Welcome back. By now you've read about my equipment selection, paper design, and construction. Now it's time to fine-tune the system. As you'd expect, there are two subsystems that are individually adjusted: audio and video.

The AVIA Guide To Home Theaters

To properly adjust these systems requires that we have signal sources that allow us to see, hear, and measure performance. Expensive test equipment works well, but is prohibitively expensive. Ideally, we'd like all the signals we need to be encoded onto one convenient and affordable DVD. The first such notable DVD was Video Essentials. It was very comprehensive in its coverage of video performance, but fell short of offering all the signals required for optimum audio performance. Enter Ovation Software's AVIA Guide to Home Theaters.

This DVD was carefully assembled by people who fully appreciate the total needs of the home theater enthusiast. It contains the most complete set of audio and video test signals one could possibly need. The basic and advanced video test patterns permit you to adjust all the parameters that affect the visual performance of your monitor or projector (some require special test equipment to get it absolutely right, like gray-scale). Colored gels are thoughtfully included for those who can't turn off individual color guns while adjusting hue and saturation. There are also a number of test patterns that allow you to assess the capability of your display and DVD player.

Video Test Patterns

I was particularly intrigued by the test pattern that demonstrated how many pixels my DVD player drops at the edges of the picture. It seems that the interrelationships among MPEG-2 encoding, DVD picture dimensions in pixels (720 x 480), and the NTSC video standard forces a few of the pixels to be discarded. This usually occurs at the right and bottom edges.

Now I have some bad news, a dirty little DVD secret that I'm sure most of you have already discovered. Even if you set up your display very carefully so that it conforms to the strictest of standards, many DVDs still don't look quite right when viewed. For example, when I used the AVIA to set my color saturation and hue, I found that I had to increase color drive on the Faroudja to obtain proper results. (The hue tracked perfectly, a testament to the accuracy of Faroudja's color processing.) Since I have a front projector in which I can turn off each gun independently, I didn't have to rely on the gels; I was assured of the most accurate setting. After these adjustments were made, I checked some of my reference DVDs (Fifth Element and Starship Troopers, to name two). I found that their color levels appeared too high. I exchanged several e-mail messages with Ovation Software's Dr. Guy Kuo who confirmed that I had adjusted my projector correctly and assured me that his test patterns were derived directly from calibrated digital sources. I was left with only one other source for the problem, the DVDs themselves. So I investigated.

Criterion includes color bars on their DVDs. I pulled their releases of Time Bandits and Armageddon off the shelf and displayed the color bars. Not only were the Criterion color bars different from the AVIA color bars, the Time Bandits and the Armageddon color bars were different from one another. So we've been gifted with the single most accurate delivery system for prerecorded home video devised to date, and the studios and their subcontractors can't seem to keep their production equipment in calibration. Pitiful. So after spending hours fine-tuning your video display, be prepared to adjust the brightness, contrast, hue, and saturation by eye to optimize each DVD. But, please don't allow this situation to discourage you from adjusting your display for best performance with a calibrated source like the AVIA DVD. This is still an extremely worthwhile endeavor that will improve the visual fidelity of your system. Let's move on to the audio.

Audio Test Patterns

The AVIA DVD contains both basic and advanced audio signals. Many are exceptionally valuable even if your ears are your only instrumentation. For example, let's assume that your surround speakers are made by a different manufacturer than your front speakers. Or perhaps the amplifier you're using for the rear channels is a different brand than the one you're using for the front channels. Not all speakers and amplifiers maintain absolute phase, so there are signals that will allow you to confirm by ear that the front and surround speakers are in phase. (When you're equidistant from two balanced speakers, the test noise should sound like it's coming from inside your head rather than coming from two different sources on either side.) Similarly, you can confirm phase between any two speakers in your room, including the subwoofer and any other speaker. There are low frequency sweeps that will help you find all those annoying buzzes and rattles.

Then there are audio tests, like absolute sound level, for which an inexpensive sound level meter (like the Radio Shack) is required. As with the video test patterns, there are many more audio tests too numerous to describe. Most important for me, this is the very first DVD that includes individual chapters of wide-band pink noise suitable for the third-octave analysis required for room/speaker equalization. (Previous such DVDs included band-limited pink noise suitable only for balancing - those signals are included here as well.) In the equipment selection chapter, I explained the need for equalization and the reasoning behind the selection of parametric equalizers in particular. To adjust them properly, each individual channel is driven with wide-band pink noise. A calibrated microphone is placed at the location of the listener and is connected to a third-octave band analyzer (I use an AudioControl Model SA-3050A). The analyzer displays the level of the noise in each of thirty third-octave wide bands from 25 Hz to 20 kHz. Because the noise is random, the analyzer has to be set to average the levels (slow integration). The process is a slow one for the low frequency bands, taking a couple of minutes to settle down into a high confidence measurement.

Each pink noise chapter lasts just one minute, but a longer duration is necessary for accuracy. You cannot program your DVD player to repeat any of these chapters; if you try, you'll get an error message explaining that program play of title six is prohibited. Fortunately, the Ovation folks thoughtfully added a feature that will repeat a chapter over and over again. But there is a quirk. The button on the screen reads Repeat Pattern On. This is not the current state; it's a command. Maneuver to the button and select it and the caption will read Repeat Pattern Off. That means it's on. A little confusing, but as long as you know the rules...

The X-Curve And Other Adjustments

The object of the equalization exercise is to manipulate the center frequency, bandwidth, and insertion gain or loss of each filter in the equalizer until the in-room speaker frequency response is linear (without lumps and bumps) for each channel.

I went through all the test patterns and signals that applied to my system and I was delighted to find that my video measures as good as it looks. Few adjustments were necessary. And the audio was pretty good even before I touched up the frequency response. A few words of caution. Explanations for several of the video patterns warn that excessive brightness and contrast will burn your display. Take this to heart. There is nothing wrong with the video levels on the AVIA DVD; you simply have to be aware that you can abuse your system with it if your display is not properly set. Also, if you choose to equalize your audio subsystem, experience has shown that a gentle slope of about -2 dB per decade sounds most natural for properly recorded orchestral music. A flat response seems a bit harsh, overly emphasizing the high frequencies.

This is not to be confused with yet another curve the motion picture industry has thrown us, the X-Curve. All motion picture theater sound systems are equalized above 2 kHz to roll off at -3 dB per octave. So soundtracks mixed to compliment this standard will sound overly bright on a relatively flat system, such as you may have at home. This is what THX re-equalization for home theater is all about; it rolls off high frequencies to conform to the ISO 2969 X-Curve equalization. I would expect an uncompensated soundtrack running at +3 dB per octave above 2 kHz to sound harsh, and for the soundtrack's voices to be excessively sibilant. I have not noticed such a general characteristic on my system. Perhaps someone could enlighten me concerning whether the X-Curve is routinely applied to soundtracks when the DVD is prepared.

I should mention that the AVIA DVD begins with an extensive set of short lectures to tutor the novice on home theater concepts and terms. My biggest disappointment with that section of the disc is that it barely mentions the 16x9 enhanced video mode. Here was an excellent opportunity to clearly define and demonstrate the nature of this important feature and its advantages. Opportunity lost. Perhaps this oversight can be corrected in the second edition. (Note that there are some helpful 16x9-specific video test patterns on the AVIA DVD.) There is then a guided section during which certain measurements and adjustments can be made. Another word of warning. Don't be alarmed when you see the first test pattern, the needle pulse pattern. As the narrator is speaking, a screen-shot of a misbehaving monitor is displayed. The problems you observe are not a reflection of the performance of your display. After the narrator completes his description of the test, the real pattern will appear. With any amount of luck, your display will respond a lot better than their demonstration unit.

Let me leave you with a hearty recommendation for the AVIA Guide to Home Theater. I found it to be a very helpful and powerful tool.

(If you have any questions or comments for the author, say hello to Mr. Blandings here.)