Take Me to Your LEDR! Page 2

Fixing problems
No loudspeaker/room system in my experience has even gotten a perfect grade on the very stringent LEDR test. Let's look at each path for problems and solutions.

First, the Up path, which is incredibly revealing of room acoustics problems. Slight errors in the Up image (curving or bowing in) are permissible, representing differences in pinna shape from individual to individual. However, large anomalies in the Up path will affect a system's accuracy, even though height information is not purposely encoded into stereo recordings. If your system gets a bad grade on the Up path, look (not surprisingly) to the ceiling and floor for reflections, which are probably controllable with proper applications of Sonex or RPG diffusors. For example, the B&W 801 monitor has considerable vertical dispersion. If the upward energy from this loudspeaker bounces off a low ceiling, it can produce comb filtering that affects frequency response and therefore the accuracy of LEDR's Up image. Solution: Sonex at a couple of critical points on the ceiling and a thick rug underneath the speakers, to at least 2' in front of them. Also look for near reflections from cabinets in front of the speakers. You may have to put a piece of Sonex on the side of cabinets facing the speakers (this is why LEDR is also called the "spouses beware" test).

Normally, if the Up path is nonexistent or very ragged, look for broken crossovers or (sadly) poor crossover design. However, one respected brand of loudspeakers (Duntechs, according to hearsay) reportedly have no Up image at all, due to cancellations between vertically placed drivers carrying identical frequency ranges. In their case, at least, the missing Up image is not due to crossover or room problems.

While LEDR-testing a pair of Apogee Divas in a poorly treated room, I found the Up and Over paths to be totally nonexistent and the Lateral path very ragged. Although Apogee does not recommend toeing in, I found that toeing in these speakers 1" restored the Up and Over paths considerably. Their sound was much improved, to the happiness of the system's owner, who did not want to make acoustic modifications that would affect the room decor.

The Over path tends to reveal problems caused by obstructions between the speakers. I diagnosed one system that had good Lateral imaging, but the Over path resembled a broken corkscrew. This was (not surprisingly) caused by a medium-size video cabinet and television monitor, which were back against the wall, but still close enough to distort the image.

The Lateral-Between path is easiest to reproduce and therefore usually quite good. If it is not symmetrical or smooth from left to right, look for obstructions between the speakers or rear-wall reflections; also look for unmatched frequency/phase responses between the channels (or your ears).

So far, no speaker system I've tested has reproduced the Lateral-Beyond path very distinctly. To my ears, all the best systems have managed to produce is a diffuse ghost image a few inches to the right of the right speaker, and the same on the left. This problem may be due to small errors in the spatial reverberation program itself. But the Beyond image should appear slightly beyond the speakers and symmetrical (left vs right).

If you have no Beyond image (footnote 5), remember that the side walls of a listening room influence the Beyond path and, if they are very close, can introduce reflections destroying the quality of the stereo image. The typical solution to this problem is matching absorbent treatment on the left and right side walls (or move to another living room).

The LEDR's power
I regret that LEDR does not contain a depth-of-image test. Although the spatial reverberator can simulate depth, calibrating distance is extremely subjective. Despite that lack, LEDR is an incredibly powerful test of stereo imaging and proper room acoustics, which we've only begun to utilize. In fact, professional acousticians acknowledge that LEDR (used by a trained listener) performs many of the functions of a $10,000 acoustic analysis computer. That's why the Editor informs me that Stereophile's reviewers will shortly begin to use LEDR tests as part of the magazine's speaker reviews. In upcoming issues, we will develop a grading scale and words to describe this new tool for evaluating imaging.

Editor's Note: Grammy-winning recording and mastering engineer Bob Katz runs Digital Domain in Florida, whose website has many more superb articles on audio. Bob also wrote about audible differences between nominally identical CD pressings in the December 1994 Stereophile.

Postscript (from January 1990, Vol.13 No.1)
Following my article in December on the LEDR diagnostic imaging recording, Doug Jones, the inventor of LEDR, has discovered that a number of CD players do not reproduce the LEDR imaging signal well. After some tests, he discovered that there was a distinct correlation between a player's LEDR reproduction capability and its interchannel high-frequency phase characteristics.

Those players whose interchannel phase difference at high frequencies is less than 5° reproduce the LEDR image quite well. Since there is a good correlation between the ability to reproduce the LEDR signals correctly and the ability to reproduce a well-defined stereo image (footnote 1) CD players that have excessive (greater than 5°) phase difference between channels at high frequencies (most likely due to a single DAC time-shared between channels, or poorly synchronized clock signals) should be avoided.

The following players, costing in the $130–$475 price range, were all reported as scoring badly on the LEDR test: Philips D6800 portable, Technics SL-PJ1, Technics SL-P116, Toshiba XR-250K, Technics SH-CDB7, and Sony CDP-550. All had from 20° to 45° interchannel phase difference at 10kHz, while a Magnavox CDB472, which uses a separate DAC for each channel, was reported as doing well on the LEDR test.—Bob Katz

Footnote 5: Please don't take this personally.