| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 22.1 | | VIKING::MARKEY | | Mon Jun 25 1984 14:08 | 12 |
| Digital Sampling is like any type of recording. The incoming musical
information is translated into equally spaced (in time) chunks which
represent the overall amplitude of the waveform at that time.
The sampled waveform is then stored in a memory. The playback is
then accomplished by stepping through the memory in the same way, and
using a digital to analog converter to build an analog output.
A good source of info on the digital music technique is an article
in the April 1980 issue of Byte (I think). Check it out.
Brian
|
| 22.2 | | NOVA::RAVAN | | Mon Jun 25 1984 17:09 | 13 |
| OK, that's the 'first level' answer. But I read an article in Computers and
Electronics (formerly Popular Electronics) on the Kurzweil which stated that
they did some kind of pattern matching on the input signals. A straight
digital playback of a recorded signal would not be able to capture all the
nuances of something as complicated as a grand piano as they have done. There
is more to their technique than simple digital playback.
If 'digital sampling' refers to the simple playback of digitized information,
then I understand (and am disappointed). However I thought that digital
sampling referred to the Kurzweilish technique of compression and 'feature
extraction' to build a mathematical model of the instrument being recorded.
Am I wrong? In any case, anyone have any idea what they (Kurzweil) do inside
their box and how they do it?
|
| 22.3 | | VIKING::MARKEY | | Tue Jun 26 1984 13:35 | 12 |
|
Bull-hockey. In fact, Kurzweil's system is just a cheap way of stuffing it
into a ROM. The correct way to do it is to analyze the individual envelope
(volume function) segments, then use a Laplace transform to isolate one
cycle of an individual waveform within the segment. The result is at least
one, preferably more, waveform tables per envelope segment. That way the
envelope period does not change regardless of pitch,
By the way. CompSync holds a patent on such a process and we plan to
introduce it as part of our own system this fall.
Brian
|
| 22.4 | | NOVA::RAVAN | | Wed Jun 27 1984 01:27 | 16 |
| Not to be a stick-in-the-mud, but I'm confused. Here's my major problem with
saying that digital sampling is digital recording:
If all the Kurzweil box is doing is digital recording, how can I
ever play a half note when all I recorded were quarter notes?
I have heard that they have a digital recording box they use in
demos but that the stuff they use to simulate natural instruments
is different. I'll admit that I have never attended a Kbox demo
and I'm only operating on second-hand information which is what started
this note in the first place.f*
And I assume you are saying your company does something the Kbox doesn't.
Sounds interesting. But since you have it patented, I won't ask you to
explain it further. Unless of course, you would like to (since you have
it patented, it should be safe to talk about it [?]...)
|
| 22.5 | | VIKING::MARKEY | | Thu Jun 28 1984 16:13 | 15 |
|
I don't feel like talking about Kurzweil forever, because frankly all
their hooplah is boring the hell out of me, but here's the info:
Their 'regular' natural instrument stuff uses a segmented group of
wave tables so that various times may be spent in the sustain portion
of the envelope, thus allowing different play lenghts.
Their sample box is like CMI's, NED's, et all. It just plain records
and plays back. The envelope period changes proportional to the pitch,
and in fact craps out at about one octave on either side of the original.
Buy my box. It's better and I won't have to talk about Kurzweil.
Brian
|
| 22.6 | | LOLITA::DIORIO | | Fri May 08 1987 17:20 | 3 |
| How much is your box going for??
Mike D.
|
| 22.7 | | AKOV88::EATOND | Then the quail came... | Fri May 08 1987 17:38 | 3 |
| RE < Note 22.6 by LOLITA::DIORIO >
Uhh,... I think the offer may be a little extinct by now...
|
| 22.8 | Where is the CompSync box? | FGVAXZ::LAING | Pipe Dreamer ... Jim Laing @DTN 261-2194 | Thu Oct 22 1987 12:00 | 4 |
| Did this CompSync box ever exist; does it exist now? Never heard
of it ...
-Jim
|
| 22.9 | Get em while they're hot! (Wanna by a watch?) | MAY20::BAILEY | Steph Bailey | Thu Oct 22 1987 17:46 | 20 |
| No. It never existed in production. (I checked).
But, there was somebody who really understood synthesis, yes, siree.
Better than one of the most respectable noters (at least I respect him,
but I respect Paul Kent, and he is only 5' 2". I think it's the
likeability that counts.) in this conference, who as we found out
in another topic, had completely the WRONG IDEA about an ideal
synthesis tool.
We all know that Kurzweil had absolutely no idea what he was doing
pawning cheap trash like that off on the masses, but it too some
of that chap's eloquent explaination to let us in on the secret.
Steph
[It's too bad I was only a read-only in the conference when this
Markey guy came through!]
|