Skip to main content ↑ ↓
Topic: Exported MIDI Malformed (Read 2620 times) previous topic - next topic

Exported MIDI Malformed

We’ve just encountered a problem we’ve never seen before in the many years we’ve used Noteworthy.

When we export a MIDI file from the attached NWC file (SleepMyLittleOneSleep.nwc), the resulting MIDI has two problems:

1. The MIDI file is twice as long as the original NWC file.

2. The second half of the MIDI is silent (no audio).


I assume something’s wrong with the NWC file, but its appearance and playback seem completely normal.

This is a huge show stopper for us. Can anyone tell what’s wrong?

Re: Exported MIDI Malformed

Reply #1 –
Interesting  :)
The reason is in the MPCs: They have an "Initial Time Offset" of 127 = their behavior starts about 127 eighths after the end of the piece. Playback and MIDI export seem to take this literally and extend the file up to that point.
Please note that playback lasts also for a much longer time (1:21) than until the final chord of the piece (about 0:36).

H.M.

Re: Exported MIDI Malformed

Reply #2 –
Ok, I changed the first two time offsets to 0 and 3, but the exported MIDI file seems to be the same. Maybe the problem is that I need to specify when the volume change ends? I read the NWC help file several times, but don't see how that’s done.

Adding to my confusion is that the NWC help file for the Multi-point Controller topic is outdated: It refers to fields that don’t appear in the MPC dialog, such as TimeOffsetMultipler & Linear Sweep Contour. And the dialog has a field that isn’t mentioned in the help file (Time Resolution).  I have NWC version 2.75a.2.

So I’m groping in the dark by trial & error.

Do you see anything wrong with the MPC definition now?

Re: Exported MIDI Malformed

Reply #3 –
The MPC in the lower staff has still Time Offset 127! ... and I get the feeling that you do not really know how MPCs work? - then, please, just delete them. The tiny decresc. at the end is not worth the hassle, I'd say  :D

Especially with your selection of instrument, Pad (4) choir / patch 91, this seems unnecessary: As one can hear in measure 4, there is a decresc. for longer notes built into that instrument anyway; so the additional effect in the last measure is not really audible.

But if you want them, some guidelines:
- The Initial Time Offset is practically always 0 - i.e., the effect starts at the point where you place the MPC.
- A typical MPC only works if at least Setting 2 is checked - otherwise, there is no target value.
- Linear Sweep, and a Sweep Resolution of 1, are typically selected. Other values here need to have a special reason.
- Typically, I limit the effect of a MPC to a single measure. So, in 6/8 piece, I would set "Time Resolution" to "Eighth", and "Setting 2 | Time Offset" to 6 (at most).
- Also, for a decresc at the very end, I'd set the "Setting 2 | Controller Value" to 0, i.e. total silence, to make the effect really audible.
When I do all that in your score, it does exactly what it should - a nice decresc at the very end, and a play time of 0:36!

H.M.

Re: Exported MIDI Malformed

Reply #4 –
You were right! All I had to do was change the corresponding MPC in the base clef, and it works perfectly now.

Actually, one of the reasons I wanted to pursue this was so I could learn about MPC’s & make better use of them in the future.

If you’d like to see this one in action on a live Web site, check out this URL:

http://www.hymntime.com/tch/htm/s/l/e/e/sleepmlo.htm

This Web site has been online for 27 years now, and has over 15,000 NWC & MIDI files. Last fall, we began adding OGG files as well.

Thanks for the help (and for being so patient!)   O:)

Dick