DSP Devices for Room/Speaker Correction


The current issue of TAS has a review of the TACT RCS 2.0 DSP, a device devoted to correcting problems in the interface characterisitcs between room and speaker. There are several other models of DCS devices, such as the SigTech, Accuphase, and Perpetual. I long ago concluded that there was little more I could do to correct some of the anomalies in my listening environment, but the TAS article really caught my attention. I'd like to hear from others who have had some experience with DSP room/speaker corrections systems.
sdcampbell
Well, equalizers in general are, theoretically, room correction devices. The problem is, in the analog domain, every time you introduce another device into the stream, you take another derivative of the signal, and therefore introduce phasing problems. So a lot of audiophiles were against it, unless their room was severely "out of whack" (pardon the technical jargon). With the advent of such devices in the digital domain, this particular problem goes away. If I could afford it, I would definitely get such a device, as I used to have an equalizer in my system, and it did make a big difference. The other problem is that the "out-of-whackness" of a room is a function of where you are in the room. In some rooms (including my current one), those differences can be quite severe. Boominess, for example, goes up as you get closer to the walls, and even more so towards the corners. I've tuned my system to the sweet spot and refuse to listen critically anywhere else. This particular criticism is probably still applicable towards even the digital system.
I have also heard that systems sound better if "you" are not a part of the room acoustics. Maybe some type of clothing or a special suit is in order, perhaps a remote listening device?
PS: Can you imagine listening to your "speakers" on a quality set of headphones?
Most DSP based room acoustic correction equipment assumed the listener sit at a fix location and throught a closed loop monitor and apply compensation techniques (microphone to digitizer to DSP to amplifier to speaker) nulling out the anomaly cause by interaction between the speakers and the room. The problem arise when the listening position is changing (listener move about in the room) then the system must be retrained to compensate for a new listening location. Currently there is not existed a powerful DSP device that is capable of doing that kind of realtime correction so there are still room for improvement. Cheers,
Well, Dekay, if what you say is true, a real audiophile would just absent himself from the room, if not the house altogether, and be satisfied that by doing so that he had enabled his system to play at its best. Not being able to hear his system would be just another sacrifice a real audiophile would be willing to make. Wouldn't you agree, or does this belong on another thread...
Jayboard: It was just one of those thoughts not unlike the ones that we all have when we are about to fall asleep, which is when it came to me. We probably don't need to carry it any further. The DSC devices do sound very interesting if they are an improvement on and easier to use than the old "slide" equalizers. I have set up a few systems using these and always felt that they had mixed results. They seemed to work better on large multiple driver speakers in larger rooms. There tendancy to add distortion and compress sound was more noticable on smaller systems in smaller rooms, this is just my experience. The setup task was also very demanding in that just achieveing the the first step of a flat response with test signals could take many many hours. I did enjoy the second phase of adjusting by ear, especially if the system owners were involved in process. The longest I ever spent at one system was 14 hours straight, it ended up turning into a party.
Dekay: you crack me up. Heisenberg's uncertainty principle, as applied to audiophilia.
I purchased an RCS 2.0 about a month ago. It is a very interesting piece and overall, a worthwhile addition to my system. The reality is that I'm still learning how to best use its capabilities. In short, it does what it's intended to do: reduce, if not eliminate, the vagaries of room acoustics as a factor in sound reproduction. Its effect is obvious even to casual listeners; non-audiophile friends immediately characterized it as "lifting a veil" from the music, which is exactly what it does. It has benefit in all parts of the spectrum, smooothing out the highs, giving greater presence and immediacy to the midrange, and controlling room boom in the bass for tighter sound. Overall it's a much more articulate and engaging musical experience. There is no doubt that it raises the quality of the sound to a higher level. In fact, the sound is so very clean that it takes a little getting used to -- we've become quite accustomed to room artifacts like room boom and mid/high range peaks and suck outs. A lot of what we think of as bass is actually room resonance. Getting to like the effect of the Tact is like learning to enjoy fine wine... it takes a bit of effort, but once you've gotten the taste for good stuff, you can't go back. I was worried that the "sweet spot" would be rather small, confined to a small volume centered where the calibration mike had been placed. In my room the sweet spot is large enough to accomodate any of three seats on a couch (although the very best effect is admittedly right dead center at the mike position). The unit is not without some downsides. It does take some practice and understanding of signal theory (e.g., impulse response, frequency response curves, and the effect of filters on frequency response) to program the unit effectively. This is not a "plug it in, let 'er rip" sort of process. It's more like learning to play a musical instrument: getting a good set of room measurements, selecting and tweaking the desired target frequency response curve take a bit of skill and care. The software doesn't prevent the user from doing dumb things; indeed the measured data hints at ridiculous possibilites such as extending the bass response down to 10 Hz or upwards to 20 kHz (absent any real capability), thereby forcing unnatural acts upon the system power amp and speakers. The user interface (a program running on a PC attached via R2-232 while the calibration is being done) is not the most obvious, user-friendly thing in the world. I would also have liked Tact to incorporate 24/96 upsampling (the current design puts out the same digital format that it's presented with at the input). Scaling of the correction filters to prevent potential output overload can lead to loss of resolution of a couple of bits. This loss of resolution can be of concern and is particularly noticeable on some disks (like orchestral music) that are recorded at relatively low level. The other thing I've noticed is that the sound is best at moderately loud listening levels; at low listening levels it sounds overly tight and constrained. Even Tact recognizes this effect and allows you to program lesser levels of correction for low listening volumes (the RCS 2.0 can store up to 9 different room correction filters for various situations). This apparently lets the room "ring out" more at low listening levels an impart a livelier feel. Would I recommend this unit to someone? Yes, provided they are willing to invest the energy to understand the problem it attempts to solve and how to use the unit effectively, just as one invests energy into issues such as speaker placement and room acoustic treatment.
Dekay, this is hardly worth mentioning, now that I've been away for a few days, but my followup comment (11-2) was meant in the same playful spirit as yours. Actually, I had the amusing thread about what it means to be an audiophile in mind. Sorry if the necessary crosstalk didn't get through the e-mail!