libcamera::ipa::FixedPointQTraits< I, F, T > Struct Template Reference

Traits type implementing fixed-point quantization conversions. More...

Public Types
using	QuantizedType = UT
	The integral storage type used for the fixed-point representation.

Static Public Member Functions
static constexpr float	toFloat (QuantizedType q)
	Convert a fixed-point integer to a floating-point value.
static QuantizedType	fromFloat (float v)
	Convert a floating-point value to a fixed-point integer.

Static Public Attributes
static constexpr UT	qMin
	Minimum representable quantized integer value.
static constexpr UT	qMax
	Maximum representable quantized integer value.
static constexpr float	min = toFloat(qMin)
	Minimum representable floating-point value corresponding to qMin.
static constexpr float	max = toFloat(qMax)
	Maximum representable floating-point value corresponding to qMax.

Detailed Description

template<unsigned int I, unsigned int F, typename T>
struct libcamera::ipa::FixedPointQTraits< I, F, T >

Traits type implementing fixed-point quantization conversions.

The FixedPointQTraits structure defines a policy for mapping floating-point values to and from fixed-point 2's complement integer representations. It is parameterised by the number of integer bits I, fractional bits F, and the integral storage type T. The traits are used with Quantized<Traits> to create a quantized type that stores both the fixed-point representation and the corresponding floating-point value.

The signedness of the type is determined by the signedness of T. For signed types, the number of integer bits in I includes the sign bit.

Storage is determined by the total number of bits (I + F) and is automatically selected, but the internal storage type is always an unsigned integer to guarantee against sign extension when storing quantized values in registers.

The trait exposes compile-time constants describing the bit layout, limits, and scaling factors used in the fixed-point representation.

Template Parameters

I	Number of integer bits
F	Number of fractional bits
T	Integral type used to store the quantized value

Member Function Documentation

fromFloat()

template<unsigned int I, unsigned int F, typename T >

libcamera::ipa::FixedPointQTraits< I, F, T >::fromFloat ( float  v )

inlinestatic

Convert a floating-point value to a fixed-point integer.

Parameters

[in]

v

The floating-point value to be converted

Returns

The quantized fixed-point integer representation

The conversion first clamps the floating-point input v to the range [min, max] and then rounds it to the nearest fixed-point value according to the scaling factor defined by the number of fractional bits F.

toFloat()

template<unsigned int I, unsigned int F, typename T >

libcamera::ipa::FixedPointQTraits< I, F, T >::toFloat ( QuantizedType  q )

inlinestaticconstexpr

Convert a fixed-point integer to a floating-point value.

Parameters

[in]

q

The fixed-point integer value to be converted

Returns

The corresponding floating-point value

The conversion sign-extends the integer value if required and divides by the scaling factor defined by the number of fractional bits F.

Member Data Documentation

qMax

template<unsigned int I, unsigned int F, typename T >

libcamera::ipa::FixedPointQTraits< I, F, T >::qMax

staticconstexpr

Initial value:

= std::is_signed_v<T>
                 ? (UT{ 1 } << (bits - 1)) - 1
                 : bitMask

Maximum representable quantized integer value.

qMin

template<unsigned int I, unsigned int F, typename T >

libcamera::ipa::FixedPointQTraits< I, F, T >::qMin

staticconstexpr

Initial value:

= std::is_signed_v<T>
                 ? -(UT{ 1 } << (bits - 1))
                 : 0

Minimum representable quantized integer value.

This corresponds to the most negative value for signed formats or zero for unsigned formats.

---

The documentation for this struct was generated from the following files:

• src/ipa/libipa/fixedpoint.h
• src/ipa/libipa/fixedpoint.cpp