High-fidelity dork-out
My listening room is comical because the TV sucks so bad. It was a freebie that I trash picked from an art show. I don't really do surround-sound either, I like two-channel and I spend a lot of time listening to music... mostly from the computer, sometimes from vinyl. I can't remember the last time I took out a CD. I've taken quite a bit of care to ensure that the vinyl setup is as close to optimally accurate as possible -- but even so I only like vinyl for the fun factor, it is NOT a technically superior medium to digital by any stretch.
I don't have a bunch of money so I put this system together with a lot of gear that I bought broken on Ebay and then fixed. It's 4-way multi-amped through a DCX2496 active crossover/processor. It does lots of different slopes and alignments, time delays, phase correction, dynamics processing, global and channel-specific parametric EQ... basically everything. I have a CX2310 that passes through the two low channels with no alteration but sums their content below about 35Hz to send to two sealed subs. Since the main speakers are passively-radiated, sealed enclosures take care of the "kick" in a small room.
My mains are RTR HPR-12s from around 1973. The woofers and passive radiators are original, and they are fantastic. 92dB sensitivity and they extend smoothly down to 30Hz. The mid and high drivers in this setup are newer -- silk dome tweeters and carbon fiber/kevlar midbass drivers. I originally biamped these speakers (with passive mid/high networks) because I realized that they were capable of much more than the original crossovers were allowing. Soon after, I dumped the paper midranges and piezo tweeters for silk and carbon fiber, and went full-active.
Active quad-amping means that every driver is run directly from its own amplifier channel. The DCX brings them all into time and amplitude alignment, and adds some EQ to address problems with the room. There's a good amount of Auralex room treatment behind the listening position and in the corners, and bookshelves in bass-critical and reflective spots, but EQ is still needed since the room is so small and all the surfaces are quite hard. With processing and measuring from the listening position, the response is within 3dB of flat from 10Hz to 22kHz. With the additional efficiency/headroom provided by multi-amping, the system can reach 113dB SPL at the listening position with typical musical signals before one of the amps clips.
I have two crossover alignments programmed into the DCX that I switch between occasionally. One is a Linkwitz-Riley 48dB/oct arrangement with basic time alignment and EQ. The other is a Butterworth 18dB/oct arrangement by Jean-Michel Le Cléac'h, which uses altered time delays and a reversed-polarity mid to achieve minimum phase variance across the spectrum (about 150 degrees, as opposed to well over 600 in the LR-48 alignment). It's important to note that these two alignments measure almost exactly the same in frequency response -- variance is less than 1.5 dB -- from the listening position in my room, yet they do seem to sound quite different. I'm agnostic on the question of phase audibility, but it's fun to be able to switch between these two.
If right now you're thinking "No! You need an all-analog signal path! Tubes! Better sources! Electrostatics! First-order passive crossovers! Better interconnects! This is so uninspiringly mid-fi..." then you should just close this browser window and go somewhere else. I like science, and I especially like how it helps me produce repeatably verifiable and measurable improvements in playback accuracy. Hey, Jonny Greenwood doesn't like audiophiles, and neither do I.
For the record, yes, all competently-designed solid-state amplifiers do sound the same, assuming that their output impedance does not react too much with the speaker (though even this is debatably audible). Tube amps do sound different, but they also measure differently (higher distortion, which last I checked, means lower fidelity). I don't like tube amps driving speakers (unless we're talking guitar amps), but I like the euphonic contribution of a tube headphone amp to certain headphones.
Rod Elliott has a great article on multi-amping on his excellent website.
HydrogenAudio is the best online forum for getting no-nonsense information from people who like science and disallow snake oil. Everywhere else I've found to be a mixed bag.
Oh and the two subs, the rack, and the turntable stand are all homemade. The turntable, phono preamp, and DAC are heavily modded; and the three NADs and SX1200 were all dead at some point and are now alive again.
Labels: Tech











