Reading Your Dirac Graphs

Last updated

Start Here

Guides

Resources

Dirac shows you graphs at two stages: after measuring, and in Filter Design. This page explains what each one shows, what a good result looks like, and what the graphs can’t tell you.

Jump to: Beginner Advanced

The graphs you’ll see

Beginner

Measurement results

After the measurements, Dirac shows the magnitude response of each speaker: how loud it is at each frequency, averaged over your microphone positions. Peaks and dips below about 300 Hz are normal in any room. They come from the room, not the speaker.

Filter Design

  • The measured curve: the average response for the selected speaker or group.
  • The target curve: the shape you’re asking Dirac to reach. You can drag it.
  • Corrected: tick this to see Dirac’s prediction of the result. It should follow the target closely.
  • With Bass Control or ART, the graph shows the speakers and subwoofers together, so you can see how they join at the crossover.

The spread graph

Dirac’s ART setup guide also shows a spread view: how much the response varies between your microphone positions. A narrower band means more consistent sound from seat to seat. Bass Control and ART aim to narrow it in the bass.

What good looks like

  • Corrected follows the target within a few dB above about 100 Hz.
  • The crossover region is smooth, with no deep notch where the subs and speakers meet.
  • Left and right look alike. A big difference usually means a measurement or placement problem.
  • Narrow dips that remain after correction are normal. See below.

Why Dirac leaves some dips alone

Most narrow dips in the bass come from sound cancelling itself: a reflection arrives out of step with the direct sound. Adding power adds it to both, so the dip stays while the amplifier and speaker work far harder. Dirac and most room correction systems limit boost for this reason. Moving the seat or subwoofer, adding a second sub, or Bass Control and ART are the ways to fill them. This is general acoustics rather than a quoted Dirac statement;

The prediction isn’t a measurement

Corrected is calculated, not measured. To see what actually happens at your seat, measure with Dirac on in REW and compare with a measurement taken with Dirac off.

Next step

Measure the result in REW

The only way to confirm what Dirac did at your seat.


Reading graphs in more depth

Advanced

Smoothing

Dirac’s graphs are smoothed, so narrow features look smaller than they are. REW lets you choose the smoothing: 1/6 octave for an overview, 1/24 or none to see the detail. Compare graphs at the same smoothing.

Averages hide seat-to-seat differences

The curve Dirac shows is an average of your positions. A flat average can still hide seats that are very different. The spread view, or separate REW measurements at each seat, show what the average hides.

Above about 500 Hz

In-room measurements above a few hundred hertz mix the speaker with room reflections. Small ripples there are expected and not worth chasing. Dirac lets you limit correction to the bass if you prefer.

Sources