Arvus H2-4D multichannel Dolby Atmos digital processor

Before I can get to what the Arvus H2-4D does and how well it does it, I have to explain why I think it's an important product. I spent years as an advocate for multichannel audio for music, based on discrete, lossless sources such as DVD-A, SACD, and high-quality downloads. All of these were either PCM or DSD and were accessible via a wide range of disc players, DACs, and local file playback apps.

I am well aware that the majority of Stereophile readers are not (yet) into multichannel, but two-channel and multichannel audiophiles have long relied on similar media and components. Some of us have added more channels. In parallel with this, the home theater world was committed to multichannel (or surround sound), but instead of DSD and PCM, they spoke multiple dialects of Dolby and DTS. Those oft-compressed formats required proprietary decoders for each dialect, which their developers were happy to license for use in AVRs and preamp/processors.

The entertainment industry noticed that the newer sound formats that support many horizontally and vertically arrayed sound sources, such as Dolby Atmos, Sony 360 Reality Audio, DTS-X, and Auro-3D, were natural for cinema and home theater, since they could create a more convincing experience of being present or immersed in the action. By applying these immersive technologies to music, the music industry could entice music buyers to purchase new equipment and music, including a back-catalog of music they already own in other formats.

I came away from the 2019 AES Convention fascinated by these new options for encoding, mastering, and distributing music. "Instead of multiple masterings and formats for binaural, stereo, and multichannel," I wrote in January 2019, "a single-format release can be played on any suitably equipped system." It would be the equivalent of a "single inventory" for the streaming and distribution sites for all formats.

It has taken almost three years, but Dolby Atmos is now permeating the music-streaming sites; note that most of the recent Grammy winners are available in Atmos.

Up to now, access to Atmos has been largely restricted to HT hardware; its success for music will depend on wide availability and non-HT options for audiophile music lovers to stream and to play discs or files in Atmos, especially in lossless versions (footnote 1). Enter the Arvus H2-4D.

What is it?
There are three basic sources for Atmos playback: web streaming from an app or a device like an AppleTV, local discs, and local files. But without an Atmos-enabled AVR or pre/pro, there has been no simple way to decode these streams (footnote 2). The Arvus (footnote 3) H2-4D is a simple solution to this problem. Its compact, nondescript 1U chassis contains an HDMI input and digital, analog, and LAN outputs. It decodes and renders all extant varieties of Dolby (including Atmos), DTS (including DTS-X), and stereo and multichannel PCM (up to 24/192). The only front-panel features are a power switch and multicolor LED that indicates power, signal lock, and data format; once set up, it should require no user attention.

The rear panel contains, on the far left, a single HDMI input and a word clock input; on the extreme right is a place to connect a wall-wart power supply. In between is an array of outputs, all active simultaneously. The outputs include (from left to right): two HDMI pass-throughs (one full A/V, one audio-only), 16 channels of AES3 outputs via a DB25 connector, 16 channels of Dante outputs and inputs on a single RJ45 connector (footnote 4); 16 channels of balanced analog output via two DB25 connectors, and a data connection on RJ45 for setup and control via the local network.

A one-page GUI (above) is accessible from any device on your LAN, for accessing the Arvus from a computer, but I found my iPad most convenient. The GUI lets you select an input (HDMI, ARC/eARC, Dante) and a loudspeaker configuration. You can choose among settings for Mute, Standby, and Global Delay (useful for video sync), LAN settings, and individual speaker delays for each channel. There is also a Global Volume control, which is necessary if you are using output devices that lack that feature, as when you're feeding an array of active loudspeakers. (Essential as this is, it is not ideal: Tapping on a screen in 1dB increments is tedious; Arvus has promised new possibilities in the next firmware update.) The GUI also displays the firmware version, allows you to update the firmware with a click, and allows for a factory reset. It also shows the codec in use, which is useful, although it doesn't show the sample rate.

The H2-4D in my system
I set up a system to audition remote streaming, disc playback, and local file playback in Atmos, DTS-X, and PCM via the Arvus. For the input end, I acquired a Marantz VS3003 3 in/1 out HDMI Switch ($199), which supports the same audio formats as the Arvus. I connected the HDMI outputs from my Apple TV 4K, my Oppo BDP-105, and my PC/Windows-based streamer to the Marantz. The Marantz's output fed the input of the Arvus, and the HDMI output of the Arvus went to my Toshiba PC monitor so that I could see each user interface as it was selected.

My regular speaker system setup is 5.1, consisting of five full-range speakers plus a trio of subwoofers run via the BassQ controller. For this review, I added PSB Imagine XA and Atlantic Technology 44-DA Dolby-enabled speakers, relocated from my Connecticut to my New York City system, as front and rear height speaker pairs, respectively, each driven by a Parasound Zamp.

I began using the Arvus's AES3 outputs because my DACs have volume controls, and when the Arvus first arrived, its volume control had not yet been implemented. With the assistance of Mogami's Phil Tennison (footnote 5), I obtained a custom Mogami Gold AES TD DB25–XLR snake to connect to the AES3 connector on the Arvus. I then plugged XLRs 1–4 to the Okto DAC8 Pro and XLR 5 into a Mytek Brooklyn+ DAC. I assumed that software would route the appropriate channels through to the power amps, as would typically happen with a pre/pro. But when I switched everything on, with the Arvus set to output 5.1.4, I got 5.1 correctly—plus Front Height sounds out of the Rear Height speakers and nothing from the Front Height speakers. If I switched the Arvus to output 7.1.4, I got the usual 5.1 plus Rear L/R channels out of the Rear Height speakers!

Here's why. The order of the channels in Atmos world is fixed, and you must connect your DAC/amp/speaker to the correct physical channel even if the intervening channel cables are left unused. The Arvus was designed for studio professionals, who would know this. For reference, here is a list of the default AES3 outputs; the asterisks indicate channels used in my 5.1.4 channel setup:

1 Left, Right *
2 Center, LFE *
3 Left surround, Right surround *
4 Left back, Right back
5 Top front left, Top front right *
6 Top side left, Top side right
7 Top back left, Top back right *
8 Left wide, Right wide

Thus, in my setup for the digital outputs from the Arvus, outputs 1–3 and 5 went to the Okto for 5.1 and the front heights and output 7 went to the Mytek for the rear heights.

Later, after the central volume control was added to the Arvus, I used direct balanced connections from the two analog output connections using a pair of DB25-to-8XLR snakes. Of course, with analog, there is one XLR per channel; the ones I used are listed below.

• Balanced Analog (1–8): 1 Left, 2 Right, 3 Center, 4 LFE, 5 Left surround, 6 Right surround

• Balanced Analog (9–16): 1 Top front left, 2 Top front right, 5 Top back left, 6 Top back right.

Streaming Atmos with the H2-4D
I acquired an AppleTV 4K some time ago (footnote 6), for the express purpose of streaming Atmos from Apple Music; since then, the amount of Atmos content has exploded. Browsing the Classical repertoire, a randomly chosen recording is almost as likely to be available in Atmos as not.

I clicked on Sol & Pat. The Arvus recognized it as "Dolby Atmos (Multi-CH PCM)" at 48kHz, but to me the sound was completely new. The transition from stereo to "immersive" was startling. From the opening tambourine fanfare on through, the experience was much more "you are there" than the "we are here" of the stereo depiction (24/96 download, Alpha ALPHA757). The recorded, reproduced ambience completely replaced any awareness of my listening room. The two performers were steadfastly stationed up front, where they should be. I have no information on how the recording was mastered for Atmos—whether it was mastered for Atmos in the studio or by some automated process (footnote 7). Apple Music's Atmos is lossy (footnote 8), but I find it a credible alternative to the high-rez stereo release (24/96 in this case), swapping the latter's clarity and immediacy for a huge enhancement in the acoustic space.

Footnote 1: Some music labels, most notably 2L, still offer discrete 7.1.4 channel files in WAV format for download. These can be played back on systems with the requisite number of channels without any decoding.

Footnote 2: Dolby offers desktop/DAW software licenses only for the professional market.

Footnote 3: Arvus is a 40-year-old independent company with deep ties to the music industry. Its earlier products include many loudspeakers and an 8-channel HDMI to AES3 digital audio converter that has long been described as "unobtanium" in the US.

Footnote 4: Due to a surprise visit from COVID, I was unable to use the H2-4D in a Dante environment, but I hope to do so soon.

Footnote 5: When I explained what I was doing, Phil swiftly realized the stock cables would not suit. "Normally, AES DB25 cables have 4 XLRs out and 4 in. Yours has 8 out, so it deviates from the normal configuration. More like a Tascam Analog cable using AES cable." He was right.

Footnote 6: In 2019, my Marantz AV8085 pre/pro was updated to support Atmos and Auro-3D, and I was greatly encouraged by the first immersive audio recordings I sampled.

Footnote 7: Mastering engineers have told me that these days an Atmos mix is an essential deliverable in most studios, which argues for the former for recent releases like Sol & Pat.—Jim Austin

Footnote 8: In Apple Music, lossless and Atmos are two distinct formats.

Arvus International Ltd.
71-75 Shelton St.,
Covent Garden, London
United Kingdom WC2H 9JQ
64 21 121 66 44

Kal Rubinson's picture

One can certainly make that argument based on "the value equation." However, I take note that you maintain "two separate front ends: one is 2-channel, the other is theater" and, I am guessing it is because you do not think the HTP-1 is up to the task for your more critical 2-channel music enjoyment. My music is multichannel as well as stereo and this review was an effort to maximize performance beyond what we expect from entry and mid-level HT processors. That effort is continuing.

BTW, I am coming to prefer dB25 outputs that, in effect, allow me to more easily manage the XLR connections in practice and, as you do, I use a selector box with them.

David Harper's picture

Hi Kal. When watching a movie with home theater surround does it ever bother you that the sound overwhelms the movie? I tried it for a while but for me the audio was so overbearing that it distracted from my enjoyment of the movie. I wound up going back to two-channel stereo.The one exception was with some new movies that are all cgi special effects, crashes and explosions the home theatre was cool since these movies don't actually have an intellegent plot anyway.

Kal Rubinson's picture

It generally does not but that may be due to program selection. We rarely watch any of the blockbusters and no "action" movies or "new movies that are all cgi special effects, crashes and explosions." As a result, our surround system's advantage over stereo is in intelligibility and ambiance.

Scintilla's picture

I do think that a higher-end front end for 2-channel is warranted and does obviously sound better in my case, in part because I am using HQPlayer to convolve filters built in acourate in contrast with the Dirac in the HTP-1. Further, I am running DSD128 out to a Holo May and Serene. This does in fact sound quite a bit better than the HTP-1 for 2-channel replay and once upon a time the systems were discrete. When I was forced to combine the sytems in a new loft, I went with split front-ends. However, over time I have come to believe that the perceptual impact of the Atmos or 5.1, 7.1 soundfield swamps any further improvements that would be wrought from an even higher-end discrete approach as you and Chris Connaker have pursued. Once you have a Dirac calibration and build filters for different purposes (such as dedicated music listening), allowing instant comparisons, I just can't imagine the real-world advantages of the discrete approach moving things that much forward. I think the brain gets inundated with so much information during multi-channel playback that it is impossible to tell the difference between any quality reduction compared to 2-channel playback. If one isn't fully satisfied with the performance of the unit for 2-channel, the answer is two front ends; if I didn't have the Holo first, I may (npi) not have even known what I was missing. To your last point, the back of the HTP-1 is a spaghetti-mess of monolith cables. Perhaps the db25s are a superior way to go... My next project is to buy Audiolense, build multichannel RC filters and drive the HTP-1 from HQPlayer with 24/192 PCM bypassing Dirac for my SACD rips. I think this may take things as far as the HTP-1 can go unless the Direct DSD of the AKM 4493 is made available through a firmware update. I think you would find the HTP-1 a groundbreaking product in sound quality comparable to discrete products with a much higher entry price. It is really a step up from mid-level AV gear.

Kal Rubinson's picture

You raise an important issue.


However, over time I have come to believe that the perceptual impact of the Atmos or 5.1, 7.1 soundfield swamps any further improvements that would be wrought from an even higher-end discrete approach as you and Chris Connaker have pursued. Once you have a Dirac calibration and build filters for different purposes (such as dedicated music listening), allowing instant comparisons, I just can't imagine the real-world advantages of the discrete approach moving things that much forward.

My only point of reference, so far, is the comparison between my CT system running via a Marantz AV8805 and my NYC system running discrete components, both with Atmos 5.1.4. There are many other differences between the systems (and rooms) but there is also an substantial difference in performance between the two. Another test will come soon when I review a very high end HT processor in NYC with the same amps and speakers I have now. I would not close the books on this issue just yet.

BluesDog's picture

Home Concert Venue Made Possible

A very outstanding, challenging but rewarding effort by Kalman Rubinson and John Atkinson. I DO agree we are at the beginnings of fully being able create a Home Concert Venue (HCV) at home.

Hopefully the hardware and software will continue to improve.

Thanks for your pioneering efforts and encouraging even die hard 2 channel audiophiles to consider new options to 2 channel sound.

Homer Theater's picture

I've messed with surround sound for music since the 1970s. I'm pretty sure I tried tried every format that has come along since the single passive rear channel and even analog processing of stereo into multi-channel. When immersive sound formats appeared, I assumed it would just be more disappointment. When decoding stereo music, Dolby Surround (what you get with Atmos to to make stereo through 7.1 sound immersive. It sounds horrible processing stereo music. DTS Neural:X sounds better than Dolby Surround, but is still not as good as stereo playback of stereo music. Then I tried Auromatic processing (now just part of Auro-3D decoding of shources with 7.1 or fewer channels) and FINALLY had my mind blown by upconversion to 12 channels that sounds better than stereo. In the years since the immersive formats have appeared, Dolby Surround is still terrible sounding--dry, dull, uninvolving, gray, and just generally unappealing. Neural:X hasn't changed either. And Auro-3D upconversion to 12 channels is still my first choice for listening to stereo music that's only available in stereo. I'll choose Auro-3D upconversion even over stereo. Dolby thought Auro-3D was such a threat to their existence, that they engineered a way to stop their TrueHD/Atmos sound tracks from being processed with Auro-3D processing (in movies). Auro-3D sued Dolby and won. Dolby has also tried to hire-away key Auro-3D employees--all in the name of trying to make Auro-3D go away. Too bad they didn't spend that time and effort trying to make Dolby Surround work better instead of attacking a competitor. For Auro-3D processing of stereo, I use 12 speakers... one directly above the front L&R channels and one directly above either the side-surrounds or rear-surrounds. The 12th channel is what Auro-3D calls the voice of god, a channel directly overhead. The other 4 immersive height channels are close to the 9-foot ceiling... the side- or rear-surround height speakers are on wall-mount brackets and are box speakers aimed at the main seat. I find down-firing in-ceiling speakers sound pretty bad compared to box speakers aimed at the audience. In-wall speakers sound less-bad than down-firing in-ceiling speakers, but still not as good as box speakers for immersive channels. Auro-3D upconversion of stereo to 12 channels FINALLY does stereo better than stereo. You might not think there would be much discrete sound in Auro-3D upconverted stereo, and that's sort of correct. When you hear music in a live performance venue, the sound mostly comes from in front of you and reflected sound comes from everywhere. That's pretty much what you get with Auro-3D upconversion. There are some sounds that are at least somewhat directional (other than left-right), but mostly you get a realistic amount of "height" immersion that makes the music seem less "flat" and more spherical with a front bias as you experience it at concerts. Auro-3D upconversion gives you fantastic sound from stereo sources, but also from 4.0, 4.1, 5.1, 6.1, and 7.1 original recordings. Auro-3D processing of Atmos or DTS:X soundtracks even sounds better because studios are not paying for a sound engineer to create Atmos or DTS:X movie soundtracks. Instead, they just use Dolby Surround on the original 5.1 or 7.1 soundtrack to get 11.1 channels and it sounds HORRIBLE with almost nothing appearing in the height channels for the whole length of the movie. You can get Auro-3D built-in to a fair amount of equipment... Denon, Marantz, Yamaha, and Arcam are installing Auro-3D in their products. Not necessarily every model of these brands will have Auro-3D. Generally it comes in mid-priced to high-priced models and NOT in lower-priced models. For those who need some sort of "magnitude of approval". Let's use a scale where a "10" is really excellent stereo music playback. Process that music with Dolby Surround, and you will be hard pressed to give anything higher than a "2" to sound quality compared to stereo. Switch to DTS Neural:X processing stereo music and you might feel like a "5" is an appropriate rating, with some possibly going as high as a "6". But process with Auro-3D and it's difficult to not want to give the music a "15" on that scale simply because the music is SO engaging. So musical, so enticing, and you don't even have to start with more than 2 channels. You don't even have to buy all your music again to enjoy this upconversion.

PS -- I've been a professional A/V equipment reviewer since 1995 and "into" stereo and surround sound since the early 1970s. I've even been published in Stereophile once, many moons ago. And I've known Kal Rubinson through online audio forums for years before either one of us were reviewing equipment professionally--and before the internet was live to the general public.

Kal Rubinson's picture

Wow! I do agree with you that Auro-Matic up-conversion is preferable to the Dolby or dts alternatives but I have no real interest any of them. And when it comes which of the three "native" immersive formats I will buy, my choice will be dictated the program content, not the format.

MNSMike's picture

The Dante output performance would be nice to know. Would be nice to know if a bitperfect output that follows the sample rate of the input, up to 24/192, is output via the Dante port. Like it does with the AES outs.

A simple way to test the Dante output would be to install the Dante virtual sound card on any Windows or Mac PC. And feed the audio to it. You would find out the sample rate capability anyways. To go further the Dante input from the unit could be fed from the computer to a USB DAC via ASIO. Where sine wave test tones could be analyzed with the AP from the DAC analog outputs. Then compared to the same sine waves played back direct from the computer hard drive to the USB DAC.

Kal Rubinson's picture

We are pursuing this now and will report on it.

MNSMike's picture

Thanks can’t wait to hear the results. There’s no owners manual to read. And the company won’t answer questions they feel might hurt sales if the public was to know. I have a feeling that if it does work with sample rates higher than 24/48, you must manually set the sample rate to a fixed setting. This means if you’re listening to a mix of music at different sample rates, you must go into the Dante controller on the computer, and manually change the sample rate the system is running at every time the music sample rate changes. Doesn’t make for a fluid listening experience. Not ideal for home audio users. Especially if the source is something like Amazon music on an Amazon fire cube. Where it’s capable of playing music up to 24/192 out of the HDMI port. If listening to music randomly, you can go from 24/192 to 16/24 to 24/96 to 24/48 with each track change.

And I’m afraid the same will be true for the upcoming Dante output only H1-D. Hopefully the actual capabilities of the unit will be shared with the public before they start accepting preorders for the beta version next month. As it stands now it’s pay first, and find out the limitations after you get it.

Kal Rubinson's picture

I am on it.

MNSMike's picture

Thanks. Consider getting an Amazon Fire-cube and an Amazon unlimited account if you want to listen to both a vast selection of Atmos, and 2 channel albums up to 24/192 out the HDMI port from a mass market black box device that just works. Alexa is far ahead of siri as well regarding voice commands. I have 2 Apple TV 4k’s and 1 Firecube, and I as much as I love my Apple TV’s, I’m not a fan of the 24/48 only sample rate limitation. Especially since their catalog has high resolution albums in it now. For music lovers the Firecube is the best option.

Kal Rubinson's picture

Thanks, I will look into the Fire-Cube.

Scintilla's picture

is for headphones, not discrete channels like Apple Music. I used a firecube for a few years and agree it has the best voice interface, but regularly, Apple beats Amazon in bitrates for content and Apple Music and Tidal are the only services offering real, discrete Atmos tracks. The Amazon content is designed for headphones, and therefore is rendering Atmos for two-channel virtual replay, not as a lossy DD stream with embedded top metadata like the other services. Amazon is very clear about this.

MNSMike's picture

Not on the new Firecube gen 3. It works up to a full 24/192 bit perfect. I’ve verified it. Apple TV can only do 24/48 max. Regardless of how much high res content they have in their library. And it can do full TrueHD Atmos as well.

AC3 (Dolby Digital)/ EAC3 (Dolby Digital Plus) /Atmos(DDP+JOC)/AC4(Atmos)/Dolby TrueHD(Atmos)

Read all specs here:

MNSMike's picture

1 more thing to try. The Firecube 3rd gen supports full Dolby TrueHD Atmos. But none of the streaming services output full TrueHD Atmos. So I looked around for ways to take advantage of this feature. And I discovered 3 apps confirmed to work with Dolby TrueHD Atmos. Plex, Kodi and Jellyfin. With these 3 apps full 24/192 True HD Atmos can be sent out from the cube’s HDMI port with a very nice user interface, voice commands, and RF remote. And I believe they all have server apps that can run on a NAS, Windows or Mac to stream the content to them.

tnargs's picture

The idea that 24/48 is audibly inferior to 24/192 is absurd. Sometimes I think that many audiophiles are becoming connoisseurs of ultrasonic tech and not of music at all.

MNSMike's picture

Depends what sample rate the track was originally recorded in, and the quality of the downsampling if it was higher. When these cheap boxes or TV’s downsample using mediocre algorithms, the degradation is very audible. I’d prefer to listen bitperfect at the native rate of the track.

BluesDog's picture

J Gordon Holt IS proud of what you have reviewed and how you reviewed it.

prerich45's picture

I actually went to the old Harman Kardon Citation 7.0 processor and hooked it up to my AV7706 through the 7.1 multi-in, as when I tried Auro3D up-mixing, it was bass heavy. I'm extremely impressed by the HK's performance (I've owned it before in the 90's). I have a SU-9N connected from my PC to the HK. I use convolution on mains through Jriver and I also have PEQ generated filters for the subs. What I was startled by was the flat measurement of the Infinity Overture 1 that I use as a center channel on it's stand. Pans are seamless - as the Overture 1 now stands about as high as the center section of my Infinity Prelude P-FR's. It's as I remembered hearing the Fosgate 3a system with the MC220 speakers - everything disappeared and sound just floated in space...around, in front of, and behind the speakers. The rears were subtle...not blaring just adding ambiance. I'm going to see if they have music that I listen to in Atmos, as I would like to hear a mix and compare it to my current up-mixing situation.

J-B-Lite's picture

Thank you for this article, it is a valuable resource:
1) I've been looking for a testing file for Atmos playback--thanks for the Immersive Audio Album link. It's a nice demo and quick confirmation of channel levels. I was hoping for something like the AIX Blu-rays that have test tracks of 5.1 & 7.1 DD/DTS/LPCM pink noise for each channel and some combined. For instance, I use the track with combined front-right + Right-Surround (and ditto for left side) to hear the fantom image on the sides, then I tweak the distance settings of the surround on my Marantz A/V processor until that image sounds solid in the side locations between the front and surround channels.
2) I was pondering purchasing that Yellow Point disc, but found the album on Apple Music which is in both Atmos & Lossless simultaneously! It sounds fantastic, and I will eventually compare to my 2-channel HDTracks download.
3) which brings me to my next point, I listen to AppleTV 4K connected via HDMI and found the sound quality of "Lossless" and "Atmos" recordings not as satisfactory compared to two-channel CD rips (or HDTracks downloads) played back from my hard drive connected to my Oppo UDP-203 and connected to the Marantz AV7702mkII via HDMI. For instance, Lorde's Pure Heroine 24/48kHz downloaded from HDTracks in 2014 sounds more dynamic and detailed (with better 'microdynamics'?) than the Apple Music versions. I am wondering if some kind of volume normalization or something else is making it seem less dynamic to me. (I also tried increasing the HDMI input level on the Marantz for the AppleTV). Two-channel comparisons with other albums (like Michael Jackson's BAD & Dangerous albums) also sounded less 'powerful' or impactful via Apple Music (I'm not an audio reviewer).
4) Another nice feature of the AppleTV is that I can browse music using the Apple Music app on my iPhone, and have it play back & control the app on the AppleTV without using AirPlay. Airplay is limited to stereo, sounds inferior, and uses the phone's volume control. But when I am sending it correctly--connecting to my AppleTV's home theater configuration name, the volume is controlled thru my preamp and it is receiving Atmos playback if the albums are in that format. When I tried Apple's new Classical music app (when it first became available), I could not get it to output Atmos either from my phone or via using only that app on the AppleTV 4K.
I'm looking forward to your follow-up with the Amazon FireTV setup.