Equipment Reviews No. 4 December 2002
CARA (Computer Aided Room Acoustics)
Test CD SRP: $19.95
8835c Columbia 100 Pkwy
Columbia, MD 21045
Main: (410) 730-2575
Sales: (800) 234-4546
Test CD designed to help discover characteristic room sound by way of test tones; (201, 187, 173, 160, 148, 137, 127, 117, 109, 101, 93, 86, 80, 74, 68, 64, 59, 54, 50, 47, 43, 40, 37, 34, 32, 25, 20, 16 Hz with distortion at <-80 dB); 20-200 Hz (27 pink noise signals at the same minus the 16 Hz frequency); 4 special pink noise test signals; all 70 seconds in duration per track.
Martin Logan SL3 speakers, Meridian 568 preamplifier, Philips CD transport, Mark Levinson No. 29 amplifier, PS Audio P300 Power Plant, DeCorp, Audioquest, Discovery cabling.
Concepts of Room Acoustics and Purpose of Testing
The booklet that comes along with the CARA test CD includes a brief background explanation of room acoustics. According to the folks at CARA, the total sound field is created by the sum of the original sound waves and all those sound waves reflected and partially absorbed at the room walls. The booklet distinguishes between the different characteristics of large and small rooms. The concern in the smaller rooms is sound coloration, due to the frequency dependence of the interference of direct and reflected sound waves.
Before getting started, there are the usual warnings about volume level, and even a warning about keeping your pets out of the room during testing due to the loud sounds that may cause strong fear in these animals. The purpose of the CD is to find better listening positions and an improved positioning of your speakers. The full CARA Rel 2.1 Plus program (which is specifically designed to determine loudspeaker placement) will be a subject of a future review.
Using the test CD is fairly simple: You play the various tracks, and walk around the room listening to the changes in sound.
Pure Sine Wave Tones
For the pure sine wave tones, the booklet suggests setting the balance to one speaker only, because when both speakers play, there will necessarily be cancellation, and it will be harder to assess the contributions by the actual room. Further, you will find the dependence of the modal structure of the standing waves on the position of the loudspeaker.
I disconnected one of the inputs to the amplifier, starting with the right speaker first and began playing the test tones. Having the CD player remote handy to change tracks would be a great help during the testing. The tracks start at 201 Hz and work down towards 16 Hz. My room is approximately 15 wide x 22 long. I moved the speakers into a different position than I usually use to listen in order to discover if, after using the Test CD, I ended up back close to the original positions.
With the first couple of tones, it was easy to start hearing commonalities that are often expressed in subwoofer set up guides and basic treatises on room acoustics. When I approached the speaker in line with it, I could hear the volume increase and decrease. With the 201 Hz tone the sound along the wall on the side of the speaker was significantly louder than away from the wall. Nearer the center of the room, and around the listening area, the sound was more uniform with the first three tones. The corner behind the speaker gave the most output (as you would expect). The corner in front of the speaker, on the same side, was another position of elevated level. So far, everything is consistent with what we know from experiencenamely, corner placement elevates the level of bass frequencies (even the upper bass), and sitting along, or placing the speaker along the wall is not a good idea. The wall away from the speaker did not result in increased level. The sound in the back of the room was less consistent, but that in itself means it would be a bad place to locate the speaker, or to sit.
When I switched to the right speaker, I expected the level on the wall on the right side to be at a peak like the left, but I soon discovered that was not the case. I can only reason that this is due to an opening (into the room) along that side. In fact, the left side of the room was just as loud as it was with the left speaker. The corner behind and on opposite sides of the speaker was also elevated in level. The area in the center was again more even or slightly lower in level than in certain other places in the room. I needed to play more tracks to try and gauge whether the frequency response was relatively flat in these areas or skewed as the frequency dropped. I already surmised that if necessary it might make sense to move both speakers to the right where the bass would not be as affected by wall proximity (due to the opening), and by doing so, possibly reduce some of the level towards the left wall. This wall has a fireplace on it, so there is not much chance someone would sit at the wall, but at the end of the couch it would be close enough to hear the level increase. The drawback of this shift to the right would be a change in the soundstage width towards the left. Definitely, more experimentation would be necessary.
Still lower frequencies were peaked in the corners and center of the room, but less so along the side walls. It was going to be hard by trial and error to really find the optimal placement, unless I was willing to go back and forth many times. I would think with a spreadsheet (like that supplied by Russ Herschelmann of Stereophile Guide to Home Theater), or a similar calculation of standing waves, might be a quicker way of determining best guess placement. However, the amount of different test frequencies on this disc will prove extremely useful for subwoofer set up and matching with existing speakers given their placement is already fixed. If you have the patience, then you could find the best locations solely by trial and error. From now on I intend to use this disc to aid in the set up of subwoofers.
I listened to a few more tones that were lower in frequency in each speaker singly, and then decided I would listen to the tones with both speakers on. What I discovered was quite interesting. With a listening position about 2-3 back, the upper tones were clearly louder in level and then fell off after about 30-40 Hz. At first I thought farther back would be a better listening position, but soon realized it would be much worse. Moving closer did not really make sense based on the width between the speaker and the set up of the room. The final position was close to where they were originally, about 3-4 away from the front wall, and about 4-5 away from the sides. The speakers are slightly more to the right in the room, making the centerline between the speakers slightly to the right of center.
Narrowband Noise Signals
The booklet explains how the last group of tests allows the analysis of the standing waves to a high degree of accuracy, and I would concur. But, in order to test the audible relevance (using normal music or speech) it is very useful to repeat the sine wave tests with the narrowband pink noise signals. I began in the same manner as before using the same test frequencies, but with the noise instead of the tones.
Sure enough, with the pink noise tones (which sound like a submarine under water), there was less difference in sound as I moved through the room. Corner and wall areas still showed peaks, but walking around the center was less noticeable in terms of shifts in level from one place to the next. Of course, the placement was already tuned, so I didnt expect to hear a dramatic difference, but it was clearly less noticeable audibly with the narrowband noise signals.
Special Noise Signals
The first set of special noise signals is designed to test for colorations and interferences in the midrange. These work in the range of 150-1000 Hz. According to the booklet, pure sine wave tones [only work] at relatively low frequencies. At higher frequencies (several hundreds of Hz) interferences of sound waves will also take place. However, their modal structure pattern is getting more and more locally fine structured, so that distinct cancellations and multiplications will rarely be audible because the sound field at both ears may be different.
The first track has two correlated channels, although the booklet instructs to audition each channel individually. This track sounds a lot like the ocean or a roaring waterfall. This track was incredibly useful for determining the toe-in of the speaker! As I walked from left to right and back again while facing the speaker, I could hear the response rolling offthis of course will be different depending on speaker (mine is an electrostatic, though I would expect a dynamic speaker to sound different as well). In a matter of seconds I was able to set the best toe-in for the exact area that I needed to cover on my couch location. I then repeated the same adjustment for the other speaker.
The next track contains the same signal, but is decorrelated, and should be listened to with both speakers playing and the balance set to the center. The booklet suggests listening to this track with headphones, and then with the speaker set to determine if coloration still exists. If so, the booklet recommends trial and error, or the CARA 2.1 program to optimize listener and speaker positions. I plugged both speakers back in and was happy with the response I got in the center of the room and slightly to the sides. The sound closer up sounded better than farther back. Right now my room is fairly empty due to a remodel, so this might account for the improved nearfield responsedue to more reflections off ceiling, walls, and floor when farther away from the speakers.
Track 58 is really a phase test. There are 6-second samples of in-phase sound with a noise signal from 20-1000 Hz followed by a 3 second sample of out-of-phase noise.
The last track is pink noise from 20-20,000 Hz. It will give you an idea of speaker coloration due to poor design or other tonal deficiencies fairly easily, although the booklet warns that it is not very useful as a test signal for room acoustics. This concludes the test signals contained in the CARA Test CD.
For those who are not very technically inclined, not willing, or do not have access to sophisticated room analysis software, this test CD can greatly help with loudspeaker and listening position location. The first group of tests helps to find standing wave locations, and obvious areas where one should NOT put the loudspeaker or the listener. With experimentation, you can find better locations in the room via these tones. The second set of tones will give you an idea of how audible the changes are from one location in the room to the next. The last group of tests is for proper phase, obvious tonal colorations (via pink noise), and will greatly help in determining midrange colorations due to room placement and/or toe-in.
Personally, I found that the wide range of test frequencies in the bass would greatly help with subwoofer set up. In addition, the midband noise made setting loudspeaker toe-in a breeze. For this alone, I think the CARA Test CD would be worth the money.
- Brian Bloom firstname.lastname@example.org
To Index of Equipment Reviews for this month
Return to December 02 Home Page