PR4 Bug Report

[oldguy]

Hi Jo,

Still running PR4 - and I like it. Yesterday was the first time I created a complete composition. Found two bugs, one more minor than other:

1. Selected a lot of different sample bits (.wav), and then tried to assign them to one mixer. Did not work - only assigned one sample to the new mixer. This functionality works with the color, but not with the mixer assignment.

2. Am running three VSTi's each into their own mixer plus one line of samples into it's own mixer. I then pipe all four mixers to the mixer 1 for final effects prior to mixdown (reverb, eq, etc.)

The problem I notice is that when one mixer is attached to another, the signal is not going through in Stereo. I know this because I've been using the Kelly Stereo Tool to pan one of the premixers, but when it goes to the final mixer, I loose the pan. When I run a premixer direct to audio, it works. I took all the effects out to see if that was the problem...it wasn't. Is this a known bug?

Again, keep up the good work on an excellent product.

Cheers,

Oldguy.

[muzycian]

Hi Jo,

Still running PR4 - and I like it. Yesterday was the first time I created a complete composition. Found two bugs, one more minor than other:

1. Selected a lot of different sample bits (.wav), and then tried to assign them to one mixer. Did not work - only assigned one sample to the new mixer. This functionality works with the color, but not with the mixer assignment.

Indeed, that does not (yet) work.

2. Am running three VSTi's each into their own mixer plus one line of samples into it's own mixer. I then pipe all four mixers to the mixer 1 for final effects prior to mixdown (reverb, eq, etc.)

The problem I notice is that when one mixer is attached to another, the signal is not going through in Stereo. I know this because I've been using the Kelly Stereo Tool to pan one of the premixers, but when it goes to the final mixer, I loose the pan. When I run a premixer direct to audio, it works. I took all the effects out to see if that was the problem...it wasn't. Is this a known bug?

I downloaded the Kelly Stereo Tools, and messed around a bit with it, including a send to another mixer strip, and it all works fine.

Can you please describe exactly what you're doing, and/or email me a session file (as basic as possible please).

[oldguy]

Hi Jo,

My apologies for wasting your time with this. I played around with this and discovered there was one VST that was blocking the stereo output -
BS-spectrum by Bismark (http://homepage.mac.com/bismark/bs-spectrum/index.html).

Luna says that it's a two channel VST, but it appears to be mono internally. Inserting it into the chain converts the stereo to mono (favoring just the left channel). Probably not worth you looking into, just the way the VST was written. Interestingly enough, just bypass doesn't solve the problem - the VST must be completely removed from the chain.

Cheers,

-oldguy

[muzycian]

Hi Jo,
Luna says that it's a two channel VST, but it appears to be mono internally.

2 channels input or output?

[oldguy]

Hi Jo,

Again, I think it's probably the plug-in, not Luna. It shows as IO = 2/2. Don't waste your time on it.

Cheers,

-Oldguy.

[guitarfrodo]

Jo, two issues. Both could be something I've got wrong on my end.
1. Can't seem to get USB's new UVI rompler going in Luna. Works fine with Cmuzys, there's a 1.02 update, still doesn't help. It has this ILOK protection that needs to be installed alongside. Could this be the issue? Both my systems do the same thing. The vsti opens, it allows choosing a bank and patch but no sound is generated.
2.This second problem I have notice for a while but was waiting to see if as the midi end of things develops it fixed itself, so had not sent a comment before this. It pertains to import of prewritten midi from a specific stand alone a midi composition tool.
The problem shows itself when importing certain otherwise functional polyphony parts. When Luna plays the part the notes do not sustain properly even though the sequence editor shows them to be sustained. These play fine on all other hosts I've tried: (Cmuzys, Samplitude se, Live 4 lite, winamp, WMP 10) In Luna they come out stuttering or in places monophonic or have breaks of silence or partial polyphony. Very idiosyncratic.
If the host uses vsti, the problem exists whichever vsti is used to play. Seems more a glitch in Luna's use of the midi data.
I can send you a folder with a sampling of the offending midi files. These would all be midi exports of my own compositions created originally in tablature or tabs I've generated listening to an artist. I tried converting them from midi 1 to midi 0, still same play problem. Also tried importing the midi into midinotate and the re-exporting without positive effect. Midi created via a keyboard straight into Luna seems to work fine.

[muzycian]

Jo, two issues. Both could be something I've got wrong on my end.
1. Can't seem to get USB's new UVI rompler going in Luna. Works fine with Cmuzys, there's a 1.02 update, still doesn't help. It has this ILOK protection that needs to be installed alongside. Could this be the issue? Both my systems do the same thing. The vsti opens, it allows choosing a bank and patch but no sound is generated.

Could it be that the VSTI is not yet connected to the Audio Mixing Desk?

Check by selecting the USB synth in the Player Panel, then click Options, CDonnect Outputs.

2.This second problem I have notice for a while but was waiting to see if as the midi end of things develops it fixed itself, so had not sent a comment before this. It pertains to import of prewritten midi from a specific stand alone a midi composition tool.
The problem shows itself when importing certain otherwise functional polyphony parts. When Luna plays the part the notes do not sustain properly even though the sequence editor shows them to be sustained. These play fine on all other hosts I've tried: (Cmuzys, Samplitude se, Live 4 lite, winamp, WMP 10) In Luna they come out stuttering or in places monophonic or have breaks of silence or partial polyphony. Very idiosyncratic.
If the host uses vsti, the problem exists whichever vsti is used to play. Seems more a glitch in Luna's use of the midi data.
I can send you a folder with a sampling of the offending midi files. These would all be midi exports of my own compositions created originally in tablature or tabs I've generated listening to an artist. I tried converting them from midi 1 to midi 0, still same play problem. Also tried importing the midi into midinotate and the re-exporting without positive effect. Midi created via a keyboard straight into Luna seems to work fine.

Yes please email me a couple of such midi files and i'll check this out.

[guitarfrodo]

1. That did not seem to change anything. But the only option that the UVI player gave for VST output in that menu was 'UniversalPlayerOutput' where all the other players tried gave the option simply 'output.' The USB UVI hangup likely is at its end as essentially all other players interact fine, still thought I'd report in case some other user had an insight.

2. Midi's for the other problem should be over to you.

[muzycian]

1. That did not seem to change anything. But the only option that the UVI player gave for VST output in that menu was 'UniversalPlayerOutput' where all the other players tried gave the option simply 'output.' The USB UVI hangup likely is at its end as essentially all other players interact fine, still thought I'd report in case some other user had an insight.

I've got a problem with UVI player too: when i try to scan it, it just quits LUNA without any error message

Not so in another music app.

To be furtherly researched.

2. Midi's for the other problem should be over to you.

Yes, thanks for that

I've imported these midi files in LUNA without any problem.

Mmm, i've just reread your original description. Do these midi files still give the same problems? Is it only with these midi files?

We'll have to elaborate on this to find out more.

[guitarfrodo]

Jo, what I'll do is is make three mp-3 files, one with Luna, one with Cmuzys, one with Creative player all with the same soundfont and using SFZ for the sequencer player and e-mail you them. Snippets will be only a few seconds, hopefully you'll easily hear the difference.

[muzycian]

Thanks for the mp3s.

I think i found the issue: in the Bach Sarabande midi file, there are several notes of the same key that end just at the same position than when the new note on for that key starts.

And LUNA sends the note-on for that new note before the note-off, and that leads to the indicated problem.
(though pure technically it's not an error, but practically it's inconvenient)

Just as a test: shorten those "just-not-overlapping" notes just a little bit and the problem goes away, because then the note offs are explicitly sent out before the note ons.

Anyway, i've updated LUNA PR5 so that note offs are always prioritized against note ons

Thanks for pointing this out!

[muzycian]

1. That did not seem to change anything. But the only option that the UVI player gave for VST output in that menu was 'UniversalPlayerOutput' where all the other players tried gave the option simply 'output.' The USB UVI hangup likely is at its end as essentially all other players interact fine, still thought I'd report in case some other user had an insight.

I've got a problem with UVI player too: when i try to scan it, it just quits LUNA without any error message

Not so in another music app.

To be furtherly researched.

Seems to crash when a debugger is active, most probably because of that PACE thing, argh

Yes, when i plug the UVI in (without the debugger) it indeed doesn't make any sound

I'm still searching for the reason.

[muzycian]

Found: The UVI engine is another plug that doesn't follow the VST protocol, doing things in its own way.

Anyway, i've adapted LUNA's VST engine so that it handles this UVI quirck.

Check out LUNA PR5 as soon as it's available

[guitarfrodo]

Whew, you made quick work of both of those items. Thanks. In my relative state of ignorance it seemed the UVI engine was not following some standard vsti protocol.
And yes, moving the prior note off or the next note start over a fraction allows all of the notes to pickup properly. The fix you propose, I for one will appreciate, as the GP5 is my note by note composing tool of choice.