david@daisy.UUCP (David Schachter) (01/10/88)
In article <1972@ncr-sd.SanDiego.NCR.COM> greg@ncr-sd.SanDiego.NCR.COM (Greg Noel) writes: >In article <17309@topaz.rutgers.edu> ron@topaz.rutgers.edu (Ron Natalie) writes: >>CD ROM roots would be bad because CD ROM's are blindingly slow. > >Currently true. But don't take the current state-of-the-art as an intrinsic >limit. They'll get faster. According to "CD-ROM: The New Papyrus" by Microsoft Press, faster data transfer from CD-ROMs is unlikely because the frequency of the data starts to approach the frequency of the servo mechanisms used to keep the optics on track and in focus. -- David Schachter (daisy!david or well!davids) Disclaimer: I'm wrong about nearly everything. Cute quote: That's not funny.
mrd@sun.mcs.clarkson.EDU (Michael R. DeCorte) (01/11/88)
>>CD ROM roots would be bad because CD ROM's are blindingly slow. >Currently true. But don't take the current state-of-the-art as an intrinsic >limit. They'll get faster. According to "CD-ROM: The New Papyrus" by Microsoft Press, faster data The reason CD-ROMS are so horribly slow has more to do with the seek times. In designing the format of CD-ROMS it was felt (I believe by Sony) that the disks should be similar to the audio disks. The audio disks do not have tracks exactly but instead have one very big spiral track. This is great for audio, no breaks in the sound when you have to go to the next track, but for computers this stinks; where does track 5 start - about there someplace, +/- a track. Last I heard seek times were something like 1/2 a second. Now this could be wrong as I haven't been keeping up for the past 2 years but I don't think they have changed the format. Michael DeCorte mrd@clutx.clarkson.edu mrd@clutx.bitnet These opinions do not necessarily represent Clarkson U.
tainter@ihlpg.ATT.COM (Tainter) (01/12/88)
In article <791@daisy.UUCP>, david@daisy.UUCP (David Schachter) writes: > According to "CD-ROM: The New Papyrus" by Microsoft Press, faster data > transfer from CD-ROMs is unlikely because the frequency of the data starts > to approach the frequency of the servo mechanisms used to keep the optics > on track and in focus. > -- David Schachter Yes. But, that is a discussion of data transfer rate, which is not what makes CD roms slow. What makes them slow is seek times. Optical disks give you very dense packing of tracks, this means you need very fine advances, which means careful movement, which means slow (relatively, have you ever used a cassette with a C64?). Also, what is to say we can't remork CD roms with alternative servos? --j.a.tainter
david@daisy.UUCP (David Schachter) (01/13/88)
In article <4603@ihlpg.ATT.COM> tainter@ihlpg.ATT.COM (Tainter) writes: >In article <791@daisy.UUCP>, david@daisy.UUCP (David Schachter) writes: >> According to "CD-ROM: The New Papyrus" by Microsoft Press, faster data >> transfer from CD-ROMs is unlikely because the frequency of the data starts >> to approach the frequency of the servo mechanisms used to keep the optics >> on track and in focus. >> -- David Schachter >Yes. But, that is a discussion of data transfer rate, which is not what makes >CD roms slow. What makes them slow is seek times. Optical disks give you very >dense packing of tracks, this means you need very fine advances, which means >careful movement, which means slow (relatively, have you ever used a cassette >with a C64?). > >Also, what is to say we can't remork CD roms with alternative servos? > >--j.a.tainter The surface of a CD-ROM isn't too flat. And the rotation rate of a CD-ROM varies as the head moves in and out. (Unlike magnetic disk drives which have constant _angular_ velocity, the Compact Disk format has constant _linear_ velocity.) And the CD-ROM isn't centered on the spindle too well, either. These factors combine to make it difficult to seek on a CD-ROM. To move to a particular sector, you must be able to read the sector id. To read the sector id, you must be rotating at the correct rate. To rotate at the correct rate must know what sector you are on. Combine this with a non-flat surface and an off-center disk and you have a non-trivial engineering endeavor. Fast seeking is hard and not needed for consumer _audio_ applications. Therefore, the price break due to volume production won't be as great as you might want. Note that changing tracks (and actually, CD-ROM doesn't have "tracks"-- it is a single 3-mile long spiral groove, on a full disk) requires changing the rotation rate of the disk, if you want to keep transferring at the full rate of the CD-ROM. There is a limit to how fast you can speed up or slow down a CD-ROM, due to inertia. (The full speed transfer rate is pitiful-- less than seventy kilobytes a second, after taking out the error correcting code for CD- ROM.)
moore@UTKCS2.CS.UTK.EDU (Keith Moore) (01/13/88)
J. A. Tainter <tainter@ihlpg.ATT.COM> writes: >In article <791@daisy.UUCP>, david@daisy.UUCP (David Schachter) writes: >> According to "CD-ROM: The New Papyrus" by Microsoft Press, faster data >> transfer from CD-ROMs is unlikely because the frequency of the data starts >> to approach the frequency of the servo mechanisms used to keep the optics >> on track and in focus. >> -- David Schachter >Yes. But, that is a discussion of data transfer rate, which is not what makes >CD roms slow. What makes them slow is seek times. Optical disks give you very >dense packing of tracks, this means you need very fine advances, which means >careful movement, which means slow (relatively, have you ever used a cassette >with a C64?). > >Also, what is to say we can't remork CD roms with alternative servos? > >--j.a.tainter > There are two major problems with improving the access times of CD-ROM drives: 1. Difficulties in positioning the laser head. This is due to the "dense packing of tracks", as well as the fact that the CD-ROM is laid out in a spiral, not concentric circles, and uses CLV (constant linear velocity) encoding. So the drive doesn't know exactly where to place the head to access a given frame of the disc; it has to use a successive approximation technique. 2. The disc rotates at a very slow speed. Even if the drive knew exactly where to place the laser, you would have to wait on average 1/2 of a disc revolution. As it is you usually have to wait longer. With frame-to-frame seek times of 1/2 second, a large portion of this delay is due to rotational latency. You can speed this up somewhat for CD-ROMs (it has to be fixed for audio players), but you start running into limitations of the optics, the decoder chips, the servos, and the control systems. These can all be redesigned, but by then your CD-ROM player doesn't even resemble an audio CD player anymore; perhaps it can't even play audio disks. So it's going to be expensive, at least until CD-ROMs become commonplace enough to make it worthwhile to design the players from scratch. Keith Moore UT Computer Science Dept. Internet: moore@utkcs2.cs.utk.edu 107 Ayres Hall, UT Campus CSnet: moore@tennessee Knoxville Tennessee BITNET: moore@utkcs1
norman@a.cs.okstate.edu (Norman Graham) (01/13/88)
in article <4603@ihlpg.ATT.COM>, tainter@ihlpg.ATT.COM (Tainter) says: > What makes them slow is seek times. Optical disks give you very > dense packing of tracks, this means you need very fine advances, which means > careful movement, which means slow [...] We must remember that a track on a CD-ROM is very different than a track on a conventional magnetic hard disk. Tracks may be different sizes on the same disk (there may even only be one track for the whole disk (or maybe there has to be two, I don't remember)). What makes CD-ROM slow is that even if you know the minute, second, and sector of the data that you want, all the drive can do is get close to where it thinks it might be and start streamming off data until it gets to what you wanted. > Also, what is to say we can't remork CD roms with alternative servos? We could, but it would defeat one advantage of CD-ROM drives... they're cheap (or at least they will be by the 2nd quarter of this year :-). Since to make a CD-ROM drive you take a CD-Audio drive (which has a HUGE market as compared to CD-ROM), take out a lot of electronics, and put a little bit of electronics back in, CD-ROM drives should approach the price of CD-Audio drives (< $200 US). > --j.a.tainter -- Norman Graham Oklahoma State University Internet: norman@a.cs.okstate.edu Computing and Information Sciences UUCP: {cbosgd, ihnp4, 219 Mathematical Sciences Building rutgers}!okstate!norman Stillwater, OK 74078-0599
paula@bcsaic.UUCP (Paul Allen) (01/14/88)
Several people have tried to explain why cd-roms are so slow, but noone has gotten it right yet. It doesn't have anything to do with the spiral track, or the fine track pitch. Cd-rom drives are basically re-packaged audio CD players. For audio, there is no requirement for fast random access, so there has been no push to develop small light read heads. Since the head is large and relatively massive, it cannot be moved very quickly. Hence, we have seek times on the order of 1 second. Improving the seek times of cd-roms will require new technology read heads and a corresponding significant increase in cost. It'll come. Patience will win in the end! Paul -- ------------------------------------------------------------------------ Paul L. Allen | paula@boeing.com Boeing Advanced Technology Center | ...!uw-beaver!ssc-vax!bcsaic!paula