Anthem is as well known to enthusiasts of two-channel stereo as it is to fans of surround sound. Anthem was formerly a subbrand of Sonic Frontiers; Paradigm, a speaker manufacturer founded in Canada in 1982 and headquartered in Mississauga, Ontario, just outside Toronto, acquired Sonic Frontiers and Anthem in the late 1990s. Many of Anthem’s electronics are now made in the same Mississauga factory in which Paradigm builds their speakers, but they’re designed in the Paradigm Advanced Research Centre (PARC), in Ottawa, Ontario, not far from where I live.
One of Anthem’s statement products is the STR Preamplifier ($3999 USD), a unique and innovative two-channel component that features a wide range of digital inputs; analog single-ended and balanced inputs, including one set for the onboard phono stage; and single-ended and balanced analog outputs. What sets the STR apart, however, is that not only does it provide full DSP room correction using Anthem’s new Anthem Room Correction (ARC) Genesis software, it also offers two sets of configurable, low-passed subwoofer outputs, and left- and right-channel high-passed main preamplifier outputs. As full of features as any preamp I’ve seen, the STR is made for someone who wants to integrate one or two subwoofers into a two-channel system.
Left to right: Mark Schuck, Diego Estan, and Peter Schuck
This most recent installment in my series focused on integrating one or more subwoofers into a stereo system began on July 27, 2019, when Doug Schneider and I sat down at PARC with Mark Schuck and his father, Peter Schuck, to chat about ARC Genesis and the Anthem STR. Peter is currently Anthem’s Technical Director; he’s been with the company since 1993, and holds a PhD in applied mathematics from Western University, in London, Ontario. Before that, he worked for Canada’s National Research Council (NRC). Mark, who began at PARC in 2006 and is now the Senior Embedded Software Developer, studied software engineering for three years at Carleton University, in Ottawa.
Doug Schneider: When did you first start working on room correction?
Peter Schuck: A long time ago, maybe 1990 or so, at the National Research Council, in Ottawa. I was working on what was known as the Athena project, which was looking at speakers and rooms, and what could be done to improve the interactions between the two.
DS: What did such systems look like back then?
PS: There was one system on the market, which was from Sigtech. They had a patent on using a finite impulse response [FIR] filter using a Texas Instrument [TI] keyboard computer. In order to implement an FIR filter in audio, it must be reasonably long, a few milliseconds -- as long as a full wavelength of sound in air -- but the computer chips didn’t have the processing power. So they figured out a way to have lots of zero coefficients, so they would avoid multiplying the zeros to save on processing, and would choose meaningful coefficients that would affect audible frequencies -- and not necessarily lower frequencies, which require longer filters. They had a very broad US patent that applied to something like “room correction at audio frequencies.” So, for something like 20 years, because of this patent, it was tough to get into the room-correction game. The original patent was issued maybe around 1980.
Diego Estan: When did the first Anthem Room Correction (ARC) system come out?
PS: In 2008. There was a lot of bench testing before then; the first product released with ARC was 2008. Between the Athena project and 2008, we had this room-correction box with four DSPs, which was from the NRC research, and we had demoed it back in 1993, but it wasn’t terribly cost-effective -- it would’ve been very complicated and expensive as a consumer product. During those intervening years I worked on changing these boxes, which were based on FIR filters, and implementing Infinite Impulse Response [IIR] filters to implement ARC like EQ [equalization], but we never came out with a finished product until 2008: the Statement D1 preamp-processor. In the development of the Statement D1, we already had one DSP chip for basic audio functions, so we decided to include a second DSP and some code for room EQ, since it didn’t add too much to the cost. It wasn’t tough to justify the inclusion of what was the first iteration of ARC in one of our products.
DE: Did the Statement D1 use IIR and FIR filters?
PS: No, just IIRs. From day one in actual products, it was IIRs only. We originally looked at using FIR filters, but they’re not very efficient. In all of our experiments, I saw no advantage in using FIR filters, only disadvantages. Besides the lack of efficiency from a processing point of view, FIRs also add a lot of delay, something like one second -- not necessarily a problem for audio-only applications, but throw in video, and it can be an issue to sync up the video on the screen to the processed audio.
DE: Could you describe the evolution of ARC to ARC Genesis, the version used in current Anthem products?
Mark Schuck: First, we got requests to support more operating-system platforms. ARC 2 was Windows only. Between ARC 2 and ARC Genesis, we did a port to the iOS mobile platform, and then, eventually, Android. We reworked the back end, made it compatible with Mac, put a new skin on it, and finally renamed it ARC Genesis.
PS: There were two parts in the evolution from ARC 2 to ARC Genesis. One was definitely to make it compatible on multiple platforms, including mobile; the other part was to get rid of PBK, Perfect Bass Kit, which was like ARC for subwoofers. It was confusing in the marketplace. We replaced it with one room-EQ system for all our products. So what we actually decided to do was not only release new products with ARC Genesis, but to also support all of our products that are currently in production, and even a few that are no longer in production, and update them to the new ARC Genesis code.
MS: The rule of thumb was that any product with an RS-232 port would not be supported and would not receive the ARC Genesis upgrade, such as the old D1 and D2 processors. So any product with a USB and/or IP connection is supported, including our line of subwoofers. This support goes all the way back to subwoofers with USB ports that ran PBK, which existed between ARC 1 and ARC 2, ca. 2009.
PS: Unlike many other room-EQ systems out there in the marketplace, where you hear testimonials from users where they try it, claim it sounds better without the room EQ, and then leave it off, we just haven’t heard these claims from ARC users.
Peter Schuck in his office
DS: What features were added to ARC Genesis that weren’t available in ARC 2?
MS: We made general performance improvements to the algorithms to speed up the use case. We streamlined the flow, so you can move through the setup, measurement, and optimization process in a more linear and user-friendly fashion. We’ve made it possible to remeasure individual positions or speakers, [and] to interrupt the measurement procedure without losing data. We’ve added new frequency-response target-curve customizations. We are not fans of just drawing an arbitrary target curve -- we feel more harm can be done by allowing such freedom. Some of the target-curve customizations include deep-bass boost and a high-frequency tilt option, such as [a] 1-2dB/decade negative slope in the top end.
DE: So as ARC Genesis is implemented now, users can’t impose or draw their own target curves?
MS: Not complete arbitrary ones, but as far as we’re concerned, there is enough customization built in to allow the user to tweak the target curve that would cover just about any use case.
DS: Is there an ARC target curve?
PS: ARC does a match to what is measured in the room. The original idea, which I still believe is valid, was to go flat to 5kHz, and then no EQ above that. People pay good money for a pair of speakers with a high-frequency response that they like, so we didn’t want to mess with that. The other issue with EQ above 5kHz is the question of how representative is the in-room measurement at those frequencies due to several factors, such as accuracy of the microphone at high frequencies, humidity in the room, repeatable microphone placement, etc. These factors can vary the response by a dB or more at 20kHz, and you want to apply EQ in this region? It doesn’t make much sense.
DE: Does ARC Genesis still restrict the user to EQ below 5kHz?
PS: Well, we’ve allowed some flexibility. By default, ARC Genesis will not EQ above 5kHz, but the user can now force EQ changes in the uppermost frequencies, up to 20kHz.
MS: When you’re equalizing, there’s always a limit at some point where you have to blend back to the in-room passband response of the speaker.
PS: There’s also the limitation of the USB microphone, which samples at 48kHz, and can therefore not theoretically exceed 24kHz in response. But back to our target curve. So it’s flat to 5kHz. Below the 200Hz hinge point, ARC Genesis looks at the measurements and the amount of natural room gain, and applies anywhere from 0 to +6dB of bass boost.
DE: Why not always apply a flat frequency response all the way down to 20Hz?
PS: There are two parts to the answer. The first pertains to the way recordings are mixed. Typically, the final EQing on a recording is done while playing [it] back through speakers in a room. The person making the judgments as to how much bass to add or take away in the recording is experiencing his/her own room gain while listening, which is not inherent in the recording. So ideally, as the consumer of the recorded music, you should be listening to it in the recording engineer’s room, which is of course impossible. So the end user needs to put something back into the bass to have a semblance of what was heard in the mastering suite. That “something” can be the end user’s natural room gain, but if the end user’s room is relatively flat in the bass, to hear what the recording engineer heard, EQ must be applied in the form of bass boost.
The second part of the answer pertains to psychoacoustics. If you hear something, human speech for example, in one room with particular acoustics, it won’t sound all that different if you hear it in a room with different acoustics. Your brain has a way of “stripping out” the effects of the room, to an extent. So ARC Genesis aims to look at what the room is doing to the bass, and instead of completely altering the character of the response, the goal is to make it smooth, meaning get rid of the big peaks and dips. So it’s not an arbitrary curve in the bass region -- it’s different for every room. ARC Genesis strives to work with the room, not against it. There’s also an algorithm to limit the amount of boost in an attempt to cure the nulls, or dips, in the in-room response, because we don’t want to stress the amps too much.
DE: What is the maximum amount of gain ARC Genesis will apply?
PS: Six decibels.
DE: Could you tell us more about ARC Genesis’s default target curve?
PS: It’s basically a shelf filter with a low-frequency bump. So a bump in the bass, then [a] transition to flat out to 5kHz. And there’s a transition region nearing 5kHz, where the target curve needs to match and blend naturally with the speaker’s in-room response. The low-frequency cutoff is also determined by the measurement -- we don’t want to stress the woofers in your speakers. If it looks like a natural fourth-order slope, the target curve will match that.
DE: What is the theoretical low-frequency limit for the target curve? Let’s say you have big subs that can play flat down to 16Hz.
PS: Fifteen hertz.
DE: What are the different ways the user can control ARC Genesis?
MS: ARC Genesis runs on Mac OS 10.12 and up and on Windows 7 and 10. There’s also a smartphone app, ARC Mobile, where you can run through the automatic process without the UI of the full software package -- just run and upload.
Mark Schuck in his office
DE: Using the example of the Anthem STR Preamplifier: You’d buy an STR, then you’d get access to the ARC Genesis software and install it on a laptop, and then all operation by the user would be done on the laptop. Is that correct?
MS: Yes. Or you could run ARC Mobile, with basically no configuration options. You select the device, it reads the configuration from the device, and then runs one series of measurements and applies them to all the output profiles. In ARC Genesis, you can make a single set of measurements, but you can also make different measurements for different positions and up to four profiles.
PS: You can also apply different target curves and save them under different profiles, so the user can easily switch.
DE: What’s the minimum number of measurements ARC Genesis requires you to take?
MS: Five is the minimum, but you can take and save up to ten. You can prioritize the measurement in the sweet spot, and then you can move the microphone around -- say, no more than 2’. For each microphone position, you can also assign up to four individual measurements.
PS: But it’s crucial to move the microphone around and take various measurements at various positions. We have people who try to fool the system with just one measurement at the sweet spot. It’s not going to sound good that way. First of all, you can’t get one microphone position to represent both your ears. And then there are the bass modes. At that one position, you will have nulls and peaks, yet these peaks and nulls will be completely different in the rest of the room. You’ve got to spread your mike positions, and you’ve got to move your mike positions up and down.
DE: Does ARC Genesis suggest specific mike positions?
MS: Yes, we recommend changing the height, moving 2’ between positions, first measurement at the primary listening position.
DE: Back to ARC Genesis as implemented in the Anthem STR Preamplifier: The STR can handle two independent subwoofers, correct?
MS: Yes. If you have two subs and an STR Preamplifier, the user needs to make a decision: Are you going to feed them both a mono signal, or feed them left and right signals individually? The ideal situation would be to measure and EQ them separately, and once that’s done, run Quick Measure on the STR, sweeping the subs together, balancing them, time- and phase-aligning them. Then, the default setting is to feed them a mono signal. However, if the crossover point is high, and the user is concerned about localizing bass mixed in the right channel out of the left subwoofer, the user can configure them as stereo subs.
DE: Does the user choose the crossover frequency?
MS: ARC Genesis will choose default crossover points based on what it measures, but the user can override them. The default recommended value, assuming the main speakers play low enough, is 80Hz. And the EQ generated by ARC Genesis guarantees that the subs and speakers cross over properly and blend seamlessly.
DS: Is the Anthem STR currently the ultimate implementation of ARC Genesis?
MS: It’s the best implementation of ARC Genesis available from Anthem, yes.
PS: On the STR, the two subs have independent phase and time-delay controls.
DE: What about the crossover slopes?
MS: We use second-order on the high-pass filter for the speakers, which assumes a natural second-order slope on the speakers, so you end up with fourth-order, effectively. And the subs have two second-order low-pass filters -- again, effective fourth-order.
PS: Bottom line, the acoustic targets for the crossovers are both fourth-order because two fourth-order filters that are in phase crossing over at the same frequency will both be 6dB down at the corner frequency, which sum to yield a flat amplitude response.
DE: Could you summarize the specific improvements for integrating subs that have been made to the version of ARC Genesis implemented in the STR?
MS: There’s the auto-phase adjustment. This is performed post-upload -- that is to say, after the measurements, the EQ target curve, crossover points, etc. The user can then run the auto-phase tool. It scans and tests the phase response of your subs and corresponding main speakers, in order to maximize response in the crossover region.
PS: So it’s essentially a phase control that’s implemented automatically, based on measurements, to blend the subs and speakers as seamlessly as possible.
DE: Once you’re done with the entire room-EQ process, you’re remeasuring the EQ’d system as a whole and applying a further performance tweak?
MS: Yes, and this is new to ARC Genesis. This may take a minute or two, but the system does a fine phase scan, with about a 5Hz sampling resolution.
There have also been improvements to the back end, in terms of ordering and sequencing the filters, which brought down the noise gain. This may not be apparent with the STR because of its significant operating bit depth, but with older subwoofers, this improvement in ARC Genesis over ARC 2 would yield real-world improvements in noise performance.
PS: With IIR filters, the way you match the poles and zeros matters, because some of them have gain and some of them have loss. And this affects the noise gain at the output. So that’s the part we optimized in ARC Genesis to reduce the noise gain.
DE: My only experience with ARC so far has been reviewing the Paradigm V12 subwoofer. In that case, I relied on my smartphone’s microphone to perform the calibration. What is your feeling about how well that works?
MS: Our good/better/best concept with respect to microphones is as follows. On Android, we can do a “good” near-field calibration on the microphone built into a phone with subwoofers, from 20-30Hz to 160Hz. Some smartphones have high-pass filters at 60Hz, but we figure that out and find the right range of frequencies where we trust the measurements. The better solution is the iPhone internal microphone, which has been a good, consistent performer for some time. We measure those and generate our own correction curves. With those, we will EQ up to 2kHz in ARC Mobile for iOS. Another, better solution is using our external microphones, which are factory calibrated, with ARC Mobile, where you can EQ up to 5kHz (Android or iOS). Finally, the best solution is using our external microphones with ARC Genesis running on a Windows PC or Mac, which you can use up to 20kHz.
DE: Thank you, Peter and Mark!
This will be my last feature article on subwoofer integration for a while, but my work on the subject is far from complete. I’m scheduled to review the Anthem STR Preamplifier in the context of my current system, which includes two SVS SB-4000 subwoofers. The combination of these subs, my digital and analog source components, and my love of room correction will allow me to test every feature of the STR and ARC Genesis.
. . . Diego Estan
diego@soundstage.com
