This really, really depends on the kind of compression you're using. VCD and SVCD use MPEG-1 and MPEG-2, which are very old technology at this point. MPEG-4 advanced simple, which is used in Divx, can provide similar quality at a half or less the data rate. And the forthcoming MPEG-4 Part 10 codec (aka H.264) will be able to do high quality standard definition (720x480) at slightly less than VCD data rates, and entertainment (movies, tv shows) at about 60% VCD data rates.
So, that'll enable putting an entire feature-length film or two hour special on a single CD-RW.
Of course, this is perfectly possible today with propritary technologies like Windows Media. The great thing about MPEG-4 is it is an international standard. And the baseline profile of Part 10 is supposed to be patent and license free, so hopefully we won't see the same confusion and delays that we did for MPEG-4 adoption.
VHS has a ton of analog noise. This means that you'll need to encode the digital copy at a substantially higher data rate to get the same effective quality, and you'll have a pretty low ceiling on maximum quality.
The difference between even S-VHS and VHS is huge.
So, grabbing off DVD or straight from a high-bitrate PVR would be quite a bit better. And if you have to go through analog, make sure you're capturing via S-Video instead of composite. Otherwise areas of saturated color will get that annoying cross-hatching effect. It's isn't so noticible on TV, but man is it obvious on a computer monitor!
I'm surprised you can even get 60 minutes on a disc
SVCD NTSC is 480x480, which is nearly three times the resolution of VCD (352x240). MPEG-2 is somewhat more efficient than MPEG-1, and you can use VBR to reduce the data rate for easy portions of the video. But still, when you're using half as many bits per pixel, MPEG-2 is going to start falling apart pretty quickly.
Still, there are lots of good new codecs on the horizon. The Advanced Simple MPEG-4 used in Divx should be supplanted over the next year by MPEG-4 Part 10 (aka H.264). It'll be an open ISO standard, and the baseline profile shouldn't (cross fingers) have any of the patent and licensing issues that held up MPEG-4 Part 2's codecs so long.
You should be able to get at least a 3x improvement in compression efficiency compared to MPEG-2, so a full-length feature at standard definition on a 700 MB disc should be quite doable.
Of course, you can do this with propritary codecs like Windows Media Video today, but this IS slashdot...
analog video "lines" and lines of digital resolution are two different things, alas.
When someone talks about lines of VHS, they're talking about how many discreet changes in amplitude you'll be able to measure on a horizontal line. And when digital video talks about lines, it's normally how many pixels high the video is.
And due to Nyquist, we know the sampling rate required to record a given frequncy is twice the frequency.
So, MPEG-1 NTSC VCD at 352 pixels wide could reproduce a frequency of 176 changes over the horizontal width of the video. So, if VHS is 250 lines, it's actually better on that measure than VCD.
Of course, VHS is plauged by horrible analog noise to the point where I can't watch it, while VCD, although low resolution, shouldn't have any noise at all. There will be some artifacts at VCD data rates, of course.
Even if the entire increase in autism was caused by vaccinations, they'd still be worth it. The infant and childhood death rate was many, many times higher in years past without vaccinations.
And it's not like they're a new technology. Benjamin Franklin wrote an essay on why he wished he had given his son the smallbox vaccination.
Well, here's the test then. Do those with autism go on to have more children or grandchilden than average?
If not, autism in itself is evolutionarily negative. However, it is possible that a milder expression of the same genes does have advantages, ala tech centers. Think sickle-cell anemia, where one gene gives you malaria resistantance, but two make you very sick.
Remeber, a human is a gametes way of making more gametes. Cool abilities don't count in evolution if they don't lead to more grandchildren.
Re:hrm, somethings amiss, me thinks
on
Itanium Problems
·
· Score: 2
But if AMD has lower margin per part, then they'll need to seel more total units to equal Intel's R&D. They'll need to gain absolute, not just relative, market share to really get a long-term leg up.
Re:hrm, somethings amiss, me thinks
on
Itanium Problems
·
· Score: 2
Oh, I have no doubt that Hammer could beat Itanium in the next generation. However, if IA-64 is a 20-year play, it doesn't need to win in the next generation (which is good for it, since it is unlikely to win in the next generation).
But, assuming Intel did their work correctly, IA64 will have long-term advantages over extensions of IA32.
Bear in mind that the orginal 8086 was for a while supplanted by the 8088 for compatibility reasons, but the 16-bit architecture won in the end.
Re:hrm, somethings amiss, me thinks
on
Itanium Problems
·
· Score: 4, Insightful
Given that Intel plans a 20 year life for the IA64 technology, they're going to go through a number of business cycles. The way to make money during the boom is to have built good products during the preceeding bust, and have them ready to sell once there is a market for them. A poor economy can gut AMD's budget just as much as Intel's, actually improving IA64's long term prospects.
This current bust is mainly just a post-bubble bust, just like "The New Economy" was mainly just a bubble. Companies will eventually start spending again, and eventually they'll even start overspending again, and then cut projects, rinse repeat.
The compression rate isn't the key. Saying 4:1 is better than 12:1 is like saying that IMA audio compression is better than AAC, which it manifestly isn't.
While a case can be made that DTS can be better than AC-3 in some cases (although way too many of the cited comparisons looked at 384 AC-3, where 448 is now standard for high-end discs), the fact that it takes up so much more space on the disc means that in many cases the video quality will need to be degraded to compensate.
While DTS might have a slight edge in terms of maximum quality, AC-3 has a huge advantage in terms of compression efficiency, which means it wins is places where file size counts significantly.
And while it might be possible to have both parties log into an ISP that supports V.92, this would hurt more than it helps. Yes, V.92 would increase bit rate somewhat (max of 48 instead of 34), but having to log in through an ISP would also increase latency a painful amount.
Communicating point-to-point helps the latency more than the lower data rate hurts the signal.
At least that's my guess of why they went that way.
If the Red HD system used MPEG-2, maybe. But since they'll use either MPEG-4 or a Windows Media derivative, they need a LOT fewer bits per pixel to delivery.
There is also a lot of cruft on current DVD's as well. For a feature length film, you'd use a dual-layer DVD-9, not offer different aspect ratios, and probably would dump the vastly overrated DTS. Probably would use AC-3 and/or AAC. You'd be able to sustain 8 Mbps average this way. And heck, there are only 160% more bits/frame with 1280x720 than 720x480, so we only need a 50% improvement in compression efficiency to be able to pull this off. MPEG-4 Advanced Simple provides this easily.
Except that Blu-ray couldn't possibly be mass market by Christmas 2003. The nice thing about a red laser system is that the physical medium of the disc doesn't have the change, which means the hardware in existing DVD players can be mostly the same, with just a different decoder chip. Fast computers will just need a software update. And, of course, replication and duplication facilities won't need to chance, so it'll cost well less than $1 to make an HD disc, which means we could start seeing mass market prices very quickly.
This is really good from the Hollywood perspective. They'll get us all to buy 1280x720 red laser HD discs from 2003-2006, and then come out with 1920x1080 Blu-ray as a mass market technology around Christmas 2006-2008, when they get all the kinks worked out. Same way we've already bought DVD and laserdisc versions of the same movie.
The article claims that the compression technology will be from Microsoft, but my contacts tell me it is much more likely to be MPEG-4, in order to have a technology not tied to any one vendor. Of course, Windows Media derived codecs would offer better compression efficiency. We shall see.
Well, from a legal perspective, you want to make BOTH crimes. Already anti-spam pressure has kept most legitimate businesses away from doing it.
In essence, we want to make it as socially and legally difficult to admit to being in any way distributing or profiting from spam as, say being in any way distributing or profiting from stolen car parts.
This is an interesting and relatively well done test (although it appears that the listeners knew which format they were listening to, so it wasn't truly double-blind, and a anti-MS and pro-Ogg bias can't be ruled out).
However, some discussions seem to be focusing on this saying AAC is bad or WMA is bad, when really it refers to the particular implementations in codecs of those formats.
For example, the Apple MPEG-4 AAC-LC encoder was used for AAC. This is a Low Complexity version of the format. Also, the Apple encoder has a strange limitation where it automatically converts 44.1 stereo to 32 stereo at that data rate. This isn't required by the AAC format. Other AAC encoders yield MUCH better results, and beat MP3 Pro in double-blind testing. I haven't seen any double-blind comparisons between AAC and Ogg.
Also, the WMA8 encoder is due to be replaced by the backwards-compatible WMA9 in early September. Of course, there may well be improved versions of the other encoders by then as well.
They aren't releasing any codecs. Streaming servers don't parse the underlying video data, they just (well, that's a misuse of the word "just" - it's hard to do right) turn the file into packets, and try to make sure that as many arrive to the client on time as possible.
For example, today's RealServer can stream QuickTime files, and they certainly don't have source to the Sorenson codecs.
One great thing about VCD is that they work in standard DVD players (NTSC and PAL VCD discs only work in players for the respective formats, of course; a limitation not seen on computers). However, lots of older DVD players can't read CD-R media, because of an incompatibiliity with the DVD laser. Duplicated VCDs on normal CD media work just fine, of course. All DVD players in the market today should work, AFAIK.
If MPEG-4 goes the way of MP3, it'll be hugely popular, widely used, and the foundation for a new industry!
The final MPEG-4 licensing terms were announced. For content providers, you have to pay $0.02 per HOUR (not minute) or $0.25 per subscriber, after the first 50,000 subscribers. So if you have fewer than 50,000 people looking at your content, you don't owe anything.
Apple has to pay a couple $1M for the encoders and decoders, but that's their problem.
Alas, MPEG-1`is a lousy format for PVR use of any quality. MPEG-2 only stores video as progressive scan. However, TV is broadcast as interlaced, where the even lines are captured 1/59.94th of a second off from the odd lines. The difference between fields includes half of the video compression information.
Since MPEG-1 can't store data like this, one of the two fields will have to be discarded before capturing. This means you'll lose half of the temporal information automatically. This will leave anything originally shot of film looking jerky on playback, and anything shot on video less "present."
Good PVR systems use MPEG-2, which can store fields. There are good MPEG-2 hardware cards for Mac, even, that they could use instead. Heck, a Dual G4 can encode MPEG-2 in software in significantly faster than real time with the DVD Studio Pro Codec.
.divx is not the same as.mp4..divx is an AVI file with MPEG-4 video and (generally MP3) audio. A real MP4 file would use AAC audio, and be in the MPEG-4 file format.
Of course, QuickTime knows how to play MPEG-4 video, MP3 audio, and read AVI files, so it seems like it would be a trivial task for Apple to have QuickTime be able to play.divx files.
Hopefully DivxNetworks will adopt.mp4 as their file format in the future. AVI is really not a good format (hence all the sync problems non-Windows Divx players have), and AAC is a lot better than MP3 and AC-3.
You do what you always do: read the fine print about how the study was run, and what it measured.
Apple lobbied Neilson because their methadology was really, really, bad, and way overcounted Real and way undercounted QuickTime. The linked article shares some details about what went wrong before, and what they changed.
Draw your own conclusions from that about how the final numbers are relevant to what you're trying to do.
Honestly, it's hard to say what decisions these kinds of studies help us understand. They tell us how many individual users watched what file types. That doesn't really tell us all that much.
First off, Apple isn't trying to manipulate statistics - they were badly, badly undercounted before. In one study, only application launches were counted for playback, so any QuickTime viewed in-line inside a web browser didn't count. Since QuickTime by default works in-browser, while WM and Real go to the player by default, this erroneously hurt QuickTIme's ranking.
Kudos to Neilson for recognizing their bad statistics, and trying to correct them.
As for your other points, have you tried a recent version of QT Player for WIndows? While QT4 had an ugly UI, QT5 is quite nice, and QT6 is better yet. Also, QuickTime has a smaller installer download than either WM or Real now, when you run the web installer. How's that bloated?
There is plenty of optimization for Intel processors in QuickTime. Sorenson Video 3, the main video codec, had a lot of hand tuning in it for Intel (bear in mind Apple didn't create it).
It sounds to me like you haven't installed QuickTime since 2000 or so. Give it a whirl today and see if you like it better.
Slashdot Mirror
This really, really depends on the kind of compression you're using. VCD and SVCD use MPEG-1 and MPEG-2, which are very old technology at this point. MPEG-4 advanced simple, which is used in Divx, can provide similar quality at a half or less the data rate. And the forthcoming MPEG-4 Part 10 codec (aka H.264) will be able to do high quality standard definition (720x480) at slightly less than VCD data rates, and entertainment (movies, tv shows) at about 60% VCD data rates.
So, that'll enable putting an entire feature-length film or two hour special on a single CD-RW.
Of course, this is perfectly possible today with propritary technologies like Windows Media. The great thing about MPEG-4 is it is an international standard. And the baseline profile of Part 10 is supposed to be patent and license free, so hopefully we won't see the same confusion and delays that we did for MPEG-4 adoption.
VHS has a ton of analog noise. This means that you'll need to encode the digital copy at a substantially higher data rate to get the same effective quality, and you'll have a pretty low ceiling on maximum quality.
The difference between even S-VHS and VHS is huge.
So, grabbing off DVD or straight from a high-bitrate PVR would be quite a bit better. And if you have to go through analog, make sure you're capturing via S-Video instead of composite. Otherwise areas of saturated color will get that annoying cross-hatching effect. It's isn't so noticible on TV, but man is it obvious on a computer monitor!
I'm surprised you can even get 60 minutes on a disc
SVCD NTSC is 480x480, which is nearly three times the resolution of VCD (352x240). MPEG-2 is somewhat more efficient than MPEG-1, and you can use VBR to reduce the data rate for easy portions of the video. But still, when you're using half as many bits per pixel, MPEG-2 is going to start falling apart pretty quickly.
Still, there are lots of good new codecs on the horizon. The Advanced Simple MPEG-4 used in Divx should be supplanted over the next year by MPEG-4 Part 10 (aka H.264). It'll be an open ISO standard, and the baseline profile shouldn't (cross fingers) have any of the patent and licensing issues that held up MPEG-4 Part 2's codecs so long.
You should be able to get at least a 3x improvement in compression efficiency compared to MPEG-2, so a full-length feature at standard definition on a 700 MB disc should be quite doable.
Of course, you can do this with propritary codecs like Windows Media Video today, but this IS slashdot...
analog video "lines" and lines of digital resolution are two different things, alas.
When someone talks about lines of VHS, they're talking about how many discreet changes in amplitude you'll be able to measure on a horizontal line. And when digital video talks about lines, it's normally how many pixels high the video is.
And due to Nyquist, we know the sampling rate required to record a given frequncy is twice the frequency.
So, MPEG-1 NTSC VCD at 352 pixels wide could reproduce a frequency of 176 changes over the horizontal width of the video. So, if VHS is 250 lines, it's actually better on that measure than VCD.
Of course, VHS is plauged by horrible analog noise to the point where I can't watch it, while VCD, although low resolution, shouldn't have any noise at all. There will be some artifacts at VCD data rates, of course.
Good point.
Even if the entire increase in autism was caused by vaccinations, they'd still be worth it. The infant and childhood death rate was many, many times higher in years past without vaccinations.
And it's not like they're a new technology. Benjamin Franklin wrote an essay on why he wished he had given his son the smallbox vaccination.
Well, here's the test then. Do those with autism go on to have more children or grandchilden than average?
If not, autism in itself is evolutionarily negative. However, it is possible that a milder expression of the same genes does have advantages, ala tech centers. Think sickle-cell anemia, where one gene gives you malaria resistantance, but two make you very sick.
Remeber, a human is a gametes way of making more gametes. Cool abilities don't count in evolution if they don't lead to more grandchildren.
But if AMD has lower margin per part, then they'll need to seel more total units to equal Intel's R&D. They'll need to gain absolute, not just relative, market share to really get a long-term leg up.
Oh, I have no doubt that Hammer could beat Itanium in the next generation. However, if IA-64 is a 20-year play, it doesn't need to win in the next generation (which is good for it, since it is unlikely to win in the next generation).
But, assuming Intel did their work correctly, IA64 will have long-term advantages over extensions of IA32.
Bear in mind that the orginal 8086 was for a while supplanted by the 8088 for compatibility reasons, but the 16-bit architecture won in the end.
Given that Intel plans a 20 year life for the IA64 technology, they're going to go through a number of business cycles. The way to make money during the boom is to have built good products during the preceeding bust, and have them ready to sell once there is a market for them. A poor economy can gut AMD's budget just as much as Intel's, actually improving IA64's long term prospects.
This current bust is mainly just a post-bubble bust, just like "The New Economy" was mainly just a bubble. Companies will eventually start spending again, and eventually they'll even start overspending again, and then cut projects, rinse repeat.
The compression rate isn't the key. Saying 4:1 is better than 12:1 is like saying that IMA audio compression is better than AAC, which it manifestly isn't.
While a case can be made that DTS can be better than AC-3 in some cases (although way too many of the cited comparisons looked at 384 AC-3, where 448 is now standard for high-end discs), the fact that it takes up so much more space on the disc means that in many cases the video quality will need to be degraded to compensate.
While DTS might have a slight edge in terms of maximum quality, AC-3 has a huge advantage in terms of compression efficiency, which means it wins is places where file size counts significantly.
And while it might be possible to have both parties log into an ISP that supports V.92, this would hurt more than it helps. Yes, V.92 would increase bit rate somewhat (max of 48 instead of 34), but having to log in through an ISP would also increase latency a painful amount.
Communicating point-to-point helps the latency more than the lower data rate hurts the signal.
At least that's my guess of why they went that way.
If the Red HD system used MPEG-2, maybe. But since they'll use either MPEG-4 or a Windows Media derivative, they need a LOT fewer bits per pixel to delivery.
There is also a lot of cruft on current DVD's as well. For a feature length film, you'd use a dual-layer DVD-9, not offer different aspect ratios, and probably would dump the vastly overrated DTS. Probably would use AC-3 and/or AAC. You'd be able to sustain 8 Mbps average this way. And heck, there are only 160% more bits/frame with 1280x720 than 720x480, so we only need a 50% improvement in compression efficiency to be able to pull this off. MPEG-4 Advanced Simple provides this easily.
Except that Blu-ray couldn't possibly be mass market by Christmas 2003. The nice thing about a red laser system is that the physical medium of the disc doesn't have the change, which means the hardware in existing DVD players can be mostly the same, with just a different decoder chip. Fast computers will just need a software update. And, of course, replication and duplication facilities won't need to chance, so it'll cost well less than $1 to make an HD disc, which means we could start seeing mass market prices very quickly.
This is really good from the Hollywood perspective. They'll get us all to buy 1280x720 red laser HD discs from 2003-2006, and then come out with 1920x1080 Blu-ray as a mass market technology around Christmas 2006-2008, when they get all the kinks worked out. Same way we've already bought DVD and laserdisc versions of the same movie.
The article claims that the compression technology will be from Microsoft, but my contacts tell me it is much more likely to be MPEG-4, in order to have a technology not tied to any one vendor. Of course, Windows Media derived codecs would offer better compression efficiency. We shall see.
Well, from a legal perspective, you want to make BOTH crimes. Already anti-spam pressure has kept most legitimate businesses away from doing it.
In essence, we want to make it as socially and legally difficult to admit to being in any way distributing or profiting from spam as, say being in any way distributing or profiting from stolen car parts.
Indeed. I wasn't critiquing your study, but pointing out that some folks were making the wrong interpretation of its subject.
This is an interesting and relatively well done test (although it appears that the listeners knew which format they were listening to, so it wasn't truly double-blind, and a anti-MS and pro-Ogg bias can't be ruled out).
However, some discussions seem to be focusing on this saying AAC is bad or WMA is bad, when really it refers to the particular implementations in codecs of those formats.
For example, the Apple MPEG-4 AAC-LC encoder was used for AAC. This is a Low Complexity version of the format. Also, the Apple encoder has a strange limitation where it automatically converts 44.1 stereo to 32 stereo at that data rate. This isn't required by the AAC format. Other AAC encoders yield MUCH better results, and beat MP3 Pro in double-blind testing. I haven't seen any double-blind comparisons between AAC and Ogg.
Also, the WMA8 encoder is due to be replaced by the backwards-compatible WMA9 in early September. Of course, there may well be improved versions of the other encoders by then as well.
Actually, that's not the case either. MPEG audio support is optional for DVD supports. Only AC-3 and PCM are mandated.
MPEG audio was in the draft spec for PAL DVD, but was dropped for the final version. It was never in NTSC.
Because the holes are round.
I mean, really, any other shape wouldn't fit...
*ducking*
They aren't releasing any codecs. Streaming servers don't parse the underlying video data, they just (well, that's a misuse of the word "just" - it's hard to do right) turn the file into packets, and try to make sure that as many arrive to the client on time as possible.
For example, today's RealServer can stream QuickTime files, and they certainly don't have source to the Sorenson codecs.
One great thing about VCD is that they work in standard DVD players (NTSC and PAL VCD discs only work in players for the respective formats, of course; a limitation not seen on computers). However, lots of older DVD players can't read CD-R media, because of an incompatibiliity with the DVD laser. Duplicated VCDs on normal CD media work just fine, of course. All DVD players in the market today should work, AFAIK.
If MPEG-4 goes the way of MP3, it'll be hugely popular, widely used, and the foundation for a new industry!
The final MPEG-4 licensing terms were announced. For content providers, you have to pay $0.02 per HOUR (not minute) or $0.25 per subscriber, after the first 50,000 subscribers. So if you have fewer than 50,000 people looking at your content, you don't owe anything.
Apple has to pay a couple $1M for the encoders and decoders, but that's their problem.
Alas, MPEG-1`is a lousy format for PVR use of any quality. MPEG-2 only stores video as progressive scan. However, TV is broadcast as interlaced, where the even lines are captured 1/59.94th of a second off from the odd lines. The difference between fields includes half of the video compression information.
Since MPEG-1 can't store data like this, one of the two fields will have to be discarded before capturing. This means you'll lose half of the temporal information automatically. This will leave anything originally shot of film looking jerky on playback, and anything shot on video less "present."
Good PVR systems use MPEG-2, which can store fields. There are good MPEG-2 hardware cards for Mac, even, that they could use instead. Heck, a Dual G4 can encode MPEG-2 in software in significantly faster than real time with the DVD Studio Pro Codec.
.divx is not the same as .mp4. .divx is an AVI file with MPEG-4 video and (generally MP3) audio. A real MP4 file would use AAC audio, and be in the MPEG-4 file format.
.divx files.
.mp4 as their file format in the future. AVI is really not a good format (hence all the sync problems non-Windows Divx players have), and AAC is a lot better than MP3 and AC-3.
Of course, QuickTime knows how to play MPEG-4 video, MP3 audio, and read AVI files, so it seems like it would be a trivial task for Apple to have QuickTime be able to play
Hopefully DivxNetworks will adopt
You do what you always do: read the fine print about how the study was run, and what it measured.
Apple lobbied Neilson because their methadology was really, really, bad, and way overcounted Real and way undercounted QuickTime. The linked article shares some details about what went wrong before, and what they changed.
Draw your own conclusions from that about how the final numbers are relevant to what you're trying to do.
Honestly, it's hard to say what decisions these kinds of studies help us understand. They tell us how many individual users watched what file types. That doesn't really tell us all that much.
First off, Apple isn't trying to manipulate statistics - they were badly, badly undercounted before. In one study, only application launches were counted for playback, so any QuickTime viewed in-line inside a web browser didn't count. Since QuickTime by default works in-browser, while WM and Real go to the player by default, this erroneously hurt QuickTIme's ranking.
Kudos to Neilson for recognizing their bad statistics, and trying to correct them.
As for your other points, have you tried a recent version of QT Player for WIndows? While QT4 had an ugly UI, QT5 is quite nice, and QT6 is better yet. Also, QuickTime has a smaller installer download than either WM or Real now, when you run the web installer. How's that bloated?
There is plenty of optimization for Intel processors in QuickTime. Sorenson Video 3, the main video codec, had a lot of hand tuning in it for Intel (bear in mind Apple didn't create it).
It sounds to me like you haven't installed QuickTime since 2000 or so. Give it a whirl today and see if you like it better.