K2 portamento/glide script

[synergy543]

Martin's suggestions are excellent and his observations on the volume are particularly intriguing and I agree with his point very much. I wonder though if one volume envelope would cover all situations? For example, during the course of a single performance, sometimes you might want a subtle effect with a strong volume envelope change (say for larger intervals); at other times you might want no volume change - essentially the way it is now (maybe for shorter intervals). I would think having the choice among several envelopes might be desirable so it would be great if the 'amount' of volume change were programmable. So maybe at a low MIDI CC there might be an extremely pronounced volume change but at a higher MIDI CC there would be almost none. This would also be a good place to add the MIDI CC 'switch' so at the lowest MIDI CC, the portamento is completely turned off.

One possible downside of the above is it could generate an excessive amount of MIDI data from this controller relative to the function. An alternative would be if you could assign a few consecutive MIDI notes to act as keyswitches where the lowest key would be portamento off and as you progressed upwards, the volume effect would go from 'strong volume envelope' to no-effect. However, a continuous MIDI controller might be more musically useful and worth the 'overhead' - unused controller steps could easily be thinned or deleted for sequencing.

Maybe a linear/exp envelope contour setting might also be useful? Maybe something like the velocity curvewhere the slope is adjustable?

Martin, forgive me if I'm distorting or misrepresenting your idea but just wanted to share these thoughts which came to mind after reading your interesting suggestion which put into words so well what I was feeling but couldn't quite place my finger on.

[Big Bob]

Thanks for the input guys, but of course most of this will be better situated in the new thread.

I will however point out something about your wishlist ideas. Obviously, each of you have certain very specific things you're trying to do and of course you'd love to have a script that would automatically solve some of the problems attendant to what you're now doing by other means. This is a good thing because 'necessity is the mother of invention' and all that jazz. However, what I hope we will be able to accomplish when we get the new thread going is to reduce a lot of such 'wishlist' items to a much smaller set of basic tools that will make it easier to do these things but not necessarily do everything that everybody wants in a fully automatic way. I'm hoping we will not only express specific needs but also attempt to arrive at a more general set of control functions that can help with (but not necessarily totally automate) those needs.

For example, extracting one of Martin's needs:

It would be great if the script could be applied only to certain groups of an instrument, and not to others. The main reason this would be needed, at least for me, is for instruments that have release triggers with the resonance of the hall (e.g. SAM brass or EWQLSO Gold). Right now, the script cuts the release trigger right at the end of the slide, when the next note begins. This really doesn't sound good, as the reverb tail of the hall gets cut abruptly

Now I read that alongside one of Synergy543's wishes:

Here's for the wishlist:

I would like to be able assign the script to keyswitches to turn the glide on/off while I play. I think this would be extremely useful.

And, then I say to myself, wouldn't Synergy's requested enable/disable function also solve Martin's cut off tail problem? Yes, I realize that you would have to insert the control code in the sequence (or activate it with your nose or something in a live situation) and that of course is not fully automatic, but, it would get the job done.

The problem with fully automatic things that do very specific functions is that we end up with either a whole bunch of special scripts or one giant script that does so many things that it may become bewildering to use. One advantage of the multi-tiered approach to designing this script will be that others will be able to write new, special-purpose scripts by only altering the higher levels of the code. Since this script will be well documented and 'open source', we can expect others to take up the gauntlet. Think of it like creating a variety of 'skins' for the same 'innards'.
************************************************** ******

Now, returning to the main topic of 'to DFD or not to DFD', maybe a compromise is in order. Instead of starting out with no DFD support, I could provide some limited support that would endow DFD with comparable-quality glides but at the expense of more resources.

Since everyone is already familiar with how Jay's script works, I'll use it as a clarifying example. Let's refer to the function of Jay's script as a Glider. Once the first note is played, each subsequent 'connected note' that's played becomes the 'current note' to further bend from (if desired). When the last played note is released, the script 'resets'. Let's call this bunch of connected notes from the first played until the last one is released a 'Bend Phrase'. When using the new 'bender engine' as a Glider, in DFD mode it becomes necessary to specify (in some way) the maximum bend range before the first note is struck. The range would generally be considered to be the maximum interval (plus or minus) from the played note that you can bend. But suppose that you glide from C upward to G, sit there a while and then glide from G upward to the next G (an octave up). If this is all part of the same Bend Phrase, the initial bend range has to be specified as 19 semitones. This will require that 19 * 2 = 38 notes be silently started with the first note.

To reduce some of the polyphony demands, I was originally going to code the engine so that the initially specified range would always apply to the last played note of a bend phrase (rather than the first played note). To do this, each time a new note is played, some of the silent notes would be terminated and other new ones would start up. This would then allow the example 'bend phrase' to be played with a polyphony of 2 * 12 = 24 instead of 38 notes. Of course. But implementing this 'dynamic range' scheme is rather tricky and time consuming. If I leave this sort of thing out (at least initially) the script will be more resource hungry in the DFD mode, but, it will operate with comparable quality to that of the Sampler mode. This is the sort of 'compromise' I'm thinking of making.

[synergy543]

Would there be a way for the script to select from two different samples?

For example, the first one can have a full attack and the second could have its start time and envelope adjusted to bypass the onset of the attack to create a smoother legato.

[Big Bob]

Sure enough, the 'inside samples' could be played from a different, keyswitched group as an example. The only problem with this idea is that in order to use the script, you first have to edit the instrument to cooperatively provide the needed, late-start samples. It's something to think about though. Of course the best answer would be for NI to make the offset parameter work in the DFD mode, no?

Right now, I'm fighting some new KSP/K2 problems relating to DFD mode for the super bender engine. It could be a bug in my code, but, so far it looks more like under certain conditions (involving the simultaneous startup of lots of samples), the KSP causes K2 to sort of go bananas and drive the CPU usage meter to the peg even after all the active voices have gone to zero! It's a weird problem and it's driving me nuts! But then, no one ever said this was going to be easy .

Bob

[Theodor]

Hello Bob,

I also work in DFD mostly but whenever i make a Fizbending patch i load them in Sampler mode.

Cheers,

[synergy543]

I also work in DFD mostly but whenever i make a Fizbending patch i load them in Sampler mode.

Why? I thought DVD was 'good'?

[Big Bob]

Why? I thought DVD was 'good'?

Probably so he can use the sample-start offset parameter that doesn't work in DFD mode.

I'm just guessing of course, so let's wait for Theo to respond.

Bob

[Theodor]

Probably so he can use the sample-start offset parameter that doesn't work in DFD mode.

I'm just guessing of course, so let's wait for Theo to respond.

Bob

That's exactly right Bob :- )

[synergy543]

Thanks TK and BigBob.
OK...here's my stupid question of the day:

How do I change a Kontakt patch from DFD to sample?

Is this it?:
Instrument Options>DVD & Load>DFD Preload Buffer?
Set to zero

Or is it in the Kontakt 2 Options>DFD
Set to zero

Do I need to save and reload?

[Diriger]

Click the wrench button to expand the instrument, then look around the top left of the green section. Just to the left of the word 'Source' you'll see a button that will enable you to choose the mode.