[directmusic] Re: wav tracks, a pain... solutions?

  • From: "Ciaran Walsh" <ciaran@xxxxxxxxxxxxxx>
  • To: <directmusic@xxxxxxxxxxxxx>
  • Date: Thu, 9 Jan 2003 18:18:00 -0000

I'm going to look into this properly before going too far down the road with
the streaming thing, cheers for the pointers Jason.

Isn't it possible to get sample accurate adjustment by using the time offset
to "coarse tune" and the sample offset to "fine tune"(I assume the time
offset you're referring to is in the Wave Properties in the Wave Track)? Or
does the time offset effectively override the sample offset? (I haven't
messed around with this stuff much yet).

Anyway, as you say, close enough is good enough for my purposes, but if it's
going to all get out of whack as it goes along, that's a different matter.

Is it just me or does this seem to be a really basic thing that should just
work? I'm very suprised...

-----Original Message-----
From: directmusic-bounce@xxxxxxxxxxxxx
[mailto:directmusic-bounce@xxxxxxxxxxxxx]On Behalf Of Jason Booth
Sent: 09 January 2003 15:26
To: directmusic@xxxxxxxxxxxxx
Subject: [directmusic] Re: wav tracks, a pain... solutions?




> By the way, I'm not using any compression at this stage - does this
have anything to do with it?

        I highly recommend you take an afternoon and play with whatever
compression option you plan to use as there are several major issues
with DMP and MP3 compression! For some reason, it always inserts a
glitch at 2 seconds into my wav track, and the wav-encoded mp3s also
insert a small amount of silence at the beginning of my track.
Unfortunately, DMP only allows you to offset the sample start point by
32000 samples or so - and because of the 2.1 seconds of tone I have to
put at the beginning of every track, I really can't use this option. My
solution, as it currently exists is as follows:

- Create a sine wav that's 2.1 seconds long with the tone generator in
Sound Forge
- Insert this tone before every track
- Load it into DMP, or another program and wav encode the track

        At this point, it should be possible to compute the difference
in length between the two wavs at a sample level. However, because you
can only offset a small amount at the sample level - you have to do your
offset in time (00:00:0000). This means your sync will always be a tiny
bit off, but within a reasonable error for most things.

- Set your offset to 2.1 seconds on the wav track
- Zoom in really close to the beginning
- Adjust the offset until you see the sine wav disapear

        If I really need things acurate, I create a click track the same
sample length as the main track (you have to do it for every length of
sample you create, since the amount of silence inserted seems to vary).
Then I play both the un-inserted, uncompressed wav track and the
offset-inserted sine wav, compressed version on top of each other and
listen for the phasing. However, because you can't get your offset down
at the sample level (because you can't offset enough at that level to
cover the 2.1 seconds of inserted tone), you'll never get the sync
absolutely perfect.

Other related posts:

  1. Home
  2. directmusic
  3. Archive
  4. 01-2003