At 01:06 PM 6/13/03 -0400, David W. Fenton wrote: >Is there something wrong with the files themselves or is my computer >causing a problem somehow? It doesn't appear to me to be related to >other activity on the computer (I've watched the CPU activity and >there is no spike just where there's a dropout).
I listened to the files -- after a tussle with the extremely slow bway server. It was serving at dialup speeds for some reason ... maybe *everybody* on this list was downloading that file! >The MP3s in quesiton are at: > http://www.bway.net/~dfenton/Collegium/HimmelUndErde/ >One piece has had it happen just before the end twice when I was >listening to it. There is no trouble with the file. I opened it also in Cool Edit Pro, and it has no problems. Finally, I back-converted it to .wav using MMWave, and still fine. >Any idea on this? My guess is the person who had the dropouts and you both have audio subsystem issues, or perhaps disk access time issues. Hard to say. I know that a few weeks ago I upgraded to a 120GB hard drive, only to discover a slowdown as my motherboard's ATA bus was too slow. I installed an ATA133 card, and audio glitches vanished. You might also have software running in the background. Again, I'm not sure if you're PC or Mac, but if PC, dump the background-scan virus program, Microsoft's "Fast Find", AOL AIM, and just about anything else you don't absolutely need to have running. Another culprit on PCs in QuickTime. Depending on your system, it can be benign or wreak havoc. I don't have it on any of my four PCs because it just won't behave. The entire audio subsystem was trashed in terms of glitches and dropouts, so I uninstalled it and combed through every system directly to get rid of its embedded files. If you're actually ripping from an audio CD, use the best extractor. A piece of freeware (again, PC only) is EAC (Exact Audio Copy). Do a web search; it's worth it, because it has very well-written features for making sure the copy is as good as possible. (Due to the nature of audio CDs, it was expected that data errors would be marginally corrected or masked; accurate data copying depends on features not present on audio CDs. Data CDs are accurate, but they also have about 15% of the disk space given over to error-correction.) If you were ripping from the CD directly to MP3, don't. :) >I ripped them at 64bits, too, and was wondering if that's normal for >this kind of situation (putting classical music on a web page). Encoding at 128K/44KHz/stereo, no VBR, is kind of the web standard (web "hi-fi" is 192K). The idea is to use the 'cleanest' compression that's still streamable at cable/ISDN rates. (It wouldn't matter from the bway server because it was so slow, at least today.) >Can one use less than 64bits and get a decent trade-off of sound >quality and size? Not a good one. If you only want musicological samples, go lower. We use 32K/mono for our radio show because it streams at 56K dialup, it sounds okay for a document, and our agreements with composers and labels limit our archives to dialup speeds. But our many listeners worldwide are quite happy to hear any of this new nonpop stuff at all! Hope that's a start. There are lots of online guides, and a CDR listserv as well (don't have the subscription info). Be sure to look for the extensive online CDR FAQ (http://www.cdrfaq.org/). Everything you need to know! Dennis _______________________________________________ Finale mailing list [EMAIL PROTECTED] http://mail.shsu.edu/mailman/listinfo/finale
