What about surround sound? Some of us are into that stuff. ;>Excellent points Steven.
It does change things slightly. USB and S/PDIF won't suffice for lossless 4.0/5.1/and beyond audio data. That leaves HDMI or analog, afaik.
Surround/multichannel playback is important and sadly rather neglected in the audiophile world. I saw recently that the TAS website posted this article on the "rebirth" of quadraphonic sound.
Hmmm, isn't this actually the "rebirth of the rebirth" of quadraphonic? As I recall, there was a rather significant thing among home theater enthusiasts and audiophiles called multichannel SACD and DVD-A around Y2K :-).
I've always thought it unfortunate that some (many!) audiophiles and magazines did not embrace multichannel audio back in Y2K but instead relegated it to the home theater hobbyists. Other than Kal Rubinson and his "Music In The Round" column in Stereophile, I've rarely seen articles on multichannel over the last decade targeted at audiophiles. Not only is multichannel neglected, but I see that some audiophiles actively oppose it with comments like "You've only got two ears!" and other such supposed justifications of that stance.
I find this kind of response rather surprising considering how audiophiles often state a desire to replicate "live" musical performances that of course take place in 3D space with potential sound sources and reflections coming from all around. And even if one is not after "live" realism, surround audio provides an opportunity for artists and producers to expand their creative options. Imagine, some audiophiles seem to believe that their ears possess such immense ability that they could hear the difference between silver and OFC copper conductors (and potentially willing to upgrade cables for thousands of dollars). Many of these audiophiles even seem to spend huge amount of time testing and "listening for" differences. Surely whatever differences one might hear doing this pales in comparison to the improved spatial representation from good multichannel set-ups and recordings! For anyone who has ever experienced good multichannel music, the difference is clearly not subtle.
It doesn't take much head scratching to guess why having >2 speakers in one's room might not have caught on for many. It's not because it doesn't sound good... Rather, it's because of resource demands to get it done right. To get high quality surround, you do need space and there is a price to pay for adding speakers, to add a few more amplifiers possibly, find high quality multichannel source devices, decent multichannel decoder/receiver with features like bass management, and everything else in between like cables.
As Steven says, lossless multichannel is a no-go with the S/PDIF interface as it was not designed to handle more than 2-channels losslessly. It can however manage a lossy multichannel bitstream like AC3 or DTS. (Remember that 5.1 24/48 requires a lower bitrate than lossless 2.0 24/192, so it's not primarily a bitrate issue for S/PDIF.)
HDMI indeed is the most common digital interface for multichannel. Although jitter was more of an issue in the past, it has gotten better with newer implementations like with the Oppo UDP-205 (remember, jitter isn't really audible unless extremely severe anyway). HDMI implementation can be a bit challenging for small companies and licensing costs can be prohibitive.
However there is actually nothing stopping USB from sending multichannel audio. In fact I think as audiophiles this is not a bad option given that it can be implemented cheaply, and the asynchronous interface should be less jittery than HDMI. In fact, USB multichannel has been done for years with the exaSound e38 8-channel DAC for example or the inexpensive MiniDSP U-DAC8. For DIY folks and OEMs, I see MiniDSP also has the 8-channel USB MCHStreamer Kit available. There are other multichannel USB devices in the pro world. Vendor-specific ASIO drivers are common but it's convenient to have the USB Audio Class 2 (UAC2) drivers built into Mac OS X and Windows 10 these days already supportive of >2 channels. Devices such as this MOTU are capable of 64 in/out channels at 44.1/48kHz through USB 2.0 so an 8-channel hi-res audio setup (like 7.1) would be no problem at all. USB 3 is ubiquitous and would probably be a great choice these days for multi-channel DACs.
As I mentioned years ago, I already have a library of multichannel FLAC files ripped from DVD-A, SACDs and Blu-Rays. Due to the high likelihood for use of DSP in the playback system like bass management and room correction, I believe that there is no reason to keep multichannel DSD files as DSD - converting them to PCM (multichannel FLAC) with a good converter for me is simply less of a hassle and potentially sounds better.
While we might imagine one day, multichannel USB DACs may show up in quantity for our audiophile listening rooms, I think we don't have many options but to just stick with HDMI for now. Blu-Ray multichannel albums (like the range of releases from AIX Records or 2L or the occasional release like The White Album) sound fantastic for those spinning disks and one could rip DVD-A/SACD/Blu-Ray and play them off PCs in the audio room.
Beyond PC playback, for those of us streaming off a NAS, I'm currently using the excellent multimedia player Kodi on a "TV box" pointed at my directory of multichannel FLAC files. Kodi does a great job with reading my directory structure and metadata (as described), allowing me to select what I want to listen to on the TV screen. Even with my TV screen off, I can control Kodi with my phone using Kore.
For those wondering, here's how I'm streaming multichannel FLAC files from my Windows Server 2016 computer over to the sound/media room by using an inexpensive 2GB/1GbE Android TV box called the H96 Pro+ - there are many others such as this newer 3GB RAM version, or similar models like this Sunvell T95Z Max:
|Amlogic S912-based TV box nestled in the equipment rack. On the left of the box is a USB 2.0 port for wireless keyboard (not seen), and on the right is a slot for microSD where an alternate OS can be booted off. (There's a recessed reset switch you hold down for a few seconds while plugging in the power connector to get the machine to boot off the microSD.)|
One writes the CoreELEC (or LibreELEC) OS image to an SD card (typically with something like Win32DiskImager) and boot the box with that SD card. I'll use it for a week or so to check stability, make sure it fits my needs and without problematic bugs. Once I've confirmed that it's working well with my network, I putty/ssh into the device's Linux terminal (for CoreELEC, default login "root" and password "coreelec") and issue the "installtointernal" command to have everything transferred into the box's internal eMMC flash (16GB is plenty) which is generally faster than the removable SD card. This will wipe out the box's stock Android OS and replace it with CoreELEC/LibreELEC. From this point on, the SD card can be removed.
BTW, if you really want to go "high end" (but not high price!) with the latest and fastest Amlogic boxes, consider the recently released HardKernel ODROID-N2 with the new Amlogic S922X hexa-core chipset with 4 Cortex-A73s. Lots of potential in that new SBC, with speeds beyond the also-recently-released Raspberry Pi 4.
Kodi will scan the music directory and does a fine job of displaying the albums with cover art. I use a wireless USB keyboard/trackpad to control:
Kore app on my smartphone/tablet:
|Android device running Kore to remote control the Kodi player. Multichannel Beatles collection displayed.|
I leave the LibreELEC / CoreELEC box turned on 24/7 and it has been stable with uptimes for months over the last year. For completeness then, my sound room multichannel setup looks like this:
Amlogic TV box --> HDMI --> Yamaha RX-V781 receiver / processor
Yamaha analogue AUDIO: pre-amp analogue output to Emotiva XSP-1 preamp for front stereo & LFE (HT bypass mode). Surround & center channels decoded and fed by the Yamaha to those speakers.
Yamaha digital VIDEO: HDMI 2.0 out to Vizio P-series 4K TVSo in my room, there are actually two "systems" in one. My standard 2-channel playback streaming with Logitech Media Server or Roon to either the Raspberry Pi streamer or Oppo UDP-205 (Roon-ready). Two-channel analogue DAC output sent to my Emotiva XSP-1 pre-amp which then feeds my Emotiva XPA-1L monoblocks with lowpass to the subwoofer. And then there is the multichannel system above that uses the Amlogic S912 TV box and the Yamaha receiver but routes the front stereo and LFE analogue channels to the Emotiva pre-amp while the Yamaha receiver amplifies the centre/rear/Atmos speakers.
While it may be a little inconvenient separating 2-channel from multichannel playback, it does work well and allows me to focus as much resources as I want on the 2-channel component system plus provides flexible multichannel playback. I would love to see a fully Roon-ready economical mainstream receiver system (like Denon, Yamaha, Marantz...) that can take multichannel lossless audio off ethernet without unnecessary HDMI cables nor piggybacked with a video stream. While not exactly the same, it would have been groundbreaking for example if the Oppo UDP-205 could have accepted multichannel Roon input and sent the data out through HDMI to the receiver; alas, the UDP-205 could only handle 2-channel Roon streams.
Let's end off this post with thoughts about the future and how maybe in time, we might find ourselves having more access to multichannel music recordings and playback capabilities.
Like with any media system, there's always a "chicken or egg" problem between hardware adoption and software availability. While multiple surround speakers in the living room isn't as economical and can be spatially cumbersome, thus commercially hasn't been a great success, is there still a way to encourage the consumption and hence production of more multichannel mixes?
Indeed, there might be. Here are a couple ways that the AV and audiophile industry could proceed forward to draw more listeners in:
1. Expand the use of surround soundbars. I know audiophiles tend to look down on soundbars, but just like Bluetooth headphones and speakers, they are popular and can still sound very good! With DSP processing and multispeaker arrangements, the potential is there for soundbars to virtually project excellent facsimiles of enhanced "space" for the listener using multichannel encoded material.
An example is the well-reviewed Sennheiser AMBEO Soundbar. Though not cheap at US$2500, it does look good, appears well built, and could overcome the space and "acceptance factor" issues in a living room.
2. Get the masses of headphone users interested with HRTF-based DSP-reconstructed binaural playback. This I think is more exciting than soundbars with potential for greater consumer participation. Back in CES2018, Creative showed off their Super X-Fi DSP system. Well guys, I got my Super X-Fi Amp (SXFI Amp) the other day shipped from Hong Kong:
To measure HRTF accurately, one would have to capture multiple binaural impulse responses. Creative instead estimates the function by having you take a photo of your ears and face using an Android phone. The app then extracts some data off that to compute your approximate HRTF to program the onboard DSP when you hook up the SXFI Amp to the phone via USB-C. From there, the device is a UAC 2.0 compliant DAC communicating at USB 2.0 speed that can accept up to 7.1 input for headphone playback (based on AKM4377 DAC chip with Creative's DSP processor).
I'm not a fan of "fake surround" with stereo content played back with the Creative DAC thus far, though vocals are pushed forward instead of sounding like they're deep inside the head. With true 5.1 material fed into it, the soundscape does open up nicely depending on the material. So far, I've mainly been listening using Windows 10 set to 7.1 multichannel playback to the SXFI Amp (after HRTF "personalization" data uploaded through the Android phone).
Multichannel movie soundtracks can sound excellent. For example, I had a listen to the recent 2001: A Space Odyssey 4K release with DTS-HDMA and the screaming chimps at the start as they battled in "surround" was rather jarring, and the soundtrack vocals when they discover the Monolith on the Moon was just plain scary in how the voices filled virtual space. Listening to the rock concert / "Shallow" performance scene in Cooper/GaGa's A Star Is Born brought a smile to my face. FLAC 5.1 multichannel music envelops the listener in a way I have not experienced before with headphones and other "spatializer" algorithms. A good album to try is the 2015 5.1 remix of Roger Waters' Amused To Death as one might expect; in fact, even the QSound-encoded CD sounded very good to me with the DSP on. Another one that sounds great is Alan Parsons' On Air decoded from DTS - have a listen to "Blue Blue Sky".
It's certainly not perfect and the effect will vary for each person, headphone used, and recording. For me, an open headset like the AKG Q701 sounded more spacious than closed cans, including the Aurvana SE (Live!) that comes in the package. At best the photos of ears and face are a vague approximation of individual HRTF with hopefully much room for improvement. It's good that they're also refining the DSP algorithm - my unit shipped with old firmware 1.19.05 and the latest 1.40.07 firmware update improved vocals significantly (vocals used to have way too much echo which was simply unacceptable). I'll report more on this device and likely give you guys some measurements of the little USB DAC in the process later. Remember that the ultimate goal of this DAC/DSP is not about accuracy or fidelity (compared to something like the AudioQuest DragonFly Cobalt based on their advertising at least), it's about creating an illusion; the adequacy of achieving this goal will have to be highly subjective! Regardless of measured/objective resolution, it's still a fun toy thus far :-).
While this Creative solution may be an easily accessible commercial product now, there are others. Things like the THX Spatial Audio Platform and JVC/Kenwood "EXOFIELD" might be of interest. Headphone enthusiasts may also already know of the Out Of Your Head software. I think things could get very exciting with headphone 3D virtualization and multichannel mixes.
Imagine if a streaming service, instead of sending out MP3 320kbps, were to stream everything in a very high quality lossy codec similar to E-AC3 at 1+Mbps and up to 7.1 channels plus Atmos-like objects (hey, Atmos-encoded Kind of Blue would be very cool!). Imagine now that all decoder software/hardware for this streaming service were capable of extracting and folding down if necessary the audio stream to whatever number of channels for playback from 2.0 to full multichannel. While most albums would still be 2.0, by making 5.1+ also available for those who can extract them in "surround" quality would be an amazing step up for music streaming as a whole and would bring multichannel to the masses! Nothing in this scenario cannot be done today at standard broadband Internet speeds, not to mention the expected huge increase in wireless speed when 5G becomes prevalent.
A streaming service like this can target all devices from a guy walking down the street listening to standard 2.0 headphones, to tricked out DSP-driven headphone playback (like the Creative SXFI Air) that can simulate the extra channels, to beam-forming soundbars like the Sennheiser AMBEO Soundbar, and finally to a full soundroom with discrete speakers and powerful surround processor. Such seamless integration of 2-channel and multichannel would be my dream music streaming service!
Using a high bitrate, high quality, lossy multichannel codec would absolutely make sense for a "next generation" kind of streaming service. Remember guys/gals, like hi-res Blu-Ray or UHD Blu-Ray video which is always lossy, there's no point arguing about lossless/lossy if the bitrate is high enough to be imperceptible. As a feature, availability of multichannel encoded material would IMO be much preferable to "hi-res" stereo which is effectively inaudible and meaningless anyway for most albums. In comparison, the ridiculous MQA codec brings nothing of value to the consumer.
While under the radar and not generally spoken about among audiophiles, we should know that MPEG-H 3D Audio (MPEG website) is already "out" and embraces the 3 methods of encoding 3D sound: multichannels (3.0, 4.0, 5.1, 7.1...), sound-objects (like Dolby Atmos, DTS:X), and ambisonic soundfield encoding. The THX link above mentions this, and it looks like Sony has demonstrated some content based on this codec with their "360 Reality Audio" this year.
While we will always have "classic" 2-channel stereo (available to consumers since 1957 with the first mass-produced stereo LPs), the path forward is 3D sooner or later. With high-resolution 2-channel audio easily achieved and mature these days, audio technology has no other option but to go 3D or remain stagnant with no real new features (beyond competing to reduce size and cost or the other direction of making expensive luxury products). While some audiophiles might never embrace the changes, don't be afraid of them or the exciting sonic opportunities! I think to maintain relevance as a hobby, it is in fact the role of audiophiles to change and embrace technology when appropriate, observe the benefits and drawbacks, while engaging in discussion and feedback to ultimately champion higher fidelity.
Hope you're all having a great July. Enjoy the music :-).