mk@wroach.cactus.org (M. Khan) (10/11/90)
Is there a hardware bottleneck that prevents digitized audio to be recorded concurrently to disk, or played back similarly, without interruption of the audio? What models of the Mac have this bottleneck? If there is no such (or on models that have no such) bottleneck does software that provides "virtual memory" (don't remember any names now) work to get around the limitations of audio-digitizing software that limits itself to available memory? Any experiences with this? Also, in case there is no bottleneck, are there any software packages that work with a MacRecorder that dont limit you to available memory?
russotto@eng.umd.edu (Matthew T. Russotto) (10/12/90)
In article <2332@wroach.cactus.org> mk@wroach.cactus.org (M. Khan) writes: >Is there a hardware bottleneck that prevents digitized audio to >be recorded concurrently to disk, or played back similarly, >without interruption of the audio? Floppies are too slow, but the SID software does it to the hard disk. >If there is no such (or on models that have no such) >bottleneck does software that provides "virtual memory" >(don't remember any names now) work to get around >the limitations of audio-digitizing software that limits >itself to available memory? Any experiences with this? VM swapping could concievably make you lose serial port data. As long as your program has enough time to read the serial port buffer before it overflows, you will be OK. (actually, I have no idea whether the software uses the buffer or goes right to the hardware for speed-- 176KBps is a lot for a machine to handle. I'd guess that the slower models (SE, plus) can't record to disk) -- Matthew T. Russotto russotto@eng.umd.edu russotto@wam.umd.edu .sig under construction, like the rest of this campus.
dan@chaos.cs.brandeis.edu (Dan Schwarz) (10/13/90)
In article <2332@wroach.cactus.org> mk@wroach.cactus.org (M. Khan) writes: >Is there a hardware bottleneck that prevents digitized audio to >be recorded concurrently to disk, or played back similarly, >without interruption of the audio? > >What models of the Mac have this bottleneck? > >If there is no such (or on models that have no such) >bottleneck does software that provides "virtual memory" >(don't remember any names now) work to get around >the limitations of audio-digitizing software that limits >itself to available memory? Any experiences with this? > >Also, in case there is no bottleneck, are there any software >packages that work with a MacRecorder that dont limit you >to available memory? The Macintosh has no inherent bottlenecks which prevent it from recording direct-to-disk or playing back direct-from-disk. Certainly the machines equipped with Apple Sound Chips have the potential to do it; machines with the Sony Sound Chip (Mac Portable, Mac SE and earlier) may be limited to the 22khz sampling rate when doing simultaneous disk operations. (When you say "disk" I assume you mean hard disk, not floppy or fileserver.) All you need is some clever software... which, as far as I know, does not exist yet. But it CAN be done, and WILL be done. Articulate Systems Inc.'s VOICE LINK(tm) digitizer product will eventually include software with this capability. For more information, write and ask for product literature. As far as using virtual memory to simulate RAM for the purposes of recording, I think you'll find the results to be unsatisfactory. You'll get glitches when the disk is accessed. This is especially true with polled recording devices like the MacRecorder(tm) which chew up large amounts of CPU time to do their work. Notice that you can't even move the mouse when recording with the MacRecorder - disk access is far more disruptive. The Voice Link is interrupt driven, so it will eventually be able to record to disk. Playback from virtual RAM may work okay, depending on the scheme that is used to play sounds. Asynchronous playback with good buffering systems would do fairly well with a VRAM system. Disclaimer: I work for Articulate Systems Inc. and have been involved in the development of the Voice Link and other products. This is not intended to be a product endorsement... (although I certainly wouldn't mind if you bought one..:-) _Dan -- | Same as it ever was | Dan Schwarz, MB 2926 Brandeis U. | RECYCLE YOUR JUNK| | Same as it ever was | I'NET dan@chaos.cs.brandeis.edu |------------------| | Same as it ever was |----------------------------------| tradetapes?mailme| | Same as it ever was...TALKING HEADS "Once in a Lifetime"| FloydRushDeadEtc.|
b3300876@rick.cs.ubc.ca (george kai yee chow) (10/16/90)
In article <1990Oct13.165359.14653@chaos.cs.brandeis.edu> dan@chaos.cs.brandeis.edu (Dan Schwarz) writes: >The Macintosh has no inherent bottlenecks which prevent it from recording >direct-to-disk or playing back direct-from-disk. Certainly the machines >equipped with Apple Sound Chips have the potential to do it; machines >with the Sony Sound Chip (Mac Portable, Mac SE and earlier) may be limited >to the 22khz sampling rate when doing simultaneous disk operations. >(When you say "disk" I assume you mean hard disk, not floppy or fileserver.) > >All you need is some clever software... which, as far as I know, does not >exist yet. But it CAN be done, and WILL be done. Articulate Systems Inc.'s >VOICE LINK(tm) digitizer product will eventually include software with this >capability. For more information, write and ask for product literature. The recently produced SID package contains a test program which does exactly this: double-buffered recording to disk. I've used that with my MacRecorder and it works. However, the SE isn't quite fast enough and so it misses a few samples whenever the buffer is switched. I've talked to the author (couldn't remember the name offhand) and he says that he is aware of this. The solution for the while: record using a faster machine. >| Same as it ever was | Dan Schwarz, MB 2926 Brandeis U. | RECYCLE YOUR JUNK| >| Same as it ever was | I'NET dan@chaos.cs.brandeis.edu |------------------| >| Same as it ever was |----------------------------------| tradetapes?mailme| >| Same as it ever was...TALKING HEADS "Once in a Lifetime"| FloydRushDeadEtc.| George
dpawson@.com (Dave Pawson) (10/18/90)
>. . .This is especially true with polled recording >devices like the MacRecorder(tm) which chew up large amounts of CPU time to >do their work. Notice that you can't even move the mouse when recording with >the MacRecorder - disk access is far more disruptive. The Voice Link is >interrupt driven, so it will eventually be able to record to disk. >| Same as it ever was | Dan Schwarz, MB 2926 Brandeis U. | RECYCLE YOUR JUNK| >| Same as it ever was | I'NET dan@chaos.cs.brandeis.edu |------------------| >| Same as it ever was |----------------------------------| tradetapes?mailme| >| Same as it ever was...TALKING HEADS "Once in a Lifetime"| FloydRushDeadEtc.| I could be wrong, but I don't think this is why you cannot move the mouse. I fooled around a few years ago with sound digitizing, and trying to write digitizing software. It turns out that there is a bit in one of the VIA registers that used to indicate if the mouse button had been pressed. However, with the advent of ADB, that bit then indicated any event on the ADB bus. I have found that a lot of old digitizing software will prompt you with something like "Press the mouse button to stop recording", but will stop if you even move the mouse. Eventually, I became quite certain that this software probably just monitors the bit in the VIA, and if it flags an event, they stop recording. They think they are waiting for a mouse click, but actually are waiting for any ADB event. Pressing a key on the keyboard stops it too. Just a piece of random info to brighten your day. . . Dave Pawson dpawson@oracle.com "Two wrongs don't make a right, but three rights make a left"