4 ms·
16x doesn't seem like that much considering they routinely went up to 52x but I guess it is up to the quality of the discs being used. If it was a 16x DVD driv
by NovaDudely 4y ago
16x doesn't seem like that much considering they routinely went up to 52x but I guess it is up to the quality of the discs being used.
If it was a 16x DVD drive then that would probably explain it.
Mostly unrelated.
Now a slight pet peeve of mine was an episode of Mythbusters where they tackled the issue of exploding discs. The issue was the Adam took the data transfer rate of 7.2MB/sec and then extrapolated this to the inner disc area and came up with 30,000 RPM. About 3 times faster than what the discs actually spin at. So naturally discs start blowing up every time they ran the experiment.
Adam assumed a constant linear when drives are actually constant angular. Drive spins, you get what ever data comes.
- dylan604 4y agowow, i'm trying to imagine how loud a 30,000 RPM CD-ROM drive would be. Did the constant angular confusion seem genuine like he just didn't know or more along the lines of it was much more compelling programming to conveniently ignore the fact?
- hatsunearu 4y agoso it's 7.2MB/sec if your data is stored near the edge? Or is data stored "closer together" when it's in the inner track?
- bityard 4y agoWith the drive spinning at a particular constant RPM (which is what they do), then the data rate is the same no matter where you are reading the disc.
- IndySun 4y agoNope. They don't (spin at a constant rpm). Perhaps that's not so surprising; after all, it's not vinyl. Furthermore, varying rpm is explicitly mentioned in the headline post. Also explained here... https://www.repairfaq.org/REPAIR/F_cdfaq5.html https://www.repairfaq.org/REPAIR/F_cdfaq5.html
- jasomill 4y agoWhile the CD data format is defined in terms of constant linear velocity[1], and CD data is always written at a fixed linear density (plus or minus a small margin of error), some drives are, under ideal conditions, capable of both reading and writing this format in CAV mode, as addressed briefly in the "Writing Modes" section of the linked article, http://www.osta.org/technology/pdf/cdr_cdrw.pdf#page=15 http://www.osta.org/technology/pdf/cdr_cdrw.pdf#page=15 In CAV modes, data rate — and laser power, in the case of writing — is a constantly increasing function of the radial position of the read/write head. [1] Though this fixed velocity — and, by extension, data density — is allowed to vary from disc to disc. Assuming the standard (4.3218 Mbit/s raw) data rate, any velocity between 1.2 m/s and 1.4 m/s is permitted, but must then remain fixed within IIRC, a +/- 0.01 m/s margin of error.
- userbinator 4y agoThe format is CLV, but drives advertise their speed as the CAV rate at the outer edge where it's the highest, and all high-speed drives use physical CAV so the data rate changes, starting at the lowest at the inner part of the disc where the data begins, and increasing with increasing LBA up to the highest at the outer edge.
- aidenn0 4y agoYes, you've got it right; the data stored near the center is read at a slower rate (because there is less data on the innermost rings, but most drives spin at a constant speed when not playing CD audio).
- lxgr 4y agoCDs are CLV, at least according to Wikipedia. But that probably doesn‘t prevent drives from operating them as CAV, as long as they vary the read/write rate accordingly.
- dumbotron 4y agoFun trivia: hard drives write from the outside in because the performance is better due to (once) being constant CAV. People would sometime "short stroke" drives by using an undersized partition, improving both latency and throughput.
- lxgr 4y agoAs far as I know, some versions of macOS have also aimed to keep frequently used files (such as core system libraries or executables) in those outer/"earlier" sectors. I think Apple even had a marketing term for this, which I unfortunately can't recall.
- aidenn0 4y agoCD-ROM drives over some speed (4 or 8x I think?) were all CAV and they used the equivalent CLV speed at the outside to get the bigger advertising number.
- generalizations 4y agoYeah, ever since hearing about the episode where they investigated whether harmonics can take down a building...and they strapped a jackhammer to a 10' I-beam...I've been taking their stuff with a massive grain of salt.
- omoikane 4y agoI think Mythbusters is meant to encourage experiments. Any actual results is bonus. https://xkcd.com/397/ https://xkcd.com/397/
- acomjean 4y agoI had a Kenwood truex 52x drive. Worked well. It basically used 7 lasers to read across the disk at a slower speed. It generally worked well. https://www.oreilly.com/library/view/pc-hardware-in/059600513X/ch10s02s03.html https://www.oreilly.com/library/view/pc-hardware-in/05960051... https://www.pctechguide.com/cd-rom/cd-rom-truex-technology https://www.pctechguide.com/cd-rom/cd-rom-truex-technology
- therealmarv 4y agoI had a 48x drive later... this thing (and more the horror stories of it) kind of scared me... always made sure my leg is a little away from it while reading full speed.