Encore - Swing Playback - removal

[Doug Kerr]

Ho, p,

...

Note: Align Playback corrected the start times, but not the durations.

My observation also. This is a bug. It used to work OK.

What did it used to do to the durations?

Best regards,

Doug

[polarbreeze]

Ho, p,

...

Note: Align Playback corrected the start times, but not the durations.

My observation also. This is a bug. It used to work OK.

What did it used to do to the durations?

Best regards,

Doug

It put them back the way they were I suppose. I never had cause to investigate whether that meant restoring them to their previous (raw) values or inserting durations "as written" but for my purposes the effect was the same: the swing was effectively cancelled.

[polarbreeze]

JUST PLEASE MAKE IT WORK THE WAY IT DID BEFORE IT GOT BROKEN!

[Doug Kerr]

Hi, p,

I wrote:

What did it used to do to the durations?

You wrote:

It put them back the way they were I suppose. I never had cause to investigate whether that meant restoring them to their previous (raw) values or inserting durations "as written" but for my purposes the effect was the same: the swing was effectively cancelled.

Thanks.

It might well be the latter. I don't know either. Maybe it worked on 4.5.5. I'll see what it did there.

Best regards,

Doug

[Denkster]

Well, I just checked what Swing Playback and Align playback did in Enc4.5.5Win.

1. In Encore 4.5.5, I created this 'score' again,
2. In the first measure I entered clean, default length eights.
3. I copied measure 1 to measures 2 and 3.
4. Applied Align Playback, to measure 3.

encore_swing-playback-test.png (10.77 KiB) Viewed 1320 times

This file, I saved as MIDI, and opened in Master Tracks Pro. Result:

Encore 4.5.5 Windows

encore_swing-playback-test-enc4.5.png (13.67 KiB) Viewed 1320 times

Compare with the result of the same procedure with Encore 5.0.2Wb3:

Encore 5.0.2 Windows beta 3

piano-roll-swing-playback-test.png (13.49 KiB) Viewed 1320 times

As far as I can see, there is no difference.
Can somebody please send me a score as pictured above, made the exact same way, in Encore 4.5.6.4 for Mac?

Thanks.
Eveline

[Doug Kerr]

Here's the same story done in Encore/Windows 4.5.5:

Swing_test_401-comp.gif (16.19 KiB) Viewed 1329 times

Evidently here again Align Playback preserved the play durations as it found them (162 ticks/72 ticks).

The original durations were all 108 ticks, as we would expect.

Note that, as in Encore 5.0.2, the second notes of each pair are "hyperextended" by the application of swing; they persist past their allocated real estate (and thus encroach on the following note).

This has probably been buggy for a long time.

I have no idea what happens/happened in the Mac version.

Best regards,

Doug

[Doug Kerr]

Hi, Eveline,

Well, I just checked what Swing Playback and Align playback did in Enc4.5.5Win.

Thanks so much. (I didn't know you were going to do that!)

Your findings are parallel to mine (just posted).

Best regards,

Doug

[Denkster]

Apparently, Encore 4.5.5Win did not what Polarbreeze expected either, as is confirmed once more, with below image (same procedure, but now with 75% swing)

swing-playback-test-enc4.5.5W-75%-PianoRoll.png (13.49 KiB) Viewed 1328 times

[Denkster]

Hi Doug,
We are doing the same things at the same time
Sorry.
I saw your posts only after hitting 'Submit' after my last post.

Your graph, more clearly than mine, shows that Swing Playback for 75% swing causes the notes to overlap.
That is bug number one.

Align Playback does not change note durations, only note start times, so it can not undo Swing Playback, as the manual states.
That manual error is bug number two.

To hear 'proper' straight eights again, from a score where 'Swing Playback' has been applied, we need to apply 'Align Playback' and 'Change duration' to 90% (or another value to your liking). Of course, this way we lose the original note start times and durations, as recorded in the underlying MIDI Data.

To be able to undo Swing Playback, and getting back the original MIDI Data, to conditions must be met:

1. the Swing Playback function uses a reversible algorithm to calculate and alter the note start times and note stop times from the original note start and note stop times
2. we need a new function 'Undo Swing Playback' (or a name like that), that applies the inverse algorithm.

Note: an algorithm like

if x > 100, y=50, else y=x/2

is not reversible. We can find a value for y, if we have x, but we will not be able to calculate x if we have only y.

kind regards
Eveline

[q]

I ran your test here Eveline. Same result on Encore 5.04 Mac. Lots of orphaned durations. And it's quite clear, in measure 3, if these were unisons Align Playback would create lots of MIDI straddle the associated clipped notes:






Thanks for you tests too Doug!

q

[Doug Kerr]

Hi, q,

Nice work. Thanks.

Could I trouble you to do the same thing with the swing parameter at 75%. (Make sure the original durations are not 100%; 90% should work fine.) I suspect you'll see some "incipient MIDI straddles" after just applying swing.

Thanks so much.

Best regards,

Doug

[q]

Could I trouble you to do the same thing with the swing parameter at 75%. (Make sure the original durations are not 100%; 90% should work fine.) I suspect you'll see some "incipient MIDI straddles" after just applying swing.

Here you go. Mac with Swing 75%. The notes were click entered, so durations are 90%.

And your prediction was correct. Bad news for swung unisons!







q

[Denkster]

Thanks a lot q,

At least it is a relief to see all versions act the same..

All graphs shows that Swing Playback for 75% swing causes the notes to overlap.
That is bug number one.

Align Playback does not change note durations, only note start times, so it can not undo Swing Playback, as the manual states.
That manual error is bug number two.

To be able to undo Swing Playback, and getting back the original MIDI Data, two conditions must be met:

1. the Swing Playback function uses a reversible algorithm to calculate and alter the note start times and note stop times from the original note start and note stop times
2. we need a new function 'Undo Swing Playback' (or a name like that), that applies the inverse algorithm.

Eveline

[Doug Kerr]

Hi, q,

Here you go. Mac with Swing 75%. The notes were click entered, so durations are 90%.

And your prediction was correct. Bad news for swung unisons!

Thanks so much. Evidently the code issues are not platform specific.

I hope that when Matt goes to work on "just put it back" he can figure out what that was. (As to the swing algorithm, I have a suspicion. But I am without a clue what "better" behavior of Align Playback might have been available in the ancien régime.

Part of the issue here is to imagine what Align Playback was originally for. I suspect it was originally for fixing up measures where editing had left the noes with anomalous start times. I suspect we have hardly any of that anymore.

In that regard, if what used to go wrong with start times during extreme editing conserved the initial durations of the notes, then for Align Playback to just do along with that makes sense.

Then a secondary use has presumably been, after import of a passage via MIDI, to make the timings match the inferred ("guessed" ) notation. It is more problematical to conclude what it's behavior should have been with regard to durations. But having no way tp discriminate as to the project for which it was being used, it followed the same rule I declared above "reasonable" in a different setting. I think it would have been impossible to define another rule for that. The only thing I can think of would be for it to put the values to the default initial duration fraction. Then, if that's not what the scorist wanted, he could adjust them (just as for a score entered directly).

Now we have something that is more of an "undo". That's a different matter altogether (as you have discussed).

But, "when all you have is a hammer, everything looks like a nail."

In any case, part of the clue is in the name of the function, "Align Playback". To me alignment doesn't necessarily suggest "improving on" the sizes of the objects being aligned.

Best regards,

Doug

[Doug Kerr]

Now, back to "unison" sequences and the infamous MIDI straddle:

Swing_test_117-comp.gif (10.94 KiB) Viewed 1382 times

Here I put harmony notes with the second notes of each pair as "markers". These allow us to see how the notes were moved even though they overlap in the "unison" line.

We note that the inappropriate Swing Playback algorithm (operating in the regime of initial durations not 100%) "overextends" the second notes of the pair. The result is a MIDI straddle in the unison notes.

As a result, the actual acoustic output only occurs up to the time of the end of the marker note (I have illustrated the mechanism of this in detail elsewhere). This is indicated on the second pair; it occurs in all pairs.

As before, we see that Align Playback does what I think it was intended to, not useful in "reversing" the swing.

Doug