MIDI control change in VSL

[Gone To Lunch]

Hi

I note the orchestral sound in the VSL Special Edition Demo 1 is very impressive, and very realistic. Is this due to their use of MIDI control change parameters ?

For example, in some of the Cello parts, there are two lines for MIDI CC 2 and 11,
which are breath control and expression.

How do these relate to the sample playback ? Why is breath control used in a string part ? What is being expressed by expression ?

I note expression is also in the Timpani part ?

thanks

Michae

[Shooshie]

Breath control is a very expressive way to perform string, wind, or brass parts. It also works well for percussion or anything else that requires sensitive voicing.

Expression control is another form of continuous control like Breath Control. BC is controller #2, and Expression is controller #11. 11 is more commonly used with pedals, faders, and mod wheels. Most MIDI Wind Controllers can output #11, as well. It depends on how sounds are programmed, but some sounds respond differently to BC than to Expression, even though they are both just a series of numbers from 0 to 127. Nevertheless, on most VI's, I can use #2 or #11 interchangeably.

Shooshie

[Gone To Lunch]

Thanks for your reply.

But what do they actually do to the sound of the sample being played back ?

[Spikey Horse]

I'm not sure what's going on in the demo you mention - and I think my templates are probably set up different to the standard mapping (but I forget how exactly) - is this a downloadable demo MIDI file you're talking about?

But anyways, Vienna Instruments has a velocity crossfade feature which blends different velocity layers. So you can choose between standard note on velocity to trigger the velocity layer (good for playing piano - I mean the instrument!) OR you can assign a CC to the velocity X-fade slider in which case VI ignores note on velocity and follows the CC instead (good for sus strings etc). You can even assign a note to switch between the two modes.

The other main CC feature particular to VI is the ability to switch between cells of its articulation matrix. VI uses standard key switches to switch between (user definable) articulations but within each key switch you can have more than just one articulation loaded - and which one plays is selected by assignable CC data value (or you can choose other parameters such as velocity or speed of repetitions).

If none of this makes any sense - and it probably won't! - I would suggest you check out the video demos on the VSL website - it's much easier when you can see/ hear what is going on together.

[Frodo]

Thanks for your reply.

But what do they actually do to the sound of the sample being played back ?

The operative term is "envelope". That's what these features do during playback.

For the moment, forget VSL. This works with almost any VI. CC7 is the overall volume controller. CC11 is called Expression, but it works WITHIN CC7's parameters. Expression really is another form of MIDI volume and/or MIDI velocity.

Say you choose to set your CC7 control number to 65 (about half). Any automation made with CC11 will be limited to a max output effect of the 65 limit of CC7. If you raise CC7 to a max 127, then expression will respond accordingly with all contours intact but expanded by a factor of (roughly) 2. That is to say, if you create a crescendo with CC11 and later increase CC7 from 65 to 127, that crescendo will be twice as "strong" or twice as prominent.

As for breath control, consider the difference between inputting notes using a keyboard vs entering notes using the power of the breath. Suddenly, with the power of breath, you control not only velocity but the ADSR (attack, delay, sustain, release) of a given note as the breath dictates. Even this works as as subset of CC7 in many ways.

If you were to utilize these features for note input, any adjustment made to CC7 will serve to expand or compress whatever contours you've created using BC or Expression. The contours would be intact, but they'd be louder/softer or stronger/weaker as you adjusted CC7 one way or another.

[Gone To Lunch]

Thanks Frodo!

Long may you reign. That was a brilliant succinct explanation of EXACTLY what I wanted to know..

[Spikey Horse]

You're welcome!

.... oh wait, you said Frodo..

But seriously, that was an interesting explanation - I always thought breath control was just like any other CC but called 'breath' out of mapping convention.

I've got to admit I'm slightly confused now ..... but will do some research....

Anyway +1 thanks Frodo!

[Frodo]

@GTL-- I only hope it helped!!

@Spikey-- yeah, there are two things at play. CC11 is indeed a subset of CC7. Just as an experiment, you might want to leave CC7 at parody and contour a melodic line in CC11. At that point, you might want to automate some CC7 data to see what effect it has on your CC11 contours.

On some level, we must consider keyboard input and BC input working in a different category entirely from everything else. If CC11 is at issue then things can not only get complicated but they can also get wonderfully sophisticated which is good news for preventing the stagnant nature of how any VI can sound. Have you ever listened to a VI and found yourself focusing on the fact that it was a VI instead of focusing on the music itself? Using BC, Expression and Velocity/Volume in combo really do bring life to a virtual realization to create a result that makes a virtual mock-up at once forgivable but acceptably pleasant.

Some I know who excel in breath controller input (Yamaha WX or Akai EWI, etc.) have had some tussles with getting a single sample to respond to every personal nuance where vibrato variances and attacks/releases and vibrato are concerned. VSL is such a huge library with samples that were recorded with many of these nuances that striking the right balance between BC and actual sample behavior can get confusing.

If the sample itself was recorded with a natural crescendo, one might want to start with a steady breath input that has little effect on the crescendo in the sample but allows the sample to do it's own thing. If one wants a more intense crescendo, then using a graduated BC input (on the increase) will intensify the sample's natural crescendo. One may also "under-breathe" crescendo samples for a more subtle effect. Once done, tweaking CC7 can play an important role in the overall mix.

At the end of the day, it's apparent that breath control becomes a very physical elemental requirement of the user which may necessitate the use of a treadmill to prevent dizziness!!

Not all of us are wind players who can make the most of the WX or EWI. There are breath control devices out there that will allow the user to breath in articulations in conjunction with keyboard input. With enough practice, using consonants such as "T" or "K" can really add to the sophistication of the sound in fast passage where variances of attacks can become all-too-predictable. Too often, the "machine gun" sound can wreck a sequence. For VSL specifically, I recommend using the Repetition samples for fast notes (where they exist outside of the full wack library). This brings important subtleties to the equation that may not exist in the SE or LE libraries. However, a BC device can add much in lieu of the actual repetition samples.

Much peace and good will,
frodo

[Shooshie]

Thanks for your reply.

But what do they actually do to the sound of the sample being played back ?

Others have told you basically what they do, now. But there are some things unique to each.
The biggest PROBLEM in using Expression control is that it is still based on Velocity in most VI's. Thus, the attack of the note sets the velocity at whatever is the initial level. That velocity level chooses the sample layer from which to play. If the velocity is 30, it chooses a soft sample. If it is 115, it chooses a loud, more strident sample. These velocity layers are intended as means of getting the sounds as realistic as possible. Since instruments generally sound different, with different overtones at different loudnesses, they sample the same instrument over and over at different loudnesses, then combine the layers into a "program" for each note on your sampler. How your sampler chooses which layer to play during performance has a lot to do with how realistic your sound is.

So, Expression control (CC#11) gives you more realistic crescendos and diminuendos, but it is often still based on velocity layers, so that whatever is your starting velocity chooses the layer. From that point you can crescendo and decrescendo, but you will be playing that same layer no matter what you do. That limits your expression… a LOT.

Breath Control was always intended to bypass envelopes and layers, allowing for starting as quietly as an instrument can play and growing as loud as it can go, or vice-versa. The velocity is not a factor, as BC sends no velocity information. That means sounds have to be programmed a little differently to respond to BC. For 20 years or so, that was a problem. Not many people used BC, so most sounds were not programmed for breath. If you tried to use BC with a sound not programmed for BC, you got into weird situations where the sounds were almost unusable, because you were stuck with the lowest velocity layer, and you could not crescendo very much.

A workaround or two was created, but they were not very successful. For BC sounds really to work, they have to be programmed to respond exactly to what a BC controller does.

VSL took a different approach. Vienna allows you to choose any set of controllers to assign to any instrument, making them essentially interchangeable. Then they let YOU program how the sampler responds to the instrument. One very important option for people wanting fluid expression without limits is to crossfade the velocity layers, so that as you grow louder and move from the level of one sample layer to the next loudest layer, then as you get quieter, it moves back down to lower layers. As it goes from one to the other, it automatically crossfades between the layers for greater realism.

Not all VI's do this. In fact, those that are envelope based rarely have an option for crossfading velocities. Samplers are doing it more often than synths, but even those are still relatively rare. The reason is because MIDI Wind Controllers like the WX-5 or the EWI are a very small niche market. But libraries like Vienna are made by people who realize that a keyboard can never make a violin sound "real" without a lot of help from continuous data. Whether it be pedals, sliders, wheels, or breath control, there are many aspects of the sound that require continuous input, not just an envelope set by the initial attack's velocity. VSL's designers realized that the wind instruments are perfect for sound input, so if they accommodated them, it was only a matter of time before people would use them, and their sounds would suddenly leap forward in realism. That, in fact, has happened. I think the age of breath control is now upon us, and we'll see more developers making it available for their instruments just for the potential of unmatched realism.

If you still do not understand what it does or how it works, ask specific questions and I (we) will answer them for you.

Shooshie

[Gone To Lunch]

OK, and big thanks for all the help !

I think I have got it, and have set up a MIDI mixer in the DP consoles window to connect to the Vienna Perform control sliders via the map control options, and it works a treat so I can now experiment musically,which is the point of all this..

But I have one more little question, what is the precise difference between CC 11 and CC 02 ? I get that CC 11 is like a nested micro volume that works within a particular velocity layer, so is CC 02 the same but across the whole 0-127 gamut ? Or are they functionally identical, but have evolved separately because of historical hardware assignment conventions ?

[Shooshie]

Basically, ALL continuous MIDI control data are the same, possibly excepting pitch bend. They are all simply a list of 128 whole numbers, with another whole number used as the identifier. The difference between any of them is in how they are programmed to be received and used. Since Expression and Breath Control work in similar ways, most sample developers treat them the same way -- continuous intensity/loudness control. Where they differ historically is that #11 (Expression) is based on velocity, while #2 (Breath Control) is based on no envelope or velocity at all.

If the instrument is programmed to receive them as they were historically intended, the BC controller would actually act as the envelope. When the sound is made louder, the elements or operators that create the sound interact with each other accordingly, making it more shrill or adding more overtones. Expression may operate in a similar way, but its envelope is limited to the initial attack's velocity. It can go no louder.

Again, that's the "historical" form usually programmed for those two controllers. Every instrument depended on the way its designers and programmers implemented the controllers. There is nothing unique or special about the controllers themselves. They are merely commands delivered in hexadecimal format such that the receiving instrument receives an identifier and a value, as well as MIDI channel info.

That's about it!

Shooshie

[Gone To Lunch]

Thank you all so very very much.

All your help and wisdom has been spectacularly useful.

Now, back to the music...