fast repetitions and notation software

[DaveTubaKing]

I'm trying to get my head round how best to use the fast repetitions with a notated score in Sibelius. (so no midi keyboard involved - all the midi instructions come from the notated score)

I have a passage where all brass are tripletongueing at mm=120. So thats six semiquavers in two groups of three (tu-tu-ku, tu-tu-ku)

Using a staccato sample sounds expectedly naff.

Using the fast repetition patches with a midi keyboard you hold down the key once to trigger the relevant repetitions.

But with a score notated for publishing or playing live providing the midi, each seperate note is notated, so you get six rapid note on note off instructions per beat presumably triggering the fast repetition sample patch six times per beat.

With experimentation I have found that as long as the speed of the repetitions is slower than the speed of the notated notes then I get a reasonable performance using the repetition samples.

But what I want to understand is, if I am in effect re-triggering the repetition sample every note, am I getting the same individual sampled note or do the repetion samples have some sort of round robin built in so that each time the sample is triggered a different note from the set is played first (thereby adding the variety that avoids the shotgun effect)

To put it another way. When a fast repetition patch is triggered and re-triggered is the order of each sampled note in the repetition the same or is the order of each note in the repetition sample automaticaly varied each time the patch is triggered.

My ears tell me using the repetition samples sounds better than the staccato patch but my brain wants to know my ears aren't deceiving me.

[evan evans]

I'd be interested in hearing the answer to this post as well.

[JWL]

Sequencing for audio production and sequencing for eventual notation for live performance has always been a bit of a doppelgänger. I had to surrender eventually to the fact that I could never fully reconcile the two. The Cube has, however, enabled fewer notation edits with the convenience of having a diversity of articulations for one instrument or instrumental group (ie: violin I) on one single track.

If there were a way to accomplish everything required within a notation app using the Cube, I could do away with my DAWs entirely.

Indeed-- if someone has a solution to offer, this would qualify for "best thread of the year" imho!

Could it be this very concern that keeps DVZ in development? It was allegedly designed for this very purpose, but it keeps "not arriving".

[evan evans]

What's "DVZ"?

[plowman]

Google DVZ string.

Are fast-reps multi-sampled?
1. I've never seen any claim of VI fast-rep round robin playback. I know that EXS fast-reps did not have them.
2. Fast-reps are intended to be used precisely when staccatos lose their credibility at high tempo. And the main reason stacs lose credibility is because each is intended to be heard discretely, even as tempos rise. But fast-reps are more of a kinesthetic effect. A threshold has been crossed; we go from syllables to words.

Why do repeated, truncated fast-reps sound different than round-robined staccatos?
1. You may hear the traces of the second note within a fast-rep sample with only one note typed, depending on a notation program's default length. And this would be compounded with each next typed note. In other words, you type one note, but you're really hearing that note and one quarter of the next note -- but it's all so fast, you can't tell.
2. Staccatos don't have release samples. Fast-reps do. You may be hearing release samples under the multiple triggering when you expressly spell out each rapid note in a notation program.

Why do fast-reps sound round-robined even when they're not?
My guess is that you've got so much aural activity and overlapping sounds, the brain can't parse out the notes anymore, the machine gun has been obscured, and your psycho-acoustic vocabulary fills in the rest.

By the way, it's possible to build a VI sound with both stacs and fast-reps. Envelope (attack or delay) the latter to come in after the staccato speaks. With stacs round-robined and fast-reps completing the sound, you have the best of both worlds. But that doesn't solve the notational challenge.

[DG]

@JWL said:

Could it be this very concern that keeps DVZ in development? It was allegedly designed for this very purpose, but it keeps "not arriving".

Nah, it's still in development 'cos it sounds cr*p. [:D]

DG

[DaveTubaKing]

@Another User said:

By the way, it's possible to build a VI sound with both stacs and fast-reps. Envelope (attack or delay) the latter to come in after the staccato speaks. With stacs round-robined and fast-reps completing the sound, you have the best of both worlds. But that doesn't solve the notational challenge.

That all sounds terribly complicated and of course that's just what I'm trying to avoid with the improved sibelius/VSL interface

Thanks for your reply

[plowman]

"...would be nice if someone from VSL could confirm." Yes. Before posting, I dug up all I could on the matter and found no hint of fast-rep multi-samples per note.

"not if I'm using a repetition sample patch which is slower than the notated repetitions surely?"

Agreed. In fact, in that case, conceivably, you may experience the opposite effect -- type one note and hear only three quarters of that note, and then its release sample (which is on its own bow).

"does the release sample kick in if I'm only triggering the first note of the patch?"

Yes. The release triggers when VI receives a note-off command. So, strictly speaking, you can never play only one note of a fast-rep patch (when "play release" is selected from the Perform page). You'll hear the first note of the patch, and then a second note which is actually the sound of the release sample. It will sound like two bows, though very high tempos may obscure this. The 150 fast-reps are distinctly two bows when release is activated.

If you de-select "play release" then you can hear only one note, but to my ears, that's a non-alternated, machine gun sound. And I don't think VI recognizes it as "only one note." VI only knows that you've released the key in mid-sample. This is a key difference between perf-reps and fast-reps.

I think you're facing two issues at once. The first is trying to parse a fast-rep sound into individual notes, and the second (per JWL's comment) is coaxing a notational program to allow you the kind of fine-tuned control that a sequencer offers.

In Logic, I'd type in a long note and let the fast-rep patch play the bows in one sample. Then I'd suppress (hide) that note in the score and type in the non-sounding individual notes to print. This, of course, is exactly the tedium you're trying to avoid. We all face this issue no matter the platforms.

[Conquer]

To the best of my knowledge: the 'fast repetition' samples are simply sampled phrases. This is the case in both the Pro Edition Performance Set and the VI's, but in the Pro Ed. you can choose between normal and 'RS' (release sample) versions, the latter giving a final note on note-off. There is no round robin programming at work in these patches in either format, they play the same first note (and same release note) every time.

The 'perf repetition' patches are a different kettle of fish. In the VI they do feature round robin programming. In the Pro Ed. Perf Set version, one could program the Performance Tool to vary the start note and the number of repetitions in a cycle. It's a bit fiddly to set up and somewhat intimidating to non-programmers, which I guess is why VSL did away with it in the VI.

Working with sampled phrases means that the notation will tend to be at odds with what you're hearing.

Apologies if I've got anything wrong!

[JWL]

@JWL said:

Could it be this very concern that keeps DVZ in development? It was allegedly designed for this very purpose, but it keeps "not arriving".

Nah, it's still in development 'cos it sounds cr*p. [:D]

DG

Alas-- as is said in some modes of the vernacular:

"Oooooo, THNAP!!"