Eval¶

This document describes the evaluation language used in the expression parameters of the Eval* nodes.

Basic operators (+, -, *, /) are supported, as well as parenthesis. Whitespaces (including line breaks) are allowed as well.

If the input resources are multi-dimensional, component can be selected with 0123, rgba, xyzw or stpq. For example, foo.z will pick the 3rd component of the 3+ dimensional input resources identified by “foo”.

Constants¶

Constant	Description
`e`	Euler’s number, base of the natural logarithm
`phi`	φ (Golden ratio)
`pi`	π
`tau`	𝜏 (2π)

Functions¶

Function	Description
`abs(x)`	absolute value of `x`
`acos(x)`	arc cosine of `x`
`acosh(x)`	inverse hyperbolic cosine of `x`
`asin(x)`	arc sine of `x`
`asinh(x)`	inverse hyperbolic sine of `x`
`atan(x)`	arc tangent of `x`
`atanh(x)`	inverse hyperbolic tangent of `x`
`cbrt(x)`	cube root of `x`
`ceil(x)`	smallest integral value not less than `x`
`clamp(x,a,b)`	constrain `x` to lie between `a` and `b`
`close(a,b)`	`1` if `a` is close to `b` within `1e-6` relative precision, `0` otherwise
`close(a,b,p)`	`1` if `a` is close to `b` within `p` relative precision, `0` otherwise
`cos(x)`	cosine of `x`
`cosh(x)`	hyperbolic cosine of `x`
`cube(x)`	cube of `x`
`degrees(x)`	convert `x` from radians to degrees
`eq(a,b)`	`1` if `a = b`, `0` otherwise
`erf(x)`	error function, `2/√π * integral from 0 to x of exp(-t²) dt`
`exp(x)`	base-e exponential of `x` (`e^x`)
`exp2(x)`	base-2 exponential of `x` (`2^x`)
`floor(x)`	largest integral value not greater than `x`
`fract(x)`	fractional part of `x`
`gt(x)`	`1` if `a > b`, `0` otherwise
`gte(x)`	`1` if `a ≥ b`, `0` otherwise
`hypot(x,y)`	euclidean distance (`√(x²+y²)`)
`isfinite(x)`	`1` if `x` is finite, `0` otherwise
`isinf(x)`	`1` if `x` is infinite, `0` otherwise
`isnan(x)`	`1` if `x` is not a number (`NaN`), `0` otherwise
`isnormal(x)`	`1` if `x` is normal, `0` otherwise
`linear(a,b,x)`	linearly remap `x` in range `[a,b]` to `[0,1]`
`linear2srgb(x)`	convert `x` from linear to sRGB (see also `srgb2linear(x)`)
`linearstep(a,b,x)`	saturated version of `linear()` (equivalent to `sat(linear(a,b,x))`)
`log(x)`	natural logarithm of `x` (see `e`), also called “ln” in math terminology
`log2(x)`	base-2 logarithm of `x`
`lt(a,b)`	`1` if `a < b`, `0` otherwise
`lte(a,b)`	`1` if `a ≤ b`, `0` otherwise
`luma(r,g,b)`	gamma encoded luma Y’ of the R’G’B’ value, as defined by BT.709
`max(a,b)`	maximum value between `a` and `b`
`min(a,b)`	minimum value between `a` and `b`
`mix(a,b,x)`	linearly remap `x` in range `[0,1]` to `[a,b]`
`mla(a,b,c)`	multiply-add in one operation (`a*b + c`)
`mod_e(a,b)`	euclidean modulo: `a - bsign(b)floor(a/abs(b))`
`mod_f(a,b)`	floored modulo: `a - b*floor(a/b)`
`mod_t(a,b)`	truncated modulo: `a - b*trunc(a/b)`
`pow(a,b)`	`a` raised to the power of `b`
`print(x)`	leave `x` unchanged but print its value on the standard output (debug purpose)
`radians(x)`	convert `x` from degrees to radians
`round(x)`	round `x` to nearest integer, away from zero
`sat(x)`	saturate `x` (equivalent to `clamp(x,0,1)`)
`sign(x)`	`1` if `x` is positive, `-1` if is negative, `0` if zero
`sin(x)`	sine of `x`
`sinh(x)`	hyperbolic sine of `x`
`smooth(a,b,x)`	hermite interpolation (`f(t)=3t²-2t³`) using `t=linear(a,b,x)`
`smoothstep(a,b,x)`	same as `smooth()` but using `t=linearstep(a,b,x)` (only the S curve remains)
`sqr(x)`	square of x (`x²`)
`sqrt(x)`	square root of `x`
`srgb2linear(x)`	convert `x` from sRGB to linear (see also `linear2srgb(x)`)
`srgbmix(a,b,x)`	mix `a` and `b` in linear space, but the inputs and output are gamma encoded
`tan(x)`	tangent of `x`
`tanh(x)`	hyperbolic tangent of `x`
`trunc(x)`	round to integer, toward zero (see also `floor()`)