Q: Mapping Attack (inter alia) to a Multiparameter?

[DarkStar]

Give this a try:

-- open a Dynamics plug-in, I used mCompressor but any will do
-- Quick Learn its Attack parameter to a Multiparameter (MP)
-- open the MP and set the Value mode to Percentages

-- drag the MP value down to 0%, the Attack parameter will be 0ms, as expected
-- drag the MP value down to 100%, the Attack parameter will be 1000ms, as expected
-- now drag the MP value to 50%, the Attack parameter will be 63ms
-- to get the Attack value of 500ms, I need to set the MP to 84.1% again.

Can someone explain what is going on?

------------------------------------
It gets weirder
In MDynamics MB, learn a band's Attack and Release full ranges to an MP. The Attack range is 1000ms and the Release range is 5000ms,
-- set the MP value to 50%
-- Attack is 63ms , as above, and Release is 313ms,
-- in the MP, Set the Release's MaxValue to 100ms
-- set the MP value to 50% (nudge it to refresh the screen),
-- both Attack and Release are 63ms.

The range of the target seems to affect the value.
For a range of 1000ms, MP = 50% gives a value of 0.50 ^ 4 * 1000 = 625
For a range of 5000ms, MP = 50% gives a value of 0.50 ^ 4 * 5000 = 312.5

Huh?

[jmg8]

The range has a built in transfer curve to give you the more useful values. Attack on a compressor usually has values less then 100ms in most cases. As its Melda and super extreme it goes all the way up to 1000ms. But if left linear, most of the knobs range would be taken up with values that dont get used often. So an internal transfer curve is used to give more resolution on the knob to the most useful values.

[DarkStar]

Thank you. Yep, I can understand that and that is what I was stumbling towards.

But that means that a 50% value for an MP does not give you an Attack value mid-way through its range.

That ^ 4 is doing the transformation, isn't it? Dos that have a proper mathematical name? Do you know if that is the only internal transformation?

Also, I would have thought that it would have been better to use the Transformation Shape graph in the Multiparameter Editor, instead of an internal hard-coded transformation, so that the designer has full control over the MP's effect on its targets.

[jmg8]

Not sure I understand some of what you are asking, sorry.

But as for the last bit. You can reverse the transfer curve. Set the MP to 50% then change the display value to "by first parameter" now open the transfer shape window and drag the curve until the value displays as 500ms.

[DarkStar]

I was wondering if you knew of any other internal transformations.

I tried the drag of the mid-point and that was OK for the 50% value. Then I had a brain-wave: I used the [Expression Evaluator] for the equation: x^0.25. And the result:


So now I know how to counteract the internal transform if I want. But that also confirms my point that using a Transformation Shape of x^4 would be better. It probably cannot be changed now for backwards-compatibility reasons.

Ah well, I've learned a bit more. Thank you.

[jmg8]

I think there are loads of parameters that have these internal curves. But I cant seem to see any disadvantages of having that curve public or hidden anyway, after all every value is still available. In my opinion (and I practice this when I make devices) 50% should always be the ideal value, then the user will be able to dial in less or more. So when I set up MPs i always have this in mind and use transfer curves to move my ideal value to 50%.
Also, I must be missing something obvious, but I can not find the expression evaluator, could you help me with that please.

[DarkStar]

Sure can:

[jmg8]

Ah, thanks. I will have to check that out tomorrow. I wish I paid attention in maths at school, it would have come in very handy right now. Oh, well, never to late to learn.

[dmbaer]

Ah, thanks. I will have to check that out tomorrow. I wish I paid attention in maths at school, it would have come in very handy right now. Oh, well, never to late to learn.

Note also just to the left of the expression entry field is the Import CSV button. If you have a calculation that could benefit from using a spreadsheet program, that's what that one's all about.

[DarkStar]

Especially as I do not know what syntax the Expression Evaluator supports.

[DarkStar]

Vojtech pointed me at the Expression parameter in MUtility:

Expressions use traditional C/C++ style formatting providing arithmetic, logical and conditional operators. The following terms are supported:

Unary operators: - (negative sign), ! (logical negation)
Binary arithmetic operators: +, -, *, /, ^, %, max, min
Binary logical operators: ==, !=, <, <=, >, >=, &&, ||, ^^
Unary functions: log, sqrt, sqr, abs, exp, sin, cos, tan, sinh, cosh, tanh, inv, asin, acos, atan, ln, log10, sgn, floor, ceil, round, rand
Ternary logical operator: a ? b : c
Constants: pi, e, sqrt2, ln2

[jmg8]

Haha, I cant understand that! Thanks. I will do some research.