antissa

Tutorials

Hands-on guides for each Mantissa mode. Read in your browser or open them inside the app's Web mode — same content, same look.

Your progress

Loading…

Porting workflow — eight standalone topics

Each step below is a self-contained reference for one piece of the porting workflow — read independently in any order. The numbering reflects the natural reading sequence for someone new to fixed-point work.

  1. 1

    IEEE-754: float & double

    Foundations — what you port from

    float = 32 bits S exponent · 8 mantissa · 23 double = 64 bits · S · exp 11 · mantissa 52 (-1)^S · 1.mantissa · 2^(e - bias)
  2. 2

    Q-format basics

    Pairs with: Fixed-point converter

    int16_t = S Q15 S 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 range = ±1.0 · LSB = 2⁻¹⁵
  3. 3

    Pick Q for your data

    Pairs with: Q auto-suggest

    2^E max|x| → E → Q
  4. 4

    Fixed-point arithmetic

    Interactive: add, multiply & truncate

    Qa × Qb Q(a+b) drop widen → compute → round
  5. 5

    Truncation & DC offset

    Interactive: the DC offset truncation adds

    0 DC offset truncation → the signal sags
  6. 6

    SFDR & quantization

    Interactive demo

    SFDR fundamental − loudest spur
  7. 7

    CIC filters

    Interactive: CIC bit-growth calculator

    ∫ × N ↓ R comb × N B_in + growth grows ⌈N·log₂(R·M)⌉ bits
  8. 8

    FIR filters

    Interactive: coefficient quantization

    quantized floor passband coeff word length → stopband depth

Hands-on practice

Two guided labs that put the workflow in front of real data. Download the sample project, open it in the desktop app, and work through the lab notes — each lab is self-contained and finishes in 20–30 minutes.

  1. Lab 1

    Debug a float vs. fixed-point port

    The sample project ships two builds of the same DSP block — a float reference and a deliberately-broken fixed-point port. Open both under the app's Debugger mode, bind float / fixed variable pairs, watch them update side-by-side, and find which variable is quantizing badly.

    Pairs with: Debugger mode

    float fixed
  2. Lab 2

    Analyse output with Wave Compare

    Take the two output .wav files from Lab 1 (or the pair shipped with this lab) into the Wave Compare mode. Run wav-diff, read the MLD / MAD / SSNR report per channel, and jump from the worst-mismatch marker straight into the offending sample.

    Pairs with: Wave Compare mode

    MLD float reference vs. fixed port

Looking for app-usage docs (how each mode's panels work)? See the Help page.