Skip to content

Spectrum Symmetry Mirror

Run the Spectrum Symmetry Mirror MicroSim Fullscreen

You can include this MicroSim on your website using the following iframe:

1
2
<iframe src="https://dmccreary.github.io/fft-benchmarking/sims/spectrum-symmetry-mirror/main.html"
        height="427px" width="100%" scrolling="no"></iframe>

About This MicroSim

Run a 16-point DFT and you get 16 complex numbers back. Only 9 of them tell you anything new.

For a real-valued input — which is every signal a microphone produces — the output has a strict symmetry:

\[X[N-k] = \overline{X[k]}\]

Bin \(N-k\) is the complex conjugate of bin \(k\): same magnitude, opposite-signed imaginary part. Click any blue bin and the arc shows you its partner in the gray region.

Two bins have no partner because they are their own mirror:

  • Bin 0 (DC) — \(N - 0 = 0\), so it maps to itself.
  • Bin N/2 (Nyquist) — \(N - N/2 = N/2\), likewise.

Both are purely real for a real input, so there is no imaginary sign to flip.

Why This Matters for Benchmarking

Counting from the diagram: bins 0 through 8 are 9 unique values out of 16. In general a real input gives \(N/2 + 1\) useful bins.

This is not a curiosity — it is a factor-of-two saving. A real-input FFT can skip computing the mirror half entirely, and a magnitude spectrum only ever needs plotting up to Nyquist. When you benchmark an FFT library, check whether it exploits this. Many do, and the ones that do not are doing twice the necessary work on the back half.

How to Use

  1. Click bin 3. The arc connects it to bin 13. Check: 16 - 3 = 13.
  2. Click bin 6, then bin 10. Confirm they are each other's partner.
  3. Click a gray bin directly. The same pair lights up — the relationship is symmetric.
  4. Click bin 0, then bin 8. Read why these two have no partner.
  5. Count the blue bins plus the two special ones. That is your unique-bin count.

How to Read the Colors

Color Bins Meaning
Orange 0 DC — the average value, its own mirror
Blue 1 to 7 Positive frequencies, unique and useful
Magenta 8 Nyquist — highest representable frequency, its own mirror
Gray 9 to 15 Mirrors of the blue bins, redundant for real input

Lesson Plan

Grade Level

Undergraduate (college junior/senior)

Duration

8-10 minutes

Prerequisites

  • A DFT of N samples produces N complex outputs
  • Complex conjugate: same real part, negated imaginary part

Learning Objective

Students will be able to interpret a full N-point DFT spectrum and explain why bins above Nyquist mirror bins below it as complex conjugates for a real-valued input.

Activities

  1. Pair them up (3 min): Students click four blue bins and record each partner, then state the rule from their data.
  2. The two loners (3 min): Students explain why bins 0 and N/2 have no partner, using the formula rather than the picture.
  3. Count the savings (3 min): For N = 1024, students compute how many bins are unique and what fraction of the output is redundant.

Assessment

Ask: "A 256-point FFT of a microphone signal returns 256 complex values. How many do you actually need to keep, and which bin numbers can you discard?"

References

  1. DFT symmetry properties — the conjugate relation stated formally.
  2. Complex conjugate — the operation relating each mirrored pair.