Embedded C for Digital Signal Processing - Signal Processing Systems - page 1241

[ Sat][short/long][unsigned]( Fixed/ Accum)

float double

float double

Fig. 6 Conversion in expressions with mixed floating point and fixed point types. Convert the
fixed point type to the floating point type

[ Sat][short/long][unsigned] Fixed

full precision

full precision

[short/long] int

[ Sat][short/long][unsigned] Accum

Fig. 7 Conversion in expressions with mixed integral and fixed point types. Compute the result in
full precision and convert to the resulting type

[ Sat][short/long]( Fixed/ Accum)

[ Sat][short/long]( Fixed/ Accum)

[ Sat][short/long] unsigned ( Fixed/ Accum)

Fig. 8 Conversion in expressions with mixed signedness of fixed point types. Convert the
unsigned fixed point type to its corresponding signed type

[short/long][unsigned]( Fixed/ Accum)

[short/long][unsigned]( Fixed/ Accum)

Sat [short/long][unsigned]( Fixed/ Accum)

Fig. 9 Conversion in expressions with mixed saturation of fixed point types. The operation and
the result are saturating