Actually, on the MPEG 1 front, mp1e has been around for quite a while.
It does real time video/audio capture direct to MPEG 1, at any bitrate, with decent audio sync and *superb* picture quality at higher bitrates.
*And* it doesn't require major hardware - I use a K6/2 475 and it rarely jumps above 20% CPU while capturing.
It is, however, rather difficult to find. (Check the V4L mailing lists).
I'm *no* coding wizard (not by a long stretch), but looking at the source this guy has done some amazing optimization (and large chunks of inline MMX assembler).
Here are a couple of links to try:
(I can't comment on any of them except PCB, which, although very simple, works fairly well and is quite intuitive to use). MUCS-PCB PCB SATCAD
Also, *most* publishing service bureaus support material brought in on Zip disks. A while back I read a survey (I'll see if I can find the source) taken of service bureaus in the Bay area (IIRC), and a large chunk of their work came in on Zip disks.
Generally any card with a BrookTree BT8x8 (BT 848 or BT878) chip should work fine.
There are varying levels of support for the onboard audio outputs and mixers, so YMMV in that respect.
I use an old IX Micro IXTV (BT848), and it works *beautifully* (except for onboard audio), which isn't a problem 'cause I run an S-Video feed from my Digital satellite receiver, and a separate audio feed into the soundcard.
Bear one thing in mid though, that the reception will *not* come close to that of a good TV, (at least in every TV card I've seen), but that's a tuner issue, not a picture quality issue.
Actually most DVD players will play VCDs, but *not* all DVD players will handle CD-R media.
Some DVD players are designed with a dual laser pickup to handle both, as the DVD laser is red (around 650 nm) and a CD-R laser needs to be near IR (750 ?? nm). CD-RWs are apparently more reflective in the 650nm region, so if your DVD player is not specced to handle CD-Rs, it might very well work with CD-RWs.
Of course it helps to have one that simply states on the box 'designed for CD-R and CD-RW';-)
Yeah, it then comes down to everyone whining about the limited hardware support, and things like 'why is my Floogledorfer mark 21 asynchronous carrier pigeon network card not supported?'
This is why NeXT (and Sun, for Solaris x86) had a booklet outlining supported hardware configurations, so that when you wanted to run the OS on an intel box, you consulted the docs, and went out and *built* a compliant machine.
They also had some farily weird ideas about what *mainstream* hardware was, I mean, no NE2000 compatible driver for NeXTstep ? (not officially, anyway).
The whole 'supported hardware' issue could have quite an impact on market penetration.
Re:It might be interesting-
on
OS X on x86?
·
· Score: 1
Touche'!
The apps were provided in the form of a 'fat binary'. The installer can strip the binary of code for unneccesary platforms, if needed.
Stay away from DAVE! It is *not* very well behaved, sometimes.
I'm also interested in alternatives, because I will shortly be implementing a system for batch image processing in a production environment, and it will need to talk to a whole bunch of Macs, and one Windoz PC. I haven't looked too closely at the file locking side of things, because I've been concentrating on network performance (throughput)
Interesting thought on the above - Appleshare over IP - throughput is faster, but is glacially slow in listing directory contents, yet over Appletalk directory listing is an order of magnitude faster - go figure!
What I thought amazing was the guy who said "Uh.. there appears to have been a malfunction", or something to that effect.
I don't know if he was what they (in the Apollo program, at least) called the "CapCom", but that was an amazing sign of professionalism and coolness under *very* adverse conditions - a sign of some *very* good selection and training.
The people & politics behind it...
on
The Challenger
·
· Score: 2
An interesting overview of what happened before and after the disaster inside NASA, look for Richard Feynman's book "What do you care what othe people think?" (ISBN 0-553-34784-5). At least half of the book is devoted to Feynman's participation in the board of enquiry, and to read it from his irreverent "march-to-a-different-drummer" point of view is quite enlightening.
This guy has a similar idea. I wonder who got round to filing patent first for the 'paper computer'?
The circuit technology sounds similar, except the phone people seem to go for a plastic substrate (going by the photo of their 'working model').
Would it not be possible to write each MP3 to a CD as a mode 2(?) track, similar to the way in which the Video CD (White Book) standard is implemented?
The player would then only need to read a raw bitstream off the disk starting at a given offset (track) position.
Forgive me if this is totally ridiculous, but I've been playing around with White Book stuff for a while, and this seems possible (plausible? dumb? plain stoopid?).
Well, I'm sitting here looking at a brand-new CD of 3.3 I got from Apple last(?) year as a Y2K upgrade.
IIRC it ran OK (bit sluggish) on a 486/66 w/ 32M RAM. According to the docs it would also run (at reduced color depth, maybe mono) in 8M.
Just for kicks I tried it on an AMD 500, and it smokes! Starting to look a bit dated, though.
For NeXTstep software, OS disks, and hardware, check out:
BlackHole Incorporated
Actually, an even *better* choice would be one of the real-time OSes designed for this sort of thing.
VxWorks (IIRC) comes to mind, theirs is the first OS to land on Mars. Pathfinder == VxWorks on rad-hardened PPC.
See Here
As for the 'golf ball' thing - IIRC the 'DEW' line installations had a similar setup?
Was it the DEW line that also ran across Greenland(?) - 'cause I've seen pics of an abandoned station out there with such a dome.
Uh.. a Beowulf cluster of these (or any similar antiques!) would probably bring down the power grid in the entire state!
Imagine the scene at the power company -
"Oh *shit* the computer museum's doing a demo again - quick! bring Three Mile Island back online - NOW!"
"...but wouldn't it be fun to ultimately travel from Tierra del Fuego to Johannesburg by train?"
Uh.. I think not! -- with the current state of railways here in (Southern) Africa, you might get as far as, say, Nairobi !
On a *vaguely* related note - what about the more popular MOOS? like LambdaMOO or BayMOO?
Are they still around, and as popular as (I thought) they were a few years back?
IIRC this was proposed quite some time ago by Rod Brooks and co. at the MIT AI lab. One of the papers was entitled "Fast, Cheap, and out of control - a robotic invasion of the Solar System".
(Co-authored I think, by Anita Flynn).
Makes for some interesting reading, along with some other papers by Dr Brooks.
Try this (the link seems to be down at the moment) Rodney Brooks' papers
Slashdot Mirror
Actually, on the MPEG 1 front, mp1e has been around for quite a while.
It does real time video/audio capture direct to MPEG 1, at any bitrate, with decent audio sync
and *superb* picture quality at higher bitrates.
*And* it doesn't require major hardware - I use a K6/2 475 and it rarely jumps above 20% CPU while capturing.
It is, however, rather difficult to find. (Check the V4L mailing lists).
I'm *no* coding wizard (not by a long stretch), but looking at the source this guy has done some amazing optimization (and large chunks of inline MMX assembler).
Here are a couple of links to try:
(I can't comment on any of them except PCB, which, although very simple, works fairly well and is quite intuitive to use).
MUCS-PCB
PCB
SATCAD
Also try Scientific Applications on Linux
Also, *most* publishing service bureaus support material brought in on Zip disks.
A while back I read a survey (I'll see if I can find the source) taken of service bureaus in the Bay area (IIRC), and a large chunk of their work came in on Zip disks.
Generally any card with a BrookTree BT8x8 (BT 848 or BT878) chip should work fine.
There are varying levels of support for the onboard audio outputs and mixers, so YMMV in that respect.
I use an old IX Micro IXTV (BT848), and it works *beautifully* (except for onboard audio), which isn't a problem 'cause I run an S-Video feed from my Digital satellite receiver, and a separate audio feed into the soundcard.
Bear one thing in mid though, that the reception will *not* come close to that of a good TV, (at least in every TV card I've seen), but that's a tuner issue, not a picture quality issue.
Actually most DVD players will play VCDs, but *not* all DVD players will handle CD-R media. ;-)
Some DVD players are designed with a dual laser pickup to handle both, as the DVD laser is red (around 650 nm) and a CD-R laser needs to be near IR (750 ?? nm). CD-RWs are apparently more reflective in the 650nm region, so if your DVD player is not specced to handle CD-Rs, it might very well work with CD-RWs.
Of course it helps to have one that simply states on the box 'designed for CD-R and CD-RW'
Yeah, it then comes down to everyone whining about the limited hardware support, and things like 'why is my Floogledorfer mark 21 asynchronous carrier pigeon network card not supported?'
This is why NeXT (and Sun, for Solaris x86) had a booklet outlining supported hardware configurations, so that when you wanted to run the OS on an intel box, you consulted the docs, and went out and *built* a compliant machine.
They also had some farily weird ideas about what *mainstream* hardware was, I mean, no NE2000 compatible driver for NeXTstep ? (not officially, anyway). The whole 'supported hardware' issue could have quite an impact on market penetration.
Touche'! The apps were provided in the form of a 'fat binary'. The installer can strip the binary of code for unneccesary platforms, if needed.
Stay away from DAVE! It is *not* very well behaved, sometimes.
I'm also interested in alternatives, because I will shortly be implementing a system for batch image processing in a production environment, and it will need to talk to a whole bunch of Macs, and one Windoz PC.
I haven't looked too closely at the file locking side of things, because I've been concentrating on network performance (throughput)
Interesting thought on the above - Appleshare over IP - throughput is faster, but is glacially slow in listing directory contents, yet over Appletalk directory listing is an order of magnitude faster - go figure!
What I thought amazing was the guy who said "Uh.. there appears to have been a malfunction", or something to that effect.
I don't know if he was what they (in the Apollo program, at least) called the "CapCom", but that was an amazing sign of professionalism and coolness under *very* adverse conditions - a sign of some *very* good selection and training.
An interesting overview of what happened before and after the disaster inside NASA, look for Richard Feynman's book "What do you care what othe people think?" (ISBN 0-553-34784-5). At least half of the book is devoted to Feynman's participation in the board of enquiry, and to read it from his irreverent "march-to-a-different-drummer" point of view is quite enlightening.
This guy has a similar idea.
I wonder who got round to filing patent first for the 'paper computer'?
The circuit technology sounds similar, except the phone people seem to go for a plastic substrate (going by the photo of their 'working model').
Would it not be possible to write each MP3 to a CD as a mode 2(?) track, similar to the way in which the Video CD (White Book) standard is implemented?
The player would then only need to read a raw bitstream off the disk starting at a given offset (track) position.
Forgive me if this is totally ridiculous, but I've been playing around with White Book stuff for a while, and this seems possible (plausible? dumb? plain stoopid?).
Well, I'm sitting here looking at a brand-new CD of 3.3 I got from Apple last(?) year as a Y2K upgrade.
IIRC it ran OK (bit sluggish) on a 486/66 w/ 32M RAM. According to the docs it would also run (at reduced color depth, maybe mono) in 8M.
Just for kicks I tried it on an AMD 500, and it smokes! Starting to look a bit dated, though.
For NeXTstep software, OS disks, and hardware, check out: BlackHole Incorporated
For those who haven't yet seen it (and some others), there's a trailer at:
This place
For those outside the USA, check out the LG 2300 - it'll play *anything* you can throw at it in the media department, except MP3s.
There is (theoretically, at least), already the capability to do some of this stuff, as outlined in
RFC1712-DNS Encoding of Geographical Location
and
RFC1876-A Means for Expressing Location Information in the Domain Name System
Actually, an even *better* choice would be one of the real-time OSes designed for this sort of thing. VxWorks (IIRC) comes to mind, theirs is the first OS to land on Mars.
Pathfinder == VxWorks on rad-hardened PPC.
See Here
As for the 'golf ball' thing - IIRC the 'DEW' line installations had a similar setup? Was it the DEW line that also ran across Greenland(?) - 'cause I've seen pics of an abandoned station out there with such a dome.
Uh.. a Beowulf cluster of these (or any similar antiques!) would probably bring down the power grid in the entire state!
Imagine the scene at the power company -
"Oh *shit* the computer museum's doing a demo again - quick! bring Three Mile Island back online - NOW!"
"...but wouldn't it be fun to ultimately travel from Tierra del Fuego to Johannesburg by train?" Uh.. I think not! -- with the current state of railways here in (Southern) Africa, you might get as far as, say, Nairobi !
On a *vaguely* related note - what about the more popular MOOS? like LambdaMOO or BayMOO? Are they still around, and as popular as (I thought) they were a few years back?
IIRC this was proposed quite some time ago by Rod Brooks and co. at the MIT AI lab. One of the papers was entitled "Fast, Cheap, and out of control - a robotic invasion of the Solar System". (Co-authored I think, by Anita Flynn). Makes for some interesting reading, along with some other papers by Dr Brooks.
Try this (the link seems to be down at the moment)
Rodney Brooks' papers
Also on 'robot collaboration'
Robotic Ants