Portamento Pain

[rbowser-]

I've tried to use Portamento in both JABB and GPO many times, but the results are always so totally unpredictable that I always have to give up and live without it. I end up doing all sorts of editing gymnastics with the pitch wheel to approximate the effects I'm needing.

I've gone through Tom's tutorial demonstrating Port with strings--Even imported his posted MIDI files so I could analyze them. I managed to get some semi-acceptable results, but I still had to semi-drown out those Port passages since every time I played the file the slides would be different.

This morning I've spent hours trying to get Port working with a JABB Trombone. Same inconsistency as when I work in GPO. Often there is no slide at all, other times when I play a file, the slide will be gigantic. Sometimes it slides up, sometimes down. Sometimes it plays the second note in a Ported series so it's the same pitch as the first one. All sorts of squeaks and squaks--I get that, or absolutely nothing.

Mostly, it just doesn't work. When I stop a sequence, and I'm at a point where Port is at zero--From that point on, all Port data will be ignored by the instrument. Only by playing a project from the beginning will the Port data kick in, but--completely inconsistently as I'm saying.

I'm using Portamento, cc20, in conjunction with notes tied via cc64.

Since I've had so much trouble with this feature, I have to say it's the one thing about the Garritan libraries that I find fault with. Every time I try Port again, I long for the old days when Portamento was just a switch flip away on hardware sequencers. You asked for Port, and there it was, gliding perfectly between notes the way you programmed it. I'm sure it must be more difficult to have Port work with samples--and considering what I've said here, it seems to me it's still not been solidly figured out.

Clues anyone?

Randy
(rbowser)

[Davidart]

Hi rbowser,
I also spent much time trying to get it to work, i have it working ok now.
basicly the portameno between note1 and note2 is the sequence of messages thus:-

cc64,127
note1
cc20,100 <- the greater the value the smaller the interval
note2
cc64,0
cc20,0

the value also depend on note durations.Its best to experiment.
This is just basic and there are other ways to get portamento.

David

[Tom Hopkins]

Randy,

Your description sounded like what I would expect if you hadn't confined yourself carefully to just the legato mode for portamento. But you say you *have* confined it to the legato mode. Kontakt is not "smart" enough to keep track of the direction or intensity of portamento data across groups. So, portamento (cc20) must be confined to just one group. Because legato is the most common situation for the use of portamento, that is the chosen group. For situations that are not amenable to portamento data, pitchbend must be used instead. That's why the tutorial dealt with the use of cc19 switching to defeat pitchbend on selected notes. I have not noticed unpredictability if these restrictions were carefully followed - although it wouldn't surprise me greatly if it was revealed that Kontakt had additional issues with this. It also wouldn't surprise me if behavior varied within different host environments or OS platforms either. In fact, nothing would surprise me anymore! But, I have not experienced it on my system.

The unusual requirements of portamento application are an unavoidable consequence (within Kontakt) of the complexity of the programming which uses multiple groups to accomplish many important things. And yes, it is very much more difficult to use portamento with samples than with synthesis for many reasons, not the least of which is the resultant formant shifting in samples of real acoustic instruments. We hope that this aspect of the libraries will become more successful and easier to use in the future (without giving up other features, of course.)

As to trombone: Personally, I use only pitchbend data for slide work. I don't believe I used portamento on any of the demos but I did use plenty of pitchbend especially between select smaller intervals to mimic typical, but usually subtle, slide movement. I think software for pitch adjustment that includes formant control would be another very interesting area to explore for trombone but I haven't tried it yet.

Tom

[rbowser-]

Thank you very much, both Tom and David.

I shall calmly try it all again. I've been using cc64 in combination with cc20, but perhaps not always placing the values where they need to be--?--I suppose that's possible.

--Sorry to say I haven't been able to make the cc19 pitch defeat work either. I was trying pitch bend on trombones yesterday, trying to stop the second note from being effected--and the results were so hit and miss that I ended up Very carefully erasing the pitch bend data so it was re-set to 0 for the second note. Sort of worked.

I'm sure I've been slipping up in some way. I'll magnify the views as wide as I can and take extra precautions to insert data where it's supposed to be!

Thanks again.

Randy B.
(rbowser)

[rbowser-]

Thank you very much, both Tom and David.

I shall calmly try it all again. I've been using cc64 in combination with cc20, but perhaps not always placing the values where they need to be--?--I suppose that's possible.

--Sorry to say I haven't been able to make the cc19 pitch defeat work either. I was trying pitch bend on trombones yesterday, trying to stop the second note from being effected--and the results were so hit and miss that I ended up Very carefully erasing the pitch bend data so it was re-set to 0 for the second note. Sort of worked.

I'm sure I've been slipping up in some way. I'll magnify the views as wide as I can and take extra precautions to insert data where it's supposed to be!

Thanks again.

Randy B.
(rbowser)

[Tom Hopkins]

Randy,

One thing to be extra careful about: the cc19 data must precede the start of the note intended to be turned on or off (PB defeat.) I tend to be generous in the space I allow ahead of the note start. Buffer settings can affect this so it's best not to cut it too close. This is a good rule for all switching controller data (sustain pedal, keyswitches, and cc19.)

Tom

[Davidart]

Tom!

Appreciate your help.

David

[rbowser-]

Thanks much for the additional advice, Tom. This could very possibly be it--that I'm parking some control data too close to notes so my system doesn't have enough processing time by the time the note hits. Interesting--I'll do more experimenting!

Randy
(rbowser)

[DPDAN]

I think software for pitch adjustment that includes formant control would be another very interesting area to explore for trombone but I haven't tried it yet.

Tom

Hi Randy, I believe your DAW just might already have some type of audio pitch automation/correction like Digital Performer does. Here is an example of a GPO Cello KS section recording with no gliss, some gliss, and a big slide.

Maybe you can record your trombone to an audio track, and use your pitch automation to draw in the portamento like I did here. I like using pitch automation as opposed to pitch bending in the midi track because the formants are maintained with pitch automation. Formants.. (keeps things from sounding like the chipmunks or Darth Vader) :)

Dan
Click here to listen (https://www.mydocsonline.com/pub/DPDAN/gliss%20test.mp3)
first phrase no portamento, second minimal, third big slide

screenshot of minimal pitch bend, red data is what I drew in.

https://www.mydocsonline.com/pub/DPDAN/portamento%201%20pic.jpg

screenshot of big slide

https://www.mydocsonline.com/pub/DPDAN/big%20slide%20pic.jpg

[DPDAN]

Randy here is a GPO t-bone slide (https://www.mydocsonline.com/pub/DPDAN/tbone.mp3).... with just midi pitch bend, with and without reverb.

https://www.mydocsonline.com/pub/DPDAN/midi%20pitch.01.jpg

[rbowser-]

Thanks, Dpdan

The audio pitch shift in the Sonar program I have isn't up to that task. The Formants are Very distorted. Pitch shift/bend in my Sound Forge is much better, but I really don't have any audio pitch bender which doesn't mess with the formants to some degree.

I find MIDI pitch bend much easier to work with. I'd rather have the performance data as much in the MIDI tracks themselves as possible, even though I do eventually work strictly with bounced audio tracks when I'm ready for mixing.

I noticed on your MP3 clips, that there was quite a loud click/whoomph sound at the end of the bend, as the new pitch settled in. Notice that?

This thread started because I was hoping to use the Portamento ala cc20 as programmed in the Garritan libraries--it's just much more "fiddley" than I'd prefer. But I'll certainly be trying it some more. Also trying the cc19 approach of stopping a pitch bend from effecting a second note. Whatever one does to get these effects, it's just going to be some careful editing, and that's fine, I'm certainly used to hours of editing.

Thanks again for all the screen shots and info.

Randy
rbowser