Boundary problems with graphic fade -Sound Forge - PC

Subject:Boundary problems with graphic fade Posted by: jlo Date:4/29/1999 7:08:09 AM Suppose you have a file with 3 contiguous regions, and you want region 1 to be 6dB softer, region 3 to be 12dB softer, and region 2 to fade down in some particular way from the region 1 level to the region 3 level. So you might do the following: * select region 1, then process->volume -6dB * select region 3, then process->volume -12dB * select region 2, then process->fade->graphic, and carefully tailor your fade so that the endpoints are exactly at -6dB and -12dB, either with dither or not. Unfortunately, something goes wrong at the boundaries of region 2, so that you get a click passing in and out of the region. If the region 2 boundaries are not at a zero crossing, you can see it when you zoom in or use the statistics tool; if they are, you can't see it but you can hear it. Smoothing at the boundaries helps a lot, but is a little hacky. I suspect that SF defines regions with both endpoints inclusive, so that region 2's endpoints are getting processed doubly by the graphic fade and the adjacent regions' volume adjustment. Then again, maybe this is the wrong way to do something like this, although I'm not sure how else to do it (graphic fade alone doesn't give you enough detail regions 1 and 3 are long compared to region 2). Ideas and help would be appreciated, James

Suppose you have a file with 3 contiguous regions, and you want region 1
to be 6dB softer, region 3 to be 12dB softer, and region 2 to fade down
in some particular way from the region 1 level to the region 3 level.

So you might do the following:

* select region 1, then process->volume -6dB
* select region 3, then process->volume -12dB
* select region 2, then process->fade->graphic, and carefully tailor
your fade so that the endpoints are exactly at -6dB and -12dB,
either with dither or not.

Unfortunately, something goes wrong at the boundaries of region 2, so
that you get a click passing in and out of the region. If the region 2
boundaries are not at a zero crossing, you can see it when you zoom in
or use the statistics tool; if they are, you can't see it but you can
hear it.

Smoothing at the boundaries helps a lot, but is a little hacky. I
suspect that SF defines regions with both endpoints inclusive, so that
region 2's endpoints are getting processed doubly by the graphic fade
and the adjacent regions' volume adjustment. Then again, maybe this is
the wrong way to do something like this, although I'm not sure how else
to do it (graphic fade alone doesn't give you enough detail regions 1
and 3 are long compared to region 2).

Ideas and help would be appreciated,

James