MakeMusic Forum

Home | Log In | Register | Search | Help

MakeMusic Forum > Public Forums > Finale - Windows - FORUM HAS MOVED! > HP Issues Continue in 2014.5

Forum Quick Jump

[ << Previous Thread | Next Thread >> | Show Newest Post First ]

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/21/2015 12:26 PM (GMT -6)

I have reported many HP issues to MM over the years, and some of them have been addressed. However, when the repeated note bug continued in 2014.5, I thought maybe we could at least document the issues by testing and reporting here so someone could then contact MM (maybe me) with a complete list.

For starters, the repeated note bug continues in 2014.5. When HP is set to incorporate Start/Stop time MIDI Data, the issue appears. Since the default is to Incorporate, many users report here from time to time that notes don't repeat. This affects VST exclusively, as I recall, and not all samples. My advice is to set that item to Ignore. However, this might have an adverse affect for other staves and samples, so it might not be ideal.

What's interesting is that applying the HP playback plugin (now changed to a MIDI Tool feature), the issue doesn't seem to appear. Thus, I think that since only live HP playback is affected, it might be easier to isolate the cause and fix it. I think this is true for some of the other HP issues as well. I haven't seen that cure the glissando bugs, at least not all of them. It does when it comes to the black key glissandos in piano music, but not to all the harp gliss problems.

I hope others will experiment and report their HP problems here so we can get a fix-it list going.

I have just tested a new workaround for the start/stop MIDI data bug. For notes entered in Speedy and Simple, all the values are uniform. I tried leaving on to Incorporate the MIDI data and then, in the MIDI Tool, choose to set the notes to 98% of their length. This seemed to work, and didn't create an audible difference, at least on Strings. Maybe others can confirm my finding.

Anyway, if we can document the continuing problems, we might get them fixed.

If I'm the only one with the problems, please set me straight. Also, this might be a Windows only thing, so any Mac users can compare Windows findings to what happens on the Mac and let us know if it is different.

Remember, if a file is posted and tested, the HP Preferences settings must match to be of comparative value, so the settings used by the poster should be mentioned. The settings can be hard wired into a file. However, in the past, I have found that loading a file with hard-wired preferences messed with my stock settings, so that might not be the best way to go.

Zuill

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

Post Edited (Zuill) : 11/21/2015 11:30:10 AM (GMT-6)

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/21/2015 12:49 PM (GMT -6)

Here's a simple example for starters. I get stuttering or muting when the start/stop values are the default step entry values and HP preferences are set to Incorporate the Start/Stop MIDI Data. However, in the second system, I have applied the MIDI Tool to reduce the length of each note to 98% of its length. This seems to resolve the issue.

Do others confirm my finding? This was done in 2014.5 but can also be tested in 2014.

Zuill

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

File Attachment :
Start Stop Time MIDI Data Bug.musx 89KB (application/octet-stream)

This file has been downloaded 271 time(s).

teacue
Registered Member

Date Joined Aug 2009
Total Posts : 85

Posted 11/21/2015 3:50 PM (GMT -6)

Trying to understand what is the problem you encounter, I downloaded your file.
The file plays in my system as it should without special HP setting.
As soon as I set Incorporate Data then stutterings or muting appear.

To better see the midi data, I exported the tracks as midi and imported them in my daw (Cubase).

The first thing that I notice is that all notes are (midiwise) a little bit longer than a quarter note.
It means the end of a midinote always overlap the beginning of the next one.
I don't know how used you are to daws and to midi but I can assure you that midi always has a problem with overlapping identical notes.
It means most of the time the overlapped note will not be played at all and the first one will stop.
I use midi since midi exists (so back in the 80s) and I vaguely remember that at the beginning it was different but since several decades Cubase acts this way.
So when I write repeated notes I always take care that there is at least one tick between two notes.
This behaviour also occurs with chords.
I would say it comes from the concept itself of a midi note which consists of a note on event and a note off event.
A C3 note off event will always stop a note on C3 event if it happens after.
And that is exactly what happens with repeated overlapping notes.
One should always take care with repeated notes that the note off event of the first note occurs before the note on event of the second note.
Only an ingenious and special mechanism that can distribute notes to the available polyphonie could achieve that two repeated notes are played even if they overlap.

In my view it is quite amazing that Finale is able to play your file without problem and without special HP setting.
I assume that some ingenious mechanism is working under the hood to achieve this!
The HP option "Incorporate Data" means to me that the midi data are then played as they really are, meaning with the overlapping value, and this is as I explained above what almost no Daw is able to safely play.
In this case FInale behaves as any Daw and stutters or so when repeated notes are overlapping

When you reduce the value to 98 % it is likely that the midi values do not overlap any more, this way they can be played without problem.

It could be of course that I absolutely do not understand what your problem is!
In this case, sorry and please forget what I wrote.

Best regards

Finale 2014.5, Windows 8.1 - 64 bit

Post Edited (teacue) : 11/21/2015 2:54:47 PM (GMT-6)

Gareth Green
Player of fine trumpets

Date Joined Oct 2001
Total Posts : 2943

Posted 11/21/2015 5:39 PM (GMT -6)

This is a wild guess, and I apologise if it wastes anyone's time, but somewhere in the playback settings there is an option to "randomize" notes. If this were selected, it could cause the effects described ... ???

Gareth J. Green

Fin2014c
Windows 7

Charles Lawrence
Registered Member

Date Joined Dec 2009
Total Posts : 3638

Posted 11/21/2015 5:47 PM (GMT -6)

I entirely agree with teacue on this one. I too have dabbled in MIDI since the early '80's. Anybody remember the Amiga and Bars 'N Pipes, a DAW before there were DAW's?

What he says is entirely accurate about one MIDI note's on event overlapping the same pitch's previously started note off event. Since the MIDI data stream is serial and a given instrument is on the same channel, the events are processed in line on the same channel. If you start one note, say a C4 and don't turn it off until after starting a second C4, then the result is that when the first C4's note off event is processed, then all sound for the C4 is stopped, even though you haven't shut off the second C4. Thus you get the "repeated note bug" as Zuill calls it. In reality, there is no bug, its just the way MIDI is designed to work.

As to why Finale creates these overlapping notes is another question. Perhaps its a quantization issue.

"Anything is possible if you don't know what you are talking about!"

Intel(R) Core(TM) i7-5930K CPU with 6 dual core processors @ 3.50 GHz (12 threads)

32 GB RAM

Realtek High Definition on board audio
NVIDIA GeForce GTX 980 GPU with 4GB ram

512 GB SSD system disk

Four 4TB and one 1TB internal SATA HD's
Microsoft Windows 8.1 Professional with Media Center x64 Edition, (06.03.9600.00)
Finale versions: 2011b.r2, 2012c.r13, 2014d.v5030

GPO4.02

Cakewalk SONAR X3

"There is a world of difference between a person who has a big problem and a person who makes a problem big." – John Maxwell

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/21/2015 6:05 PM (GMT -6)

The point is, before 2014, the HP problem didn't occur. This behavior is new to 2014. The MIDI data is what either Speedy or Simple enter when inputting. When preferences are set to Ignore the start/stop time, there is no problem. Also, when hard-wiring with the HP "plugin", the problem doesn't occur. It is only a real time issue. I would rather not turn that preference off, but without modifying the default lengths, there is this problem. That's why I did a 98% length reduction in the 2nd system.

By the way, MM has already acknowledged the issue, as far as I recall from my discourse with them. This started in 2014 and continues to 2014.5. We get posts quite regularly here and I suppose I could do a search and find those threads for you. And, by the way, this doesn't affect MIDI playback. Just VST.

Zuill

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/21/2015 7:05 PM (GMT -6)

Okay. Here's the same file in 2012 format (legacy). If you have 2012 and 2014 and 2014.5, and you have HP preferences set to Incorporate Start/Stop MIDI Data, you'll say that it plays fine in 2012 but not 2014 or 2014.5. I refer to the first system, as that has the Speedy/Simple default note lengths.

So, between 2012 and 2014, something got broken.

Many of you might remember when I reported here on the forum about Layers not playing back MIDI data properly in 2014. You'll see that in the readme file, that was one of the fixes. I guess I didn't report enough about the Start/Stop data bug.

Zuill

P.S.: From the readme file:

"Human Playback/MIDI tool

• MIDI key velocity information is now played back in all layers.

• Stop and start values are now interpreted correctly by Human Playback."

That second item I think was referring to all Start/Stop Data. It was not addressing the specific issue I am reporting.

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

Post Edited (Zuill) : 11/21/2015 6:09:01 PM (GMT-6)

File Attachment :
Start Stop Time MIDI Data Bug 2012.mus 43KB (application/octet-stream)

This file has been downloaded 320 time(s).

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/21/2015 10:21 PM (GMT -6)

How about Harp Glissandos? Here is a 2012 file created by Michel R. E. some time back when we were discussing 2014 glissando problems (brought up by Ron). Try it in 2012, then try it in 2014 and then 2014.5.

Zuill

P.S.: For those that just want to hear the difference, here are the audio generated by 2012 and 2014.5. The 2012 file plays properly. The 2014.5 file is extremely lame. What was done to mess things up? This was first discovered in 2014 but continues in 2014.5.

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

Post Edited (Zuill) : 11/21/2015 9:30:48 PM (GMT-6)

File Attachment :
harp_gliss_demo 2012.mus 42KB (application/octet-stream)

This file has been downloaded 318 time(s).

File Attachment :
harp_gliss_demo 2012 in 2012.mp3 184KB (audio/mpeg)

This file has been downloaded 128 time(s).

File Attachment :
harp_gliss_demo 2012 in 2014.5.mp3 185KB (audio/mpeg)

This file has been downloaded 95 time(s).

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/21/2015 10:41 PM (GMT -6)

Another glissando situation brought to our attention by dfwmd a bit of time ago.

The original file (2014 now resaved in 2014.5 by me) has playback issues. In the revised file, HP is set to none and the HP Playback plugin was applied. This is now a hard wired feature in the MIDI menu. You can see that HP correctly interprets glissandos when it hard wires them into MIDI data, but live playback is interesting, if not funny.

Zuill

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

File Attachment :
Glissando Test 2014.5.musx 86KB (application/octet-stream)

This file has been downloaded 273 time(s).

File Attachment :
Glissando Test Revised 2014.5.musx 88KB (application/octet-stream)

This file has been downloaded 286 time(s).

Ralph L. Bowers Jr.
Polymathist

Date Joined May 2012
Total Posts : 883

Posted 11/21/2015 10:52 PM (GMT -6)

Thanks for posting the audio demo Zuill....doing our leg work for us..... That second audio demo is horrible.
This needs to have a fix.

Finale 2010b, 2011b, 2012c, 2014d, TGTools Pro, Patterson plugins, JW plugins (current)
Sibelius 6.2, 7.1.3, 7.5.1, 8.0.1, Write Score Sound Sets, TMT Publisher Bundle Plugins, Bob Zawalich plugins, Dolet 6.4
Print Music 2004, 2010a, 2011a, 2014a
Progression, Progression 2, Progression 3, Notion 4, [Notion 5, (bought but not installed)]
Pro Tools 9.5, Reaper
Kontakt 5
GPO4, World Instruments
SmartScore X Pro, SmartScore X2 Pro, PhotoScore Ultimate 6 & 7 & 8.04 ( 7 has some utility----best of those available, 8 has some issues that need fixing)
M-Audio "Oxygen 25" Midi input keyboard (recent addition 2014)
Systems (4) // Windows XP Pro, 2@ Windows 7 Pro, 8.1 Pro, 64 bit, 4GB - 16GB RAM
Paper & Pencil

BMus, MM (Musicology)

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/21/2015 10:54 PM (GMT -6)

Recently tzalia on the Mac forum posted a file that caused the volume to progressively drop every time a fall-off occured. After a quarter rest, the volume restarted at the proper level.

You can see in my revised file it plays back properly because HP is set to none, and the data is hardwired. Unfortunately, the applied HP playback had the erroneous data that progressively dropped the velocity data on the notes to create the same fade-out effect the original file displayed (that's how I discovered the method by which HP was misbehaving). I went into the MIDI Tool and cleared that erroneous data.

Zuill

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

File Attachment :
ArticulationTest.musx 87KB (application/octet-stream)

This file has been downloaded 307 time(s).

File Attachment :
ArticulationTest Revised.musx 87KB (application/octet-stream)

This file has been downloaded 259 time(s).

nordensten
Registered Member

Date Joined Jul 2000
Total Posts : 419

Posted 11/22/2015 9:02 AM (GMT -6)

I can confirm the START/STOP time bug in 2014.5 - and it also happens when Finale is in MIDI mode!

In 2012c merely switching on Incorporate, has no additional effect in itself.

In 2014.5 (and probably any 2014...) switching on Incorporate has a DISASTROUS (and seemingly random) effect on MIDI NOTE OFFs on REPEATED notes. (This is easily observable graphically by recording the finale output into any sequencer/DAW and open the piano roll view)

There are (at least) two workarounds: either 1) adding an articulation that affects Note Duration on the repeated notes (This will, BTW, often enhance the "naturalness" of the playback anyway) or 2) use the Midi tool to reduce the note length as Zuill describes.

(Finale 1.0 (1988) -> 2014d) using 2012c (on Win since 1997)
Cubase 8 Pro & Reaper - Vienna Ensemble Pro - Ircam SPAT - Hauptwerk 4
Libraries: VSL, Spitfire, Orchestral Tools, Xsample, Embertone, Sample Modeling, EWQL +++
Garritan Steinway - Garritan CFX pianos
Multiple Windows computers W7/W8/i7/32GB

Post Edited (nordensten) : 11/22/2015 8:09:39 AM (GMT-6)

Ralph L. Bowers Jr.
Polymathist

Date Joined May 2012
Total Posts : 883

Posted 11/22/2015 10:56 AM (GMT -6)

Zuill as you suggested clearing the MIDI date and reapplying Human Playback corrects this issue.

But as you have stated we should not need to do this with even legacy files.

File Attachment :
harp_gliss_demo 2012 (2014.5 corrections).musx 89KB (application/octet-stream)

This file has been downloaded 256 time(s).

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/22/2015 2:28 PM (GMT -6)

I followed your advice and the result was not the same. Which style did you apply?

Zuill

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

Ralph L. Bowers Jr.
Polymathist

Date Joined May 2012
Total Posts : 883

Posted 11/22/2015 2:37 PM (GMT -6)

I moved (copied) my Functioning "Human Playback" (fxt) plugin from my F2012c plugins folder to the Plugins folder in F2014.5 and used that to make corrections to the offending files.

Worked for me.

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/22/2015 2:42 PM (GMT -6)

No wonder I didn't get the same result. MakeMusic should be reading this thread and pack their 2012 Apply HP plugin in 2014.5.

Thanks for the workaround. Too bad they seriously broke HP, which was not that great to begin with.

Zuill

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

Ralph L. Bowers Jr.
Polymathist

Date Joined May 2012
Total Posts : 883

Posted 11/22/2015 2:53 PM (GMT -6)

Zuill said...
No wonder I didn't get the same result. MakeMusic should be reading this thread and pack their 2012 Apply HP plugin in 2014.5.

Thanks for the workaround. Too bad they seriously broke HP, which was not that great to begin with.

Zuill

I don't see why they can not document and make available the old fxt in the knowledge base?

Zuill
"The Troll"

Date Joined Oct 2003
Total Posts : 29077

Posted 11/25/2015 5:43 PM (GMT -6)

I just submitted a support case regarding the harp gliss problem and the suggestion of the use of the 2012 Apply HP plugin. I suggested they make that plugin available to non-2012 users who wouldn't have that already until they get 2014.5 fixed. I also suggested they read the posts about the HP issues here for more items needing fixing. I'll probably send support cases about those as well.

Zuill

"When all is said and done, more is said than done."

Win 7 64bit, 2011b, 2012c, 2014d, 2014.5

Favorite Forum quote: "Please, everybody, IGNORE THE TROLL!"

gogreen
Registered Member

Date Joined Dec 1999
Total Posts : 1618

Posted 9/14/2016 8:20 AM (GMT -6)

I can confirm that the repeated note bug exists in Finale 2014.5 and in Finale 25.

Does the human playback plugin work in Finale 2014.5 and Finale 25? I don't have Finale 2012, but I do have Finale 2011.

Arthur J. Michaels

/www.facebook.com/composerarthurjmichaels

Finale 2000, 2001, 2003, 2005, 2006, 2007, 2011b.r2, 2014d, 2014.5, Finale 25
Core i7 860 @ 2.80 GHz, 8.0 GB RAM, Windows 10 Home x64
Garritan COMB2, GPO4, GPO5, Aria Player 1.872, Audacity 2.1.2
Dell 2408 WFP, 1920x1200
M-Audio Delta Audiophile 2496
Casio WK-3000
M-Audio AV40 monitors
AKG K-240 Studio headphones
Brother MFC-L2740DW printer

nordensten
Registered Member

Date Joined Jul 2000
Total Posts : 419

Posted 9/14/2016 11:04 AM (GMT -6)

Some have better luck than others, then... on my system, the repeated note bug is fixed in V25, but not in 2014.5. By all means, there are still several HP bugs that need mending. Side bar: so far I haven't seen a single (playback related) crash testing the V25 demo ... I'm seriously considering to pay for the upgrade now. (Tobias and Jari, please hurry :))

(Finale 1.0 (1988) -> 2014.5) using 2012c (on Win since 1997)
Cubase 8.5 Pro & Reaper - Vienna Ensemble Pro - Ircam SPAT - Hauptwerk 4 - Overture 5
Libraries: VSL, Spitfire, Orchestral Tools, Xsample, Embertone, Wallander, Pianoteq, Sample Modeling, EWQL +++
Garritan Steinway - Garritan CFX pianos
Multiple Windows computers W7/W8.1/i7/32GB

Forum Information

Currently it is Tuesday, December 19, 2023 7:06 PM (GMT -6)
There are a total of 403,820 posts in 58,165 threads.
In the last 3 days there were 0 new threads and 0 reply posts. View Active Threads

Need Support? | Forum Help Manual