Subject: Re: [linux-audio-dev] memory management
From: Paul Davis (pbd_AT_Op.Net)
Date: Mon Nov 26 2001 - 23:20:17 EET
>I would be interested to know how this differs from schemes used by other
>editors, if anyone is willing to briefly explain. For example I have
>never heard of EDLs and I couldn't find a good explanation on the net.
ardour (almost) never modifies data on disk. all edits are done using
a playlist that consists of a list of time-ordered regions. each
region describes a starting point in an audio file and a duration.
all FX are applied in real time, or applied on a per-region basis, and
written back to disk in a new audio file. all temporal-based edits
just consist of modifying region definitions and playlist composition.
this is consistent with the way that most DAWs operate.
an EDL is an "Edit Decision List", and just consists, bizarrely
enough, of a list of editing decisions made by the user. There are,
however, two ways to store an EDL. one way is as the list of edits
themselves:
cut from X to Y
insert foo at N
and so on. the other way is to store the result, which is typically
called a playlist in many programs:
at time N: file foo, start at sample M, continue for O samples
at time P: file bar, start at sample T, continue for R samples
ardour does the latter (it used to do the former). when you take one
or more playlists and add automation data for gain levels, fx
parameters, and so forth, you end up with a complete specification for
the playback mix. when all this data exported as a file, industry
people sometimes call the result "an EDL". OMF is probably the most
popular format for this, though its proprietary (owned by Digidesign).
the approach outlined here does not equate to waveform editing, though
there are some overlaps in the kinds of things that can be done with
both approaches.
--p
This archive was generated by hypermail 2b28 : Mon Nov 26 2001 - 23:19:06 EET