Acoustics, Calibration & Asset Architecture
Reducing First Reflections and Flutter Echo in Small Desk Rooms
A small desk room can send a speaker reflection back to your ears within a few milliseconds, ring between bare walls, and bounce your voice off the desktop into the mic. Mark each path, calculate a panel gap that fits the room, then compare reflection levels in decibels and record dry voice takes to hear what changed.
Find out which sound you are hearing
Start with the room in its working configuration. Leave the screen, keyboard and chair where they usually sit. Mark the chair and mic stand positions with tape, then measure from the floor to seated ear height and from the desktop to the vocal mic capsule. Write down the speaker locations and the distance between the two largest parallel hard surfaces. A change in chair position or an open door can spoil a useful before-and-after comparison.
Play a short mono percussion loop through one monitor at a time. A single early reflection can blur an attack or color its tone as it combines with direct sound. A handclap between hard parallel surfaces may expose flutter as a rapid metallic train. In a room 2.6 m wide, an extra round trip across the width is 5.2 m, or roughly 15.2 ms at 343 m/s. Real peaks can interleave. Treat that spacing as a clue and test the suspected wall pair with temporary absorption.
Mark the mirror points for both speakers
Sit at your taped listening position and have someone slide a small flat mirror along one sidewall, flush and untilted, near ear height. Mark where each monitor’s tweeter appears, then repeat on the other wall. Check again with your head about 10 cm forward and back and slightly to each side. If the tweeters sit well above or below your ears, move the mirror vertically too. Use the spread of marks to size the panel. A 600 by 1,200 mm panel is a practical starting size when the marked zone fits comfortably inside it. Give the ceiling a separate check if it remains bare and close to the listening position.
To work alone, use the image-source geometry shown in the figure. In plan, put the wall at x = 0 and measure y along it. With speaker and ear at the same height, place the speaker at (0.60, 0) m and the ear at (1.20, 1.20) m. Reflect the speaker to (−0.60, 0) m and draw a straight line from there to the ear. It hits the wall at y = 0.40 m. The ray meets the wall at about 34° to the wall normal on both sides of the bounce. The reflected path is 2.16 m and the direct path is 1.34 m. The extra 0.82 m predicts an arrival about 2.4 ms after the direct sound. Allen and Berkley’s image-source method.
Calculate the rear gap from your depth budget
Measure the clear gap from the back of the porous core to the wall. Total acoustic depth is core thickness plus that gap. Allow separately for anything proud of the front face, such as a frame lip. Keep the face and the rear open to airflow with breathable fabric over the core. A solid rear board can block the core from the cavity you intended to use.
For a first layout, use the quarter wavelength in air as a depth reference. Against a rigid wall, air particle velocity for a normally incident standing wave reaches its first maximum a quarter wavelength out. Putting some porous material farther from the wall can improve absorption at longer wavelengths. For a front-face reference at 500 Hz, calculate 343 ÷ (4 × 500) = 0.1715 m. Subtract a 100 mm core from 171.5 mm to get about 72 mm of gap. A 70 mm spacer gives 170 mm total. The material occupies a range of depths, so use this arithmetic to size the assembly. Predicting absorption requires the material’s airflow resistivity, layer thickness, backing and angle of incidence. Arup’s porous absorber model.
| Porous core | Clear rear gap | Total depth | Placement trade-off |
|---|---|---|---|
| 50 mm | 25 mm | 75 mm | Fits a tight passage. Try it for bright flutter where wall coverage matters. |
| 100 mm | 70 mm | 170 mm | Fits the 500 Hz depth example. Test it first at the sidewall marks. |
| 100 mm | 120 mm | 220 mm | Allows a deeper cavity where a rear wall has more clearance. Verify the benefit in place. |
At 250 Hz the same reference depth is 343 mm. If a doorway or chair limits projection to 170 mm, use the depth you have and judge the measured improvement in the bands you need. A wide enough panel over both mirror marks is a better first trial than a deep narrow panel that misses one of them. For flutter, temporarily cover a patch on one of the facing walls and clap again. If the ring persists near the back of the room, shift or extend coverage where those repeated paths can still travel between bare surfaces.
Calm the bounce from the desktop into the vocal mic
Check the tabletop separately from the speaker reflections. With the mouth 20 cm above the desk, the capsule 30 cm above it and 35 cm of horizontal separation, the direct distance is about 36 cm. Imagine the mouth 20 cm below the desk for the reflected route. That gives 50 cm of vertical separation and a reflected distance of about 61 cm. The extra 25 cm adds about 0.72 ms. A strong bounce with little phase shift can introduce a first dip near 700 Hz. Its depth depends on the level of the bounce and the mic’s response to sound arriving from below.
For that geometry the bounce lands 14 cm horizontally from the point below the mouth toward the mic. Cover that patch temporarily with a folded thick towel and record the same line twice at fixed gain, once bare and once covered. Use closed headphones and listen for a change in the hollow tone of sustained vowels as well as the brightness of consonants. A towel may take the edge off the upper frequencies while the lower coloration remains. Try a boom that brings the capsule 15 to 20 cm from the mouth with a pop filter. If there is room, move the mouth and capsule together beyond the front edge so the predicted bounce point falls off the desk. Keep the mic aimed at the mouth and listen again before reaching for EQ.
Verify the change before you mount anything
Capture a baseline sweep and a treated sweep from each monitor independently, with the measurement mic fixed at ear height and gain unchanged. In the energy-time curve, align each direct arrival at 0 ms and measure the candidate reflection relative to its own direct peak. Start around the predicted arrival time and cover or uncover one surface at a time to identify it. A peak that moves from −12 dB to −19 dB relative to direct has dropped 7 dB. Use that kind of repeatable change alongside listening. Overlapping arrivals can hide a successful treatment of one path. Room EQ Wizard’s impulse-response guide.
- Repeat the baseline sweep to establish normal variation, then save each trace with the speaker, panel position and gap in its label.
- Check the predicted peak at the main seat and 10 cm forward. If the gain disappears with a small move, revisit coverage before increasing depth.
- Clap from the chair and the standing vocal position. Listen for a shorter metallic tail and check for a surviving parallel path if only one position improves.
- Compare the dry voice takes at matched playback loudness. Keep the mic position that gives the clearest vowels while leaving enough working space to perform comfortably.