Suggestion for Herb

[duoartc]

Hi Herb,

Just wondering whether in future updates you can add an extra two variations to those articulations which have variations - this would give four different up/down bows for strings and four variations to the woodwinds, brass. SAM has four to five for some of its libraries and I note that EWQLSO has three for some strings. This would make your alternating tool more powerful (perhaps you could add a random mode). I'm sure you have already recorded various takes anyway so it wouldn't mean re-recording.

Craig

[duoartc]

I thought this was a good idea. Anybody have any opinions?

[evan evans]

Absolutely. Please. I was under the impression that it came with 16 variations. But I guess I was misinformed by the general public.

Evan Evans

[William]

I mentioned something similar on another thread, I forgot which. I also think this is very important to do. I personally would like more variants of notes rather than more new instruments each with fewer notes sampled. For example - just violins with five thousand notes, instead of violins, Irish fiddle, Appalachian fiddle, medieval, and renaissance period instruments each with one thousand notes.

The sample sets of keyboards, even today, have hundreds of instruments but each with only a few samples total. That is the opposite extreme. By the way I am always struck by the HUGE gap that has arisen between keyboards and software samplers. This is relatively new. If you think about how ten or fifteen years ago, the most advanced samplers were keyboards - like the Kurzweil or Emulator. But now the hardware manufacturers have completely given up even trying to compete, as if it is impossible. Why? Maybe the the large manufacturers of keyboards are like the car companies - ultra conservative. ("Sorry," said the CEO. "You can't exceed 128 mb. That thing will explode.") Korg brags about having a few hundred MEGAbytes of samples - for all of its instruments. What? This to me is a bizarre by-product of technological advance: that those instruments have been completely left behind.

The main reason to have more variations of the same sampled notes is to capture the basic fact of a live recording - EVERY note is a variation. The ideal would be to have so many variants that, combined with the repetition tool, you would never hear the same note twice - long, medium, short, pp, ff, legato, whatever. That of course seems impossible right now, but think of how data storage is becoming more and more gigantic every few months. If you have a multi terabyte sample set, it would be no problem (except for the musicians who played them and are six feet under as a result).

[NG1]

@Another User said:

Just wondering whether in future updates you can add an extra two variations to those articulations which have variations

You can add my vote too. I wonder if this would even be all that difficult to do - with some careful editing of the already recorded "_rep" samples, we could have a huge number of new alternations to use with many lengths to choose from as well.

[William]

That is exactly what I'm thinking of doing - extracting the repetitions so they can be used in the more lazily applied alternation tool. There is a huge amount of data in those repetitions.

Another thing is legato performance variations, which if you have a lot of similar leaps or intervals can sometimes sound repetitious with no variation. Though of course this is getting into really impractical amounts of sampling and beyond state of the art.

[duoartc]

@Another User said:

I personally would like more variants of notes rather than more new instruments each with fewer notes sampled.

I don't think you need all that many. The ear, I think, mainly hears the difference between two variations when played one after the other rather than the absolute sound of each sample. If you only have two variations, 1 & 2, then you only have two combinations 1-2 and 2-1. If you have three variations you have 3! (factorial) or 3x2x1=6 combinations and for four variations 4! or 24 variations. Obviously, the alternating tool would need a cycling mode to go through all the variations.

With the woodwinds and brass, I think 3 variations is enough ie one more than what is provided in the Cube. With the strings, you may need four variations if you want to keep the order of repeating up and down bows eg for variation 1 up-bow, 2 down-bow, 3 up-bow, 4 down-bow you would have four different combinations 1-2, 3-4, 1-4, and 3-2. Mind you there are then quite a few combinations on how you can combine these four pairings in sequence.

Of course there comes a point when you could use the repetition mode.

Hope this makes sense. Maybe I'm wrong.

[duoartc]

Sorry, messed up the quote William.

[William]

Thanks for the response. You're right that practically speaking you usually don't need that many variations. But oddly enough I've had some situations where two alternations were glaringly obvious - almost more obvious than no variations at all which in the past you would try to help by velocity changes, filtering, etc. It is as if the ear instantly discerns a pattern almost like a musical motif. Obviously it would be better with three, and about perfect with four or five.

Though all this is why the repetition tool was created.

It reminds me of doing footstep foley FX on a movie production. I found that if I had five separate steps, the variations possible made it easy to get a perfectly natural effect. Any fewer, and it sounded artificial.

[NG1]

@Another User said:

Another thing is legato performance variations, which if you have a lot of similar leaps or intervals can sometimes sound repetitious with no variation. Though of course this is getting into really impractical amounts of sampling and beyond state of the art.

I think Herb is actually addressing this issue for the upcoming Symphonic Cube release. If I understood correctly, he's sampling interval repetitions, such as C - E - C - E, etc.

I think this will be a very important addition for some types of writing.