The advent of HDMI brought to the digital arena all the bells and wistles of the high tech video and audio industry. Particularily in the audio sector, HDMI makes it possible to go a step closer to the original analog signal.
Wait! Analog!? Didn’t everybody say Digital was better?
Well, it is, in a way. See, technically, video is represented on screen pixel by pixel. Until LCDs came around, pixels weren’t quite squares on the screen but interpreted stuff from analog signals. So yes, in terms of visual accuracy, digital is better. Square for square, pixel for pixel, data for data, simply the pure stuff.
However, it is still argued whether a digital cinema camera can beat a film camera because of the natural grain on the various kinds of film, but that has nothing to do with our case. Why? Because DVD and Blu-ray, both digital formats, have visual data stored in pixels, 100% digital 1s and 0s. And since your LCD (HDTV or not) uses a physical grid to represent pixels, digital is better (no conversion, just pure video straight out from the DVD/Blu-ray player, provided you are using a digital output such as HDMI).
But the sound realm is something completely different. Sound is composed of waves. The way speakers and headphones, well, anything that makes sound actually, gets to output what you hear is by vibrating some membranes, a lot like your eardrum vibrates to the sound it receives, which is then transmitted to the brain for analysis.
So whatever you do, however sophisticated your system may be, the end of the line is always analog because the speakers are. Now, to represent sound in digital form, as analog signals cannot be stored on digital media like CDs and Blu-rays, you have to interpret them into bytes. This is where it gets geeky. But basically, all you have to know is that no digital audio format is able to interpret at 100% the analog signal. This is negligeable though because an analog signal degrades over distance and what not. Digital on the other hand still has the benifit of being lossless, even if carried accross multiple devices. The closer you can bring your digital stream to the speakers without converting it, the better, which is just what HDMI does.
There are, in general, two big different ways to interpret audio digitally: PCM (Pulse-Code Modulation) and DSD (Direct Stream Digital). Here is an image from Wikipedia, explanations afterwards.
Hmm, fancy graphics…
So, the white line you see is the famous audio wave, more specifically called Waveform in this case because it is the form the audio wave makes, clever. The green lines are how the data is represented digitally.
PCM is fairly easy to understand. Just looking at its graphic provides basic understanding. Common flavors of PCM come in 16 bit resolution. Each bit represents either a 1 or 0, this is why the total amount of possibilities is 2 and X number of bits (2X bit). This means that for each sample, there are 65′536 steps, represented from top to bottom. Think of a sample as a frame in video, it’s multiple images stringed together to make video, but in this case, it’s multiple bits of different sound frequency values stringed together to make sound!
PCM is also defined by its Hz. This is how many samples there is per second. Common flavors of PCM come in 44.1 kHz (k = 1000). So, 44′100 samples per second at 16 bit resolution.
The PCM graphic shown above shows a largely imprecise drawing of how PCM works. Here is a more accurate representation in 4 bit.
While looking at a PCM graphic though, its weakness jumps to the eyes. Sinusoidal equations (the waveform) represented by steps that can be thought as a bar graph, means that some precision is ultimately lost and some frequencies, especially pulses, can only barely be represented by the highest audio resolution and sample rate, which is where DSD comes in.
Oh, another fancy graph (brought you by an analysis from Merging Technologies). As you can see, PCM, the three bars in the middle, is not very good at representing impulses. Whether this really helps DSD attain better quality is highly debated. But the theory is DSD is closer to the original analog signal than PCM. Why is that so?
Ok, if we remember correctly what we were going over, audio or Waveforms are mathematical sinusoidal calculations. Well, in reality, much more complex audio is a combination of those or other basic functions. But that is not the point.
DSD is represented over only 1 bit and is sampled at a much higher rate at 64 times that of 44.1 kHz stereo audio: 2.8224 mHz (m = 1′000′000) or 2′822′400 samples per second. The single bit per sample represents either a 1 or 0, or positive or negative value in this case. Along with an algorithm to interpret it, the mix match of positive and negative values makes it is possible to store the audio data as shown above. The advantage of DSD is that audio data is made back from algorithms that more closely represent the amalgamy of sinusoidal equations that form an audio wave. This less visual and more mathematical way to store information makes it possible to convert the digital data back to a signal closer to the original signal that came from the microphones.
However, DSD has less hardware capable of transferring it over digital streams. The reason is simple. While the industry was finding ways to digitize audio transport to lose less quality over the multiple devices required for surround home theater technology with efforts like optical audio over Toslink (SPDIF, Sony/Philips Digital Interconnect Format), Sony (ironically the same company who co-developed SPDIF) introduced the SACD with DSD technology (DSD was also co-developed by Sony and Philips) in an analog fashion.
The motive to make SACD a primarily analog media was simple: make it compatible with current receivers. In 1999, a time where DVD was just starting out, next to no-one had a digital-capable receiver, it wasn’t quite around yet. This and the fact SACD requires new hardware to convert into an analog signal is why even multi-channel SACD was being outputted on analog. Additionaly, SACD doesn’t allow unencrypted streams to be played back digitally, so only i.Link and HDMI can transport DSD streams.
Today, the HDMI specification allows the transport of SACD streams but even Sony’s only SACD stand-alone player still sold in America doesn’t support passing SACD streams through HDMI (some higher end models from Pioneer, Onkyo and others do support it). Instead, DSD streams are converted and then passed as Stereo CD quality content (16 bit @ 44.1 kHz), far from SACD’s potential. SACD fans hoping to pass DSD streams through the Playstation 3’s HDMI port are probably out for luck for ever.
But, luckily somehow, SACD is a primarily analog format. Theoretically, any SACD-capable player can output DSD through analog, which, simply put, is outputing the signal only, not the actualy digital stream.
Since we’re talking analog, the PS3 can output the full signal of any of its audio formats, including SACD and Dolby TrueHD. However, one of your devices (either the player or the receiver) must provide bass management if you’re using a satellite + sub system. The PS3 does not provide bass management. But the PS3 can output your video through HDMI, so you have full 1080p HD quality and full audio quality through analog.
Keep in mind the PS3 is limited to Dolby Pro Logic II when outputting surround through analog. I couldn’t find out whether Dolby Pro Logic IIx is available on the PS3 (for 6.1 and 7.1 surround sound) but one thing’s for sure, no matter how far from discrete channels Pro Logic is, you’ll still get the full audio signal from stuff like Dolby TrueHD Audio, which is interesting to know for Blu-ray playback on older non-HDMI equipped systems.
And yes, the PS3 really outputs all of its audio formats in full audio resolution on the analog outputs (the signal is directly converted by the PS3 from source), Dolby TrueHD Audio included. The quality is nothing short of amazing if you have the proper receiver, cables and speakers of course.
