Getting SACD sound to your speakers isn’t obvious… – FarMedia

or
The Perils of DSD…

“The road is long, with many a winding turn…”

(incept date: 2023.08)
Second round edits.
-Added terms section after the main text.
-Distinguish between multi-channel content, delivery method, and presentation for surround sound.

tl; dr:
Questions:
First, do you want to listen to just stereo SACD’s, or also SACD multichannel?
Second, do you want to listen to other multi-channel sound formats, like DVD-A or Dolby Atmos?
Third, consider how much hassle you’re willing to go through to set up and operate the system. And how to explain all this nonsense to your housemates, so they can still operate the system.

If the answers are “also SACD multichannel”, “yes”, and “little to no hassle”, then:
FIRST – Buy a modern multi-disc player that plays all the shiny discs (CD, SACD, DVD-Video, DVD-Audio, 4K Ultra HD, etc.), and can output all the various audio formats (PCM, AC-3, dts, DSD, DTS-HD Master Audio, Dolby TrueHD, dolby:atmos, dts:X, etc.) as bitstreams over an HDMI interface. (As an example, I personally use a Sony UBP-X800M2.)
AND – Next, buy a modern A/V receiver or pre-pro with DSD and Dolby Atmos decoding by accepting all the audio bitstreams (including DSD) over the HDMI interface and handles the bass management issues for you. This is most of the current generation Dolby Atmos A/V receiver or pre-pro units.
AND – Lastly, plug it all together with lots of wiring and lots of speakers, properly configure the multi-disc player to output all the pure bitstreams with no conversions, configure the pre/pro to accept all the pure bitstreams for decoding, then do the usual A/V surround setup and calibration processes. And enjoy. 🙂

And that’s the easy way (!)

Here’s the hard way: understanding the subtleties. (Refer to the list of terms at the end of this article if this topic is new to you.)

Specification changes:
The first thing to understand is the specs for DSD (the audio data format that is on a SACD disc) has evolved since inception, and so has the interconnects used to transfer digital signals between devices, eg HDMI/HDCP/EDID. Enhanced features like multichannel hybrid/dual layer and bitstream support are only supported in later versions of these specs, which means only in later product offerings.
If you use older equipment, you will soon discover “you can’t get there from here” situations, like the very annoying “Yes, the SACD player has a HDMI connector, but it will not output a DSD bitstream over it” problem.

And our eternal foe, digital rights management (DRM via HDCP) weaves it’s tentacles throughout. Sony’s somewhat arbitrary copy protection rules for DSD, coupled with HDMI’s flaky copy protection schema (HDCP) make things even more bizarre to understand, and difficult to troubleshoot. For example, hot plugging HDMI cables doesn’t always “reset” the connection, you have to unplug both system units from the AC line *and* also the cables from both system units and wait for a few minutes to reset them.
And that’s not to mention the previously mentioned generational specification issues to contend with when you mix n match old and new equipment.
(ugh!)

Stereo vs Multi-channel:
The first SACD specification was for stereo only, and it took (and still takes) a special enhanced player to play the SACD discs. Since CD’s were still king, SACD playback was offered on combination CD/SACD players for convenience. [EDNOTE: research history of when SACD/CD hybrid discs happened]

Time marches on, and discrete surround sound was again a thing. Dolby AC-3 (later renamed “dolby digital”) and dts improved the movie experience a lot, and these movie centric surround encoding systems started getting used for audio only surround use too. Cool.

Content vs delivery

That brings up an interesting point: A surround sound delivery system also needs surround sound source material, and over the years there have been a variety of methods employed to produce this multichannel content.

One such way is “accidental”, in the sense that the original recording done was on a multi-track recorder with the various parts separated out on different tracks. Although those multi-channel tape tracks were intended to be mixed down to a stereo master, then distributed as stereo content, it’s often possible to go back to those multi-track recordings and remix for a surround format (eg dts 5.1, or mutli-channel SACD, or Dolby Atmos) and create a new surround sound master for distribution.
(Opinion: The results of this “re-mix” can range from wonderful to weird, depending on the music, the tastes of the engineer, and the artist.)
Steve Wilson is well known for doing a good job on this kind of remixing.

For example, many of the vintage but still well regarded “Living Stereo” recordings were done using a 3 track tape recorder with mics at Left, Center, and Right. At the time, these recordings were mixed down and released on LP as stereo, but it was an easy transition to remix for a 3 channel SACD release. (The LFE and surrounds are not used in this case.)

Another set of examples are the “multitrack source to quad” mixes from the early 70’s. Recordings like the Alan Parson quad mix of Pink Floyds “Dark Side of the Moon” and Mike Oldfields “Tubular Bells” mix were intended for delivery in one or more of the multiple competing quad formats of the time. Neither of these “quad” releases use the LFE or the center channel in order to stay true to the original artistic intent of the times.
(For more on these now obsolete LP formats, reference CD-4, SQ, and QS).

Note that these examples also show that not all multichannel playback channels will necessarily contain audio, either.

Ahem.
Another way to get multichannel content is to record it with the intent of a multi-channel distribution from the beginning. This is typically used for “live” recordings of classical music in halls with good acoustics, and the surrounds are usually the hall ambiance.
And Pink Floyd actually produced concerts in surround, some of their recordings reflect that concept.

Lastly, you can get pseudo surround sound content in “real time” from 2 two channel sources via various synthesis or “upmixing” techniques. An example of this is the dolby pro-logic II upmixer, pioneered by Jim Fosgate, still in use today.
dts sports their Neo:6 upmixer, and there are lesser known others.

Anyhow, to compete with the other surround delivery formats, a follow on SACD spec was created which allowed for a maximum of 6 channels… but without specifying how those 6 channels are mapped to physical speaker locations.
Most SACD discs nowadays assume those 6 channels will be mapped to the now standard 5.1 “movie” speaker layout, but there are some discs that assume some oddball speaker mappings(1), like what Telarc, Chesky and MDG have experimented with. Playing these uniquely mixed discs properly will require repatching your signal path for the changed assumed speaker layout.

Let’s start with the simplest complicated case, 2 channel analog outputs from the player

2 Channel Analog connections:
There are quite a few players that can output 2 channel (or stereo) analog audio read from the SACD stereo area (or possibly downmixed from the multi-channel area.) If the DACs in your player are good, this can be a very good solution for a stereo only system. Beware of menu settings in the player, some players will default to playing the PCM CD layer on hybrid SACD discs!
For example, I have a well regarded Denon DVD player I use for “stereo only SACD” on my 2 channel audio reference system.

Multi-channel Analog connections:
Although there are indeed a few players that can output multichannel from the SACD multi-channel layer over the analog connections, results can be less than optimal depending on where and how your bass management is applied for your speakers.
And, as of 2023, I know of no combi players that have Atmos or dts:X decoding built-in, and there are very few with even DTS-HD Master Audio and Dolby TrueHD.
So using a players analog multichannel outputs for SACD will wind up excluding playback of some of the other surround formats. Also, be aware some early DVD player mutli-channel analog outputs may not actually output SACD multi-channel audio, another one of those generational spec misalignments. (ack)

Digital connections:

As of 2023, I am aware of three ways to get a digital DSD64 file out of a SACD player.

One: via HDMI cabling, typically to an AV pre-pro or receiver’s HDMI input. This uses the pre/pro’s processing to do the decodes and conversions. It works quite well for both stereo and multichannel SACD’s and the various other audio formats too. This is the simplest and most versatile connection method.
One “point five:” There is a HDMI “fringe” sub case, where you can use an HDMI Audio extractor to extract the DSD signal from the HDMI connection and send it off via S/P-DIF to a DAC that supports DSD decode from it’s S/P-DIF inputs. Unfortunately, the lower cost HDMI audio extractors do *not* generally support DSD extraction, but the fairly expensive GeerFab DBoB does. There a lot of subtleties in getting this to work and it has the drawback of being stereo only, AFAIK. If interested, best read up carefully on this topic.
Two: via “ripping” the SACD’s DSD64 data to a computer file, usually suffixed with .DSF or .DFF, see: https://www.psaudio.com/blogs/copper/down-the-rabbit-hole-of-sacd-ripping-and-dsd-extraction on how this can work.
Three: via I2S, a chip level DAC interconnect method occasionally hacked into service (which confusingly, may use the HDMI connector for output but is not compatible with a general use HDMI signal) to a DAC with a possibly matching I2S over HDMI input. (See PS Audio’s site for more on this.) I2S will work to send two-channel native DSD64 (the digital SACD signal type) to a stereo DAC. I am unaware of I2S equipped devices that decode SACD surround sound, much less the other multichannel A/V sound formats, like dts, Dolby, Atmos, etc.
Although theoretically possible, I am not aware of any SACD players that will *output* a DoP signal over USB via emulating a USB host interface.
(Opinion: No big loss, USB is a poor way to transport audio anyways.)

A note on S/P-DIF: (Sony/Phillips Digital Interface Specification)
The optional S/P-DIF digital output (optical, coaxial, or it’s close relative, AES over XLR) ) on a SACD capable player causes a lot of confusion. Due to content management restrictions enforced by Sony licensing, S/P-DIF is not permitted to pass a DSD bitstream, it will only output the SACD’s CD layer PCM digital data. So if you are playing a disc with only a SACD layer (not hybrid) and the only connection between your player and receiver/DAC is via a digital coax or optical cable, you won’t hear anything.
UPDATE: Confusingly, some newer players allow you to do a DSD64 to PCM 176.4 /16 down conversion internally, and that *converted to PCM* signal may then be output over a S/P-DIF or HDMI connection. I believe this conversion will only be 2 channel, however. (References, anyone?)

I did tell you this was complicated. And it’s all so very annoying when you just want to listen to your music!

Terms:

– PCM: “Pules Code Modulation”
A baseline, standard, uncompressed, digital audio signal format. (The .WAV file format is basically PCM data with a descriptive header in front.)
Other digital audio signals can be imbedded inside a PCM signal, the resultant signal is referred to as a “bitstream”. This is the technique is what dolby digital, dts, Atmos, etc. use for transport.
A key point is a bitstream must be passed along “bit-perfect” (i.e. not be modified in any way), otherwise the encapsulated audio information will be corrupted and unusable.

– DSD – “Direct Stream Digital”
This is the audio signal format used by, and on SACD’s.
SACD discs are bandwidth limited to DSD64, but later versions of the DSD spec allow a higher sample rate for use with computer files and USB, eg DSD128, DSD256, etc.

– DoP – “DSDxx over PCM”
A way to encapsulate DSDxx data inside a PCM data frame for transport. Originally created for use for USB transport, theoretically it could also be used for a S-P/DIF over optical or coaxial connection if DRM wasn’t in the way. This packaging technique is also used for DSD computer files, .DSF or .DFF.
[EDNOTE: HDMI may also use DoP encapsulation, need to research]

– S/P-DIF – “Sony/Phillips Digital Interface Specification”
S/P-DIF is a now ubiquitous method for transporting digital data between devices. It’s basically a consumer version of the original 1985 AES/EBU spec from the pro audio world, differing primarily in the lower cost connectors and cabling used.
Of special note, S/P-DIF (and AES/EBU) connections have bandwidth limitations, which impacts the max sample rate and use for multi-channel applications.
For convenience, I’ll use the term S/P-DIF to cover all three cabling types and the protocol.
If you’re fascinated by this bit of arcana, look up the relevant IEC specifications.

– HDMI – ” High-Definition Multimedia Interface” or “One cable to rule them all”
A one cable solution for querying system features (EDID), transporting Audio and Video signals, enforcing copy protection (HDCP), and transporting oddsnends, like ethernet, CEC, ARC/eARC, and a few I’ve forgot.
Some notes:
HDMI cables, plugs, and jacks are all notoriously fragile, take care to not stress them or bend the cables tightly.
The generational version changes of this collection of specifications make for interesting times when hooking up devices. An oft overlooked issue is the cables themselves have undergone feature and bandwidth changes, ie early cables simply don’t work at higher resolutions and frame rates. For example, supporting a 4K video signal requires all devices in the chain support a minimum of 18GBps, HDMI 2.0, and HDCP 2.2, even the cabling. (!) And issues with CEC are common.

– EDID – “Extended Display IDentification“
A standard thats been around since the early VGA display days, and it still live on in HDMI. Its job remains the same, it allows a display device to publish it’s feature set up to a renderer. The relevance here is audio features are also negotiated by this protocol, and a low feature monitor can disable surround sound if the receiver isn’t configured properly.

– HDCP -“High-bandwidth Digital Content Protection”
The digital rights management schema used by HDMI, among several other standards.
It too has had multiple generational changes, and is yet another reason why things may not work. A HDCP error usually presents as a screen full of confetti, but can also be a blank screen or no sound.

The key factor here is HDMI & HDCP use a challenge/handshake metaphor that starts with the source querying the HDMI connection chain all the way to the endpoint (typically a monitor). This process determines the feature set and security of the system, and it is why turning on your system monitor first, and then proceeding back up the HDMI connection chain to your source is so important. It is also why different components spec level support may have odd effects on how things will function.

Bass management

Bass management is tricky, both to understand and to implement in mastering and playback. Unfortunately, the same terms are loosely used for both the source content mastering and the playback system and settings, leading to more confusion.

The first common misconception is about the “.1” subwoofer channel in the 5.1 nomenclature. The .1’s proper name is LFE, for “Low Frequency Effects”, and it was (and is) designed to convey loud sound effects for movies (Boom! Crash!), and not be a part of the soundtrack.

The second common misconception is that bass management and the LFE channel is the same thing, it is not.

As part of the authoring process, surround encoded music discs can (and do) ignore the LFE channel. Or the mastering engineer can choose to place a summed low passed bass track there, either way can be considered technically correct.

As part of the playback process, the processor can choose to pass all channels straight through (often called direct or pure mode), bypassing any speaker eq or bass management.

Or the playback processor can implement bass management, where the low frequencies are removed from some or all of the main channels and redirected to a subwoofer output, usually using the LFE subwoofer for double duty.

The key point here is the authored media and the playback system can have differing ideas on what to do with bass, and sometimes you need to adjust things for best results.

(end of article)

References:

https://en.wikipedia.org/wiki/Direct_Stream_Digital

https://forums.stevehoffman.tv/threads/dsd-over-pcm-dop-idiots-guide.1174925/

https://en.wikipedia.org/wiki/S/PDIF

https://en.wikipedia.org/wiki/High-bandwidth_Digital_Content_Protection

https://en.wikipedia.org/wiki/HDMI

The “SACD & DVD Audio Enthusiasts” forum on FB

Decades of audio, video, and surround sound experimenting and hard knocks

Content vs delivery

Let’s start with the simplest complicated case, 2 channel analog outputs from the player

Digital connections:

Terms:

References:

Other fringe topics to discuss/research