Articles written either for MIX
Magazine or Recording Engineer/
Producer Magazine or Modern
– by Bob Hodas
Gain insights into studio listening environments and rooms.
Speaker Placement and Acoustic Environment
Effects on Nearfield Monitor Systems
MIX Magazine, November 1994 – by Bob Hodas
As an independent recording engineer, I have been using Nearfield monitors for the past fourteen years. The decision to purchase my own nearfields was made in 1980 after a disastrous recording project in Japan. Studio selection was handled by the recording company and everything was set in stone when I walked off the plane. Three separate studios were used for tracking, overdubs, and mixing. All three studios sounded completely different since they had different monitors and acoustic designers. I brought along a tape that I was familiar with, (these were the days before CD) and tried to grasp a reference in each room. Well, this turned out to be quite difficult, especially in the mix room.
All of the staff engineers were quite proud of their rooms and I was very uncomfortable suggesting that there were acoustical problems. One studio's staff was absolutely beaming because they had not paid a design fee, even though this was obviously a room done by a famous designer. They had taken the design drawn up for their European parent company and used it for their own room. The only problem with this was that the Japanese room was much smaller and they had simply shrunk the design dimensions using a ratio (not to mention that they installed different monitors). I didn't have the heart to tell them that it doesn't work that way. The perfect finale occurred at the mastering studio whose dimensions were a perfect cube. Standing waves were a nightmare.
So upon my return I purchased a set of speakers that was to become my reference standard no matter where I went. At this point I would like to give credit where credit is due for those of you that don't know your audio Hall of Fame history. Ed Long's Calibration Systems in Oakland CA holds the trademark for the term Nearfield as well as other industry sweeping innovations such as Time Align. Out of respect for Ed, engineers should know that today Ed's term "nearfield" has become synonymous with close field monitors systems just like Kleenex is synonymous with tissue paper. (Wait a minute, wrote that one already)
Within two years, nearfields became a standard in the industry as staff engineer positions vanished and the independents roamed. Engineers found themselves fooled by in house monitor systems that were inaccurate or not properly maintained. I must admit that having a standard helped me to make better records, but I still found that my speakers could sound different in a variety of studios. This history led me into the field of room measurement as I had a desire to quantify exactly what I was hearing, all in the quest of making a better record. I hope, in this article and those to follow, to share with you the knowledge that has been gathered in the many rooms I have analyzed and voiced.
I want to make it perfectly clear at this time that I firmly believe in large room/transducer interface designs. Large soffit mounted monitors can sound fantastic and at the same time be more fun to work on than small speakers. If properly designed, they may also be more accurate than untuned console top speakers. I have several room voicing clients who are very fastidious and proud of their monitoring environments. They keep their rooms regularly voiced, recone woofers and replace diaphragms on a regular schedule. Future articles will deal with many aspects of soffit mounted speakers and their room interactions.
I would like to address one main issue in this article. That is the pervasive belief that if you use a console top speaker, you are not affected by the control room acoustics and will get a more accurate frequency response. This line of thinking has also led many people to believe that home studios can get away without acoustical planning or treatment since the speakers are in your face. In a word, WRONG. I plan to show in the charts below that nearfield monitors can be accurate only if care is taken in the placement of the speakers and room issues are not ignored.
Allow me to take some space to tell you about the measurement system I use. This is important as the charts you will be seeing in the articles to come need some explanation. I utilize a Meyer SIM 3 System and consider it to be the state of the art in acoustic test instruments. SIM won the prestigious R&D 100 award placing it in a class with MIT, Bell Laboratories, and Los Alamos National Laboratory among others. SIM allows me to gather large amounts of information in real time which aids in diagnosing problems quickly. In one screen I can display a room response pre and post EQ as well as the EQ curve applied to the room along with an analysis of the system coherence. Other screens allow me to look at room and phase response in real time. Room reflections may be identified and time alignment of components are clearly displayed (20sec.). The system gathers information at 1/24th octave resolution (245 frequencies, 8Hz-22kHz) which gives me the ability to look very deep into a room. Test signals include impulses, tones, noise and even music may be used.
Though many of you may have seen similar charts in articles by Roger Nichols (we both use SIM), in the articles to come I will be introducing other parameters in addition to frequency response and phase. One such parameter is Coherence, a 245 point signal to noise ratio (on a per frequency basis) of the system under analysis. It compares the test signal source to the signal received at the microphone. This can show direct vs. reflected sound as well as distortion in the system. The Delay Finder (you will see graphs in articles to come) displays an impulse response that shows time alignment and room reflections.
The charts should tell most of the story so I will dispense with longwinded descriptions. Also, for ease of interpretation in this tiny display size, these charts have been smoothed to 1/3rd octave. To make specific points in future articles I'll petition for more space and show you 1/24th octave resolution. The frequency chart is scaled for 6dB per division on the vertical (indicated on left side) and 20-20kHz on the horizontal (indicated on bottom). The frequency response is the line in the middle of the chart. The line at the top is the coherence factor and is scaled into the top half of the chart (indicated on right side). In a good situation, this line would be fairly flat and nestled into the top quarter of the chart, occasionally dipping down as far as the 0 line. The lower the line, the worse the coherence and wide dips indicate serious problems.
So now let's start with a look at a typical console top mounted speaker response. I want to make clear that the charts to follow do not represent isolated experiments in a single room with a specific brand of monitor, but are typical of many of the rooms around the country which I have analyzed. Three different popular speakers are represented in these charts, all of which display fairly flat free field response.
Figure 1: Cancellation pattern caused by monitor sitting horizontally on console bridge
Figure 1 shows the theoretical problem of placing a monitor on its side on top of the console bridge. Reflected signals off the console surface should combine with the direct signal creating a comb filter and canceling certain frequencies that correspond to the path length difference. Some low frequency anomalies may also occur as the console bridge could act as a baffle for the woofer. So, can we prove this theory? You bet! Figure/Chart #1 shows the response of an average studio nearfield placed to replicate fig. 1. Please note large holes in the response at about 1150 Hz (-8dB) and 3kHz (-10dB). Also note that the coherence at these frequencies is pretty bad, indicating reflections. If we were to look at the delay finder charts (sorry not enough space) we would see the reflections within the first few milliseconds that are associated with those holes. Also note that there is a +6dB bump at 150Hz most likely as result of console baffling. Looks pretty bad doesn't it! But don't despair, let's see if we can correct the problem.
Figure 2: Proposed solution to fig. 1 reflections.Speaker positioned vertically on stand a short distance behind console.
Figure 2 shows the theoretical solution of placing the monitor on a stand a short distance behind the console and standing it up vertically (the way it was designed incidentally). This should change the reflection angles so that the critical frequencies are below the listening position. We would also hope to eliminate the console baffle. Does this solve the problem? (For the answer send $50 to PO Box..... oh what the heck, I'll give it away for the love of music). Figure/Chart #2 shows the same speaker as Figure/Chart #1 positioned vertically, 8" behind the console on a stand. A respectable difference, eh Watson? The holes have filled in and the coherence is dramatically improved. The bump at 150Hz is gone and bonus time, the bottom end response (38-60Hz) has even improved by several dB. This looks like a desirable solution to me.
Now it's time to address the low end interaction of nearfields in the control room environment. Measurements were taken in a different studio from that used above. Figure/Chart #3 represents a soffit mounted main speaker system. Note that the chart scale only displays 10-200Hz. In this control room, modes cause cancellations of about -10dB at 50Hz and -8dB at 100Hz. A look at the nearfield response in Figure/Chart #4 demonstrates that the room modes are also effecting our supposed reference. The nearfield is effected by cancellations at 58Hz and 100Hz and in fact has some major bumps up as well at 82Hz and 135Hz.
I am not suggesting that nearfields will always display the exact same problems as the soffited speakers. Position in the room can make a big difference. I have solved certain problems by moving speakers a mere six inches. The important point here is that the room acoustics do affect the nearfields, especially if the problems are gross. We cannot expect to avoid a room problem just by putting a speaker up close to our face.
These low end anomalies are usually room dimension problems but sometimes can be diaphragmatic as well. An example of a diaphragmatic problem would be an unreinforced wall that vibrates at a specific frequency and cancels it out of the room. What are the solutions to these problems? There are two ways to go about it. I often recommend contacting an acoustician once the problem has been identified. You need an expert to give you cost effective solutions. Solutions may include such things as bass traps, resonators, diffusers, or even moving walls. Some dimensional problems and placement loading problems may also be solved cost effectively using a minimum phase parametric equalizer. You need a minimum phase EQ because the room problems described here are minimum phase phenomenon and must be corrected as such. So there are your two solutions, not simple but either is the problem. These solutions are also not necessarily mutually exclusive. The combination of acoustic solutions combined with judicious use of a high quality minimum phase EQ can produce stellar results.
Commercial studios often address their low end problems with acoustics and EQ. Many still have some problems but the majority of gross problems I have run into have been in project studios. Because project studios are often in homes, they share certain dimensional restrictions such as an eight foot ceiling height. Figure/Chart #5 demonstrates a response that represents problems often found in the project studio. Looking at the upper mid range we see that this room exhibits some mild coherence problems. Cancellations are present at those frequencies and are the product of a combination of console, ceiling and wall reflections in the listening position. The large, broad bump in the low mid and upper bass regions is typical of speakers being positioned too close to a wall or corner (1/4 or 1/8 space loading). Many small rooms suffer from an abundance of 150-300Hz. The main signature that seems to proliferate in the project studio are modes that make the low end look like the Alps. In this room the peak is at 58Hz with a dip at 80Hz. Once again, this kind of response appears to be the norm.
Allow me one more example which for me is a pet peeve. It is the addition of home stereo subwoofers to near fields with the expectation that all low end problems will be solved. Figure/Chart #6 demonstrates what happens when an improperly placed subwoofer combines with the floor/ceiling mode to create a gaping hole in the low end response. I have also seen several subwoofers where the cross over point does not meet the manufacturers specification. This can cause some significant problems as well. Can you believe that Chart 6 is someone's idea of a mastering room? I certainly hope none of my records get mastered there! Aside from the 70 Hz crater, side wall, ceiling, and tabletop reflection create severe mid range holes. Coherence in this room is very poor. The sad part is that this is a professional studio who thought they would get into the mastering business just because inexpensive digital mastering programs are available.
Pardon my soapbox but don't think you can open a mastering room just because you can buy cheap digital mastering programs. Especially if you don't pay attention to room response! I hope that this is not a growing trend. There is a lot to be said for relying on an expert's ears and abilities in a room where many records have been mastered.
Care and attention need to be taken when setting up any listening environment whether it is an existing professional control room or a new project studio. I hope the information above will dispel some myths and help you to make better records. I'll be back soon with some more interesting charts that should be SIMply revealing.