The Well-Tempered Studio

Feb 1, 1999 12:00 AM, By Geoffrey Goacher

           

Improve the sound of your personal studio in three easy steps.

Identifying the problems. The best way to tell if you have a problem with standing waves or speaker/boundary interference is through a combination of listening and measurement. First, listen to a finely engineered CD through your monitoring system at a decent mix level. The CD you select for this exercise should have a tight bass sound and minimal reverb. Some of my favorites are Tchad Blake’s mixes on Crowded House’s Woodface album and Sheryl Crow’s two latest albums, Sheryl Crow and the Globe Sessions. Tchad Blake uses interesting imaging and minimal reverb, which makes his mixes great for critical listening exercises. (But then, that’s the style of music I mix; you may prefer something else).

FIG. 3: The omnidirectional nature of low frequencies causes them to reflect from all nearby room boundaries.

As you listen to these CDs through your system, notice whether the bass sounds tight, smooth, and consistent in volume. If the mix sounds full and warm, then your room naturally promotes good bass response. Larger “small rooms,” rooms with lots of windows, and rooms with lightweight walls tend to balance bass frequencies nicely.

However, if the room rumbles and booms with the music, or if the bass either sounds mushy or alternates between high and low levels, then you may have a problem with resonance or speaker/boundary interference. At this point, it’s a good idea to take a precise measurement of your room’s resonant characteristics.

Simple measures. Measuring for room resonance and speaker/boundary interference requires a high-resolution frequency analyzer, rather than the usual octave or 1?3-octave real-time analyzer. Octave and 1?3-octave analyzers average out too much information to be useful for this task. Fortunately, some new software programs allow you to perform high-resolution acoustical analysis affordably from a computer (see the sidebar “Acoustical Programs for the PC”).

For example, Figure 4 depicts a high-resolution, low-frequency response graph made with AcoustiSoft’s ETF 4.0 software. Measurements such as these can help you better identify the problem areas in your studio. Given the high resolution of the measurement, narrow notches in the response aren’t that bad, but you should pay attention to the general frequency-response trends. Notice, for instance, that the average signal level is around -16 dB, with a boost of 8 dB centered around 300 Hz, a 12 dB boost around 125 Hz, and one sharp 16 dB boost around 55 Hz.

Based on this measurement, you can guess that our monitoring system would sound muddy in this room because of the 300 Hz boost, too bassy because of the 125 Hz boost, and too boomy because of the peak at 55 Hz. Given that the specifications for the loudspeakers used in this test are flat throughout their low end, it can be assumed that these low-frequency response deviations result from room influences.

FIG. 4: Shown is a high-resolution, low-frequency response graph made with AcoustiSoft’s ETF Room Acoustics Analyzer 4.0 software.

Fixing the problems. A good way to smooth out resonance and boundary-reflection problems is by optimizing the location of the speakers and the listener in the room. Resonance and boundary reflection are less pronounced in certain areas of a given room. In fact, changing the location of the loudspeakers and listening position often results in a drastic change in sound from the previous location.

Optimizing speaker and listener locations used to be a process of trial and error. Now, however, there are several PC-based programs that can model the acoustics of your room and help you find a location where room resonance and boundary reflections are minimized. I used a speaker/listener optimization program on the same monitoring system shown in Figure 4. Then, after changing the location of the listening station, I remeasured the low-frequency response (see Fig. 5). As you can see, the optimization program worked well: the average signal level is still about -16 dB, but the significant boosts shown in Figure 4 have been smoothed out. The only exception is that the boost at 55 Hz remains.

The frequency response in Figure 5 still appears to have a lot of variation because of the high resolution of the measurement. If measured using a lower resolution of 1?6-octave or 1?3-octave, however, the frequency response would appear as a nearly flat line.

After optimizing the placement of the speakers and listener, you can do several other things to further reduce any problematic boosts in the low frequencies. Applying normal acoustical foam works well to dampen high- and mid-frequency energy but doesn’t adequately absorb low frequencies. You can absorb low frequencies by using a bass trap, which is any acoustical device that absorbs low-frequency energy in a room.

Often, the best place to put bass traps is in the corners of rooms, because that’s where low-frequency energy collects. As the low-frequency energy is absorbed, the various peaks (as exemplified in Figures 4 and 5) are reduced, resulting in a smoother bass sound overall. The average listening room benefits from having about 1 percent of its total volume dedicated to bass trapping.

Many companies manufacture broadband bass traps, but one in particular, Acoustic Sciences Corporation (ASC), also offers affordable acoustical consulting for treatment of low frequencies using a test that they developed called the M.A.T.T. (Musical Articulation Test Tones) test.

Step 2:

REDUCING EARLY REFLECTIONS

The second step deals with frequencies of 500 Hz and up. This range has a critical effect on the accuracy of the monitoring system’s imaging and its mid- and high-frequency tonality. The biggest detriment to mid- and high-frequency accuracy is the presence of early reflections.

FIG. 5: This is a low-frequency response graph of the same room after speaker/listener placement was optimized.

Early reflections. When listening to your monitors, you hear a combination of the direct sound from the speakers followed by the reflections of the direct sound from the room’s boundaries (walls, ceiling, and other hard surfaces). Reflections that hit the ear within 20 milliseconds of when the direct sound is produced are heard as part of the direct sound and are called early reflections. Because sound waves travel at a rate of about one foot per millisecond, most of the first reflections that make their way to the listening position in a small room qualify as early reflections (see Fig. 6).

Early reflections often add audible comb-filter distortion to the direct signal, tainting the frequency response with a variety of boosts and dips. Early reflections also tend to blur the stereo imaging between the speakers, making it difficult to accurately hear the exact position of sounds within the stereo field.

Identifying early-reflection problems. The best way to determine whether you have a problem with early reflections is to listen for and measure them. For this exercise, play a well-mixed CD that has clear and precise imaging, such as one of those mentioned in Step 1.

As you listen, notice whether the locations of the instruments are clearly identifiable in the stereo spread or whether they blend between the speakers. You should be able to hear the various instruments coming from specific points in the stereo field. Problematic early reflections, however, will degrade the aural clues that help us identify stereo imaging, and the resulting mix will sound blended and fuzzy.

Measuring for early reflections requires a high-resolution analyzer that can generate an impulse response, or an energy-time curve, of your environment. Or you can use one of several acoustics analyzer programs mentioned in the sidebar “Acoustical Programs for the PC” to generate an energy-time curve. This kind of measurement will give you a clear idea about any problems you might have with early-reflection levels.

In Figure 7, the direct sound from the speakers is shown at 10 milliseconds, with the early reflections occurring between 10 to 30 milliseconds—a period of 20 milliseconds. In a balanced acoustical environment, the reflections between 10 and 30 milliseconds would be 15 to 20 dB below the level of the direct sound. In other words, early reflections should be virtually inaudible.