Re: [Audacity-devel] Rate envelope

[Dr W. B. <wj...@ab...>]

On Fri, Dec 06, 2002 at 12:25:51AM -0500, Shane Mueller wrote:
> 
> Hey, I finally got a chance to check out the rate envelope. Cool! I
> can't wait to hear it working.  Here are a couple issues you might want
> to think about:

Great!  Thanks for trying it!  There's a new version in the same place,
which will hopefully fix some things you don't like (see below).

> * Pretty cosmetic: the color of the envelope should match the color of
> the icon.    Anyway, I like the red in the icon better than the red in
> the envelope. And, there may be a better icon to use--maybe some type of
> clock or watch? Molasses? A Bunny/tortoise?

Ok, this no longer matters (I think) since there is no longer a separate 
icon - you just use the standard envelope icon.  Unless people want the
speed envelope in a different colour to the volume envelope?  That's easy
to do - just let me know.

> * You may have gone over this with Dominic already, but it seems to me
> that per-track sample-rate adjustment can't be done by warping the
> ruler.  Actually, I don't even think that project-wide sample-rate
> adjustment should be done this way. To me, the ruler as a measurement
> device is sacrosanct--An elastic ruler doesn't do a lot of good.

You can now see both the warped and the unwarped ruler at the same time -
hopefully that makes things at least a *bit* nicer for you ;-)

> I personally think that the track samples should do the stretching.

As do I... in a *perfect* world.  Unfortunately this stuff is hard, and we
haven't even started on the back-end yet!  My plan at the moment is:
1. Make a GUI that works.
2. Get the back-end working.
3. Enable separate envelopes for each track.
4. Re-visit the GUI and see if wave-stretching is possible/practical.
There are lots of reasons why wave-stretching, if possible, is the best
thing to do.  One of the main reasons I can think of is that if your waves
get stretched, you can visually line up beats, so you can *see* when
two tracks are in time with each other.

> Plus, playing with your warping code got me all confused, because I kept
> thinking that moving the envelope up was slowing stuff down since the
> ruler was expanding.

Hopefully being able to see both rulers will make this a bit better?

> But I think per-track adjustment is preferable to project-wide
> adjustment, and that will only be possible with track-specific warping.

Agreed.  I'm just trying to do the *easiest* thing first, and build up to
the Right Thing.

> But, there are apparently two main problems with stretching the samples
> rather than the ruler: (1) the redrawing code might be too slow, and (2)
> wierd graphical movement might occur while adjusting the rate.

and (3) it's more programming work.  IMHO, (3) outways (1) and (2) since
I'm currently the only person working on this ;-)

> Anyway, I personally don't think it is worth your while to continue
> working on the ruler warping.

I think that's going a bit far! ;-)
I accept that ruler-warping probably isn't the best way to go in the long
term, but it's there and (for me at least) it works.

> Others may not agree, and my opinion is not that important. But, I think
> that you may have greater success by  figuring out an efficient way to
> drawing a track according to a 'warped' mapping.

Your opinion is very important to me, as one of the target users of this
feature.  Thanks for the comments, and for testing.

Best wishes,
		Bill.
-- 
Dr. William Bland.                          Computer Programmer, UK.
www.abstractnonsense.com