Shoveling Something…But Not Snow!
I grew up in a suburb outside of Detroit. I can appreciate the cold temperatures and piles of snow that are displayed nightly on the national news. Poor Boston…storm after storm has been dumping the fluffy white stuff on the city as crews try to keep streets open by hauling it away or even melting it by the ton. And now I live along the coast in Southern California. The temperature today is in the mid 80s, the sun is out and hordes of people from the valleys are crowding the beaches just down the hill from my home. It gets hard to relate to my friends in the Midwest…why continue to suffer through terrible winters when you can move to a city with year round great weather?
The recent challenge by Cookie Marenco of Blue Coast Records and Paul McGowan of PS Audio is obviously going nowhere. While I believe that Mario Aguilar and David Pogue wrote really bad articles based on outdated facts and very poor research about high-resolution audio and specifically Pono, their conclusions weren’t that far off the mark. The “myth of high-resolution audio” persists because there is so much misinformation being spread around as fact.
One of the tabs on my browser is till open to an informational page on the SHARP electronics website (Click Here to visit “The Benefits of 1-bit”). I happened upon the page after participating in the roll out of the Sharp high-resolution wireless system a few months ago. You got to love the paragraph that is presented on the page:
“A CD’s PCM multi-bit technology records and reproduces sound frequencies up to 20 kHz, the audible limit for humans. But we can actually perceive sounds above 20kHz, and some sounds even reach frequencies of up to 100kHz. To record and reproduce the full range of natural sounds, then, a 1-bit amplifier samples the original analogue signal an incredible 2,822,400 times per second (2.8224MHz) — 64 times faster than PCM (44.1kHz). 1-bit thus achieves far superior sound reproduction and transient characteristics. And, with our latest generation 5.6448MHz 1-bit amplifiers, the bar has been raised higher still.”
Figure 1 – Sharp illustration showing the “benefits” of 1-bit high-resolution encoding.
You have to love the marketing pitch by Sharp. While I agree that hearing may extend past 20 kHz, it has certainly not been proven. And even the McGill research “Sampling rate discrimination: 44.1 kHz vs. 88.2 kHz” states that “While we observed audible differences between sample rates of 88.2 and 44.1 kHz, they remain very subtle and difficult to detect. This statement about 100 kHz is technically true but just how high and at what amplitude do you think is appropriate for music recording and reproduction? Do you really want a 75 kHz tone at -120 dB as a “low level detail” in your playback?
The advocates of DSD like to boast about the 100 kHz frequency range…but they don’t tell you that to achieve the extended range you have only 6 dB of dynamic range. When you push the noise out to the “audio band” (we’re back to 20-20 kHz) to get decent dynamics, the high frequency extension goes away. Let’s talk facts for a while.
According to the paragraph, the analog to digital conversion is done using a “1-bit amplifier” sampling the incoming analog waveform at an “incredible 2.8224 MHz, which just happens to be 64 times the sample rate of the Redbook CD spec). They want us to believe that 1-bit therefore “achieves far superior sound reproduction and transient characteristics.” It doesn’t. If any of this were true, don’t you think that the record industry would be swapping out their lousy PCM systems for the latest DSD workstation? They aren’t and they won’t be. It’s simply impossible to buy into DSD when you can’t work with it in its native form and it doesn’t deliver anything better than what we already have. Sorry Sharp…time to talk to the engineers and producers that make the records you’re reproducing.
They continue…
“1-bit’s reputation is based on its ability to reproduce sound with unprecedented accuracy. 1-bit heralds nothing less than a whole new era in audio.”
Cookie told me that there is a firmware update to your Pono player that allows DSD playback. Now I suppose Neil will re-encode his 2.1 million tracks to DSD to extract additional dollars from DSD lovers.
I have always been mystified by the idea of 1-bit sampling. PCM, at 16 or 24 bit, encodes how loud the signal is at that moment in time. What can 1 bit encode? The package inserts with early SACDs explained that it asked, millions of times a second, “Is it louder than it was last time?”
If this is the case, the system allows for two answers: “Yes, it’s louder” or “No it’s not.”
How much louder? What if it’s quieter?
That explanation is clearly nonsense, and I’ve never figured out how a two-valued sample – which is all that a 1-bit system can encode – can provide anything but square waves, which is probably why there’s so much supersonic noise that the decoder needs to include a “brick wall filter” to protect the power amplifier and speakers (and our ears).
Can you hear me now?
Phil, using 1-bit samples at very high sample rates does work…in fact, virtually all modern ADC and DACs use Delta Sigma in their designs. The two states essentially monitor the direction of the waveform (with a feedback loop). Because the samples are taken so frequently, there question of amplitude…or how much…is moot. The system simply describes the original waveform. That’s why advocates for DSD often say…it’s almost like the analog wave. DSD is a completely impossible method for producing commercial recordings. Most (85%) of the DSD/SACDs that exist are not native DSD.
Curious as to your thoughts on this post by Charles Hansen @ Pono:
“Hello All – FLAC is an efficient way of storing PCM (Pulse-Code Modulation) files. PCM samples the analog signal at a rapid rate, and assigns a value (the higher the number of bits, the more resolution in that value). When played back, it can regenerate the analog signal.
The higher the resolution, the more closely it can track the value. The sampling rate is a bit more nebulous. Everyone knows that a higher sampling rate will allow higher audio frequencies to be captured.
The question is, “Why capture frequencies beyond the range of human hearing?” There are two answers:
1) We are just finding out now THAT THE TRADITIONAL UNDERSTANDING OF HUMAN HEARING IS WOEFULLY INCOMPLETE. It turns out that throughout the animal kingdom that there are QUANTUM-BASED sensors are used extensively.
This for example is what allows birds to migrate thousands of miles and arrive at their precise destination — in other words, THINGS THAT OUR NORMAL “SCIENCE” CANNOT EXPLAIN. The exact same is true for our hearing.
Don’t believe anybody that flat-out claims “Humans simply cannot hear this or that.” You all know what you can hear…
2) The real benefit of high sampling rates is that it gives us much more flexibility in designing the filters required for digital audio. Traditional filters that have been used in ALL digital audio equipment “smear” the time information.
The human ear/brain is EXQUISITELY sensitive to time information, as this has what has allowed us to survive. Identifying the location of either prey or predator by sound alone is key. Humans can differentiate time delays as small as 1 to 10 micro-seconds.
Normal digital filters completely smear this time information, destroying the naturalness of the sound. Ayre has worked for years developing digital filters that preserve as much time information as possible. These went straight into the PonoPlayer.
~~~~~~~~
DSD is not an engineering term. It is a marketing term invented by Sony. So it means whatever Sony wants it to mean. They have defined it as a 1-bit digital signal recorded at 64x the CD sampling rate (later expanded to include 128x CD also).
DSD uniformly sounds very good because there is NO FILTERING on the record side, and very little filtering on the playback side. The difficulty with DSD is that it requires the recording and mastering studios to purchase expensive new equipment and learn new ways of using it.
~~~~~~~~
With the PonoPlayer, the advanced digital filters used allow for OUTSTANDING sound from PCM, even at the single sample rate. Double sample rates are better and quad is the best (virtually no time smear).”
Charles get most of this right and I agree with his sentiments on ultrasonics. However, he’s not completely on the mark when talking about DSD. His comment, “The difficulty with DSD is that it requires the recording and mastering studios to purchase expensive new equipment and learn new ways of using it,” makes not sense. The problem with DSD is no matter what equipment professional studios acquire, there are no tools that allow one to work natively with the format.
DSD uniformly sounds very good because it has got much better impulse response as sampling rate hugely increased
Actually, until quite recently, it was still unknown that oversampling recreates impulse response after the event {recording}. Thus far, any fully digital CD track is potentially a much more natural sound than master tape, LP, etc., and Infinite Oversampling should purport the Ultimate Sound Quality. Hence, it is insane to record at higher than 44.1 kHz as it would always mean less quality increase than that of software oversampler.
Jay…I’m going to stick with other professionals and keep recording at higher sample rates…call me insane.
I keep wondering if DSD is really just a delta modulator. I built one as a senior project in 1979, and although mine worked to a certain extent, the voltage comparators of the day didn’t allow for a very high clock speed (I don’t remember what mine maxed out at, but I did pass recognizable voice).
A delta modulator is basically an up/down counter with its output going to a DAC, and that output going to one input of a comparator with the other input being the signal being digitized. There’s only one bit doing the controlling, telling the counter to go up or down based on how the input and output compare. It seems like this is a similar scheme, but I haven’t been able to dig up anything that tells me how DSD compares to a delta modulator.
It is Pulse Width Modulation, a delta sigma modulator spewing out a stream of digital ones and zeros that can accurately represent an analog waveform. The problem comes when you try to do anything with the native stream.
I’m huddled up inside this evening in Boston, while we are getting yet ANOTHER 12-15 inches of snow. Because apparently the first 3 storms weren’t enough. MAKE IT STOP!!
I wonder how much more rediculous the lies will become over the next 10 or 15 years as the snake oil peddlers look for ever new make believe technology to sell us recordings that sound little improved over what my beloved 60-80s rock does on the CDs I bought 30 some years ago. MFSL did some good sounding remasters back in the day, but today’s remasters really are only super compressed horrible sounding loudness war releases, or some snake oil BS tech who’s only reason for existence is to get deep in your pockets selling you some greatly overpriced POS.
Mark here and a few others due some incredible sounding recordings that are well worth the price if you can find anything in their catalogs that meets your musical taste.
Mark, It never fails to amaze me how damaged my brain must be that I actually spent 60 years living in the frozen hell known as Chicago. The last 5 years here in Central Florida have been like paradise. Tonights NASCAR 75 in Daytona was a gas. WOOT
I lived in Southern California for 10 years. I actually ended up getting sick of all the sunshine and really missed the change of seasons. Ended up moving to central Ohio and never looked back. LA was always this bizarrely surreal place where nothing ever seemed quite as it was. Probably explains why all the big studios and record labels are there! 😉
Sick of all the sunshine? Not me. I live near the ocean and love the foggy morning…knowing it will clear off later. I can appreciate the beauty of the seasons but I choose to visit the seasons not experience them everyday.
A.S. ‘On July 21st, 2012, at 8.20 a.m., the city of Jask {Farsia} experienced an unprecedented heat in known history: +34°C air temperature at 100% relative humidity .’
“why continue to suffer through terrible winters when you can move to a city with year round great weather?”
The warmest place on Earth is Pulo Anna Island {Palau} .
Now, to the point:
* 2048+ Oversampling should be applied during recording at 44.1 kHz to completely avoid anti-alias filtering
* Extreme Oversampling can be applied to a properly noise-shape-dithered recording during playback in order to not only exclude anti-alias filtering again but make the sound incomparably more natural than any analog source could ever deliver
Seems that Paul of PS Audio has even one-upped Sharp in his forever novelistic techno-blurb marketing.
Writing about his new DSD DAC: ” talk about providing multiple versions of the Pikes operating system for DirectStream. Good conversations. Stimulating talk. The idea centers around tuning the OS to each person’s system. Create multiple OS releases, some warmer, some cooler, others with stronger top ends, still others with better bass. It’s an enticing idea but one with a number of problems.¨
Obviously no concept whatever of high fidelity – just personal tweaking with factory incentive, but OMG, of the OPERATING SYSTEM of a DAC !?! This is just so hilarious if not sad – but reading Paul always makes me re-quote Billy Joel on Pianoman:
“Paul is a real estate novelist” => “Paul is an audiophile noveiist” , though he certainly seems to have the knack of capturing clients money in exchange for snake oil… meanwhile writing science fiction all round.
Oh boy…tweaking the operating system? Hard to imagine how far some will go to pitch their products. I’ve seen articles on different sounding “NAS”.
Speaking of shoveling, off topic: what do you think about this from Mr. Hansen at Pono:
“Cardas is my personal favorite cable. Just be aware (that like the PonoPlayer) Cardas cables take a good week (or two!) of playing on repeat to break-in.”
I’ve never bought in to the “break-in” thinking for equipment and/or cables.
Hooray! As an audiophile, I have to bite my lip while fellow-audiophiles go on about cable break-in (and capacitor break-in, preamp break-in, DAC break-in) incessantly.