Now it is well known I have been highly skeptical of automated room correction and Eq systems. I did have a demonstration of Audyssey from Kris Kyriakakis of USC. He was using SVS MTS-01 speakers to demonstrate Audyssey at SOTU. In all honesty I was underwhelmed by the demonstration. I thought the bass was improved, but that overall the sound was worse. Kris canceled all except low frequency correction and then things were improved.
So after doing a careful manual set up and calibration of my updated rig, and living with it for a few days, I unleashed Audyssey on my very unusual system not expecting it would make any sense of this highly unconventional system.
The most unconventional are the front three, especially the main front left and right speakers. So it is necessary to give a description of operation.
As many know I’m a full ranger at heart, and regard all crossovers as a necessary evil. So the concept is to keep the system as close to a full range driver as possible. In fact there is only one complete electronic crossover in each three fronts. These are between bass/mids and tweeters. The mains contain independent aperiodic transmission lines. The center is a single line.
The 7″ drivers in the mid line are allowed to roll off in the bass acoustically. F3 is 42 Hz, 44Hz for the center. The bass line has an F3 in the room of 23 Hz.
An active crossover sends the bass and step response correction to the upper of the two 10″ drivers.
The LFE signal is low pass 60 Hz and fourth order roll off. This signal is sent to the amps powering the lower of the 10″ drivers. The LFE signal is also sent to mono sum buffer amp and then sent to the amps powering the upper 10″ drivers along with the diffraction compensation signal. This is to keep the diffraction compensation stereo. Diffraction compensation is performed by the 10″ driver and not through the 7″ drivers, to off load them and prevent large cone excursions.
There is a low pass crossover that provides diffraction compensation to an amp supplying the upper of the two drivers in the center speaker. One of the biggest reasons for an active crossover is to have precise control of diffraction compensation which varies greatly with speaker boundaries and room placement. This is in no way possible with passive crossovers and is a big advance in accurate reproduction. The design concept minimizes phase shifts and time aberrations due to crossover artifact.
The surrounds are sealed speakers with Dynaudio drivers with a Fs of around 50 Hz second order roll off.
The rear surrounds are biamped dual transmission lines.
The system is optimally configured with all speakers set to large, and the speaker bass management set to both with the LFE crossover at 60 Hz.
A careful level calibration was done prior to utilizing Audyssey.
So I set up the rather cheap feeling calibration mic in the first position on a tripod. I plugged the miserable small gauge mic cable into the jack on the front of the AV 8003. With not much optimism I started the calibration, of the staccato tones played via each speaker in turn and at the end all three fronts. There were two false attempts from problems such as a noisy fly that got in the room. However third time lucky. I measured in six positions.
So then I checked results expecting to see nonsense. Well I was dumbfounded that it had done the speaker set up as designed. It did not change any speaker levels from my calibration, and imagine my surprise when I saw it had figured out all speakers should be set to large and the LFE crossover at 60 Hz! All just as intended.
The only confusion concerned the woofer distance. It placed the woofer 5 ft behind the mains, and yet they are the same driver. I thought about this and realized the program has averaged the driver and port arrival times. This I had to reverse and manually put woofer and main speakers in the same place as there would have been phasing trouble with the upper diver as it is receiving a mix of signals. It also altered the distances of the other speakers from measured. However I think this has had a highly beneficial effect on getting a more even result though out the room.
Here are the results of the Audyssey frequency response correction. In the main except for the surrounds between 1 and 2 kHz the mid band correction was insignificant.
However for all speakers except the rear surrounds there was a first order lift put in place above 8 kHz. I think this was because in three of the testing positions the microphone was close to the rear surrounds. There is a fall off of HF with distance and this in my view should not be corrected. Listening test confirmed this.
The system sounded better with the frequency response correction canceled.
However I was intrigued by this, so I did and experiment purposely introducing a frequency response error. I boosted the level of the bass amps to the rear backs below 150 Hz by 7db. I did a run of Audysey and it caught the problem and corrected it.
So in summary I think this program is useful and may well take the edge off loudspeaker errors. I suspect this is much more a loudspeaker correction program than room correction in a lot of cases. For instance I know my surrounds have a slight null between 1 and 2 kHz that was identified by Audyssey.
I feel the most useful aspect in my studio was the re timing of the speakers. I think this made a difference in two areas. The sound is more uniform over the seating area.
However I think the greatest benefit was the improved sense of space achieved in 7.1 reproduction, especially. I was definitely hearing different spaces captured. In essence the reproduction was to a very large extent freed from the boundaries and constraints of the room.
I will make more comments about this when I report on some of the discrete multichannel discs I have listened to.
My biggest issue with the system is what I feel to be inappropriate last top octave equalization. I would encourage adopters to be alert to this.
My preference would be for Audyssey to only make cuts in the top octave and leave alone any high frequency loss above a first order roll off.
I have to say I was fascinated evaluating this system and regard it as a work in progress. I think we are early in total system/room Eq and suspect there is still a lot to learn.