When this patch receives a harmony, it spits out the constituent notes in quick succession instead of all at once. If I play a C major chord, I will get a C, then an E will join it, and finally a G will pop in to complete the triad. I'm using a program that sends out all the notes at exactly the same instant.

[notein]
| \
| \
[poly 2]
| \ \
[pack f f f]
|
[route 1 2]
| |
[unpack 0 0] [unpack 0 0]
| | | |
[number boxes] <----- Here's where the problem appears. One box after another displays a note until the full chord is written out. It happens almost in the blink of an eye, but that isn't good enough for my purposes. I need the notes to arrive simultaneously.

Also, this is a little off-topic, but there's some kind of bug with the division object. It receives a 1019 in the left inlet and a 3 in the right inlet, and then declares that 1019 divided by 3 is 509.5 (it seems to dislike the number 3, as it will do the math correctly when it receives a 4). It only does accurate division when the numbers reach it by certain routes of an unknown nature or happen to be numbers which, for mysterious reasons, it is capable of dealing with. As far as I know, it shouldn't matter what routes the numbers reach it by, and there's certainly no reason it should cut a number in half when told to cut it into thirds.

Maybe because you're using poly 2 so you can't have a three or more notes polyphony, just 2 ... try at least a [poly 4 1] for chords.
As for the division question I have no idea. Isn't it a trigger problem? Are you using a [t f f]?

I'm using a 6-note poly with 6 [unpack]s and number boxes, and keeping the harmonies within 6-note limitations, so the problem isn't that there aren't enough places for the outputs to arrive at. I left out the other 4 voices because it was too much to write out. The problem isn't that there aren't enough places for the outputs to arrive at.

No, I'm not using [t f f]. I did figure out a solution to the problem, though. The problem stemmed from another case of messages being put out in super fast succession instead of all at once, so that when the division object was triggered it was receiving a wrong number to its left inlet, a fact which was concealed by the correct number arriving there in the blink of an eye, after the object had already been triggered and was finished doing its math.

Edit: To be clear, the main problem for which this thread was made has not been solved. Also, the above solution doesn't really work because for some reason it causes Puredata to play notes that were never sent to it.

Maybe if you post the patch we can investigate ...

anyway midi data don't happen at the same time but there will ever be a sequence of data so if you are listening it will sound ok but if you just want to see each note in a number box I reckon you have to split the list first

My first post gives all of the relevant information about the patch--the issue is occurring entirely within that compass.

Investigation with the list-lastx object reveals that [poly] or [notein] are sending along the notes in a very specific order: from highest to lowest pitch (I strongly doubt that [route] has anything to do with it, although I suppose it could). Normally poly sends out notes on the sole basis of the order in which they arrive, so unless it uses pitch height as a secondary sorting method when multiple pitches arrive together, the problem is most likely with [notein]. This makes it seem possible that the problem is, as you say, inherent to the way that MIDI data is transferred. However, I have slight doubts about that, because the note data is coming in slowly enough that I can sometimes hear notes in a harmony coming on one after another, whereas with normal MIDI playback in other programs, notes that are supposed to come on simultaneously always sound like they're coming on simultaneously. At the very least, PD is making the problem far worse than it should be, which makes me suspect it's creating the problem (I'm not sure that MIDI data being produced in succession would necessarily translate to a chord's notes not being played all at once. More likely it would translate to there being a slight delay before the full chord comes on. Just as MIDI would necessarily wait for information about a note's velocity to funnel through before producing a sound, I would expect it to wait for information about all of the notes in a chord to funnel through before making a sound, otherwise it would be playing something it wasn't directed to play).

I don't think the problem has to do with needing to split a list. [route] basically does that already. Maybe I'm not understanding what you mean, though.

EDIT: If [noteout] is designed to send note values prematurely, the only thing to do is to delay the arrival of its output to [poly] by a fraction of a second with a setup that will store all of the notes and then release them all at once when the time is up. I tried to use [speedlim], but the object doesn't seem to work correctly. It allows messages to pass almost immediately, and for some reason it keeps note-offs from arriving.

ok, good luck then,
sorry if my answer bothered you, just trying to help.

maybe put a timer after notein to make sure they are coming in at different times? also idk why this would matter but you could also try specifying the midi channel to receive on in note in
it really helps to have a patch when troubleshooting

@alistair_blunt: No, your answer didn't bother me. : )

@sebfumaster: I tried fooling with the MIDI channels, but unfortunately the results are still the same.

My thought was that I could have the notes sent to an object which stores them, and then after a certain time long enough for the correct pitch value to arrive, have the correct pitch isolated and sent along while the other notes are blocked (this is an adequate solution because what I need is not the correct pitch values as such but the correct transformation of them after they're funneled through my plugin). I'm not sure how exactly to do that, though. I'll need to search through the available objects for the necessary tools, unless someone here can tell me what to do.