Digital Room Correction - Friend or Foe?
Room interaction plays a huge part in how sound is perceived at the listening position. Even the best speakers are hampered by in-room resonance and reflections that bloat or obscure the sound at certain frequencies. In an ideal world, our rooms would be capable of a reasonably flat line audio response between 20Hz-15KHz (the audible range for most of us); the reality however is often far from this proposed optimum. Digital correction software works by playing a full-range frequency sweep, recording the response with a measurement microphone, and then inverting the measured response, thus smoothing room response artifacts.
Initial experiments were performed on the free open source DRC program called ACXO that unfortunately does not appear to support Vista 64 properly. The measuring aspect works fine on the Vista 64-bit OS; however, attempting to apply correction causes the program to hang without ever making a change. If you're using Vista 32 or XP 32-bit, we'd recommend you check it out.
As attempts to use ACXO with Vista 64 failed, I decided to go with the full version of Audiolense 3.0 for two-channel setups. The cost for the basic single user license is around $190, which is not exactly cheap for most of us, but we must recognize the level of work that goes into creating and supporting such products. There are a few alternatives to Audiolense varying in complexity, ease of use, and of course cost. For example, Acourate is a standalone package which seems to have more complexity to its GUIs than Audiolense, although there are video tutorials available on the Acourate website to help users get started.
The beauty of both Audiolense and Acourate is that both companies are new budding enterprises, which leads to support being handled directly by the authors of the software. We hope this continues as emails are currently handled within hours of being sent. Another software option is Art Teknika's Console that can be used with the Voxengo Curve Eq plug-in and used for DRC. The expansion possibilities with Console are huge, with plug-ins supporting digital crossovers and rerouting of input and output signals as long as the hardware used is capable.
Note that all of the software mentioned here features 30 day trial versions with limited functionality to give you a taste of what DRC is all about, so if you're interested but don't want to plunk down hundreds of dollars on something you might not like that's a nice option. You'll also need a measurement microphone and preamplifier. We're using the affordable Behringer ECM8000 and UB802 (now discontinued) to take room response measurements. In all, expect to fork out an additional $150 for these two components. The microphone preamp needs to support "Phantom 48V power" if you're using the ECM8000, so make sure you buy one that's compatible with this requirement.
114 Comments
View All Comments
notanakin - Wednesday, December 3, 2008 - link
Generally I found the article interesting, but why on earth was Vista 64 used for the review? The author admits he couldn't find USB-ASIO drivers, and that ACXO doesn't support Vista 64 properly. Surely first you check what would be interesting for the review and then choose an appropriate OS? Also, I'm pretty sure that Vista 32 or Win XP are far more likely to be used by most readers of the article.Thank goodness he wasn't running Win ME on his PC.
milosz - Wednesday, December 3, 2008 - link
Single Ended Triode amps add a considerable - and DEMONSTRABLE- amount of low number even-order distortion to reproduced sound. To their devotee's, this sound is "full," "rich," "harmonically complex" - etc. But what they're hearing has been ADDED by the electronics. If you listen on speakers like the single-cone types used in the review, adding some additional harmonics to the overtone structure can compensate for the lack of upper treble and poor transient presofrmance of these single-driver systems. These single driver systems store a LOT of energy in the low treble- just look at their waterfall plots. When the signal stops the cone keeps moving, from the high mass storing energy. A ribbon tweeter has a moving mass of 0.01 g while a single-driver speaker like these has a mass of about 10 grams. You tell me which will be able to faithfully folow the motions demanded of complex high frequency reproduction, and which one will have some much inetria that it will keep moving long after the signal has stopped. But if you use a SET on these things, it generates a lot of extra low order even harmonics to sweeten things up - it masks the problems introduced by the speaker to some degree.I think it preferable to use electronics that add as little of their own coloration as possible. That's why I try to use the best electrostatic or ribbon speakers I can afford, with the best class-A amps (tube or solid state, a good amp is a good amp) I can afford, as well as high-slope digital crossovers and top-notch D-to-A converters. I've also tried some SACD and DVD-A audio and frankly it doesn't sound any different to me from CD. Vinyl also has a lot of good music in it if you can tolerate the light hiss and crackles that even the best vinyl playback has. Of course the thing I've spent the MOST money on is the MUSIC. For a n old guy like me who has made a little dough, $75,000 for CDs and LPs, $25,000 for a 2-channel audio system is about the right ratio.
Carnildo - Wednesday, December 3, 2008 - link
I noticed one word missing from this review: "double-blind". To evaluate something as subjective as audio quality, you need to do double-blind testing to make sure the results aren't being influenced by your expectations of what the sound should be like.One thing I'd like to see is an all-digital sound system: the sound is read off the disc, converted to 64-bit floating point, processed for playback (mixing, volume adjustment, room compensation, frequency separation if you're not using full-range speakers, etc), sent out over the wires still digital, and then converted to analog just before it hits the coils in the speakers. Keeping it digital will avoid most noise sources, 64-bit floating-point will give you a noise floor far below what anything analog or integer digital can manage, and you've got the full processing power of a modern computer to adjust for things like an imperfect room layout.
RagingDragon - Saturday, December 6, 2008 - link
Floating point leads to rounding errors, and processing the floating point signal would increase those rounding errors. Todays top DAC/ADC are mostly 24-bit (unsigned) integer, so using 32 or 64 bit integers would be a more than sufficient, and without the rounding errors inherent to floating point math.headbox - Wednesday, December 3, 2008 - link
So you tested audio in a bare room with hardwood floors. Hardly a room for an "audiophile."strikeback03 - Wednesday, December 3, 2008 - link
IMO that is the kind of thing they should focus on - what can be done with computers to get the best sound possible in normal rooms? Kinda like Anand's home theater, music rooms are interesting tech but not something most people have the time, money, or space to implement. So what can be done to get the best sound in a dorm room, or multi-purpose family room, etc.pcfxer - Tuesday, December 2, 2008 - link
"One of the chief perpetrators of this revival was a fellow named Throsten Loesch. "You mean Siegfred Linkwitz, yes?
ccd - Wednesday, December 3, 2008 - link
Sigfired Linkwitz has been a major proponent of dipole speakers, not open baffle speakers.ccd - Tuesday, December 2, 2008 - link
Like many here, I think the DRC discussion is best suited for a website like Anandtech. As for some of the other issues, I highly recommend that you visit www.linkwitzlab.com Not only is the Orion the best speaker that I have personally heard, but Sigfried Linkwitz is an engineer who designed a line of very well reviewed commercial speakers. At his site, you will find a number of papers written by him on speakers and how we hear. You will find some good science there, not snake oil or myths passed among audiophiles as truth. You will find that the speakers he offers at his site as well as the literature there runs counter to many of the assumptions made by this author.
royboy66 - Tuesday, December 2, 2008 - link
Now to the meat of this audio is not something we know everything about -battery power isolates all power line noise if done properly, tube amps and tube microphones and recording equipment -when properly maintained have made magnificent lifelike recordings -i have many of them -close your eyes and you are there! I have a large hifi collection and had 19 amps at one stage -i got sent over a mint 1966 Fisher X10D valve amp owned by a guy who looked after it -original box, valves everything and maintained by a valve technician –it had not had a hard life – i never ever thought it was going to sound any good hooked it up and nearly fell over the sound was alive –i sold almost all the amps i had after that experience –they were sterile crap! I still had some very fine solid state though eg Electrocompaniet for example and it too made beautiful music .Now you guys who just blast your ipods at 120db in your ears have no idea about sound quality –if you know anybody passionate about audio go listen to their setup –chances are it will sound great because they have put the effort in to make it so. I do not believe in voodoo or bullcrap –but never underestimate power line noise and rf contamination etcAlso i can comment on speakers and dacs –they all make a difference –if i ever have to listen to a tda1543 again it will be too soon !the cymbals sound dull because that dac is barely 15 bit in quality –would you use a commodore 64 now to surf the net – of course not. This dac is so far outdated it is not funny –listen to cymbals on a good record they do not sound like the tda1543 reproduces them –it simply does not have the resolution in the mid to upper frequency to sample properly the lifelike sound of a cymbal –but the other did because it was a better computer for those same frequency
I have as one of my system the NHT XDA digital active system –are we supposed to believe it is the last word in audio because it is digital and has one of the flattest responses irrespective of room –it basically does not need that. I asked a few friends how they thought it sounded compared to my 20 year old Infinity Kappa9 and not one of them liked it over 20 year old speakers driven by valve or solid state. what does this mean? –there is more to audio than we think –so guys keep an open mind and hopefully through these articles we will end up with great computer sound –i personally at the moment use a squeezebox with lossless Flac and or digital out to it or a city pulse dac11 for computer audio –amongst other options
Pc noise is definitely the big problem – i live in the country it is quiet so my background noise is lower and the pc is loud –totally silent or in another room is the go –but my power is crap due to being in a rural area –hence i say if everything could be run off battery or solar it would make a big difference –listen to Jack Johnsons solar powered album –how good does that sound –awesome! Lastly guys I am no audio snob –i have no car but plenty of audio equipment and cds ,records, dvdaudio, sacd, upsampled music etc – i get given bucket loads of mp3 and use that to find music worth listening to –the pc can and is going to be the source of quality music for the future so let the articles continue