional bits per pixel you need to achieve "good enough" quality. RealVideo 9 needs fewer bits per pixel, so it can use higher resolutions. For example 192x144 is just fine with lots of content with RV9 over modem.
At low bitrates, RV9 is quite a bit better than MPEG-4 implementations like Divx. This may change with MPEG-4 part 10/H.264, but we don't have any tools that use that yet.
Slashdot Mirror
Doh! Of course. I was thinking of 4:2:2, which is used for lossless editing of standard def content. HD uses 4:2:0 instead.
Well, if you wanted to use this for HD editing...
1920x1080 pixels
30 frames a second
16 bits per pixel*
That's be 949 Mbps, or 118 MB per second.
Or about 70 minutes of uncompressed editing on this at max resolution.
Of course, being FireWire, it'll have a lowly peak data rate of 400 Mbps. We'd need the 1394b 1600 Mbps standard for this to be useful for uncompressed HD editing. This is why honkin' Ultra-160 RAID systems are used for this kind of work!
The good thing is that over the air HD transmissions are a measly 19.2 Mbps. That'd give you 58 hours or so.
* (it's YUV with chroma sampled at 4:2:0, so there is one luma bitmap at 1920x1080, and two chroma bitmaps at 960x520, all at 8 bits per channel).
For smooth motion, capture frame rate should be an integer division of the source frame rate. So, for NTSC, 60, 30, and 15 are good options, but 24 isn't.
Of course, if you do an inverse telecine with film source originally shot at 24p, restoring it to 24, 24 is a fine option.
Why 384x288 for NTSC? That's quarter PAL screen. Since NTSC only has 480 lines, quarter screen is normally 320x240.
DivX 5 handles interlaced video, I believe. 640x480 DivX 5 could give you better quality, and the file size wouldn't be too much worse if you turn on all those Advanced Simple features (you'll need a fast box though).
Ah, but in fact 29.97 is an approximation itself. The actual value is 30000/1001.
:).
As long as we're being geeky
Plus it's progressive scan, so actually losing 75% of the original 60 frame interlaced NTSC source. This system was definitely designed by a computer guy, not a video guy (as honorable and deep a form of geekhood as any).
:).
He really wants something that can do interlaced capture, like MPEG-2 or some of the higher profile versions of MPEG-4. I believe this is supported in the current CVS of Xvid, at least experimentally. MPEG-4 would give a LOT smaller file sizes than MPEG-2 at the same quality.
Or, since VCD is his goal, he could capture straight to ffmpeg in a VCD compatible profile. Or use SVCD MPEG-2, which would be higher quality, and could still fit a half-hour show on a 800 MB CD-ROM.
Windows Media 9 has great, free, integrated capture that can do interlaced, but that'd be hardly Linux friendly
Okay, if each one is double the one before, that means you'll have a 2^100 ratio between lowest and highest data rate. Thus, if your lowest is 1 Kbps, the highest would be... Not going to happen that way.
Also, you assume the sweet spots are 2x the one before. In fact, jumps of more like 1.25 are likely to be optimal (albeit with a lot fewer jumps!).
Scalable techniques like this are very cool, but hardly novel.
MPEG-4's scalable profiles provide a similar effect (albeit in the other direction, with enhancement layers). Some of the higher end audio codecs (beyond CELP and AAC), like ER BSAC (Error Resistant Bit Sliced Arithmetic Coding) do exactly this. The idea in this case is that the server will in real-time only provide as many bits as the connection can currently provide. Very nice for wireless.
QDesign's QDX format does almost exactly what is described for Ogg, with arbitrary bitrate peeling down to the 1 Kbps level. The idea is that you could copy as much data as you want to your mobile player, and it'd dynamically thin to the data rate that would fill up your device.
And still image codecs like JPEG have used progressive modes for years, where additional data adds more detail to the image.
Specifically, MacOS X and PocketPC don't have a decoder for the ACELP.net voice audio codec. Microsoft licensed that from a third party who didn't provide updated binaries. WM9 uses a new voice codec developed internally that they'll be able to deploy cross-platform.
Some technical points on Mac DRM:
m 7/ drm/offering.asp
Windows Media for MacOS only supports WM DRM v1, which only supports the older WMV7 codec, not the WMV8 MovieLink is using. Presumably they're using DRM 7.1 (7.0 was cracked). However, MovieLink will run on Windows 98, which doesn't support the Secure Audio Path, so there isn't a huge technical DRM difference here.
http://www.microsoft.com/windows/windowsmedia/w
Real's subscription service is available for MacOS X with the full functionality of Windows, so their DRM is presumably feature complete cross-platform. And I believe for Linux as well, but I haven't checked.
This is all rather besides the point. Even if Apple doesn't provide system-level DRM, application-level DRM works just fine in the formats MovieLink is using (RealMedia and Windows Media). And Windows, while they talk about system-level DRM eventually with Palladium, doesn't have it either today.
So, whatever MovieLink might claim is their reason, they aren't technical. They probably don't want to do it for marketshare reasons, and are using Apple's DRM statements (which are really rather mild) as an excuse/flogging horse.
Me, I love trade shows. More specifically, I love the trade shows I love.
I work out of my house or on client locations most of the year, so trade shows are where I can actually get some networking done. With a focused, industry specific show, a significant portion of my collegues, clients, and potential clients are in one place. Business gets done.
For folks who only get the free exhibit pass and do the show floor, you're missing about 90% of the action, and the 10% you've got left has been rendered somewhat irrevelant by the internet. Sure, trolling the show floor is nice, and you can occassionally see some surprises, or see a product close up and grok it in a way written descriptions didn't work. But, in a four day trade show, I might spend four hours looking at the exhibits.
One thing a good trade show will have is good sessions, taught by people who know what they're talking about. In the dot-com era, there were way too many shows where it was the VP's of marketing up on the stage, but the ones that are left focus much more on people with real-world experience telling their stories and, if they're good, answering questions. War stories can teach a LOT, and an expert can often answer a question in 30 seconds that might take a full day on Google to get straight.
Having a conference pass also helps beyond just doing the conferences. It gets you mingling with the other attendees. There is often free snacks and coffee, and sometimes full meals for attendees as well. Wonderful networking at those. The conference pass also gets you a lot more attention at the booths, because the vendors know that someone with a pass paid money for the show, and immediately consider them a more serious prospect. I've worked both sides of the booth, and the sales managners always stress this point in pre-show prep for those working the booths.
These days, I only go to shows that I'm either speaking at (I do lots of sessions about video compression), or that I have press creditials for (I'm a contributing editor of DV Magazine). Either pass is nice, since you can get into the speaker or press room, which is GREAT networking, plus they normally have broadband, drinks, and snacks, and not just at designated snack time. Wearing a press badge on the show floor can be almost dangerous if the marketing guys see you (the sales guys typically couldn't care less).
The big problem with Comdex is that it is so diffuse, it's hard to imagine it having a focused enough audience to have a good chance of bumping into people into the same stuff you are. The computer industry is so broad, it'd be like having a trade show on "transportation." It underlies so many things, it can't be really treated as a unified whole.
But in general, just going to a show for the exhibit floor is only scratching the surface. Try to get a conference pass, or even try to get a speaking gig if possible. But if you can't swing either, at least try to track down the free vendor classes, and any relevant free Birds of a Feather sessions (generally run in the evening). The one thing the internet can't give us is actually talking to 3D people, so focus on that aspect to get value out of a show.
And if you do go to conferene sessions, ask questions! And it's perfectly expected and accepted to go up to the speaker after the session for followups.
Trade shows I love (being a compression nerd) are:
DVExpo. Lots of classes by practicing video people, very enthusaistic audience. Probably the highest consistant quality of sessions of any show I do.
QuickTime Live: Geeky when it needs to be, but man does Apple know how to throw a party! Also excellent session quality. And catered by Odwalla!
NAB: A huge show for video professionals. Amazing exhibits, and enough different conference tracks to keep things interesting for everybody. The geeky stuff used to be done by DVExpo, who alas don't handle that anymore. Still a fun show.
MacWorld: Verging on diffuse, since people use Macs in so many different ways, but the great Esprit de corps. The Stevenotes really are best experienced in person for maximum RDF impact (and you often get gifts under the seats). I was at the infamous Lou Gestner 3 hour marathon one a few years ago, and man is that a telling contrast!
WEMP: This is put on by the MPEG-4 Industry Forum. I've only gone once, but it was the best in codec nerd love. Truly excellent sessions - it's one thing to read a standard, it's another thing to hear the person who wrote it tell you why it's a certain way.
Well, it would depend on the rights that were assigned to you by the content holder. I imagine a company could assign a "works on all CD Players" right. You might also be able to go to their web site and have them issue you a new rights package for a different machine.
Intertrust doesn't define the business rules, it just enforces them. It's up to the content providers to figure out how much a pain in the arse to be.
As long as it targets MPEG-1 in VCD compatible mode, that means only 352x240 resolution. The only way that's going to look good on a Cinema display is if it just plays in a small window. Even a good 720x480 DVD can show with problems digital scaled up that much. It isn't as bad with CRT, since the video card can lower the resolution, softening out some of the scaling artifacts you get with a fixed resolution display.
an ATSC (USA standard) HDTV stream goes up to about 19.2 Mbps, so no, USB 1.1 doesn't have the bandwidth needed. And it's MPEG-2, not MPEG-1.
If you want to convert it to a low resolution MPEG-1, sure. But why?
Sorenson has been pretty blase about this kind of stuff so far. They couldn't distribute the decoder to other folks due to their contract with Apple. But they make money selling ENCODERS, so they don't have much of a motivation to come down hard on folks who are increasing their audience.
What's wrong with the mail server? My needs aren't complex, but it's been running without a hitch for me for ages.
As for t"just tinker with via & the zone files," that is rather antithetical to whole reason why I'd use MacOS X Server in the first place! Back in the summer of 1989, I worked at a company that did banking software, and the whole company was on SCO. vi was the standard word processor!
I'm a pico man now.
Ah, but I bought 10.0 server over a year ago, and am using the $20 10.1 upgrade to that. So I can't really complain. It's just that 10.2 doesn't do enough different than 10.1 for the stuff I do (file server, FTP, email).
I look forward to Apple figuring out compelling resons for me to give them my money (something they have an excellent record of with laptops...).
There are lots of neat things that .mac does (I paid my $50), but it'd be great if I could do them locally. Backup is a HUGE one - 100Base-T is a lot better than DSL! But being able to use the pretty slick webmail for my own mail domain would be very useful as well, instead of having to forward to my mac.com address. Local iCal would be very nice as well.
.mac services to my local users. Sure, there are variety of ways to hack it together, but if it all "just worked" that'd be better yet.
I can certainly understand why Apple doesn't want to make these available everywhere for free, but it'd be great if MacOS X Server 10.3 or whatever made it possible to provide some
This would certainly give me a reason to pay $1K to upgrade from the 10.1.5 server I'm running right now.
And on an unrelated note, 10.3 REALLY should include a graphical DNS admin. It's really jarring to have all these great, simple controls for the whole server experience, except DNS. Webmin works, but still, that's hardly the MacOS X vision!
Of course, if you're implementing a digital stream in the first place, it is MUCH better to just stuff the metadata into the stream as binary data as well. Which is what all the digital formast do. The domain for this solution is only full-bandwidth uncompressed signals, which are going to increasingly be a thing of the past.
Manual prebinding is no longer needed in 10.2. The first time a non-prebound app is launched, the OS will quietly prebind it behing the scenes, so the second launch will be at full speed.
Of course, many installers will still do it on install. This is kind of irritating if you have to do a lot of installs at once, like update a stock install with all the updates.
The thing about perceptual codecs is that they only try to preserve the information that is perceptible. Thus, any imperceptable information, like white noise or watermarking, will tend to get stripped out. And even if a system can carry through a particular compression technique, it might not work with other, future, more advanced techniques.
This whole technology is based around sending full-bandwidth signals, which is definitely NOT the trend in digital communications.
Stegography-like stuff requires lossless compression ala GIF. Doesn't work well with JPEG!
Not only do I think this will be ineffective, I think in many cases it'll be self-defeating.
I've got a toddler in the house, which means that CD cases left in the open get opened and covered in peanut butter fingerprints. C'est la vie, so I went ahead and ripped my library via iTunes to a pair of 80 GB drives, and now I've got a wonderful, searchable, kid-proof music library.
I simply can't imagine going back to having to deal with physical CD media anymore. I'm happy to rip the disc when I get it and put it in the storage room, but that's about it.
So, if I really wanted music that was on a copy-protected format that was effective, I'd HAVE to pirate it to listen to it.
Other folks are in the same boat - everyone who listens to music on systems not compatible with this protection. The presumption behind this copy protection is that users will replace their in-dash CD players with a compatible one. Instead, I think it is MUCH more likely users will return the CD to the store, and download the tracks from a P2P site.
It only takes one user to crack the copy protection to make the content available online. But EVERY case where the copy protection works is a lost sale for the record company.
They need to understand that the effectiveness of a copyright protection scheme is inverse proportion with how difficult the copy protected version is to use compared to a cracked version.
This is one of the reasons dongles have been disappearing in the software industry - users would crack a legit copy just to use the software on a laptop!
Yes, I meant "efficient", thanks.
Newer codecs certainly do use more CPU cycles per pixel. One of the biggest areas where they get improvements in compression efficiency is through using more efficient but computationally expensive techniques.
They mediate these in a number of ways. First, aggressive SIMD optimization is always used now. RealVideo 9 is highly optimized for AltiVec, which your G3 doesn't have.
Also, they can do optional postprocessing, like deringing and deblocking. These can improve the quality of the decoded video. More expensive but higher quality techniques will be used on faster processors. This means faster machines might not see the CPU load of decode drop much, but the quality of the final video will improve.
Ah, but the fundamental measure of compression efficiency is how few fractional bits per pixel you need to achieve "good enough" quality. RealVideo 9 needs fewer bits per pixel, so it can use higher resolutions. For example 192x144 is just fine with lots of content with RV9 over modem.
At low bitrates, RV9 is quite a bit better than MPEG-4 implementations like Divx. This may change with MPEG-4 part 10/H.264, but we don't have any tools that use that yet.