Slashdot Mirror
Triple-Density CD-RW From TDK & Friends
Houndogk writes: "I came across this reading the news of the day at Tomshardware. This [article] talks about a new generation of CD-RW that promise to be 3x as fast and have 3x the capacity as current drives. It is also expected to scale to 4x and 5x." From the article: "[T]he premise of ML technology is the use of gray-scale disc encoding, with 3 bits per spot giving eight shades of gray. Under a microscope, the disc surface appears as a continuous blending of light to dark shading, versus the traditional disc appearance of either dark or bright spots." And what happens when we go to 24 bits per spot? ;) This announcement seems to partly answer GeoffM's quest for dual- or quad-density CD-Rs, and handily top Sony's moves to double-density.
So ??
How many discs are you burning that you dont expect to hold a very long time ??
lets see what data are usualy put on discs
The article talked about putting this into camera's and mp3 players / "walkmens" / "discmen" / "minidiscman" /
Sure a long lifetime would be nice, but some people, i know i do, usualy just listens to the lastest music the radio plays. So a short lifetime doesnt matter if the disc is cheap.
So, what does a marketing dude see from this... Cool 2 markets, one for casual data where the lifetime doesnt matter that much, and one for importent data where the lifetime is importent. So, you sell the 1-3 years cheap, knowing you'll sell lots of these. The others have longer lifetime. And since they offer an extra value, you can sell these for more money. Though they might be more expensive to make, they shouldnt be that much more expensive. Imagien having cd's which could hold data until my kids gets old... i bet people wouldnt mind paying 10$ for a disc that lasts this long or perhaps even 20$. Remember you dont have to use the long term discs. I doubt it would matter much if they only hold the usualy 650MB for the long term discs. (at least for the next decade or so)
Read my other replies, i think i happened to write it twice, about why these discs can also hold 1bit instead of 3bit.
ion++
My first concern when reading this article was "what happens to the data as the plastic ages?".
Does the data change if the plastic starts to darken or yellow? Could make for some interesting aat data recovery. ("Well, you just have to subtract one from every byte".)
"Trademarks are the heraldry of the new feudalism."
One problem with this technology is that it is incompatible with the methods used to produce CD-ROMs and DVDs. A stamper produces pits on a blank disk. It can't produce shades of gray.
Mea navis aericumbens anguillis abundat
Analog means infinitely variable values: in other words, decimal values.
Digital means discrete values: in other words, only certain values are permissible.
Tri-state is digital: there are only three possible values, and there are no in-between values.
There are some excellent arguments to be made for using digital technology that goes beyond simple on/off. Easier to build fuzzy-logic devices, f'rinstance. Two tri-states provide more than twice what you get from two binary states, without requiring twice the circuit complexity, and the savings increase exponentially the greater the number of sensors/indicators/controls needed.
--
--
Don't like it? Respond with words, not karma.
I had fun several years ago when the high-speed (greater than 4x) CD-ROM drives came out. I had one in my new machine at work that almost make the CPU case shake with some disks. I got to thinking 'hmmm, disk out of balance...' so I started fooling with the drive, putting progressively more scotch tape on a disk to see how loud it would get. Then I taped a small metal washer onto a disk. Big mistake. The CD-ROM drive made so much noise (I was in my cubicle) that I had to shut off the whole machine to keep people from wondering what the hell I was doing (embarassing questions would have been asked, I suspect). The faster, ever-faster CDROM drives definitely result in reduced reliability over time. CD media isn't always perfect balanced. At my new company they even put 'taped labels' that throw off the whole balance of the disk on official CD releases (internal master copies) which I immediately recognized as a bad idea.
Hay thar.
You're all way off with this tri-state/base 3 discussion. Just because the article talks about "3x" drives doesn't mean that there's base 3 maths involved. You're letting the 3 confuse you.
The 3x refers to the fact that the drives hold 3x as much information. If you want to talk about the number of states of each dot, it's not 3, but *eight*. So if you want to talk in bases, you should really be talking base-8.
But in any case, this is all abstracted to a high enough degree that to all intents and purposes one could consider the drive to be emitting a stream of binary digits, just like the old days. Only at the very lowest levels does the firmware need to know it's not dealing with binary states.
Windows is shit.
--
If RGB color on a CDRW were possible, a really cool hack would be to generate an ISO that would burn on the CD and create a full-color image on the data side ;-)
the real at&t mix
binary digit ==> "bit"
ternary digit ==> "tit"
Now we can have "megatits", "tit compression schemes" (= corsets?), "parity tits", "titwise logical operators", "tit rotation", "tit buckets", and "128 tit encryption".
"big endian" and "little endian" will remain unchanged.
If your interests run toward utility rather than purience, you can notice that 8 tits (a "tyte"?) will store 3^8 = 6561 distinct values.
--
Sheesh, evil *and* a jerk. -- Jade
So would you only like one OS provider, for greater interoperability (at cost of service, performance, and capability), or one phone service provider, or one ISP, or one music provider, etc?
The price of competition, incompatibility, is balanced by the value of competition, which is each of the 5 standards trying to outdo each other, on the ground of price, performance, capability, reliability, useability, etc.
If there were only 1 standard, why would you think we'd get any improvement or innovation? The same forces that would bring about 1 standard *should* also force the system to never ever change or improve; the minute someone comes up with a better idea, you'd have 2 competing incompatbile systems, and the problem starts again!
Geek dating!
GPL Deconstructed
I agree that, philosophically at least, more error correction should exist to handle more data. There's plenty of info and science around error correction, so I'm not too worried about that.
Your problems with accuracy can be handled in manifold ways; multiple lasers (3, ostensibly) to handle the burn accuracy. Read accuracy would probably be no different than what you described, using 12.5% intensity deltas between values, instead of 50% intensity values.
There's no reason to slow down the burn to increase accuracy; just use better receptors, and higher tolerance devices. That technique seems to work fine for Intel, which keeps cranking out faster and faster CPUs with better and better processes and technique!
So IDE doesn't cut it? That's why there is... SCSI, Firewire, SerialATA and USB2. My preference is for Firewire to take the lead, as it is the cheapest and most established of the 4 listed technologies...
So, given that we can get around all those technical difficulties... Why can't we see a faster read and burn rate?
Geek dating!
GPL Deconstructed
You need to pay more attention in your history class, bub.
EDSAC, in 1949, used 35-bit words. The same year, BINAC used 31-bit words. 1951, the EDVAC with 44-bit words; and the IBM "Defense Computer" with 36-bit words.
Now, granted, these were binary bits; but it does show that powers-of-two are not necessary and, indeed, weren't even the norm back in the beginning.
There's no reason not to use three bits. It's fifty-percent more complicated to detect than binary, but twenty-five percent less complicated than detecting a four-state system. And tri-state electronics are plenty common, whereas four-state electronics aren't.
I think that as a proof-of-technology, 3 bits is the logical choice. As the technology advances, we'll undoubtedly see more bits-per-recording-pit.
--
--
Don't like it? Respond with words, not karma.
That's kind of the idea behind quantum computers. Each quantum bit would have 32 possible values, which is equivalent to 5 binary bits.
No, it's not. The idea of quantum computing is to have bits which are 0 and 1 at the same time. A "byte" made of 8 quantum bits is a superposition of all 256 states a normal binary byte can represent. So if you want to check which one out of the numbers 0-255 is the key in an encryption scheme, you can either test 256 values one after the other conventionally or check them all at once with a quantum byte.
For the current price of a DVD-ROM drive, you could buy a CD-ROM drive and have enough money left over to tack on another 20GB to your hard drive,
I doubt that I could find a cd-rom and 20gb hard drive for $50
This Wiki Feeds You TV and Anime - vidwiki.org
No, instead of 1 bit (2 states) you store 3 bits (8 states). So that gives you three times as many bits, hence three times the capacity. But to encode 3 bits, you have to be able to produce (and later read) 8 gray levels.
EagerEyes.org: Visualization and Visual Communication
650 is a hard number to multiply in my head.
Keeping
In reality, one of the selling points of CDRWs is that they can be played on almost any regular CD-ROM (although some cheap players certainly have problems reading CDRWs). At this point, these CDs seem to be in a race with DVDRWs (which will have more density than even 3xCDRWs). Especially damming for this technology is that you will need a special player (the burner) to read the discs one you make them.
I read the internet for the articles.
The fact that these drives/media won't be subject to DVD consortium technology licensing fees would be the biggest reason I'd see for their development.
Timothy, 24 bits per spot means that it has to differ between 2^24 (16777216) different shades of grey... it's very hard to reach that level of precision, so I don't see it coming any time soon.
Yes, multiple colors are exactly what's on the way. . .
http://www.c-3d.net/tech_frameset.html
I thought the point of 1 bit per spot, also known as "digital", was to reduce the amounts of errors. Now they want to go back to storing data analog?
This sounds like that story in which RAM was made that could store 4 values instead of 2 by using the same technology described here.
That didn't make it. Will this?
Sanjuro
do we really have a need for this drive?
more specifically, do we really have a WANT for this drive? before you call me -1 troll, ust consider this...
there are a number of possible endings the removable media story could have from this:
1) everybody buys these, and they become standard. media prices plummet, everybody is happy.
2) there becomes a "standards-war" between these drives and all of the other removable media types, prices get slashed, and the consumer has no real standard, but prices are cheaper as each company tries to out-do the other.
3) there becomes a "standards-war" between these drives and all of the other removable media types, and because nobody is buying enough of any one type of drive, there is no standard, and prices are higher because of production costs.
4) Not enough people will buy the drive, so it will die quickly. The few people that do own them will have to pay high prices for the discs, and they won't be able to give them to anybody else, as maybe 5 people will have them.
sorry if it sounds like FUD, but that's all i think when i see new removable media...
"I hope I don't make a mistake and manage to remain a virgin." - Britney Spears
I found absolutely nothing in Google with the word "chlarodium" in it. It looks like you're BSing us with the intent of karma-whoring.
I hereby declare Shenanigans on you unless you can fess up with a reference or a spelling correction.
Sure, tri-state is digital, and very cool, but 3 bits per spot does not mean it's tri-state. Tri-state would presumably mean it's got black, white, and grey values. 3 bits means it would need 2^3, or 8 different values.
"The question of whether a computer can think is no more interesting than that of whether a submarine can swim" -EWD
Damn. That's what I get for pretending to be an electronics engineer!
--
--
Don't like it? Respond with words, not karma.
Increasing the bit depth from one bit(2^1= 2 shades of gray) to three bits(2^3 8 shades)increases the complexity of the read and write process. More error correction would have to be present on these discs, and although at the same rotational speed these discs have three times the data rate(3x density increase), we can run into problems. Our current burners have enough time at 1/4 of the complexity of these discs. When burning current discs, 52% intensity is rounded to 100%, and 49% is rounded to 0%(in a perfect world). With 8 shades we can't do this, and would have to slow down the burning process to insure accuracy, and on an IDE bus(consumer equipment) couldn't supply 3x the bitrate of current burners anyway(anyone ever had buffer-underruns). So we would have to decrease the burn rate to that of current burners because of bus-throuput(SCSI doesn't have this problem), and decrease it further for accuracy. Reads still get pulled off a little faster, but seek time would stay the same. I would say 2x faster sequential reads and .5x burns in this "faster" technology.