Currently the largest LED is Luxeon's 5 watt emitter, which gives off around 100-120 lumens. No one has been able to manufacture anything larger easily yet. There's a rumored 10W LED coming from another company, although its light output is specced as LESS than Lumileds' Luxeons (i.e. it's less than half as efficient.)
5W units are $40/each in small quantities, dropping to $24 or so each in larger quantities (100+) They are VERY hard to obtain.
The way fluorescent lights work is that the gas in the tube emits UV light, which is absorbed and re-emitted as white by the phosphors in the light. A decent amount of this UV leaks through...
White LEDs are similar, except that they use blue light (much safer) to charge up the phosphor.
You don't mention that the EB lights happen to also use 3 N cells, which are much more expensive and harder to find than AAs. None of the LedLensers (The EB lights are rebadged "LEDLenser" lights from a German company named Zweibruder - I think their website is www.zweibruder.de) have DC/DC upconverters, so if they run on AAs or AAAs, they need 3 cells, and for 2AA form factor they must use N cells.
If you want BRIGHT LED flashlights, look around on http://www.candlepowerforums.com/ - The creations of forum users ElektroLumens, Lambda, McGizmo, dat2zip, and Mr. Bulk are simply incredible. dat2zip sells an excellent drop-in for MiniMags that includes a DC/DC converter circuit and a 1 watt Luxeon LED. (Equivalent to 10-20 of Nichia's best 5mm units.)
You might want to try using more recent drivers (Such as the most recent Detonators plus modified.INF hack - Best place to find info is the community forums on Dell's support site.) - They might be more predictable.
I've never been very happy with the state of multiple monitor support under Windows. It seems very unpredictable.
Check out www.avsforum.com - The WinTV-D/HD cards suck. (Design-wise, not so bad, but Hauppauge's software support is HORRENDOUS.)
Most popular cards out now are the Telemann HiPix and the MIT (not the educational institution in MA) MyHD. The MyHD is the cheapest, at $300. (Others are $400). Another Korean company, DVICO, is releasing an HDTV tuner card for approx. $160 that depends on software for the MPEG decoding. (The other cards have hardware MPEG-2 MP@HL compliant decoders, which is a significant portion of the cost. Cheapo decoders such as most RealMagic ones only support MP@ML. Of course the hardware cards have been used successfully in a 233 MHz Pentium while the SW card needs approx. a P3-1 GHz)
The Alienware systems use desktop CPUs, not mobile CPUs. No SpeedStep and less efficient GHz for GHz (usually the mobile chips have more recent manufacturing process technology than the equivalent desktop version)
And go update your drivers. You're obviously NOT running the latest ones.
My dad had bluescreen problems with his I8000 and obtained new drivers from *Windows Update* of all places. This was months ago.
I was using the latest Dell drivers from their website with no problems whatsoever. I don't think my 8200 has ever bluescreened even once.
And if you bothered to do ANY research at all, you would've found the D-Force (and related) modified INFs that are regularly maintained so that you can use your latest Detonator release with "Go" series of GeForces. Yes, I'm running the 41.09 Detonator release on my 8200 with full functionality.
BTW, Dell has some excellent user-to-user support forums if you go to their support website.
Some of the business communications vendors (Such as Avaya) have microcellular systems designed for in-building and short-range communications using the same ideas as cellular. Some microcell phones are designed to automatically hand off to a proper cellular network (At least in theory - I've never seen such a system actually installed and working.)
If you want longer range - Find another solution. Your GSM phone won't cut it. If you're in the USA, get a real provider. The GSM operators in the US have HORRENDOUS coverage. One other poster suggested Verizon, I would too. Verizon's network and AT&T's old pre-GSM network are the only ones in the U.S. that can claim options of true nationwide coverage. (And at this point I don't think you can sign up for AT&T's "old" service, leaving VZW as the only game in town.) VZW's 100% coverage plan is somewhat more expensive than their America's Choice plans, which still cover 60-70% of the landmass of the US, significantly more than any other provider can claim for non-roaming service.
Yes, Verizon's plans are significantly more expensive as far as minutes per dollar, but you get what you pay for. Verizon is #1 in coverage, one of the top providers as far as customer service, and one of the top if not #1 in quality control of both their network AND the phones they release. (This is why the selection of phones available on Verizon is usually a bit more limited/behind other CDMA providers - It takes a LOT longer for a phone to pass Verizon QA testing if it passes at all. Nokia hasn't been able to even get a phone through the process successfully since the 5185. They make good GSM phones but are not nearly as competent when it comes to CDMA.)
Another option for you is two-way business-band VHF or UHF NBFM. This will require an FCC license in most cases, but one that's a hell of a lot easier to obtain than a PCS spectrum license. There are a few unlicensed frequencies, max range on those is 5-10 miles.
I believe all CF implementations are at least 10,000 to 100,000 write cycles minimum. (I'm 99% positive it's at the 100,000+ range at this point.)
I don't know of any flash memories limited to 1,000 write cycles at this point except for the program flash of Atmel AVR processors, but the only time such a unit should ever exceed (or even come close to) 1,000 flashes is during the software development cycle. (I think the average life of an AVR in Cornell's EE 476 lab is 2-4 weeks, due to the fact that it gets flashed every 2-3 minutes or more for 4-5 hours/day. In a production system the program flash should never need to be altered and in fact CAN'T be altered without external programming hardware in most AVR models.)
There are some terms you're obviously unfamiliar with:
Mass production/volume economics Supply and demand
Oh, and recordable DVDs can be found for as low as $1 each if you search. Hell, even at Best Buy recordables are only $3.50 each, not $5.
Now, if a company can sell a recordable DVD that they produce for 10 cents for $5 or sell it for $1, which do you think they're going to charge? $5 - More profit for them. They'll charge as much as they can get away with charging. Also, because recordable CDs have been around much longer than recordable DVDs and CD-R drives are much more common, CD-Rs are produced in much larger quantities and hence cheaper to produce.
Also, you seem to miss that I mentioned that for prepressed DVDs, they are the same materials but the manufacturing process must be more precise (More data bits per unit of surface area). Same goes for recordable DVDs - Manufacturing tolerances become more and more critical as data density increases, so even if they were made from the same materials (which they aren't for recordables, because of the different laser color different dyes are needed, ones which might be more expensive to obtain/produce.)
Now, if you use a blue laser at DVD densities, it'll work, and you only have to worry about the manufacturing tolerances of a DVD, but: a) Blue lasers are EXPENSIVE b) You still have to pay for the cost of whatever new dye is used for recording with a blue laser.
In other words, if you apply this technology to lower data-density mediums, you will increase costs with no benefit.
You can't go the other way around - Using an IR laser but increasing precision and manufacturing tolerances won't work because physics dictates a minimum feature size that can be used with IR.
Those time estimates are pretty far off, although it does depend on a few things. Even at a given resolution, you have some freedom to adjust video bitrate.
That said - 23 GB can hold 2.5 hours or so of 1080i video at the bitrate used for ATSC (USA HDTV) programming. An ATSC transport stream is 19.2 Mbps, or approx. 8.5 gigabytes/hour. 1080i video usually takes around 85-90% of a transport stream. (It's somewhat variable - Most people see a 10-15% filesize reduction when using HDTVtoMPEG2 to strip a 1080i stream out of a transport stream.)
A higher-bitrate 1080i source might take more space, but you're not likely to see such sources in the USA.
As to why MPEG-2 - Backwards compatibility with DVDs is one reason, but more importantly, compatibility with current HDTV standards. ALL digital TV broadcast standards use MPEG-2. Why ruin the quality by recompressing with a different codec when you can simply save the stream without recompression?
Plus MPEG-2 is a lot easier to implement in dedicated hardware/is already implemented in lots of hardware. I have only seen a handful of hardware MPEG-4 decoders and zero dedicated hardware MPEG-4 encoders, while hardware MPEG-2 codecs are widely available.
As to MPEG-2 being "pure crap" - You must have been using a crappy encoder and low bitrate. MPEG-4 was designed to excel at low bitrates, while from what I've heard, it loses most of its advantages and in some cases is worse at higher bitrates. (Note that this is the case with other encoders even within the same format - For example, I believe LAME is optimized for high-bitrate MP3 while Fraunhofer usually optimizes their encoders for low-bitrate MP3. As a result the Fraunhofer codec blows away LAME below 96-128 kbps or so, while LAME wins hands-down beyond 128-160 or so.) I've noticed this limitation of DivX, which is just one implementation of MPEG-4. At lower bitrates it blows away MPEG2, but for many videos, MPEG2 blows it away at bitrates approaching that of DVDs.
Because these are neither cheap nor easy to manufacture.
Red lasers are now available in comparably high power (5 mW is a LOT compared to that used in media readers - Burners might be a different story) dirt-cheap - A few bucks for a complete pointer. Partly because the red can be directly generated.
As far as I know, almost all (if not all) green semiconductor lasers are NOT truly a green laser. They are an IR-pumped IR laser combined with nonlinear optics that double the frequency. I believe it's something like: 800something nm IR pumps a crystal that lases at 1000something-nm IR, which is then frequency doubled to green visible light. I don't remember the exact wavelengths, but that's the general idea behind how DPSS green lasers work. Not cheap or easy to manufacture, or easy to put into a small optical drive.
In case you haven't noticed, DVDs are just an evolution of the CD. Same for this technology.
This won't allow for 5x the recording on the exact same media - It'll require new recordable media, or in the case of prepressed discs, a new manufacturing process that can make smaller pits.
It's the exact same evolution from DVD->Blu-Ray as CD->DVD
Name: CD->DVD->Blu-Ray Laser color: IR->Red->Blue (The main difference) Media size: approx 5.25" - Same size for all three The only difference in the media is that the dyes used for recordable media must be different depending on the laser wavelength used. Prepressed media will likely be made from the exact same materials, just with greater precision.
In short: If you apply this technology to CDs, you will get the exact same end product. The laser color is the primary differentiator between the three formats. (And of course, blue-laser devices need higher-precision motors/optics.)
Don't forget that HD programming will look pretty good on any decent monitor. (i.e. any CRT capable of 1280x1024 or higher) - It won't be huge, but the resolution is what really matters. You can get HDTV tuning capability for $300 if you already have a decent display for your PC.
And you severely underestimate the number of HD stations in the USA. Something makes me think that you live in the NYC metro area or are severely misinformed. In most places, a significant fraction of stations have HD transmitters online. (3/4 of the Philly stations have been broadcasting HDTV since 1998, and at this point all of the Philly network affiliates are broadcasting HD. This is the same throughout most of the country. The biggest exception is NYC - It used to have excellent HD availability, but unfortunately it's been a challenge for broadcasters to even get their analog transmitters back on the air after 9/11 - Every NYC station that wasn't based off of the ESB to begin with is operating at severely reduced power right now. The only HD transmitters in NYC are the ones that were not located on the WTC to begin with. Thank God the Bayonne tower has finally been approved for construction...)
http://www.antennaweb.org/ and http://www.100000watts.com/ have great listings of HDTV availability.
The 4-pin connector is as much of a standard across all names for 1394 as the 6-pin connector is. There's nothing special about it. The 4-pin connector is just a smaller one for devices that don't need bus power or for which bus power is insufficient or can't easily be provided. (For example, every camcorder I've seen, regardless of brand, had a 4-pin 1394 connector. The 1394 connector on Dell Inspiron 8x00 laptops is also 4-pin.)
This will most likely work with no problems at all in the following configuration:
PCI HDTV tuner for OTA signals PC with Firewire port this drive DVHStool
A similar setup is already used to record HDTV streams to D-VHS decks (Mainly the JVC HM-D30000U, I think there are 1-2 other DVHS decks now but I'm not sure.)
Note on DVHS: For archival of MPEG-2 Transport Streams (as used for HDTV), 40-50 gigs per unit of cheap media (high-grade VHS tape, I've seen ones specifically marketed as DVHS tapes for $8-9 each, and regular VHS tapes will probably work.) is here now. DVHS decks run around $1000. Unfortunately, the HM-D30000U has a not-so-stellar reliability rep - The decks break a lot.:(
Unfortunately no PC-based HDTV tuner can tune in cable HD signals - Cable companies use 64QAM or 256QAM modulation, while OTA HDTV uses 8VSB modulation. SOME cable companies will rebroadcast locals in 8VSB format, but most demod the 8VSB stream and remodulate it as QAM. There are no QAM-capable PC tuners right now. Hauppauge claimed it for the WinTV-HD at initial release, but it never happened and they no longer make such claims. A Korean company (DVICO) is coming out with an 8VSB tuner that relies on software decoding of the stream soon, and the next version they're releasing will include QAM capability. (Will not likely work for encrypted digital cable, but I believe there's a law that states that cable companies may not encrypt local station feeds.)
I have a PCI HDTV tuner - Once the price of this comes down I'm really looking forward to it. Yes, it'll be too small for a full hard drive backup, but it will ease the strain on my hard drives because it will be practical for near-direct archival of HDTV recordings. After commerical editing, I could easily fit three one-hour TV shows onto one disc. Right now a one-hour show won't even fit onto a DVD after commercials are edited out.
YES. This is definite... A bad sample rate conversion can REALLY kill your music.
Case in point: esd (aka eSound), a soundserver common on Linux boxes. esd SUCKS if it needs to do samplerate conversion. 48 kHz MP3s sound HORRIBLE on any system I've used if played via esd.
Same goes for trying to play a 44.1 kHz file at 48 kHz.
I think sox has a decent converter, so do most sound editing programs. Heck, depending on your school you might even be able to use Matlab. (It's not intuitive by any means, nor is it fast, but you'll learn a LOT about audio processing in the process.)
And yes, I've processed audio files in Matlab, mainly at a time when I was playing with some DSP tricks for amateur radio use.
Yup. When dealing with distortion, you definately have to increase your bandwidth, in order to accurately measure this distortion. 24/96 would be a minimum for accurately modeling the distortion in guitar amps, maybe even a higher sample rate. (In this application, sample rate would be significantly more important than ADC/DAC res. Probably most people wouldn't be able to tell the difference between 16 and 24 bits if the sampling rate were high enough. I routinely push the limits of a $120k spectrum analyzer with a 16-bit DAC. Of course said DAC is operating at 350+ MHz...)
I'm not talking about those cheesy "Use your VCR as backup" devices, those can only be described one way: Cheap hacks. They are limited by all the circuitry in the VCR that mangles the input signal. (The AGC circuitry, for example. Note that this is the same circuitry that allows Macrovision to work.)
I'm talking about D-VHS. (Such as the JVC HM-D30000U, and 1-2 other DVHS decks in existence.) These are designed from the ground up to record digitally, and given the bitrates specified and the recording times, it comes out to something like 40-50 gigs per DVHS tape. Not sure how well "standard" VHS tapes work, they're probably exactly the same except for a cost premium for the DHVS-specific tapes.
Current limitations: a) DVHS decks record MPEG-2 transport streams, not DV video streams. Which means dvbackup won't work. DVHStool for Windows is about halfway there - But it only will send prerecorded or premade transport streams. The VCR might be unhappy if the TS doesn't have a real MPEG stream in there, but one could embed a low-bitrate 480i stream and fill the rest of the transport stream with data. (MPEG-2 Transport Streams as used in HDTV broadcasts allow for arbitrary data. For example, a Fox station broadcasting 480p with no subchannels is padding their 19.2 MBit/sec stream quite a bit.) b) Current DVHS decks are rare and the ones that do exist don't have the best reliability record.
He didn't say that it was lossless, but it was essentially lossless.
i.e. the compressionr ratio is so low that it's uncompressed for all practical purposes.
Just like it's practically impossible to tell the difference between an SHQ 2048x1536 JPEG from my Olympus C-3000 and a TIFF from the same camera. SHQ JPEGs from that camera equate to using a quality level of 98% or 99% when saving using GIMP or ImageMagick filesize-wise. (Default seems to be 80% or 90% - Even at those quality levels it's hard to see a difference.)
24 bits @ 96 KHz is beyond the ability of the human ear to discern any differences.
In fact, with a good antialiasing filter, 16 bits @ 44.1 KHz will put you below the noise floor of all but the best amplifiers and cover the whole range of human hearing. Problem is, an antialiasing filter suitable for 16/44.1 is VERY tough to design without causing distortions in the range of human hearing. 24/96 is easy to develop a suitable antialiasing filter for.
Problems ensue when you are processing the data, though. If you process 16-bit data in the DSP with 16 bits of precision, then at every step in the processing chain you'll likely have rounding errors. Such errors accumulate.
For 16-bit data, I believe most people use DSPs with 24-bit internal precision at a minimum. For 24-bit DSPs, 32 is probably the minimum. I don't know what the likes of Line6's products use. A floating-point DSP would do VERY well for eliminating rounding errors, but those cost $$$.
Interestingly enough - These amps try to use a model of another amplifier's nonlinearities to emulate the nonlinearities of said amp. In my line of work, we do the exact opposite. (Correcting for nonlinearities in RF amplifiers to minimize distortion of any form.)
Slashdot Mirror
http://www.luxeon.com/
Currently the largest LED is Luxeon's 5 watt emitter, which gives off around 100-120 lumens. No one has been able to manufacture anything larger easily yet. There's a rumored 10W LED coming from another company, although its light output is specced as LESS than Lumileds' Luxeons (i.e. it's less than half as efficient.)
5W units are $40/each in small quantities, dropping to $24 or so each in larger quantities (100+) They are VERY hard to obtain.
The way fluorescent lights work is that the gas in the tube emits UV light, which is absorbed and re-emitted as white by the phosphors in the light. A decent amount of this UV leaks through...
White LEDs are similar, except that they use blue light (much safer) to charge up the phosphor.
You don't mention that the EB lights happen to also use 3 N cells, which are much more expensive and harder to find than AAs. None of the LedLensers (The EB lights are rebadged "LEDLenser" lights from a German company named Zweibruder - I think their website is www.zweibruder.de) have DC/DC upconverters, so if they run on AAs or AAAs, they need 3 cells, and for 2AA form factor they must use N cells.
If you want BRIGHT LED flashlights, look around on http://www.candlepowerforums.com/ - The creations of forum users ElektroLumens, Lambda, McGizmo, dat2zip, and Mr. Bulk are simply incredible. dat2zip sells an excellent drop-in for MiniMags that includes a DC/DC converter circuit and a 1 watt Luxeon LED. (Equivalent to 10-20 of Nichia's best 5mm units.)
It's an issue with anything running NVidia's binary drivers.
It's possible to get APM working with those drivers with a minor source tweak and disabling AGP.
Everything else under the laptop works flawlessly.
You might want to try using more recent drivers (Such as the most recent Detonators plus modified .INF hack - Best place to find info is the community forums on Dell's support site.) - They might be more predictable.
I've never been very happy with the state of multiple monitor support under Windows. It seems very unpredictable.
Quake 3 engine based games are known for being quite well suited to widescreen resolutions, same for UT and its descendants.
Other games - It depends.
Check out www.avsforum.com - The WinTV-D/HD cards suck. (Design-wise, not so bad, but Hauppauge's software support is HORRENDOUS.)
Most popular cards out now are the Telemann HiPix and the MIT (not the educational institution in MA) MyHD. The MyHD is the cheapest, at $300. (Others are $400). Another Korean company, DVICO, is releasing an HDTV tuner card for approx. $160 that depends on software for the MPEG decoding. (The other cards have hardware MPEG-2 MP@HL compliant decoders, which is a significant portion of the cost. Cheapo decoders such as most RealMagic ones only support MP@ML. Of course the hardware cards have been used successfully in a 233 MHz Pentium while the SW card needs approx. a P3-1 GHz)
The Alienware systems use desktop CPUs, not mobile CPUs. No SpeedStep and less efficient GHz for GHz (usually the mobile chips have more recent manufacturing process technology than the equivalent desktop version)
And go update your drivers. You're obviously NOT running the latest ones.
My dad had bluescreen problems with his I8000 and obtained new drivers from *Windows Update* of all places. This was months ago.
I was using the latest Dell drivers from their website with no problems whatsoever. I don't think my 8200 has ever bluescreened even once.
And if you bothered to do ANY research at all, you would've found the D-Force (and related) modified INFs that are regularly maintained so that you can use your latest Detonator release with "Go" series of GeForces. Yes, I'm running the 41.09 Detonator release on my 8200 with full functionality.
BTW, Dell has some excellent user-to-user support forums if you go to their support website.
Oh yeah, and it runs Linux beautifully too.
Some of the business communications vendors (Such as Avaya) have microcellular systems designed for in-building and short-range communications using the same ideas as cellular. Some microcell phones are designed to automatically hand off to a proper cellular network (At least in theory - I've never seen such a system actually installed and working.)
If you want longer range - Find another solution. Your GSM phone won't cut it. If you're in the USA, get a real provider. The GSM operators in the US have HORRENDOUS coverage. One other poster suggested Verizon, I would too. Verizon's network and AT&T's old pre-GSM network are the only ones in the U.S. that can claim options of true nationwide coverage. (And at this point I don't think you can sign up for AT&T's "old" service, leaving VZW as the only game in town.) VZW's 100% coverage plan is somewhat more expensive than their America's Choice plans, which still cover 60-70% of the landmass of the US, significantly more than any other provider can claim for non-roaming service.
Yes, Verizon's plans are significantly more expensive as far as minutes per dollar, but you get what you pay for. Verizon is #1 in coverage, one of the top providers as far as customer service, and one of the top if not #1 in quality control of both their network AND the phones they release. (This is why the selection of phones available on Verizon is usually a bit more limited/behind other CDMA providers - It takes a LOT longer for a phone to pass Verizon QA testing if it passes at all. Nokia hasn't been able to even get a phone through the process successfully since the 5185. They make good GSM phones but are not nearly as competent when it comes to CDMA.)
Another option for you is two-way business-band VHF or UHF NBFM. This will require an FCC license in most cases, but one that's a hell of a lot easier to obtain than a PCS spectrum license. There are a few unlicensed frequencies, max range on those is 5-10 miles.
I believe all CF implementations are at least 10,000 to 100,000 write cycles minimum. (I'm 99% positive it's at the 100,000+ range at this point.)
I don't know of any flash memories limited to 1,000 write cycles at this point except for the program flash of Atmel AVR processors, but the only time such a unit should ever exceed (or even come close to) 1,000 flashes is during the software development cycle. (I think the average life of an AVR in Cornell's EE 476 lab is 2-4 weeks, due to the fact that it gets flashed every 2-3 minutes or more for 4-5 hours/day. In a production system the program flash should never need to be altered and in fact CAN'T be altered without external programming hardware in most AVR models.)
There are some terms you're obviously unfamiliar with:
Mass production/volume economics
Supply and demand
Oh, and recordable DVDs can be found for as low as $1 each if you search. Hell, even at Best Buy recordables are only $3.50 each, not $5.
Now, if a company can sell a recordable DVD that they produce for 10 cents for $5 or sell it for $1, which do you think they're going to charge? $5 - More profit for them. They'll charge as much as they can get away with charging. Also, because recordable CDs have been around much longer than recordable DVDs and CD-R drives are much more common, CD-Rs are produced in much larger quantities and hence cheaper to produce.
Also, you seem to miss that I mentioned that for prepressed DVDs, they are the same materials but the manufacturing process must be more precise (More data bits per unit of surface area). Same goes for recordable DVDs - Manufacturing tolerances become more and more critical as data density increases, so even if they were made from the same materials (which they aren't for recordables, because of the different laser color different dyes are needed, ones which might be more expensive to obtain/produce.)
Now, if you use a blue laser at DVD densities, it'll work, and you only have to worry about the manufacturing tolerances of a DVD, but:
a) Blue lasers are EXPENSIVE
b) You still have to pay for the cost of whatever new dye is used for recording with a blue laser.
In other words, if you apply this technology to lower data-density mediums, you will increase costs with no benefit.
You can't go the other way around - Using an IR laser but increasing precision and manufacturing tolerances won't work because physics dictates a minimum feature size that can be used with IR.
OK, a few things:
Those time estimates are pretty far off, although it does depend on a few things. Even at a given resolution, you have some freedom to adjust video bitrate.
That said - 23 GB can hold 2.5 hours or so of 1080i video at the bitrate used for ATSC (USA HDTV) programming. An ATSC transport stream is 19.2 Mbps, or approx. 8.5 gigabytes/hour. 1080i video usually takes around 85-90% of a transport stream. (It's somewhat variable - Most people see a 10-15% filesize reduction when using HDTVtoMPEG2 to strip a 1080i stream out of a transport stream.)
A higher-bitrate 1080i source might take more space, but you're not likely to see such sources in the USA.
As to why MPEG-2 - Backwards compatibility with DVDs is one reason, but more importantly, compatibility with current HDTV standards. ALL digital TV broadcast standards use MPEG-2. Why ruin the quality by recompressing with a different codec when you can simply save the stream without recompression?
Plus MPEG-2 is a lot easier to implement in dedicated hardware/is already implemented in lots of hardware. I have only seen a handful of hardware MPEG-4 decoders and zero dedicated hardware MPEG-4 encoders, while hardware MPEG-2 codecs are widely available.
As to MPEG-2 being "pure crap" - You must have been using a crappy encoder and low bitrate. MPEG-4 was designed to excel at low bitrates, while from what I've heard, it loses most of its advantages and in some cases is worse at higher bitrates. (Note that this is the case with other encoders even within the same format - For example, I believe LAME is optimized for high-bitrate MP3 while Fraunhofer usually optimizes their encoders for low-bitrate MP3. As a result the Fraunhofer codec blows away LAME below 96-128 kbps or so, while LAME wins hands-down beyond 128-160 or so.) I've noticed this limitation of DivX, which is just one implementation of MPEG-4. At lower bitrates it blows away MPEG2, but for many videos, MPEG2 blows it away at bitrates approaching that of DVDs.
Because these are neither cheap nor easy to manufacture.
Red lasers are now available in comparably high power (5 mW is a LOT compared to that used in media readers - Burners might be a different story) dirt-cheap - A few bucks for a complete pointer. Partly because the red can be directly generated.
As far as I know, almost all (if not all) green semiconductor lasers are NOT truly a green laser. They are an IR-pumped IR laser combined with nonlinear optics that double the frequency. I believe it's something like:
800something nm IR pumps a crystal that lases at 1000something-nm IR, which is then frequency doubled to green visible light. I don't remember the exact wavelengths, but that's the general idea behind how DPSS green lasers work. Not cheap or easy to manufacture, or easy to put into a small optical drive.
In case you haven't noticed, DVDs are just an evolution of the CD. Same for this technology.
This won't allow for 5x the recording on the exact same media - It'll require new recordable media, or in the case of prepressed discs, a new manufacturing process that can make smaller pits.
It's the exact same evolution from DVD->Blu-Ray as CD->DVD
Name: CD->DVD->Blu-Ray
Laser color: IR->Red->Blue (The main difference)
Media size: approx 5.25" - Same size for all three
The only difference in the media is that the dyes used for recordable media must be different depending on the laser wavelength used. Prepressed media will likely be made from the exact same materials, just with greater precision.
In short: If you apply this technology to CDs, you will get the exact same end product. The laser color is the primary differentiator between the three formats. (And of course, blue-laser devices need higher-precision motors/optics.)
$1900 will get you a pretty nice HDTV.
Don't forget that HD programming will look pretty good on any decent monitor. (i.e. any CRT capable of 1280x1024 or higher) - It won't be huge, but the resolution is what really matters. You can get HDTV tuning capability for $300 if you already have a decent display for your PC.
And you severely underestimate the number of HD stations in the USA. Something makes me think that you live in the NYC metro area or are severely misinformed. In most places, a significant fraction of stations have HD transmitters online. (3/4 of the Philly stations have been broadcasting HDTV since 1998, and at this point all of the Philly network affiliates are broadcasting HD. This is the same throughout most of the country. The biggest exception is NYC - It used to have excellent HD availability, but unfortunately it's been a challenge for broadcasters to even get their analog transmitters back on the air after 9/11 - Every NYC station that wasn't based off of the ESB to begin with is operating at severely reduced power right now. The only HD transmitters in NYC are the ones that were not located on the WTC to begin with. Thank God the Bayonne tower has finally been approved for construction...)
http://www.antennaweb.org/ and http://www.100000watts.com/ have great listings of HDTV availability.
i.Link = Firewire = 1394
The 4-pin connector is as much of a standard across all names for 1394 as the 6-pin connector is. There's nothing special about it. The 4-pin connector is just a smaller one for devices that don't need bus power or for which bus power is insufficient or can't easily be provided. (For example, every camcorder I've seen, regardless of brand, had a 4-pin 1394 connector. The 1394 connector on Dell Inspiron 8x00 laptops is also 4-pin.)
This will most likely work with no problems at all in the following configuration:
:(
PCI HDTV tuner for OTA signals
PC with Firewire port
this drive
DVHStool
A similar setup is already used to record HDTV streams to D-VHS decks (Mainly the JVC HM-D30000U, I think there are 1-2 other DVHS decks now but I'm not sure.)
Note on DVHS: For archival of MPEG-2 Transport Streams (as used for HDTV), 40-50 gigs per unit of cheap media (high-grade VHS tape, I've seen ones specifically marketed as DVHS tapes for $8-9 each, and regular VHS tapes will probably work.) is here now. DVHS decks run around $1000. Unfortunately, the HM-D30000U has a not-so-stellar reliability rep - The decks break a lot.
Unfortunately no PC-based HDTV tuner can tune in cable HD signals - Cable companies use 64QAM or 256QAM modulation, while OTA HDTV uses 8VSB modulation. SOME cable companies will rebroadcast locals in 8VSB format, but most demod the 8VSB stream and remodulate it as QAM. There are no QAM-capable PC tuners right now. Hauppauge claimed it for the WinTV-HD at initial release, but it never happened and they no longer make such claims. A Korean company (DVICO) is coming out with an 8VSB tuner that relies on software decoding of the stream soon, and the next version they're releasing will include QAM capability. (Will not likely work for encrypted digital cable, but I believe there's a law that states that cable companies may not encrypt local station feeds.)
I have a PCI HDTV tuner - Once the price of this comes down I'm really looking forward to it. Yes, it'll be too small for a full hard drive backup, but it will ease the strain on my hard drives because it will be practical for near-direct archival of HDTV recordings. After commerical editing, I could easily fit three one-hour TV shows onto one disc. Right now a one-hour show won't even fit onto a DVD after commercials are edited out.
It dynamically adjusts the throttling depending on the "spamminess" of a connection, evaluated line by line.
i.e. it won't affect your mail unless the server thinks that your messages look like spam.
Are you spamming?
YES. This is definite... A bad sample rate conversion can REALLY kill your music.
Case in point: esd (aka eSound), a soundserver common on Linux boxes. esd SUCKS if it needs to do samplerate conversion. 48 kHz MP3s sound HORRIBLE on any system I've used if played via esd.
Same goes for trying to play a 44.1 kHz file at 48 kHz.
I think sox has a decent converter, so do most sound editing programs. Heck, depending on your school you might even be able to use Matlab. (It's not intuitive by any means, nor is it fast, but you'll learn a LOT about audio processing in the process.)
And yes, I've processed audio files in Matlab, mainly at a time when I was playing with some DSP tricks for amateur radio use.
Yup. When dealing with distortion, you definately have to increase your bandwidth, in order to accurately measure this distortion. 24/96 would be a minimum for accurately modeling the distortion in guitar amps, maybe even a higher sample rate. (In this application, sample rate would be significantly more important than ADC/DAC res. Probably most people wouldn't be able to tell the difference between 16 and 24 bits if the sampling rate were high enough. I routinely push the limits of a $120k spectrum analyzer with a 16-bit DAC. Of course said DAC is operating at 350+ MHz...)
I'm not talking about those cheesy "Use your VCR as backup" devices, those can only be described one way: Cheap hacks. They are limited by all the circuitry in the VCR that mangles the input signal. (The AGC circuitry, for example. Note that this is the same circuitry that allows Macrovision to work.)
I'm talking about D-VHS. (Such as the JVC HM-D30000U, and 1-2 other DVHS decks in existence.) These are designed from the ground up to record digitally, and given the bitrates specified and the recording times, it comes out to something like 40-50 gigs per DVHS tape. Not sure how well "standard" VHS tapes work, they're probably exactly the same except for a cost premium for the DHVS-specific tapes.
Current limitations:
a) DVHS decks record MPEG-2 transport streams, not DV video streams. Which means dvbackup won't work. DVHStool for Windows is about halfway there - But it only will send prerecorded or premade transport streams. The VCR might be unhappy if the TS doesn't have a real MPEG stream in there, but one could embed a low-bitrate 480i stream and fill the rest of the transport stream with data. (MPEG-2 Transport Streams as used in HDTV broadcasts allow for arbitrary data. For example, a Fox station broadcasting 480p with no subchannels is padding their 19.2 MBit/sec stream quite a bit.)
b) Current DVHS decks are rare and the ones that do exist don't have the best reliability record.
He didn't say that it was lossless, but it was essentially lossless.
i.e. the compressionr ratio is so low that it's uncompressed for all practical purposes.
Just like it's practically impossible to tell the difference between an SHQ 2048x1536 JPEG from my Olympus C-3000 and a TIFF from the same camera. SHQ JPEGs from that camera equate to using a quality level of 98% or 99% when saving using GIMP or ImageMagick filesize-wise. (Default seems to be 80% or 90% - Even at those quality levels it's hard to see a difference.)
Hard drives came without keyed power connectors.
THAT was asking for trouble... Toasted a hard drive that way.
24 bits @ 96 KHz is beyond the ability of the human ear to discern any differences.
In fact, with a good antialiasing filter, 16 bits @ 44.1 KHz will put you below the noise floor of all but the best amplifiers and cover the whole range of human hearing. Problem is, an antialiasing filter suitable for 16/44.1 is VERY tough to design without causing distortions in the range of human hearing. 24/96 is easy to develop a suitable antialiasing filter for.
Problems ensue when you are processing the data, though. If you process 16-bit data in the DSP with 16 bits of precision, then at every step in the processing chain you'll likely have rounding errors. Such errors accumulate.
For 16-bit data, I believe most people use DSPs with 24-bit internal precision at a minimum. For 24-bit DSPs, 32 is probably the minimum. I don't know what the likes of Line6's products use. A floating-point DSP would do VERY well for eliminating rounding errors, but those cost $$$.
Interestingly enough - These amps try to use a model of another amplifier's nonlinearities to emulate the nonlinearities of said amp. In my line of work, we do the exact opposite. (Correcting for nonlinearities in RF amplifiers to minimize distortion of any form.)