Skip to main content ↑ ↓
Topic: bug report (Read 6682 times) previous topic - next topic

bug report

I have transcribed a phrase in common time consisting of a quarter rest, followed by four 16th grace-notes and then two eighth-notes, then a half-note. Why is it that Noteworthy plays back the first grace-note simultaneously with the eighth-note that follows the grace-notes? If this is a known issue, how is it fixed?

Mal

Re: bug report

Reply #1 –
I see this bug has been reported:
"NWC treats all grace notes as occurring on the beat. This is (more or less) correct behavior for music before Beethoven, but not for more recent styles."

Noteworthy is a poor performance software, but I guess I got my money's worth for software that is over 200 years out of date.

Re: bug report

Reply #2 –
The response to this same bug report from over a year ago:

"NWC treats all grace notes as occurring on the beat. This is (more or less) correct behavior for music before Beethoven, but not for more recent styles"

does not even accurately reflect the bug. It was never the style that the first grace note is played simultaneously with the big note that follows. I understand that I have to try to fix the problem myself with 32nd notes played not as grace notes. But the fact is, this bug has been present for years and years, the software designers know about it, and nothing has ever been done.

Re: bug report

Reply #3 –
What you are saying is that in NWC1.75, under some circumstances (grace notes), the printable appearance and the MIDI export do not coincide. There is, of course a work-around, and I imagine that you've discovered it yourself. If not, try layering (this allows one staff to have invisible notes that sound, with a companion staff having visible notes that don't sound).

Re: bug report

Reply #4 –
You can put it in technical language if you like, but all I'm saying is that this is an acknowledged bug that has never been fixed and should be fixed. The purpose of playback is to hear what the printable music sounds like, and in this case it does not play back correctly, not even close to it. The simultaneous playing of the first grace-note and the main note is not right even "before Beethoven's time," to the contrary of one tech's suggestion over a year ago.

Re: bug report

Reply #5 –
A Modest Proposal*

Since there seems to be such disagreement about how grace notes should be treated, what durations they should occupy in time, how much time they should steal from the preceding and subsequent notes, whether or not there should be any overlap with the preceding and subsequent notes, and other such sticky aspects for which it's impossible to achieve consensus, I suggest that grace notes in NWC be done away with entirely. All notes should be spelled out exactly as they are to be played. After all, NWC has the facility to do tripletted double-dotted sixty-fourths, any musician worth his salt should be able to read these.

* cf. Jonathan Swift

Re: bug report

Reply #6 –
A modest reply*

No doubt the music would play better if the richer notes were allowed to keep more of their time, rather than having grace notes tax them so much.

* cf. Bush tax plan.

More seriously:
Let us, for the sake of discussion, suppose that everyone agrees that the NWC1.75 timing of grace note playback is not the way it should be.

My point is that the way NWC permits layered staves, and hidden notes, allows for an easy work-around. The work-around permits the user to achieve virtually ANY playback effect, without disturbing the printed appearance of the score.

To get grace notes to play earlier (for example), the auxiliary staff would have the preceding regular note written with shortened duration. The grace notes, written as regular notes of very short duration, would intervene prior to the note they modify. All of this would be hidden but sounded. meanwhile, the printable notes on the real staff would have properties specified as visible, but muted.

In summary, it is not that NWC1.75 cannot do grace notes the way users would like them. It is that the default behavior is not the way users would like them, but this can be user-modified.

Re: bug report

Reply #7 –
Re: A modest proposal

Grace notes come in handy in emulating a struck chord on a guitar, although there are certainly workarounds.

I wonder if options could be offered as to the grace notes (played on the beat, after the beat, before the beat, etc.) much as in the case of text expressions and the like?

Re: bug report

Reply #8 –
As someone pointed out in another thread about grace notes (can't seem to find it now), even if NWC played them "the right way," there would be many cases where the user might need to modify them anyway, in order to get a more "musical" interpretation.
So, the sound is fine.  Just use a hidden/aural staff and a visual/muted staff, as mentioned above.
The problem I have with grace notes is that the accidentals are the same size as regular notes, not reduced in size.  Also, I use grace notes for cues, but half- and whole-notes cannot be "graced" in NWC.  Too bad for me.

Re: bug report

Reply #9 –
Chris: I use cue notes (for singing) all the time, but I don't use grace notes for the reasons you mentioned.

One of my free music fonts has pre-formed notes (all kinds) that can be placed as text and scaled the way you like. If you can't find it on the Scriptorium, the topic is mentioned in another recent thread.

Re: bug report

Reply #10 –
One problem with using text items for cue notes is that eighth notes cannot be beamed together.  Another problem is the lack of the ability to tie notes together.  And then there's the leger line problem.  And lest we not forget the accidentals problem.  And the slurs problem.

Re: bug report

Reply #11 –
I've waited a couple days to see if further replies even come close to understanding my point, but it seems hopeless. A grace note should not be played *simultaneously* with the note it modifies, like a chord. This is not merely a matter of style as some would suggest, but an egregious and disappointing problem with the program itself. It's not a question to be answered by dancing around the theory of grace notes. Typically, they take half the time of the preceding note. But they are NEVER played simultaneously with the note they modify, as Noteworthy playback has it.

Re: bug report

Reply #12 –
The nearest I could get to reproducing a "chord" effect was to insert a pedal mark before a grace note when using a non-percussive instrument - e.g. clarinet but the first grace note definitely plays alone before the following note.  It DOES sound rather odd but then my clarinet doesn't have any pedals!

Could the feature that Malenor describes be hardware-related?

Stephen

Re: bug report

Reply #13 –
A grace note lasts the same length of time as a demisemiquaver (AFAICT) so if you put twelve of them before a crotchet then four of them actually sound after the note has finished.

But I agree it would be nice to have some control. You can fake it to some extent by tying grace notes together (two tied ones take the same time as two untied but only sound once. If you make all but the last invisible then it looks right on print-out too.

Re: bug report

Reply #14 –
The hanging question is, “Does the grace note play simultaneously with the note that follows in NWC?”  I tried a little experiment on my equipment that gives a resounding answer of “NO!”, at least for my system (Windows 98, Soundblaster sound card) and for single grace notes.  I made the top staff two bars of 32nd notes on the beat followed by double-dotted 8th note rests, MIDI channel 10; that’s just two bars of very short CLICKS to establish the ictus (the BEAT).  The second staff had a bar of quarter notes and one of half notes, with all variations of grace notes that I could think of.  NWC allowed quarter, 8th, 16th, 32nd, and 64th  notes as grace notes, none of which had slashes thru them.  At the bar line, I placed the grace note at the end of the first measure with a slur to the half note that started the second measure.  I slowed it down to MM eighth note = 20, the slowest tempo available.  Each of the six examples clearly played the grace note FIRST, ON the beat, followed by the note that it was gracing (via a slur).  The duration of the grace note was the same for all 6 examples, regardless of the appearance of the grace note itself (from quarter to 64th) or that of the following note (quarter or half); it was roughly one-eighth of a single beat, making it roughly the duration of a 32nd note.  The grace note at the end of the first bar sounded on the downbeat of the second bar.  There was absolutely no hint of two notes playing at the same time.

Then I added a staff with the problem posed by Malenor, namely a quarter rest, four 16th grace notes, two 8th notes, and a half note.  The supposed “bug” did occur, exactly as described; the first grace note and the first 8th note sounded together on the second beat.  The reason, at least algorithmically, is found by successively removing the grace notes, one at a time.  With three grace notes, all three sound BEFORE the first eighth note, as expected; the timing sounds like four 32nd notes followed by the second 8th note.  With only two grace notes, the timing is like two 32nd notes (the grace notes) followed by a 16th note (the remainder of the first 8th note), followed by the second 8th note.  With only one grace note, the behavior is consistent with my first example.  It is pretty clear that the presence of four grace notes, each of which is to be played with the duration of a 32nd note, overrides the timing value of the poor 8th note that follows (the first one), and the only way that note can sound is to do it “au naturale”, as if the grace notes weren’t there!  Is this really a “bug”?  Doesn’t the composer have some responsibility to keep the timing demands within reason?  How would you suggest that the timing be corrected in order to have all 4 grace notes sound followed by the first 8th note, all 5 notes sounding within the duration of the single 8th note?  The limiting factor appears to be the 32nd note duration assigned to each grace note by NWC.

Pursuing this a little further, using 5 grace notes does, indeed, fit into my above explanation.  The first 4 grace notes sound simultaneously with the first 8th note, while the 5th and final grace note sounds with the second 8th note.  Converting the 8th notes to 16th notes (and adding an 8th rest) completes the analysis:  the first two grace notes sound with the first 16th note, the 3rd and 4th grace notes sound with the second 16th note, and the 5th grace note sounds alone during the rest.

The treatment of ornaments in general is almost a no-win scenario as far as trying to establish the perfect algorithm for the NWC program to follow.  For grace notes, each of the following is possible:  the duration of the grace note itself; whether it borrows from the preceding or the following note; whether or not is has a slash through it; whether it is accented or not; whether the slur (if used) is FROM the preceding note or TO the following note; whether the slur is across the bar line; etc.  A lot of these items have to do with the period and the style of the composer.  I’ve sung a lot of Mozart and Handel, for instance, and their grace notes would follow the NWC scheme to some extent, being accented, on the beat, and borrowing from the note that follows; these are a subset of grace notes in general, and are properly called appoggiatura’s; the MIDI durations are all wrong, however, as the appoggiatura generally uses HALF the note value of the following note.  (Modern publishers often write out the correct treatment above the staff for clarity of interpretation.)  That interpretation would be totally wrong in more “recent” compositions such as Sousa marches, for instance, where the grace notes come BEFORE the beat and borrow time from the preceding notes; these are often written across the bar line, which makes the intention of the composer quite clear.  There was a time when I thought that the slash thru the grace note would clarify its usage, but those days of idealism are long gone.  I recall playing a band piece last summer that had both slashed as well as un-slashed grace notes; at least that gave the band members a chance to be consistent; we just had to figure out which was which!  Fred has the right idea -- write the notes out the way you want them to sound!

Finally,  regarding cue notes,  it would be really nice if NWC could extend the process to include half and whole notes as well so that grace notes could serve double-duty as cue notes.  On many of my projects, I have to place text above the staff saying “Clarinet Cue Notes -- Do NOT Play”, etc.  Couldn’t we just have another option under “Notes” that would do essentially what “=” does for grace notes, but for all durations of both notes as well as rests?  I don’t really care how the MIDI playback treats these notes (play??  or  mute??), but the VISUAL representation would be a nice addition to an already fine program.

Re: bug report

Reply #15 –
To Stephen Randall:

I had originally posted the exact group of notes that created the chord effect. But I neglected to mention that the instrument I used was Acoustic Grand Piano, and the pedal-down notation at the beginning of the bar.

Here is the situation I described:

"I have transcribed a phrase in common time consisting of a quarter rest, followed by four 16th grace-notes and then two eighth-notes, then a half-note."

For what it's worth, the key is F major, and the run of 4 grace notes after the 1/4 rest is AGFE, followed by the main note 8th D, 8th C, half-note D, pedal-up, bar line. There is also a slur mark over the grace notes. Removing the slur mark has no effect on the playback error. (No, it's definitely not a tie.)

This is the first time I've ever heard the chorded effect since I first started using this program back in, oh, 1999.

Re: bug report

Reply #16 –
Thank you PeterJon for that cogent analysis of the "bug." I doubt if I could have worked it out on my own. The solution to the problem would be for the program to take a little more time away, a 1/32, from the preceding rest to make up for the extra grace note(s), don't you think? In this case, it would require a 1/32 rest on the bottom line just after the 1/4 rest. But I think that would be beyond the capacities of the software. I imagine it would require a ground-up rewriting of the thing just to fix it.

Hearing the song actually played by the composer would be best. But with grace notes the effect I commonly hear is for the beat to take a short break to make room for the grace notes to play. Thus they are not restricted to a half-beat or whatever the time of the song requires.

I guess I expected too much from Noteworthy Composer, and yet it was simply not expected from a program that is so miraculous and provides me many hours of enjoyment.

Re: bug report

Reply #17 –
PeterJon:

If I told you the piece was late 20th-century New Age, would that help determine the proper interpretation of those 4 grace notes? Wouldn't they then take the time from the preceding beat? Does the fact that they are 16ths indicate the amount of time to be taken from the beat? Does it indicate that there should be a slight pause at that point in the beat as is often the case with modern pieces?

Re: bug report

Reply #18 –
Never mind, Pete. I worked out the grace-note problem based on your educated reply. Then using the hidden staff feature I retained the original look of the song.

Mal