30 Sep
2004
30 Sep
'04
2:47 p.m.
On Wed, Sep 29, 2004 at 08:33:03PM -0400, Peter Lutek wrote:
I don't have good answers for this. I see that a typical work flow is
very different from my own. But, there may be some hope in ecasound's
ewf control files.
I do see that having unlimited levels of undo-redo builtin to the app
would make what you describe much less difficult. One thought I had
thinking about all this earlier this morning ... Much of this could be
managed by keeping ecasound chainsettups in scripts. You could set up a
script with all of your original input files, tweak the effect settings,
do a trial run only outputting to your monitors, lather, rinse, repeat
... and then only add the mixdown file outputs when you are satisfied
with the effect settings. Instead of keeping many different versions of
the actual audio files, you would keep different iterations of the
mixdown/effect settup scripts. Would that be more manageable? I think
I've read examples in the docs and mailing list where Kai describes this
sort of workflow in his own use of ecasound.
-Eric Rz.
On Wed, 2004-09-29 at 18:38, Eric Dantan Rzewnicki
wrote:
ok. so I am complete nut after all. :) I'm realizing in this discussion
that most people don't work the way I work. That's ok.
But, there have been a number of attempts at providing higher level
interfaces to ecasound via the ECI interface. I've just this morning
finished reading the August 2003 ecasound-list archive. In August and
September of last year there was much discussion, and even some trial
implementations, of features that come close to providing a lot of what,
I think, typical multitrack editor users want to have. Check out the
SuperEcasound and related threads from that time frame.
For instance the emacs lisp interface implements position markers. I
believe a number of these interfaces had rudimentary support for cut,
copy, paste operations. Admitedly this stuff is not all fully
implemented and I'm not sure if it is all actively developed. However,
the OP (was that you?) did see building a tool himself as an option.
With ecasound much of the ground work has already been laid ... I
suspect what ever else is needed could be made to work by building on
what's already been done, or even just learning from it.
Maybe it's not the best answer, but is at least worth a look.
(disclaimer: take this all with a grain of salt.
I've been a bit giddy
all week. 8) -edrz)
On Wed, Sep 29, 2004 at 06:02:34PM -0400, Peter Lutek wrote:
i am not for a minute suggesting that a commercial-look-alike
GUI is
important, and i assure you that i do appreciate the value of a text
interface for many, many things -- that is a huge part of why i am
almost entirely "linuxized" at this point.
i've spent some time with ecasound (no, not enough to be called fluent,
but still...) and love the flexible manner in which it operates and the
efficiency with which you can do many things. however, i found myself
constantly running up against one big problem (and perhaps you can offer
a solution?): accurate location of edit points is nearly impossible
without seeing a waveform display. yes, i know one can edit analog tape
by "rock-n-rolling" the reels to find edit points, but we can't even do
that, and we demand finer control in the digital environment. the other
issue is still that of being able to roll back changes or modify
settings in a multi-layered task like audio editing. more on that later,
re SND... thanks for your thoughts, greg!
i am in a very similar position to you, and agree with most of what you
have to say. irrespective of ardour's current status, it IS a shame that
there is not at least one other viable professional
multitracking/editing package nearing completion.
Am I just a complete nut?
(which I'm ready to admit is entirely
possible ;) ) or does no one else see ecasound as an alternative for
getting work done? it is stable and fairly complete, reasonably well
documented ... it's just different ... Why is it that having a "viable
professional <foo> package" means having a GUI that looks like something
from the commercial world? Just because it's the dominant paradigm for
96, or whatever, % of computer users doesn't have to mean it has to be a
requirement for "professional" use. I much prefer mutt to any lookout
clone, for example. ... it is
well, yes, but......say, for example, you have three tracks,
each of
which you're modifying in three ways (i.e. amplitude, EQ, compression).
if you wish to retain the ability to undo any one of those three
changes, you must maintain three backups of each track (each with only
two of the changes applied). so, nine backups for EACH change of all
three characteristics on all three tracks (retaining detailed records of
what changes were applied at each pass). then, if you mix it all down,
and decide you then want to change the EQ on track two, you must dig up
the track-2 backup with only amplitude and compression applied, apply a
different EQ, and re-write the mix. and this is a VERY simple example.
surely it is more efficient to have an environment (and, yes, i would be
much happier with a text UI with a bunch of cells of values -- a mix
spreadsheet, if you will) in which all of these changes to all of the
tracks are applied non-destructively in realtime. that way you can
change any given parameter on any given track at will.
on another front, say you make a track by taking bit of takes and
editing them together. surely it is a lot more efficient and flexible
(in terms of editability) to have that track exist as a bunch of
adjustable regions which refer, on the fly, to base takes, rather than
as a single, written composite which must be re-assembled and re-written
for any change.
again, please let me know if there are ways of approaching these issues
which i have overlooked! i would love to be using SND and/or ecasound
fulltime! near-impossible to step back and tweak an EQ you
adjusted a few days ago
on a particular component of your mix.
can't one simply back up a file and
work on a copy? ...