16 Jun
2004
16 Jun
'04
4:52 p.m.
Version 2.6.0 of the cost free musical score editor NoteEdit is available:
http://rnvs.informatik.tu-chemnitz.de/~jan/noteedit/noteedit.html
New features:
+ The main improvements concern the MIDI import:
- distribution of MIDI data onto multiple
voices per staff during MIDI import.
This is especially useful if your MIDI file
contains the equivalent of such a score:
|\ |\ |\ |\ |\ | |\
| | | |\ | | |
---|--/----/----/----/----/.-------/----/---|----
---|----------------------------------------|----
---|----------------------------------------|----
---|----------------------------------------|----
---|--\---------\-------------\.--------\---|----
| | | |
| | | |/
- Drum note recognition
- Stacatto recognition
- there is a new slider: "Volume step" to determine
the volume difference of 2 adjacent notes needed
to insert a volume signature (default: 10)
This is important, because otherwise NoteEdit
sets a volume signature between all adjacent
notes who's volume differs. If the MIDI file is
recorded from MIDI keyboard this is practically the
normal case.
+ Almost all is exported to ABC music including:
- drum notes
- guitar chord diagrams
+ export/import of octaviation (va lines) (Tanks to Leon Vinken)
+ Note shifting by mouse wheel:
- Wheel + Ctrl: shift note by 1 line
- Wheel + Shift: shift note by 1 octave
- Wheel + Shift + Ctrl: shift note by 1 semi tone
+ Cursor up/down + Ctrl: shift note by 1 semi tone
+ bug fixes
--
J.Anders, Chemnitz, GERMANY (ja(a)informatik.tu-chemnitz.de)
16 Jun
16 Jun
5:12 p.m.
New subject: [linux-audio-user] NoteEdit 2.6.0: MIDI-->multiple voices per staff
On Wednesday 16 Jun 2004 7:43 pm, Joerg Anders wrote:
Version 2.6.0 of the cost free musical score editor
NoteEdit
Any particular reason you say "cost" free? That phrase was enough to
make me go off and check it was still under the GPL!
+ The main improvements concern the MIDI import:
- distribution of MIDI data onto multiple
voices per staff during MIDI import.
Is this based on pitch?
Chris
6:07 p.m.
New subject: [linux-audio-user] NoteEdit 2.6.0: MIDI-->multiple voices per staff
On Wed, 16 Jun 2004, Chris Cannam wrote:
Is this based on pitch?
I'm not quite sure whether this gives always good results
(or at least in most cases.)
It is a mix: I divide the midi data into chunks. A junk is
a portion of adjacent midi notes at leats 4 quarters long.
The start and stop of every chunk is a non-overlapping moment
(I hope it is always possible to find such a moment in a
distance shorter than - say - 32 quarters):
|-- event 1 ---| |--event2---| |-event3--| |---|
|-----event 4-------| |--event 5-| |-event6--|
^ ^
| start of chunk; first possible end of chunk |
After that I create a fully meshed graph with all events as
nodes:
|------- e1 \----------
| / | |
| -- e2--|-- e5 ---- |
| | | \ | / | | |
| | | / - \ | | |
|--|- e3 -|-- e6-----|-|
| \ | / /
\ --- e4 -------
For every edge I compute costs in both directions (thus
every egde are actually 2 edges):
if (MidiOnTime(j) - (MidiOffTime(i) < 0) {
costs(i, j) = infinity
}
else {
costs(i, j) = PITCH_DIST_COST_FAC * |(pitch(i) - pitch(j))| +
START_DIST_FAC * (MidiOnTime(j) - (MidiOffTime(i))
}
(see miditimescale.cpp: initialize_desicion_tree)
Whereby PITCH_DIST_COST_FAC and START_DIST_FAC are empirically
choosen values. Thus, the notes close together regarding the
pitch and the time distance get the lowest costs.
Then I select 2 unmarked notes:
first note: The note with lowest midiOnTime;
last note: The note with higest midiOnTime
and I compute the shortest path from first note to last note
using Dijkstra's shortest path algorithm.
I mark all notes from first to last note belonging to one voice end
repeat the selection and the Dijkstra algorithm till all notes are marked.
Thus, every note is assigned to a voice.
Till now I sort the voices. Again I use a cost function:
cost (voice_i) = PITCH_FAC * sum_of_all_pitches / num_pitches + COUNT_FAC * num_pitches;
(see miditimescale.cpp: sort_voices)
The voice with highest costs is the first voice. The formula shall
ensure the voice with the higest pitches is the first voice but
avoid that a voice with only 2 notes becomes the first voice and
a voice with 1000 notes becomes the 2nd voice.
Perhaps it is an error to sort the voices after the Dijkstra algoritm. I'll test
other possibilities.
As a last (or better first) step I have to take into account
the triplets. Unfortunately, the triplet recognition does not
yet work (is commented out). But for successful triplet
recognition it is certainly necessary to recognize the
triplets before the voice distribution takes place.
I had an algorithm but it failed in these case:
3
-----------------
| |
|\ |\ |\ |\ |\
| | | | |
| | | | |
/ / / / /
\-----/ \-----/
--
J.Anders, Chemnitz, GERMANY (ja(a)informatik.tu-chemnitz.de)
17 Jun
17 Jun
6:58 a.m.
New subject: [linux-audio-user] NoteEdit 2.6.0: MIDI-->multiple voices per staff
On Wednesday 16 Jun 2004 8:58 pm, Joerg Anders wrote:
The formula shall ensure the voice with the higest
pitches is
the first voice [...]
A nice algorithm. Presumably this is intended for uses like choral
music where there is a clear division into parts and each part is
essentially monophonic -- I can imagine it might not work so well
with something like a piano played in a single MIDI track, where the
parts can have quite different "shapes" (e.g. widely separated chord
clusters in the left hand, running scales in the right).
How do you decide whether to run this algorithm at all? I mean, how
can you know at the outset whether a track is expected to contain
one, two, or several voices? I imagine you could find a division
into "voices" of pretty much any MIDI track with chords in it.
(pause while I try out the feature)
Yes, it looks like it runs on every MIDI file -- so if I import piano
music, I might get two staffs with eight or nine voices on each. Not
ideal for piano music, then. That's not too problematic because the
default mode in NoteEdit appears to be to edit all voices at once,
right? Can I merge selected voices as well?
I have several MIDI files that cause a crash on import, with an error
saying '"NMidiTimeScale::findPathsInChunk" Error Code is: 1' -- do
you want any examples? It'd be nice if it abandoned the voice
algorithm but continued to import the file.
btw, how does NoteEdit quantize on MIDI import? Is it simply to a
grid, and if so, is the grid configurable?
Chris
9:34 a.m.
New subject: [linux-audio-user] NoteEdit 2.6.0: MIDI-->multiple voices per staff
On Thu, 17 Jun 2004, Chris Cannam wrote:
I have several MIDI files that cause a crash on import, with an error
saying '"NMidiTimeScale::findPathsInChunk" Error Code is: 1' -- do
you want any examples?
That means there are more than 9 voices necessary ;-)
I guess the problem is different. Before answering the other questions:
Is it possible the MIDI file contains piano pedal marks ?
--
J.Anders, Chemnitz, GERMANY (ja(a)informatik.tu-chemnitz.de)
9:44 a.m.
New subject: [linux-audio-user] NoteEdit 2.6.0: MIDI-->multiple voices per staff
On Thursday 17 Jun 2004 12:26 pm, Joerg Anders wrote:
I guess the problem is different. Before answering the
other
questions:
Is it possible the MIDI file contains piano pedal marks ?
Assuming you mean MIDI controller 64, then yes, it's very likely --
although I have successfully opened other MIDI files that also use
that controller.
Chris
9:59 a.m.
New subject: [linux-audio-user] NoteEdit 2.6.0: MIDI-->multiple voices per staff
On Thu, 17 Jun 2004, Chris Cannam wrote:
Assuming you mean MIDI controller 64,
Yes, excatly this!
then yes, it's very likely --
although I have successfully opened other MIDI files that also use
that controller.
Hmm, this is big problem. At:
http://rnvs.informatik.tu-chemnitz.de/~jan/noteedit/noteedit.html
now is the "tse3_sustain_patch.patch" that sould be applied
to the TSE3(!!!) library (tse3-0.2.7).
The background is: If the MIDI is:
|----|
|----|
|----|
|----|
|----|
|-----|
|-----|
Sustain On Sustain Off
The TSE3 library removes the Sustain On/Offs and constructs:
|----|
|---------|
|--------------|
|-------------------|
|------------------------|
|------------------------------|
|------------------------------------|
This is certainly good for MIDI players but bad for
sequencers and other related software, because
this can only be solved by creating 7 voices.
I already contacted Pete Goodliffe <pete.goodliffe(a)pace.co.uk>
and asked him to make this configurable. He promised to change this
in next version. Till then the patch is the only solution :-(
--
J.Anders, Chemnitz, GERMANY (ja(a)informatik.tu-chemnitz.de)
12:37 p.m.
New subject: [linux-audio-user] NoteEdit 2.6.0: MIDI-->multiple voices per staff
On Thu, 17 Jun 2004, Chris Cannam wrote:
music, I might get two staffs with eight or nine
voices on each. Not
ideal for piano music, then. That's not too problematic because the
default mode in NoteEdit appears to be to edit all voices at once,
right? Can I merge selected voices as well?
Just now I created Beethoven's Sonata No.8 in C-, Op.13 'Pathetique'
(530 measures) from:
http://www.stormloader.com/users/beethoven/midi.htm
The result is: 2 staff with 4 voices each.
I used the TSE3 --> Merge... feature because the MIDI file has
5(!) tracks.
(Unfortunataly, this feature doesn't work in noteedit-2.6.0. I'll fix
this together with Dave Phillips' problems)
The main parts are playable. The problem is the MIDI file:
Beethoven wrote:
|\ |\ |\ |\ |\ |\
---------|\------------|\------------|\------------|\------------|\------------|\---
---------|-------------|-------------|-------------|-------------|-------------|----
--|\----/-------|\----/-------|\----/-------|\----/-------|\----/-------|\----/-----
--|\------------|\------------|\------------|\------------|\------------|\----------
--|-------------|-------------|-------------|-------------|-------------|-----------
/ / / / / /
But on MIDI file is:
|\ |\ |\ |\ |\ |\
---------|\------------|\------------|\------------|\------------|\------------|\---
----\----|-------------|-------------|-------------|-------------|-------------|----
---\/---/-------------/-------------/-------------/-------------/-------------/-----
---/--------------------------------------------------------------------------------
------------------------------------------------------------------------------------
\ \ \ \ \ \
| | | | | |
| | | | | |
|-------------|-------------|-------------|-------------|-------------|
actually distributed over 2 staves. Of course, NoteEdit cannot comprehend this
and computes 2 voices :-(
The reason for the:
NMidiTimeScale::findPathsInChunk" Error Code is: 1"
is certainly the the TSE3 sustain problem, because the 'Pathetique'
is certainly a good test example.
How do you decide whether to run this algorithm at
all? I mean, how
can you know at the outset whether a track is expected to contain
one, two, or several voices? I imagine you could find a division
into "voices" of pretty much any MIDI track with chords in it.
(pause while I try out the feature)
Yes, it looks like it runs on every MIDI file --
Yes, indeed: But if all works(?) there shouldn't be unnessesary
voices. Imagine the events are so:
pitch 4: |---ev2--| |--ev6--|
pitch 3: |--ev1 --| |--ev5--|
pitch 2:
pitch 1: |--ev4----------|
pitch 0: |--ev3--|
| | | | | | | --> time
0 4 8 16 24 32 40
And PITCH_DIST_COST_FAC = 2 and START_DIST_FAC = 20 then
according to:
if (MidiOnTime(j) - (MidiOffTime(i) < 0) {
costs(i, j) = infinity
}
else {
costs(i, j) = PITCH_DIST_COST_FAC * |(pitch(i) - pitch(j))| +
START_DIST_FAC * (MidiOnTime(j) - (MidiOffTime(i))
}
the cost matrix is:
To: | 1 | 2 | 3 | 4 | 5 | 6 |
--------|------|------|------|------|------|------|
from: 1 |infty | 2 | 86 | 244 | 560 | 722 |
--------|------|------|------|------|------|------|
from: 2 |infty |infty | 8 | 166 | 482 | 640 |
--------|------|------|------|------|------|------|
from: 3 |infty |infty |infty | 2 | 326 | 488 |
--------|------|------|------|------|------|------|
from: 4 |infty |infty |infty |infty | 4 | 166 |
--------|------|------|------|------|------|------|
from: 5 |infty |infty |infty |infty |infty | 2 |
--------|------|------|------|------|------|------|
from: 6 |infty |infty |infty |infty |infty |infty |
--------|------|------|------|------|------|------|
The first event is evt1, the last evt6. As you can see the shorest path
from evt1 to evt6 is:
evt1 --> evt2 --> evt3 --> evt4 --> evt5 --> evt6
The sum is: 2 + 8 + 2 + 4 + 2 = 18
There is certainly no shorter path from evt1 to evt6
in the graph:
|------- e1 \----------
| / | |
| -- e2--|-- e5 ---- |
| | | \ | / | | |
| | | / - \ | | |
|--|- e3 -|-- e6-----|-|
| \ | / /
\ --- e4 -------
And this computes Dijkstra's shortest path algorithm. Because all
notes are marked in first loop there is no need for a 2nd
loop and all events belong to the one-and-only (and 1st) voice.
As explained above: (If all works correctly) the algorithm computes
multiple voices only if the MIDI events
overlap. If so, such score is produced:
|\
-----|--------------
-----/--------------
--------------------
--------------------
-----\--------------
|
|
How can I merge this ? Such a way:
|\
-----|----------------
----/|---------|\-----
-----|---------|------
-----|---------|------
----/---------/-------
\-----/
Ok, but how to merge this:
|\ |\ |\ |\ |\ | |\
| | | |\ | | |
---|--/----/----/----/----/.-------/----/---|----
---|----------------------------------------|----
---|----------------------------------------|----
---|----------------------------------------|----
---|--\---------\-------------\.--------\---|----
| | | |
| | | |/
btw, how does NoteEdit quantize on MIDI import?
Dynamically! The snap value is computed from MIDI note length.
--
J.Anders, Chemnitz, GERMANY (ja(a)informatik.tu-chemnitz.de)
18 Jun
18 Jun
1:41 p.m.
New subject: [linux-audio-user] NoteEdit 2.6.0: MIDI-->multiple voices per staff
On Wed, 16 Jun 2004, Joerg Anders wrote:
Perhaps it is an error to sort the voices
after the Dijkstra algoritm. I'll test
other possibilities.
Indeed, I changed this in version 2.6.2 and the result is much
better. The distribution now works chunk-wise.
Also the stem directions are better choosen.
I made a score from a MIDI file containing
Smetana's "Moldau" (17 staves, 450 measures): There
are only a few crossing beams because of bad
distribution of the notes or bad stem directions.
-----------
On Thu, 17 Jun 2004, Dave Phillips wrote:
It appears that changes to the MIDI input have killed
it. I tried
recording notes from my CZ101 but they were all crammed into a single
chord. NE then issued an error dialog that could not be removed, and I
ended up having to kill the app.
There was a big bug. I don't know how to connect a MIDI keyboard
to my Audigy2 sound card. I went to a box with AWE64 and got
the same error, too.
The MIDI record from MIDI keyboard should work (again) in version 2.6.2
------------
Furthermore the note insertion with "C", "D", "E", ... key
works (for the first time). The next note is now (as actually
planned) in a fourth distance.
--
J.Anders, Chemnitz, GERMANY (ja(a)informatik.tu-chemnitz.de)
7440
days inactive
7442
days old
linux-audio-user@lists.linuxaudio.org
8 comments
2 participants
participants (2)
-
Chris Cannam -
Joerg Anders